CN104580143A

CN104580143A - Security authentication method based on gesture recognition, terminal, server and system

Info

Publication number: CN104580143A
Application number: CN201410625738.7A
Authority: CN
Inventors: 李若斌; 臧存勋
Original assignee: Individual
Current assignee: Individual
Priority date: 2014-11-09
Filing date: 2014-11-09
Publication date: 2015-04-29

Abstract

The invention belongs to the technical field of identity authentication and discloses a security authentication method based on gesture recognition, a terminal, a server and a system. The security authentication method based on gesture recognition comprises the steps of registering user feature information and certifying user identity, wherein user feature information registering comprises the following steps: acquiring a preset palmprint image of a user, extracting palmprint features from the preset palmprint image and storing the preset palmprint features; user identity certifying comprises the following steps: initiating an authentication request by the user, generating a palmprint hand shape sequence according to the preset palmprint image, sequentially showing the user a hand shape image, sequentially performing corresponding gestures by the user according to the shown hand shape image, acquiring the authentication hand shape image with palmprint of the user, extracting the handshape and palmprint features from the authentication hand shape image, matching the hand shape and palmprint features with the preset hand shape and palmprint features, and feeding back the matching results. According to the invention, the possibility that the identity information is hijacked is reduced to a greater extent, and the identity authentication security is improved.

Description

A kind of safety certifying method based on gesture identification, terminal, server and system

Technical field

The present invention relates to a kind of safety certifying method, terminal, server and system, particularly a kind of safe-conduct method based on gesture feature identification, terminal, server and system, belong to communication security authentication techniques field.

Background technology

Along with the fast development of communication network, each corner that mobile Internet is more and more deeply social, various service is also increasing.The things that the shopping of video, mobile terminal, mobile-phone payment etc. have become people and know seen by surfing Internet with cell phone, mobile phone.How to carry out reliable long-distance identity-certifying in a communication network to seem and become more and more important, the long-distance identity-certifying of high safety has become prerequisite and the necessary condition that Internet Service Provider provides various service.

At present, the main method of existing long-distance identity-certifying comprises: the various combinations etc. of text password, dynamic password (or password calculation device), smart card, biological characteristic or said method.When adopting the mode of text password to carry out long-distance identity-certifying, text password sends to the certificate server of communication network to carry out certification by terminal, and which not only requires that user remembers various loaded down with trivial details text password, and text password is also easily lost and stolen.In order to avoid this complete problem, have developed the mode of verifying dynamic password, because dynamic password is dynamically generated by server end when each authentication, therefore can prevent lawless person from carrying out authentication when obtaining user cipher, but there is the problem of easily losing in dynamic password card (or password calculation device), still can not safety certification when dynamic password card or (or password calculation device) are separated with user.

In order to improve fail safe further, it is found that and adopt the intrinsic biological characteristic of human body or behavioural characteristic to be relatively safe authentication method, wherein face, iris, fingerprint, refer to that vein and vena metacarpea identification etc. have been widely used in gate inhibition, the fields such as finance.At present, the payment authentication system commonly used at mobile terminal adopts the mode of short message verification code to prevent user cipher stolen, but when lost terminal or SIM card are replicated, the fail safe of user profile and property is guaranteed being difficult to.A lot of lawless person had occurred copies SIM card at present, and then makes the checking short message sending in mobile payment process to the terminal of lawless person, finally steals the case of user's wealth.

Be in the Chinese invention patent application of 201410138350.4 at application number, disclose a kind of Mobile payment identity verification method based on somatic fingerprint and iris feature.The method is safer compared to the independent verification method based on fingerprint or iris, but when certification terminal is held as a hostage, still there is the risk that user identity is forged or palms off.

Summary of the invention

The technical problem to be solved in the present invention is: provide a kind of safety certifying method based on gesture identification of high security, terminal, server and system, effectively can prevent the risk falsely used and be held as a hostage, improve the fail safe of authentication.

For realizing above-mentioned goal of the invention, the invention provides a kind of safety certifying method based on gesture identification, terminal, server and system.

On the one hand, the invention provides a kind of safety certifying method based on gesture identification, comprise the step of registered user's characteristic information and the step of authenticated user identity;

The step of described registered user's characteristic information comprises:

Gather the preset palmprint image of user;

Palm print characteristics is extracted to preset palmprint image and saves presets palm print characteristics;

The step of described authenticated user identity comprises:

User initiates authentication request;

Generate palmmprint hand shape sequence by preset hand shape image and sequentially show hand shape image to user;

User sequentially does corresponding gesture motion by the hand shape image shown and gathers the certification hand shape image of user with palmmprint;

Hand shape, palm print characteristics mating with preset hand shape, palm print characteristics are extracted to certification hand shape image;

Feedback matching result.

Wherein more preferably, it is characterized in that: the step of the preset palmprint image of described collection user specifically comprises:

Smooth palm facies palmaris is positioned over image capture position by user;

Gather user's palm facies palmaris image;

To palm facies palmaris Image semantic classification.

Wherein more preferably, the described step to palm facies palmaris Image semantic classification specifically comprises:

To the palm facies palmaris image coarse segmentation process gathered;

Palm facies palmaris display foreground after coarse segmentation process and background are distinguished and equilibrium treatment;

To the palm facies palmaris image median filter process after equilibrium;

Binary conversion treatment is carried out to the prospect hand shape region of the palm facies palmaris image after medium filtering;

Make connected region to the palm facies palmaris image after binaryzation to detect, obtain the hand shape image of band palmmprint;

Rim detection is carried out to the palm facies palmaris image of the hand shape image of band palmmprint and obtains the hand shape region after binaryzation.

Wherein more preferably, the step of the described palm facies palmaris image coarse segmentation process to gathering specifically comprises:

Palm facies palmaris image is carried out color space conversion, is converted to HSV space from rgb space;

Utilize the area of skin color of H component extraction hand.

Wherein more preferably, described palm facies palmaris display foreground to gathering and background distinguishes and the step of equilibrium treatment specifically comprises:

Palm facies palmaris image after coarse segmentation is converted to gray level image; Region beyond area of skin color is set to background area;

Histogram equalization process is carried out to palm facies palmaris image and obtains the gray level image after equilibrium.

Wherein more preferably, described preset palmprint image or certification hand shape image are by camera collection.

Wherein more preferably, the step of described extraction palm print characteristics specifically comprises:

Locate the facies palmaris region in preset palmprint image;

Cut the facies palmaris region in preset palmprint image;

Extract the palm print characteristics in preset palmprint image.

Wherein more preferably, the step in the facies palmaris region in the preset palmprint image in described location specifically comprises:

A), to preset palmprint image horizontal threading edge image from top to bottom, one group of threading including 8 boundary points is found; Using adjacent 2 boundary points as one group, totally four groups; Using these four groups as the initial threading group being four fingers;

B), for each in four fingers refer to, with initial threading group for original position, search for respectively up and down, until search up-and-down boundary;

C), in the same way, horizontal threading edge image from bottom to up, finds one group of threading including 4 boundary points;

D), same to step b), refer to, for original position, search for respectively up and down with the initial threading of thumb, until search up-and-down boundary, and then orient thumb key point.

Wherein more preferably, the step in the facies palmaris region in the preset palmprint image of described cutting specifically comprises:

Find out the key point c between the key point a between forefinger, middle finger, the third finger, little finger of toe;

Find out the mid point b that key point a and key point c forms line segment;

With b point for initial point, make the normal f of line segment ac;

Then with c point for starting point, along normal f, offset the some e of a distance towards volar direction;

Centered by e, generate a length of side and be P and the rectangular area parallel with line segment ac;

Extracting P × P region is facies palmaris region.

Wherein more preferably, described palmmprint hand shape sequence is from hand shape candidate collection, select the hand shape image of multiple different hand shapes to generate.

Wherein more preferably, described palmmprint hand shape sequence is stochastic generation.

Wherein more preferably, described to user show hand shape image with hand shape instruction illustrated form show at user terminal.

Wherein more preferably, the step of the certification hand shape image of described collection user with palmmprint also comprises:

When gathering the certification hand shape image of user with palmmprint, adopt hand shape area tracking algorithm opponent shape area tracking, determine that whether palmmprint hand shape image is continuous;

If the discontinuous existence of the palm of pickup area, points out user authentication failure;

If the palm continued presence of pickup area, continue to gather other palmmprint hand shape images.

Wherein more preferably, the step of the preset palmprint image of described collection user also comprises the step verified preset hand shape image, specifically comprises:

Preset hand shape image is showed user successively;

User shows corresponding gesture according to the preset hand shape image shown and gathers user's hand shape image;

Extract user hand shape image hand-shaped characteristic and with preset hand shape image comparison;

If coupling, then think that user accurately can show current preset hand shape image;

If do not mated, then think that user can not show current preset hand shape image;

Save presets the result of hand shape image checking.

Wherein more preferably, also comprise the pretreated step of certification hand shape image after the step of the certification hand shape image of described collection user with palmmprint, specifically comprise:

To the certification hand shape image coarse segmentation process gathered;

Certification hand shape image prospect after coarse segmentation process and background are distinguished and equilibrium treatment;

To the certification hand shape image medium filtering process after equilibrium treatment;

Binary conversion treatment is carried out to the prospect hand shape region of the certification hand shape image after medium filtering;

Make connected region to the certification hand shape image after binaryzation to detect, obtain the hand shape image of band palmmprint;

Rim detection is carried out to the hand shape image of band palmmprint and obtains the hand shape region after binaryzation.

Wherein more preferably, the step of the authentication image encryption to band palmmprint is also comprised after the step of the described collection certification hand shape image of user with palmmprint;

The step of the authentication image deciphering to band palmmprint is also comprised before the step of described extraction hand-shaped characteristic.

Wherein more preferably, the step of described extraction hand-shaped characteristic comprises:

The certification hand shape image of band palmmprint is compared as matching degree with all hand shape masks pre-set, finds out the hand shape mask matched, using hand shape corresponding for this hand shape mask as the hand shape identified.

Wherein more preferably, the step that described all hand shape masks with pre-setting compare as matching degree specifically comprises:

The hand shape image of extraction is filled the hand shape mask that preset hand shape image is corresponding;

Fill result and determine corresponding matching degree.

Wherein more preferably, the step that the described hand shape image by extraction fills hand shape mask corresponding to preset hand shape image also comprises:

Calculate the length and width of each finger and palm in the palmmprint hand shape image extracted;

The length and width of hand shape mask are adjusted by the length and width of finger each in palmmprint hand shape image and palm;

The certification hand shape image of band palmmprint is filled the length and width of hand shape mask;

Corresponding matching degree is determined according to filling result.

Wherein more preferably, also comprise before described feedback matching result:

To judge after certification palmprint image identification matching treatment whether the palmmprint hand shape sequence generated gathers, identifies, mated;

If do not complete palmmprint hand shape sequence, then to the next hand shape image of user's transmission user, user repeats above-mentioned steps until the palmmprint hand shape sequence generated all completes;

If palmmprint hand shape sequence completes, then search the palmprint information matching result of record, if there is the result that the match is successful, performs corresponding business operation for user and point out user authentication success by user terminal; If without the palmprint information matching result that the match is successful, point out authentification failure by user terminal to user.

On the other hand, the present invention also provides a kind of safety certifying method based on gesture identification, comprises the step of registered user's characteristic information and the step of authenticated user identity;

The step of described registered user's characteristic information comprises:

User terminal gathers the preset palmprint image of user and reaches service end;

Service end is extracted palm print characteristics to preset palmprint image and is saved presets palm print characteristics and feed back registering result to user terminal;

The step of described authenticated user identity comprises:

User terminal initiates authentication request to service end;

Service end generates palmmprint hand shape sequence transmission by preset hand shape image and also sequentially shows hand shape image to user to user terminal;

The hand shape image that user shows by user terminal sequentially does corresponding gesture motion, and user terminal gathers the certification hand shape image of user with palmmprint and transfers to service end;

Service end extracts hand shape, palm print characteristics mating with preset hand shape, palm print characteristics to certification hand shape image;

Service end feeds back authentication result according to matching result to user terminal.

Wherein more preferably, the step of the preset palmprint image of described collection user specifically comprises:

Smooth palm facies palmaris is positioned over the image capture position of user terminal by user;

User terminal gathers user's palm facies palmaris image;

User terminal is to palm facies palmaris Image semantic classification.

User terminal is to the palm facies palmaris image coarse segmentation process gathered;

User terminal is distinguished and equilibrium treatment the palm facies palmaris display foreground after coarse segmentation process and background;

User terminal is to the palm facies palmaris image median filter process after equilibrium;

Binary conversion treatment is carried out in the prospect hand shape region of user terminal to the palm facies palmaris image after medium filtering;

User terminal is made connected region to the palm facies palmaris image after binaryzation and is detected, and obtains the hand shape image of band palmmprint;

The palm facies palmaris image of user terminal to the hand shape image of band palmmprint carries out rim detection and obtains the hand shape region after binaryzation.

Palm facies palmaris image is carried out color space conversion by user terminal, is converted to HSV space from rgb space;

User terminal utilizes the area of skin color of H component extraction hand.

Palm facies palmaris image after coarse segmentation is converted to gray level image by user terminal; Region beyond area of skin color is set to background area;

User terminal carries out histogram equalization process to palm facies palmaris image and obtains the gray level image after equilibrium.

Service end locates the facies palmaris region in preset palmprint image;

Service end cuts the facies palmaris region in preset palmprint image;

Service end extracts the palm print characteristics in preset palmprint image.

A1), service end to preset palmprint image horizontal threading edge image from top to bottom, find one group of threading including 8 boundary points; Using adjacent 2 boundary points as one group, totally four groups; Using these four groups as the initial threading group being four fingers;

B1), service end for four refer in each refer to, with initial threading group for original position, search for respectively up and down, until search up-and-down boundary;

C1), service end in the same way, horizontal threading edge image from bottom to up, finds one group of threading including 4 boundary points;

D1), with step b1), service end refers to, for original position, search for respectively up and down with the initial threading of thumb, until search up-and-down boundary, and then orients thumb key point.

Service end finds out the key point c between key point a between forefinger, middle finger, the third finger, little finger of toe;

Service end finds out the mid point b that key point a and key point c forms line segment;

Service end for initial point, makes the normal f of line segment ac with b point;

Service end then with c point for starting point, along normal f, offset the some e of a distance towards volar direction;

Service end, centered by e, generates a length of side and is P and the rectangular area parallel with line segment ac;

It is facies palmaris region that service end extracts P × P region.

Wherein more preferably, described palmmprint hand shape sequence is that service end selects the hand shape image of multiple different hand shapes to generate from hand shape candidate collection.

Wherein more preferably, the described hand shape image shown to user is shown at user terminal with hand shape instruction illustrated form.

When gathering the certification hand shape image of user with palmmprint, service end adopts hand shape area tracking algorithm opponent shape area tracking, determines that whether palmmprint hand shape image is continuous;

Wherein more preferably, the step of the preset palmprint image of described collection user also comprises the step to the preset hand shape image checking of service end, specifically comprises:

Preset hand shape image is showed user by user terminal by service end successively;

User shows corresponding gesture by user terminal according to the preset hand shape image shown, and gathers user's hand shape image by user terminal and reach service end;

Service end extract user hand shape image hand-shaped characteristic and with the preset hand shape image comparison sending to user terminal;

Service end saves presets the result of hand shape image checking and feeds back to user terminal.

User terminal is to the certification hand shape image coarse segmentation process gathered;

User terminal is distinguished and equilibrium treatment the certification hand shape image prospect after coarse segmentation process and background;

User terminal is to the certification hand shape image medium filtering process after equilibrium treatment;

Binary conversion treatment is carried out in the prospect hand shape region of user terminal to the certification hand shape image after medium filtering;

User terminal is made connected region to the certification hand shape image after binaryzation and is detected, and obtains the hand shape image of band palmmprint;

The hand shape image of user terminal to band palmmprint carries out rim detection and obtains the hand shape region after binaryzation.

Wherein more preferably, the step of user terminal to the authentication image encryption of band palmmprint is also comprised after the step of the described collection certification hand shape image of user with palmmprint;

The step of service end to the authentication image deciphering of band palmmprint is also comprised before the step of described extraction hand-shaped characteristic.

The certification hand shape image of band palmmprint compares as matching degree with all hand shape masks that service end pre-sets by service end, finds out the hand shape mask matched, using hand shape corresponding for this hand shape mask as the hand shape identified.

Wherein more preferably, the described step compared as matching degree with all hand shape masks that service end pre-sets specifically comprises:

The hand shape image of extraction is filled hand shape mask corresponding to preset hand shape image by service end;

Fill result and determine corresponding matching degree.

Service end calculates the length and width of each finger and palm in the palmmprint hand shape image extracted;

Service end presses the length and width of the length and width adjustment hand shape mask of each finger and palm in palmmprint hand shape image;

The certification hand shape image of band palmmprint is filled the length and width of hand shape mask by service end;

Service end determines corresponding matching degree according to filling result.

Service end will judge after certification palmprint image identification matching treatment whether the palmmprint hand shape sequence generated gathers, identifies, mated;

Again on the one hand, a kind of safety certifying method based on gesture identification, comprises the step of registered user's characteristic information and the step of authenticated user identity;

The step of described registered user's characteristic information comprises:

The step of described authenticated user identity comprises:

User terminal initiates transaction request to transactional services end;

Transactional services end sends authentication request to service end;

Service end feeds back authentication result according to matching result to user terminal and transactional services service end;

Transactional services end performs transaction by authentication result, and to user terminal feedback trading result.

User terminal gathers user's palm facies palmaris image;

User terminal is to palm facies palmaris Image semantic classification.

User terminal utilizes the area of skin color of H component extraction hand.

Service end locates the facies palmaris region in preset palmprint image;

Service end cuts the facies palmaris region in preset palmprint image;

Service end extracts the palm print characteristics in preset palmprint image.

A2), service end to preset palmprint image horizontal threading edge image from top to bottom, find one group of threading including 8 boundary points; Using adjacent 2 boundary points as one group, totally four groups; Using these four groups as the initial threading group being four fingers;

B2), service end for four refer in each refer to, with initial threading group for original position, search for respectively up and down, until search up-and-down boundary;

C2), service end in the same way, horizontal threading edge image from bottom to up, finds one group of threading including 4 boundary points;

D2), with step b2), service end refers to, for original position, search for respectively up and down with the initial threading of thumb, until search up-and-down boundary, and then orients thumb key point.

It is facies palmaris region that service end extracts P × P region.

Fill result and determine corresponding matching degree.

Wherein more preferably, described transactional services end performs transaction by authentication result, and comprises to the step of user terminal feedback trading result:

Service end transmits authentication result to after authenticating user identification to transactional services end;

Transactional services end is according to authentication result process transaction request;

Pass through if authentication result is certification, then transactional services end performs corresponding transactional operation;

If authentification failure, then feed back the information of certification again or Fail Transaction to user terminal.

Again on the one hand, the present invention also provides a kind of security certification system based on gesture identification, comprises service end and user terminal;

Described service end is connected with described user terminal communication;

Described user terminal gathers the preset palmprint image of user and is sent to described service end, and described service end is extracted palm print characteristics by palm grain identification method and saved presets palm print characteristics;

Described user terminal initiates authentication request to described service end, described service end generates palmmprint hand shape sequence and transfers to described user terminal, described user terminal sequentially shows hand shape image, the hand shape image that user shows by described user terminal sequentially does corresponding gesture motion, described user terminal gathers the hand shape image of the band palmmprint of user and transfers to described service end, described service end is extracted palm print characteristics by palm grain identification method and is mated with preset palm print characteristics, and described service end presses matching result to described user terminal feedback authentication result.

Wherein more preferably, described user terminal comprises: data transmit-receive module, business logic modules image capture module;

Interactive interface module, data transmit-receive module, image capture module communicate to connect with business logic modules respectively, and image capture module and data transmit-receive module communicate to connect;

User initiates instruction by interactive interface module to business logic modules, business logic modules controls to send registration or authentication request by data transmit-receive module to service end, and business logic modules is controlled image capture module and gathers palmmprint hand shape image and transmitted to service end by described data transmit-receive module;

Business logic modules control data interactive module receives the palmmprint hand shape sequence of service end transmission and transfers to interactive interface module and sequentially show hand shape image to user.

Wherein more preferably, described user terminal also comprises data encryption module, and business logic modules, image capture module and data encryption module communicate to connect, and data encryption module is connected with data transmit-receive module;

The image of image capture module collection is transmitted by data transmit-receive module after being encrypted by data encryption module.

Wherein more preferably, described user terminal also comprises image pre-processing module;

Business logic modules, image capture module and image pre-processing module communicate to connect;

Business logic modules controls image pre-processing module and the image information of image capture module collection is done preliminary treatment, is transmitted by pretreated image by data transceiving unit.

Wherein more preferably, described service end comprises:

Safety certificate server with identity authentication function and database, database comprises palm print characteristics database and hand shape setting data storehouse;

Palm print characteristics database and hand shape setting data storehouse and safety certificate server communicate to connect;

Safety certificate server comprises data transmit-receive module, hand shape identification module, personal recognition module and dynamic hand shape generation module;

Described data transmit-receive module is connected with hand shape identification module, personal recognition module and dynamic hand shape generation module respectively, described hand shape identification module, dynamically hand shape generation module are connected with described hand shape setting data storehouse respectively, described personal recognition module and palm print characteristics DataBase combining;

During registered user's characteristic information:

The preset palmprint image received is transferred to personal recognition module and extracts palm print characteristics and be kept in palm print characteristics database by data transmit-receive module, and sends feedback registering result to user terminal by data transmit-receive module;

During authenticated user identity:

Dynamic hand shape generation module generates palmmprint hand shape sequence by preset hand shape image and transfers to user terminal by data transmit-receive module;

The certification hand shape image of reception is transferred to hand shape identification module by data transmit-receive module, personal recognition module extracts hand shape, palm print characteristics mating with the preset hand shape in hand shape setting data storehouse, palm print characteristics database, palm print characteristics;

Hand shape identification module, personal recognition module feed back authentication result by data transmit-receive module to user terminal according to matching result.

Wherein more preferably, described service end also comprises the data decryption module corresponding with user terminal;

Data transmit-receive module is connected with data decryption module respectively with personal recognition module;

Described safety certificate server is sent to data decryption module by data transmit-receive module, by reprocessing after data decryption module decryption processing after receiving the image of user terminal encryption.

Wherein more preferably, described service end also comprises palm tracking module:

Described palm tracking module is connected with described data transmit-receive module;

Described palm tracking module, when gathering the certification hand shape image of user with palmmprint, adopts hand shape area tracking algorithm opponent shape area tracking, determines that whether palmmprint hand shape image is continuous;

If the discontinuous existence of the palm of pickup area, points out user authentication failure by data transmit-receive module;

If the palm continued presence of pickup area, continue to gather other palmmprint hand shape images by data transmit-receive module prompting user.

Wherein more preferably, described service end also comprises recognition decision module:

Described recognition decision module respectively with described hand shape identification module, described personal recognition model calling;

Described service end is by after certification palmprint image identification matching treatment, and described recognition decision module judges whether the palmmprint hand shape sequence generated gathers, identifies, mated;

If do not complete palmmprint hand shape sequence, then described recognition decision module is to the next hand shape image of user's transmission user, and user repeats above-mentioned steps until the palmmprint hand shape sequence generated all completes;

If palmmprint hand shape sequence completes, the then palmprint information matching result of described recognition decision module searches record, if there is the result that the match is successful, for user performs corresponding business operation, then by data transmit-receive module to the successful message of user terminal certified transmission; If without the palmprint information matching result that the match is successful, by data transmit-receive module to user terminal certified transmission failure.

Again on the one hand, the invention provides a kind of security certification system based on gesture identification, comprise service end and user terminal;

Described server is connected with described user terminal communication;

Also comprise the transactional services end be connected with described server and described user terminal communication;

User initiates transaction authentication request by described user terminal to described transactional services end, described transactional services end initiates identity energy authentication request to described service end, described service end generates palmmprint hand shape sequence and transfers to described user terminal, described user terminal sequentially shows hand shape image, the hand shape image that user shows by described user terminal sequentially does corresponding gesture motion, described user terminal gathers the hand shape image of the band palmmprint of user and transfers to described service end, described service end is extracted palm print characteristics by palm grain identification method and is mated with preset palm print characteristics, described service end is to described transactional services end user terminal feedback matching result, described transactional services end starts transaction results information according to the matching result of feedback by user terminal.

Wherein more preferably, described service end comprises:

During registered user's characteristic information:

During authenticated user identity:

Again on the one hand, the invention provides a kind of safety certificate server based on gesture identification,

Described safety certificate server is used for the authority of authenticated user in security certification system;

Described safety certificate server is connected with described user terminal communication;

Described safety certificate server receives the preset palmprint image of described user terminal collection user, and described safety certificate server extracts palm print characteristics by palm grain identification method and saves presets palm print characteristics;

Described safety certificate server receives described user terminal and initiates authentication request to described safety certificate server, described safety certificate server generates palmmprint hand shape sequence and transfers to described user terminal, described user terminal sequentially shows hand shape image, the hand shape image that user shows by described user terminal sequentially does corresponding gesture motion, described user terminal gathers the hand shape image of the band palmmprint of user and transfers to described safety certificate server, described safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, described safety certificate server presses matching result to described user terminal feedback authentication result.

Described safety certificate server and described user terminal and trading server communicate to connect;

User initiates transaction authentication request by described user terminal to described trading server, described safety certificate server receives described trading server and initiates authentication request to described safety certificate server, described safety certificate server generates palmmprint hand shape sequence and transfers to described user terminal, described user terminal sequentially shows hand shape image, the hand shape image that user shows by described user terminal sequentially does corresponding gesture motion, described user terminal gathers the hand shape image of the band palmmprint of user and transfers to described safety certificate server, described safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, described server is to described trading server user terminal feedback matching result, described trading server starts transaction results information according to the matching result of feedback by user terminal.

Again on the one hand, the invention provides a kind of safety certification user terminal based on gesture identification,

Described user terminal is used for the authority of authenticated user in security certification system;

The preset palmprint image that described user terminal gathers user transfers to described safety certificate server, and described safety certificate server extracts palm print characteristics by palm grain identification method and saves presets palm print characteristics;

Described user terminal and described Security Authentication Service and trading server communicate to connect;

Described user terminal initiates transaction authentication request to described trading server, described safety certificate server receives described trading server and initiates authentication request to described safety certificate server, described safety certificate server generates palmmprint hand shape sequence and transfers to described user terminal, described user terminal sequentially shows hand shape image, the hand shape image that user shows by described user terminal sequentially does corresponding gesture motion, described user terminal gathers the hand shape image of the band palmmprint of user and transfers to described safety certificate server, described safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, described server is to described trading server user terminal feedback matching result, described trading server starts transaction results information according to the matching result of feedback by user terminal.

Safety certifying method based on gesture identification provided by the invention, terminal, server and system, dynamic gesture sequence and the identification of palm feature are combined, neither need user's keeping and carry extra identity keepsake, reduce again the possibility that identity information is held as a hostage to a great extent, improve the fail safe of authentication.

Accompanying drawing explanation

Fig. 1 is registered user's characteristic information flow process schematic diagram of the present invention;

Fig. 2 is authenticated user identity schematic flow sheet of the present invention;

Fig. 3 gathers preset palmprint image schematic flow sheet in registered user's characteristic information flow process of the present invention;

Fig. 4 is the key point schematic diagram that the horizontal threading method of the present invention detects between finger;

Fig. 5 is the key point schematic diagram that the horizontal threading method of the present invention obtains leader;

Fig. 6 is the facies palmaris area schematic that the present invention locates in preset palmprint image;

Fig. 7 is authenticated user identity detailed process schematic diagram of the present invention;

Fig. 8 is the preset hand shape image exemplary plot of the present invention;

Fig. 9 is user rs authentication of the present invention preset hand shape schematic flow sheet;

Figure 10 is the hand shape instruction legend that the preset hand shape image of the present invention is corresponding;

Figure 11 is the preset hand shape mask exemplary plot of the present invention;

Figure 12 is the present invention's authenticated user identity detailed process schematic diagram when concluding the business;

Figure 13 is the security certification system structural representation that the present invention is based on gesture identification;

Figure 14 is user terminal one example structure schematic diagram of the present invention;

Figure 15 is another example structure schematic diagram of user terminal of the present invention;

Figure 16 is service end one example structure schematic diagram of the present invention;

Figure 17 is another example structure schematic diagram of service end of the present invention;

Figure 18 is the safety certification transaction system structural representation that the present invention is based on gesture identification.

Embodiment

Below in conjunction with drawings and Examples, the specific embodiment of the present invention is described in further detail.Following examples for illustration of the present invention, but are not used for limiting the scope of the invention.

The invention provides a kind of safety certifying method based on gesture identification, comprise the step of registered user's characteristic information and the step of authenticated user identity; As shown in Figure 1, the step of registered user's characteristic information comprises: the preset palmprint image gathering user; Extract the palm print characteristics of preset palmprint image and save presets palm print characteristics; As shown in Figure 2, the step of authenticated user identity comprises: user initiates authentication request; Generate palmmprint hand shape sequence and sequentially show hand shape image to user; User sequentially does corresponding gesture motion by the hand shape image shown and gathers the certification hand shape image of user with palmmprint; Extract the hand shape of certification hand shape image, palm print characteristics mating with preset hand shape, palm print characteristics; Feedback matching result.Below in conjunction with multiple embodiment, detailed description is launched to safety certifying method provided by the invention.

Embodiment 1

The present embodiment provides a kind of safety certifying method based on gesture identification, comprises the step of registered user's characteristic information and the step of authenticated user identity; As shown in Figure 1, the step of registered user's characteristic information comprises: user terminal gathers the preset palmprint image of user and transfers to service end; Service end is extracted the palm print characteristics of preset palmprint image and is saved presets palm print characteristics; As shown in Figure 2, the step of authenticated user identity comprises: user terminal initiates authentication request to service end; Service end generates palmmprint hand shape sequence and sequentially transmits to user terminal and show hand shape image; User sequentially does corresponding gesture motion by the hand shape image shown, user terminal gathers the certification hand shape image of user with palmmprint and transfers to service end; Service end extracts the hand shape of certification hand shape image, palm print characteristics mating with preset hand shape, palm print characteristics; Feedback matching result.Detailed description is launched to the safety certifying method that the present embodiment provides below.

Part I, introduces the step of registered user's characteristic information.

During registered user's characteristic information, need the preset palmprint image first gathering user, then extract palm print characteristics by palm grain identification method and save presets palm print characteristics.Illustrate below.

As shown in Figure 3, during registered user's characteristic information, user terminal initiates registered user's feature request to service end; Smooth palm shape (such as palm is stretched flat, and the five fingers open) is positioned over the image acquisition region of user terminal by user, starts the preset palmprint image that image acquisition procedure gathers user; Transfer to service end after collecting preset palmprint image, service end utilizes palm grain identification method to extract the palm print characteristics of the user in preset palmprint image, preserves palm print characteristics after extraction.Since then, registered user's characterization step terminates.

In one embodiment of the invention, gather in the preset palmprint image step of user, or extracting palm print characteristics and saving presets in the step of palm print characteristics, also comprise the pretreated step of preset palmprint image.To can some unnecessary graphical informations of filtering after preset palmprint image preliminary treatment, reduce volume of transmitted data, improve compression ratio, substantially increase data transmission efficiency.The pretreated step of preset palmprint image is comprised:

1), first to the palm facies palmaris image coarse segmentation process gathered.First, palm facies palmaris image is carried out color space conversion, is converted to HSV space from rgb space; Secondly, the area of skin color of H component rapid extraction hand is utilized.By can reduce the pixel quantity of successive image process further to the coarse segmentation process of palm facies palmaris image.In the present invention, closest to the effective range of area of skin color when finding to get H ∈ [0.22,0.48] in HSV space by inventor's test of many times.H in HSV space is chrominance component, only uses H component can reduce the impact of illumination.In this step to gather palm facies palmaris image procossing and distinguish hand shape region and non-hand shape region.

2), the palm facies palmaris display foreground gathered and background are distinguished and equilibrium treatment.First the palm facies palmaris image after coarse segmentation is converted to gray level image; Secondly, the region beyond area of skin color is set to background area; Finally histogram equalization process is carried out to gray level image and obtain the gray level image after equilibrium.

Preferred logarithmic transformation method, specific as follows:

G(x，y)＝21.6×ln((g(x，y)+1)

Wherein, g (x, y) is the gray value of gray level image, and G (x, y) is the gray value after logarithm change.By the impact of illumination on palm facies palmaris image can be reduced as far as possible after histogram equalization process.

3), to the gray level image medium filtering process after equilibrium, more stable prospect hand shape region is obtained.More accurate by the hand shape region obtained after medium filtering process.

4), to the prospect hand shape region of image carry out binary conversion treatment, thus generate the prospect binaryzation region (binaryzation hand shape area image) of hand shape figure image.In the present invention, preferred ostu method (maximum variance between clusters) is to prospect hand shape region binary conversion treatment.

5) hand shape image of band palmmprint, is obtained.In the present invention, preferred floodfill (unrestrained water is filled) algorithm does connected region detection to binaryzation hand shape area image, and getting the maximum connected region of pixel number in image is hand shape region, thus filters out other background noise points.

6), based on Sobel (Sobel, Sobel operator) operator, rim detection is carried out to binary image and obtain the hand shape region after binaryzation.

\begin{matrix} f_{x} = [\begin{matrix} - 1 & 0 & 1 \\ - 2 & 0 & 2 \\ - 1 & 0 & 1 \end{matrix}] & f_{y} = [\begin{matrix} 1 & 2 & 1 \\ 0 & 0 & 0 \\ - 1 & - 2 & - 1 \end{matrix}] \end{matrix}

Wherein, f _xfor horizontal filter, f _yfor vertical filter.

In the present embodiment, can complete at user terminal the pretreated step of preset palmprint image, also can carry out in service end.Can enable when network is not good and perform the pretreated step of preset palmprint image at user terminal, transfer to service end again to after preset palmprint image preliminary treatment.In the good situation of network environment, or when user terminal processes ability is bad, directly can be sent to service end by user terminal, perform the pretreated step of preset palmprint image in service end.Can select flexibly to perform at user terminal or service end according to actual conditions to the pretreated step of preset palmprint image, shorten registered user's characteristic information time as far as possible.

In one embodiment of the invention, palm print characteristics is extracted and the step saving presets palm print characteristics specifically extracts palm print characteristics to preset palmprint image by palm grain identification method.Below the step of pressing palm grain identification method extraction palm print characteristics is described in detail.

First, the step in the facies palmaris region in the preset palmprint image in location is introduced.In the present invention, preferred levels threading method detects the key point between finger, describes in detail below to point the preset palmprint image opened.

As shown in Figure 4, a), horizontal threading edge image from top to bottom, one group of threading including 8 boundary points is found.Using adjacent 2 boundary points as one group, totally four groups; Using these four groups as the initial threading group being four fingers.

B), for each in four fingers refer to (not comprising thumb), with initial threading group for original position, search for respectively up and down, until search up-and-down boundary.The strategy of search is based on same connected region, carries out region growing up and down, till the border finding connected region.Whether judging of up-and-down boundary is same connected region as foundation, and then orients key point 601 to 607.

C), in the same way, horizontal threading edge image from bottom to up, finds one group of threading including 4 boundary points.Wherein the horizontal coordinate of rear 2 boundary points is positioned at the right half area of image, and later 2 boundary points are as the initial threading group of thumb.

D), same to step b), refer to, for original position, search for respectively up and down with the initial threading of thumb, until search up-and-down boundary, and then orient key point 608 and 609.

Secondly, the step in the facies palmaris region in the preset palmprint image of cutting is introduced.

In the present invention, preferably the facies palmaris region in preset palmprint image is cut in employing based on Harris (Corner Detection Algorithm).Harris Corner Detection Algorithm is a kind of interest point detect operator based on signal, and its principle is that image window to be processed is moved micro-displacement (μ, ν) to any direction.In order to improve anti-noise ability, carried out Gaussian smoothing filter to image window, the Gauss's window selected is as follows:

G_{x, y} = \exp [- \frac{1}{2} (x^{2} + y^{2}) / σ^{2}]

Then gradation of image knots modification E _x,ybe defined as:

E _x,y| _(x,y)＝ΣG _x,y[I(x+μ,y+μ)-I(x,y)] ²

In formula, E _x,yfor moving the brightness changing value that a wicket (μ, υ) occurs at (x, y) place; I is the palmprint image that will extract angle point.

As shown in Figure 5, a point, b point, c point in the key point between use Harris (Corner Detection Algorithm) to find out from preset palmprint image forefinger that threading method in Fig. 4 obtains, middle finger, the third finger, little finger of toe.Wherein, a point represents above-mentioned threading and sends out key point 607, the c point detected and represent key point 605.B point for a point and c point composition line segment mid point.With b point for initial point, make the normal f of line segment ac; Then with c point for starting point, along normal f, towards volar direction skew certain distance (length of line segment ac) place some e, centered by e, generate a length of side and be P and the rectangular area parallel with line segment ac.Extract P × P region, palm central area.

Finally, the step of the palm print characteristics extracted in preset palmprint image is introduced.

Extract in above-mentioned steps in P × P region, preset palmprint image central rectangular region and make palmprint feature extraction.The method be preferably based in the present invention in " Kong A W K; Zhang D.Competitive coding scheme for palmprint verification.In:Proceedings of the 17th Inter-national Conference on Pattern Recognition.Washington D.C.; USA:IEEE, 2004.520-523 " carries out feature extracting and matching.Use the real-valued Gabor filter of six direction to palmprint image filtering in the method, and to the minimum direction encoding of amplitude, be called Competition coding (Competitive code).In the present invention, to can with reference to following method during text feature extraction: such as: " based on the personal recognition new method of the bi-directional matching of major line features. Journal of Computer Research and Development, 2004,41 (6): 996-1002 "; " an implementation of biometric technology.In:Proceedings of the14th International Conference on Pattern Recognition.Brisbane, Australia:IEEE, 1998.219-221 "; " A novel approach of palm-line extraction.In:Proceedings of the 3rd International Conference on Image and Graphics.Washington D.C., USA:IEEE, 2004.230-233 ".

As shown in Figure 3, in order to ensure the accuracy of the preset palm print characteristics preserved further, palmmprint can be verified smoothly during authentication, need the preset palmprint image of multi collect when registered user's characteristic information and extract user can multiple palm print characteristics information and preserve.Specific as follows: constantly to repeat said process, until the palm print characteristics quantity collected reaches M; Preserve the palm print characteristics of user to database; Service end is to user terminal message transfer prompting user palm print characteristics collection success.

Part II, introduces the step of authenticated user identity.

As shown in Figure 7, user, when transaction or other need authentication of users scene, needs by this method authenticated user identity.During authenticated user identity, user initiates authentication request by user terminal to service end; Service end generates palmmprint hand shape sequence and transfers to user terminal, sequentially shows hand shape image to user; User sequentially makes corresponding gesture motion by the hand shape image shown, user terminal gathers the certification hand shape image of user with palmmprint and transfers to service end; The certification hand shape image of service end to band palmmprint extracts hand shape, palm print characteristics mating with preset hand shape, palm print characteristics; Service end feeds back matching result to user terminal.Detailed description is launched to the step of authenticated user identity below.

First, the step that user initiates authentication request is introduced.As shown in Figure 7, when initiating authentication request, user terminal sends after authentication request information package.Preferably send in xml format.Authentication request information comprises ID, the Customs Assigned Number of user terminal.In this step, in order to ensure the fail safe of data verification further, after initiating authentication request, service end also needs the legitimacy of further authentication verification request, in authentication request information, therefore also comprise the random coded of generation.The data packet format example that user terminal encapsulates when initiating authentication request is as follows:

After service end receives the authentication request packet of user terminal transmission, service end reads authentication request packet, by verifying the legitimacy of the mode authentication verification request of random coded.If the verification passes, then start a certification thread and carry out follow-up certification; If checking is not passed through, then transmit result to user terminal in the form of a message, inform that user authentication request is illegal, certification terminates.

Secondly, the step generating palmmprint hand shape sequence and also sequentially show hand shape image to user is introduced.

As shown in Figure 7, service end presses Client-initiated authentication request stochastic generation a set of palmmprint hand shape sequence, sequentially transmits hand shape to user terminal and shows at user terminal.Palmmprint hand shape sequence is the sequence that multiple hand shapes of different hand shape form at random.When service end is to user terminal transmission hand shape, send after the hand shape information package of stochastic generation.Preferably send in xml format.The hand shape information of stochastic generation is comprised the hand shape information such as authentication request random coded, hand shape numbering.The hand shape information packet format sample of service end transmission is as follows:

Generate palmmprint hand shape sequence in service end and sequentially show in the step of hand shape image to user terminal transmission, due to the sequence that palmmprint hand shape sequence multiple hand shape images that are different hand shapes form at random.And in authentication procedures, need the gesture feature and the palm palm print characteristics that gather user.Multiple hand shape image is prefixed for stochastic generation hand shape image sequence in the present invention.Various different hand shape in palmmprint hand shape sequence is selected from the hand shape candidate collection in service end.Such as: for hand shape candidate collection PC={p ₁, p ₂... p _m, 5≤M≤7; Wherein, p ₁-p _mrepresent the sequence number of different hand shape, M represents the quantity of the actual operable hand shape of user.In the present invention, in order to ensure the fail safe of user authentication process, ensure that the hand shape in the hand shape sequence provided to user can not very little, M gets and is at least 5.Generate the hand shape sequence PC that a length is N _n, PC _n={ pc ₁, pc ₁... pc _n.Wherein, pc ₁-pc _nrepresent various different hand shape, for each palmmprint hand shape sequence pc _s=random{PC), and random (AI) }, 1≤s≤N.Wherein, random (PC) represents random hand shape image, and random (AI) represents the random anglec of rotation.In the present invention, random function random uses the random number generator based on time seed.

Below various hand shape image during service end generation palmmprint hand shape sequence is illustrated.As shown in Figure 8, at this, only make respective description to part hand shape image, the figure not doing in other figure to show also still can realize palmmprint hand shape sequence in the present invention.The gesture now enumerating several hand shapes wherein in the present invention corresponding illustrates, is described as follows: gesture 1: the five fingers open flatly stretches formula, can see in Fig. 8 500.Gesture 2: the five fingers close up flat stretches formula, can see in Fig. 8 501.Gesture 3: " shooting formula ", under the basis of gesture 2, draws middle finger in, nameless and little finger of toe, can see in Fig. 8 502.In gesture 4: four refers to, button thumb out formula, namely under the basis of gesture 2, opens thumb, simultaneously deduction all the other four to refer to, can see in Fig. 8 503.Gesture 5: four refers to flat stretch thumb internal fastening type, and namely under the basis of gesture 2, interior button thumb, stretches all the other four fingers simultaneously, can see in Fig. 8 504.Gesture 6: the five fingers internal fastening type, is namely keeping under the smooth prerequisite of palm, all fingers all in button, can see in Fig. 8 505.Gesture 7: four refers to close up and refers to flat stretch thumb out formula, and namely under the prerequisite keeping palm smooth, four refer to flat stretching, and thumb out, can see in Fig. 8 506.

In the present invention, consider the individual physical difference of people, not all people can both make various gestures corresponding to preset hand shape in above-mentioned Fig. 8 accurately.Therefore, in the step of registered user's characteristic information, be necessary that can adding users adapt to the verification step of above-mentioned preset hand shape.Particularly, user needs to make corresponding gesture by the hand shape image experiment of prompting and gather corresponding palmmprint hand shape image successively to do corresponding checking.Determine whether accurately to put preset hand shape by the result, user accurately can not put the corresponding gesture of the hand shape image of prompting.Service end makes corresponding record.User shows the corresponding gesture of the hand shape image of prompting, simultaneously for active user, in the random palmmprint hand shape sequence generated in authentication procedures, palmmprint hand shape sequence generates according to the result of preset hand shape, and palmmprint hand shape sequence does not comprise those accurately to show the corresponding gesture of hand shape image.Below the step of user's enrollment status information is remarked additionally.

Particularly, as shown in Figure 9, the process of user rs authentication preset hand shape mainly comprises the following steps: that preset various hand shape images are transferred to user terminal and present to user by service end successively.User shows identical gesture and by the corresponding image transmitting of user terminal collection to service end according to each the hand-type image shown.The image recognition hand shape image that service end sends client also feeds back comparison result (this step is identical with comparison process with hand shape image identification in the step of authenticated user identity, and concrete steps are described in more detail below) with the hand shape image comparison transmitted to user.If user repeatedly cannot show identical gesture (number of times >=F1 namely made mistakes), then point out user " current gesture mistake needs to readjust to put posture ".If user continues repeatedly to show identical gesture (number of times >=F2 namely occurred, and F2>F1), then user is pointed out " whether to abandon current gesture ".If user selects to abandon current gesture, then continue to get the gesture that next user does not attempt putting; With the current gesture of tense marker for user abandons state; Otherwise, then the current gesture of display is continued.If identical gesture has been shown in user's success, then continue to get the gesture that next user does not attempt putting; With the current gesture of tense marker for user can selection mode.If user has attempted all preset hand shapes, then judge whether at user option gesture quantity is greater than M.If be not more than M, then prompting user cannot successful registration.If be greater than M, then point out user that corresponding logical operation function is set, then preserve all gesture features, point out user's gesture to succeed in registration simultaneously.

In the present invention, when choosing the generation of the hand shape image in hand shape candidate collection palmmprint hand shape sequence and showing to user, user needs the hand shape image shown in the user terminal by palmmprint hand shape sequence do corresponding gesture and gather certification hand shape image.Consider palm-print image capture and make other complexity and other factors on the impact of palmprint image, accelerating verification process as far as possible, improve authentication efficiency.Therefore, need to do certain restriction to during generation palmmprint hand shape sequence choosing the hand shape image chosen in hand shape candidate collection.Such as, in the present invention to PC _nincrease a restrictive condition, namely at least must comprise " the five fingers open to put down and stretch a formula " hand shape.Due in the certification hand shape image that the various hand shape images in hand shape candidate image geometry are corresponding, can open to put down see 500 the five fingers in Fig. 8 and stretch formula.It is the most complete that the five fingers open flat formula palmmprint region of stretching, and the impact of action on palmmprint deformation of finger is minimum.In personal recognition program, recognition speed is the fastest, saves user's authenticated time.If really do not comprise " the five fingers open to put down and stretch formula " hand shape after stochastic generation, then by " the five fingers open to put down and stretch formula " the insertion PC that hand shape is random _nin sequence.Restriction like this, guarantees the reliability that palmmprint detects further.It should be noted that at this, the restrictive condition of palmmprint hand shape sequence be not limited only to this, also can be other ways to restrain (by choose in hand shape candidate collection in other hand shapes Fig. 8 any one).Certainly can understand, also can not arrange the restrictive condition of the hand shape image chosen in palmmprint hand shape sequence.Accelerate the user authentication time by other means, improve authentication efficiency.

Again, introduce user sequentially do corresponding gesture motion by the hand shape image shown and gather the step of the certification hand shape image of user with palmmprint.

As shown in Figure 7, after user terminal receives the authentication requesting of service end transmission, the hand shape that user shows by user terminal makes corresponding gesture motion in appointed area, and the band palmmprint hand shape that user terminal gathers corresponding hand shape transfers to service end.Gather the accuracy of palmmprint hand shape image to improve user terminal further and improve the speed gathering user's palmmprint hand shape image in verification process, user terminal can draw out corresponding hand shape instruction legend, and points out user.As shown in Figure 10, when user terminal gathers certification palmmprint hand shape image, show corresponding hand shape instruction legend at user terminal, the palmmprint hand shape image of in Figure 10 in hand shape instruction legend 510 to 516 corresponding diagram 8 500 to 506.User needs the position of the hand shape image adjustment hand shown by user terminal by hand, be positioned over by hand in the collection hand shape instruction legend of user terminal, and user terminal sends to service end at backstage timing acquiring user present image simultaneously.After user terminal collects the palmmprint hand shape image of certification, after being packed by the palmmprint hand shape image of certification, be sent to service end.For ensureing the fail safe of palmprint image transmission, need to be encrypted by palmmprint hand shape image further, the palmmprint hand shape image after encryption sends with the form of packet.Certification palmmprint hand shape image preferably sends in xml format.This packet comprises the random number of request and the coding of encrypted image.Can select whether pretreated step is carried out to certification palmmprint hand shape image according to network condition before send certification palmmprint hand graphic data bag to service end.Identical with to the step of preset palmprint image and process to the pretreated step of certification palmmprint hand shape image, just repeat no more at this.The palmmprint hand shape authentication data packet format sample encrypted when user terminal thinks that service end sends palmmprint hand shape image is as follows:

3rd, to the certification hand shape image of band palmmprint do hand shape identify and with the step of preset hand-shaped characteristic comparison.

In invention, first hand shape is done to the certification hand shape image of band palmmprint and identify; When the identification of hand shape is done to certification hand shape image, be to be with the certification hand shape image of palmmprint to compare as matching degree with all hand shape masks that service end pre-sets, and find out the hand shape mask matched, using hand shape corresponding for this hand shape mask as the hand shape identified.The corresponding hand shape mask that pretreated palmmprint hand shape image and service end pre-set compares as matching degree, and what mate with hand shape mask then thinks corresponding hand shape image, if unmatched, not thinks it is corresponding hand shape image.Particularly, the length and width of each finger and palm in pretreated palmmprint hand shape image are calculated.By the length and width of the length and width adjustment hand shape mask of finger each in palmmprint hand shape image and palm, the certification hand shape image of band palmmprint being filled the length and width of hand shape mask, determining corresponding matching degree according to filling result.By the hand shape after identifying and corresponding preset hand-shaped characteristic comparison during generation palmmprint hand shape sequence, if correspondence, think coupling, otherwise then not think and mate.

A palm area is divided into different regions by diverse location by the present invention.Such as: four refer to region (forefinger, middle finger, the third finger, little finger of toe), thumb area and palm area, respectively with F, T, P represent that four fingers are divided, thumb part and palm portion.The feature group FeatureList of a palm _{feature_type}represent.The feature group of palm is shown below:

FeatureList _{feature_type}＝{F _w,F _h,T _w,T _h,P _w,P _h}

Wherein, F _wrepresent the width in four finger regions; F _hrepresent the height in four finger regions; T _wrepresent the width of thumb area; T _hrepresent the height of thumb area; P _wrepresent the width of palm area; P _hrepresent the height of palm area.

In the present invention, based on above-mentioned 9 key points, calculate every numerical value of Feature, four refer to regions, the length and width of thumb area and palm area are calculated as follows:

tp ₁＝max(distance(pt ₆₀₅,pt ₆₀₆),distance(pt ₆₀₅,pt ₆₀₆))

tp ₂＝2×pt ₆₀₇-pt ₆₀₆

F _h＝f ₀×distance(center(pt ₆₀₆,pt ₆₀₇),pt ₆₀₃)

F _w＝f ₁×tp ₁

P _w＝f ₂×tp ₁

P _h＝f ₃×(distance(tp ₂,pt ₆₀₈))

T _w＝f ₄×(distance(tp ₂,pt ₆₀₉))

T _h＝f ₅×(max(distance(tp ₂,pt ₆₀₉),distance(tp ₂,pt ₆₀₈)))

In the present invention, f ₀~ f ₅coefficient, and f ₀~ f ₅can rule of thumb get certain fixed value, i.e. f ₀=0.56, f ₁=1.15, f ₂=4.25, f ₃=1.92, f ₄=0.66, f ₅=0.45.

In order to improve accuracy of identification and discrimination further, adjusted the length and width of hand shape mask by the length and width of finger each in palmmprint hand shape image and palm.

As shown in figure 11, the present invention is provided with the hand shape mask corresponding with hand shape image shown in Fig. 8 in service end, and often overlapping mask and be made up of 6 parts, is 4 fingers, a palm and a thumb respectively.Each finger is divided into upper and lower 2 parts, and a thumb is divided into 4 parts.Such as, P Regional Representative palm area in a set of hand shape mask, T1 to T4 represents thumb area; F0U, F0D represent forefinger area; F1U, F1D represent middle finger region; F2U, F2D represent nameless region; F3U, F3D represent little finger of toe region.

As shown in figure 11, when the hand shape image generation palmmprint hand shape sequence that service end is chosen in hand shape candidate collection is shown to user, each hand shape one_to_one corresponding hand shape mask in the hand shape candidate collection chosen, namely the hand shape mask in the corresponding Figure 11 of the hand shape image in Fig. 7, corresponding relation is as shown in table 1:

	Hand shape 1	Hand shape 2	Hand shape 3	Hand shape 4	Hand shape 5	Hand shape 6	Hand shape 7
								Hand shape	500	501	502	503	504	505	506
Mask	700	701	702	703	704	705	700

As shown in figure 11, in service end, the hand shape mask that each hand shape is corresponding, comprises 13 minimum units altogether.Shown in the unit formula composed as follows of each hand shape mask:

UnitList＝{P,T1,T2,T3,T4,F0U,F0D,F1U,F1D,F2U,F2D,F3U,F3D}

The hand shape mask that each hand shape is corresponding has corresponding regional ensemble, is shown below:

RegionList _{feature_type}＝{FR _{feature_type},BR _{feature_type}}

FR _{feature_type}＝{F ₁,F ₂,…F _n}

BR _{feature_type}＝{B ₁,B ₂,…B _m}

m+n＝13

Wherein, each hand shape mask RegionList _{feature_type}comprising two class regional ensembles, is foreground area set FR respectively _{feature_type}with background area set B Region _{feature_type}; For each foreground area F _nwith background area B _m, then can be single minimum unit, such as F _n={ P}, i.e. single palm district; Also can be the combination of multiple minimum unit, such as B _m={ F1U, F2U, F3U}.

The regional ensemble of each the hand shape mask in Fig. 8 in the corresponding Figure 11 of each hand shape, the corresponding relation of hand shape and hand shape mask foreground area is as shown in table 2:

FR ₅₀₀	{P},{F0U,F0D,F1U,F1D,F2U,F2D,F3U,F3D},{T2,T3}
		FR ₅₀₁	{P},{F0U,F0D,F1U,F1D,F2U,F2D,F3U,F3D},{T1,T3}
FR ₅₀₂	{P},{F0U,F0D},{F1D,F2D,F3D},{T2,T3}
		FR ₅₀₃	{P},{F0D,F1D,F2D,F3D},{T2,T3}
FR ₅₀₄	{P},{F0U,F0D,F1U,F1D,F2U,F2D,F3U,F3D},{T3}
		FR ₅₀₅	{P},{F0D,F1D,F2D,F3D},{T3}
FR ₅₀₆	{P},{F0U,F0D,F1U,F1D,F2U,F2D,F3U,F3D},{T2,T3}

The corresponding relation of hand shape and hand shape mask background area is as shown in table 3:

BR ₅₀₀	{T1,T4}
		BR ₅₀₁	{T2,T4}
BR ₅₀₂	{F1U,F2U,F3U},{T1,T4}
		BR ₅₀₃	{F0U,F1U,F2U,F3U},{T1,T4}

BR ₅₀₄	{T1,T2,T4}
		BR ₅₀₅	{F0U,F1U,F2U,F3U},{T1,T2,T4}
BR ₅₀₆	{T1,T4}

In the present invention, when pretreated palmmprint hand shape image is mated with hand shape mask, the filling rate (filling rate=loading/region area) of each minimum unit in each foreground area of first comparison pretreated palmmprint hand shape image opponent shape mask and background area.If the filling rate of current minimum unit is greater than predetermined threshold value, be considered as, when remote holder shape can fill minimum unit fill area when remote holder shape mask, returning Filling power 1; If the filling rate of current minimum unit is less than or equal to default filling rate threshold value, be considered as, when remote holder shape can not fill minimum unit fill area when remote holder shape mask, returning Filling power 0.Particularly, by the filling rate of each minimum unit in each foreground area of pretreated palmmprint hand shape image and hand shape mask and background area respectively (such as, the index finger portion of hand shape image and the foreground area minimum unit FOU of hand shape mask and minimum unit F0D make packing ratio pair respectively, if minimum unit FOU can fill, return 1, otherwise return 0, if minimum unit FOD can fill, return 1, otherwise return 0; Make packing ratio pair, if can fill, return 1, otherwise return 0; The foreground area minimum unit T1 of the thumb part in hand shape image and hand shape mask, minimum unit T2, minimum unit T3, minimum unit T4 make packing ratio pair respectively, if minimum unit T1, minimum unit T2, minimum unit T3, minimum unit T4 can fill, return 1 respectively, otherwise return 0).

Foreground area in pretreated palmmprint hand shape image and hand shape mask and the packing ratio of background area are shown below to relation:

{Value}_{n} = \{\begin{matrix} 1, FilledPercent (R_{n}) > {Threshold}_{n} \\ 0, FilledPercent (R_{n}) \leq {Threshold}_{n} \end{matrix}

Wherein, FilledPercent (R _n) represent the filling percentage in region, Threshold _nrepresent and preset filling rate threshold value; The filling percentage FilledPercent (R in region _n) be greater than corresponding threshold value Threshold _ntime, score value is 1, otherwise is 0.

Also need each minimum unit comparison respectively according to the foreground area of hand shape mask and background area when certification hand shape image and preset hand-shaped characteristic comparison, concrete comparison step is as follows: the hand shape region of certification hand shape image and hand shape mask foreground area and background area each minimum unit respectively comparison do respectively to mate; Do to fill according to the Shou Xing district (comprising palm area and multiple finger areas) of certification hand shape image mate with multiple parts (comprising each minimum unit comparison respectively of foreground area and background area) of multiple hand shape mask; Certification hand shape image and prefabricated hand-shaped characteristic comparison result is determined according to the result of the Filling power returned.Particularly, the palm area of certification hand shape image and multiple finger areas are done to fill with multiple foreground area of hand shape mask and each minimum unit of background area respectively and mates, if the palm area of certification hand shape image and multiple finger areas are mated with each minimum unit of the foreground area when remote holder shape mask and non-hand shape region is mated with each minimum unit of the background area when remote holder shape mask, then the preset hand-shaped characteristic being considered as current authentication hand shape image corresponding with when remote holder shape mask mates; Otherwise be considered as not mating.Describe in detail for the characteristic matching of hand shape mask 702 in Figure 11 below.The hand shape region (comprising finger areas and palm area) of current authentication hand shape image is compared as matching degree respectively with multiple foreground area of hand shape mask and each minimum unit of background area, if the hand shape region of certification hand shape image is mated with when the foreground area of remote holder shape mask and each minimum unit of background area are done to fill respectively, 8 minimum unit P of foreground area, F0U, F0D, F1D, F2D, F3D, T2, the Filling power that T3 returns is all 1, and 5 of background area minimum unit F1U, F2U, F3U, T1, the Filling power that T4 returns is all 0, the preset hand-shaped characteristic being then considered as current authentication hand shape image corresponding with the feature of hand shape mask 702 in Figure 11 mates, if the Filling power having any one region to return in 8 of foreground area minimum units P, F0U, F0D, F1D, F2D, F3D, T2, T3 is not 1, or the Filling power having any one region to return in 5 of background area minimum units F1U, F2U, F3U, T1, T4 is not 0, then the preset hand-shaped characteristic being considered as current authentication hand shape image corresponding with the feature of hand shape mask 702 in Figure 11 does not mate.

In order to improve precision further, in the present invention, when pretreated palmmprint hand shape image is mated with hand shape mask, each foreground area of first comparison pretreated palmmprint hand shape image opponent shape mask and the filling rate (filling rate=loading/region area) of background area.If the filling rate of current region is greater than predetermined threshold value, be considered as, when remote holder shape can fill fill area when remote holder shape mask, returning Filling power 1; If the filling rate of current region is less than or equal to default filling rate threshold value, be considered as, when remote holder shape can not fill fill area when remote holder shape mask, returning Filling power 0.Particularly, each foreground area of pretreated palmmprint hand shape image and hand shape mask and background area being compared respectively, ({ F0U, F0D} make packing ratio pair to the foreground area of such as, the index finger portion of hand shape image and hand shape mask, if can fill, return 1, otherwise return 0; The foreground area of the thumb part in hand shape image and hand shape mask T2, T3} make packing ratio pair, if can fill, return 1, otherwise return 0).

{Value}_{n} = \{\begin{matrix} 1, FilledPercent (R_{n}) > {Threshold}_{n} \\ 0, FilledPercent (R_{n}) \leq {Threshold}_{n} \end{matrix}

Also need the foreground area according to hand shape mask and background area comparison respectively when certification hand shape image and preset hand-shaped characteristic comparison, concrete comparison step is as follows: the hand shape region of certification hand shape image is done to mate with each region of hand shape mask respectively; Do to fill with multiple parts (comprising foreground area and background area) of multiple hand shape mask according to the Shou Xing district (comprising palm area and multiple finger areas) of certification hand shape image and mate; Certification hand shape image and prefabricated hand-shaped characteristic comparison result is determined according to the result of the Filling power returned.Particularly, the foreground area of the palm area of certification hand shape image and multiple finger areas and multiple hand shape mask and background area are done to fill respectively and mates, if the palm area of certification hand shape image and multiple finger areas are mated with the foreground area when remote holder shape mask and non-hand shape region is mated with the background area when remote holder shape mask, then the preset hand-shaped characteristic being considered as current authentication hand shape image corresponding with when remote holder shape mask mates; Otherwise be considered as not mating.Describe in detail for the characteristic matching of hand shape mask 702 in Figure 11 below.The hand shape region (comprising finger areas and palm area) of current authentication hand shape image is compared as matching degree with multiple parts of multiple hand shape mask, mate if the hand shape region of certification hand shape image is done to fill with the foreground area and background area of working as remote holder shape mask respectively, 4 foreground area { P}, { F0U, F0D}, { F1D, F2D, F3D}, { T2, the Filling power that T3} returns is all 1, and 2 background area { F1U, F2U, F3U}, { T1, the Filling power that T4} returns is all 0, the preset hand-shaped characteristic being then considered as current authentication hand shape image corresponding with the feature of hand shape mask 702 in Figure 11 mates, if { P}, { F0U, F0D}, { F1D, F2D, F3D}, { T2, the Filling power having any one region to return in T3} is not 1, or { F1U, F2U, F3U}, { Filling power having any one region to return in T1, T4} is not 0, then the preset hand-shaped characteristic being considered as current authentication hand shape image corresponding with the feature of hand shape mask 702 in Figure 11 does not mate.

It should be noted that at this: it is identical that hand shape 500 defines with the region of hand shape 506, therefore the differentiation of the two also needs the connection between detection four finger (except thumb), namely based on horizontal threading method testing result, if exist at least one threading result and comprise 5 sections of continuous print line segments, be then hand shape 500, otherwise be hand shape 506.

Finally, introduce by palm grain identification method extraction palm print characteristics and the step of mating with preset palm print characteristics.

As shown in Figure 7, service end is deciphered this packet after receiving certification palmmprint hand graphic data bag, after deciphering, preferentially judges whether the certification hand shape image in certification palmmprint hand graphic data bag has palm to exist.If there is palm to exist, then carry out next step palmprint image identification; If without palm, then to user feedback authentification failure.Service end detect band palmmprint certification hand shape image after to certification hand shape image identification palmprint information.To the step of certification hand shape image identification palmprint information with preset palmprint image is coexisted mutually by the step that palm grain identification method extracts palm print characteristics to this has just repeated no more.

As shown in Figure 7, after the palmprint information identifying certification hand shape image, the palmprint information after identifying is mated with the preset palmprint information prestored in database, record palmprint information matching result.

As shown in Figure 7, judge whether the palmmprint hand shape sequence that service end generates gathers, identifies, mated further after the certification palmprint image identification matching treatment transmitted by user terminal.If do not complete palmmprint hand shape sequence, then to the next hand shape image of user's transmission user, user repeats above-mentioned steps until the palmmprint hand shape sequence generated all completes.If palmmprint hand shape sequence completes, then search the palmprint information matching result of record, if there is the result that the match is successful, performs corresponding business operation for user and point out user authentication success by user terminal.If without the palmprint information matching result that the match is successful, point out authentification failure by user terminal to user.Identity authentication result sends to user terminal with the form of packet by service end, preferably sends in xml format.This packet comprises the random number of request, authentication result and hand shape numbering etc.The palmmprint hand shape authentication data packet format sample encrypted when user terminal thinks that service end sends palmmprint hand shape image is as follows:

As shown in Figure 7, in order to ensure the fail safe of verification process, before pressing palm grain identification method extraction palm print characteristics, when gathering the certification hand shape image of user with palmmprint, the palm confirming pickup area is needed to be continued presence, if the palm of pickup area is discontinuous, point out user authentication failure (such as in authentication procedures, user is when performing palmmprint hand shape image sequence, hand is shifted out the image acquisition region of user terminal by user, then think that this ID authentication request is illegal, the failure of prompting user authentication request, needs authenticating identity again).

In one embodiment of the invention, when gathering the certification hand shape image of user with palmmprint, when confirming whether the palmmprint hand shape image of pickup area is deposited continuously, adopt hand shape area tracking algorithm opponent shape area tracking, determine that whether palmmprint hand shape image is continuous.Hand shape area tracking optimal algorithm selection CamShift algorithm realizes the tracking in opponent region.Below the step of the tracking adopting CamShift algorithm opponent region is described in detail.

Adopt CamShift algorithm to realize the tracking in opponent region in the present invention.Idiographic flow is as follows:

1) after first time detects correct hand shape, using the region corresponding to the minimum unit P in hand shape mask as tracing area.

2) carry out back projection (Back Projection) to the image of tracing area, the concrete practice is: first image is converted to HIS space, then gets the existing 1D histogram of H component extraction, i.e. the probability distribution graph of each pixel value in region.

3) tracing process adopts MeanShift (average drifting) method.

4) for the new image arrived, CamShift (being called the MeanShift algorithm of continuous adaptive) method is adopted to upgrade.

As shown in Figure 7, in order to ensure the fail safe of verification process, in one embodiment of the present of invention, in the step of authenticated user identity, the present invention also provides and puts and the step verified by specified angle user's gesture.Technical scheme of the present invention is also prefixed multiple angles degree AI, gets 0 °≤AI≤90 °.

As shown in Figure 8, in Fig. 8,500 expression angles are " the five fingers open to put down and stretch formula " of 0 degree; 508 represent that angles is " the five fingers open to put down and stretch formula " of 15 degree.500 to 507 is all the right hand of people as shown in Figure 8; For the right hand, angles 90 ° is equivalent to the right hand 500 half-twist in the counterclockwise direction; Left hand then in contrast, namely 500 of left hand is rotated in a clockwise direction 90 °.

In the present invention, the granularity of hand angles is 1 °, so for each hand shape, all there are 91 kinds of angles., there are 637 (=7 × 91) and plant possibility in all hand shapes for each the unknown; For the hand shape sequence that a length is N, then have 637 ⁿplant combination.In the present invention, N>=5, namely for each group hand shape sequence, at least exist 104881082626957 (=637 ⁵) plant combination.

As 508 in Fig. 8, several angles are wherein exemplified below respectively:

1) 0 degree: i.e. reference position, palm is flat to be stretched, and facies palmaris is parallel with user terminal screen; Finger tip vertically upward.

2) 15 degree: take reference position as original position, 15 degree are rotated counterclockwise.

3)-15 degree: be original position with reference position, turn clockwise 15 degree.

4) 30 degree: take reference position as original position, 30 degree are rotated counterclockwise.

5)-30 degree: be original position with reference position, turn clockwise 30 degree.

6) 45 degree: take reference position as original position, 45 degree are rotated counterclockwise.

7)-45 degree: be original position with reference position, turn clockwise 45 degree.

The anglec of rotation of above-mentioned various gesture can be selected to open or close according to actual conditions, simultaneously can in the step of authenticated user identity, the different hand shape of different gesture collection and the above-mentioned various anglec of rotation corresponding palmmprint hand shape sequence of composition that cooperatively interacts is shown to user.User can put the corresponding hand shape image of corresponding gesture collection by above-mentioned authenticating step in appointed area and participate in verification process.

The selection of above-mentioned angles, in order to further consider the physiological property of staff, i.e. the complexity of palm in the anglec of rotation; The mode of the preset gesture of user rs authentication in registered user's identity characteristic step can be adopted to verify preset angles.Just repeat no longer one by one at this.Take into account when palm is all presented in user terminal screen collection frame, the distance of palm and user terminal is not excessive simultaneously.

In addition, also relate on the screen of the subscriber terminal in technical scheme of the present invention, palm puts the display of instruction frame.As shown in Fig. 8, Figure 10, the present invention have chosen distance up and down each 10% rectangle frame carry out indicating user palm put to this rectangle frame.As in Fig. 8 in 508, Figure 10 shown in 518, this rectangle frame also can do corresponding rotation along with current angles.Adopt the form of rectangle frame can also reduce the data processing time gathering palmmprint hand shape image.Can only cut when gathering palmmprint hand shape image figure in rectangle frame as in requisition for palmmprint hand shape image.As in Figure 10 518, user terminal can be drawn out corresponding hand shape instruction legend and also can do corresponding rotation according to corresponding angles simultaneously.

In another embodiment of the present invention, in order to improve the fail safe of verification process further, in the step of authenticated user identity, corresponding logical operation can also being provided to each gesture further, doing corresponding selection for user according to self needing.Be exemplified below below:

A) logical operation 1: no matter current gesture is any, all shows appointment gesture preset in advance.

B) logical operation 2: for whole (or part) gesture, show gesture in contrast.

C) logical operation 3: for whole (or part) gesture, show another gesture that predefined is good.

In the step of personal recognition, if do not recognize the correct gesture by logical operation, and there is not palm area in image, then feedback terminal user authentication failure.If do not recognize the correct gesture by logical operation, but there is palm area in image, then Returning process 1, continued to gather image by terminal.If recognize the correct gesture by logical operation, then the extraction carrying out palm print characteristics with mate, and preserve matching result.

The object that the invention provides logical operation is the fail safe in order to improve authentication further, for logical operation 1, rule " first gesture occurred in for each authentication is provided with in advance user, no matter service end requires that the gesture shown is any, all show shooting formula gesture (in Figure 10 512) " when, if despiteful aggression, so when not understanding this rule in advance, user just probably shows the gesture of mistake, and this wrong gesture, can process as a kind of warning message in the present invention.

Embodiment 2

The present embodiment is substantially the same manner as Example 1, and its difference is, the present embodiment is applied to the authentication to transaction request in trading environment.The present embodiment provides a kind of safety certifying method based on gesture identification, it is characterized in that: comprise the step of registered user's characteristic information and the step of authenticated user identity; As shown in Figure 1, the step of registered user's characteristic information comprises: the preset palmprint image that user terminal gathers user transmits to service end; Service end is extracted palm print characteristics and is saved presets palm print characteristics; As shown in Figure 2.The step of authenticated user identity comprises: user terminal initiates transaction request to transactional services end, and transactional services end initiates authentication request to service end; Service end generates palmmprint hand shape sequence and sequentially transmits to user terminal and show hand shape image; User sequentially does corresponding gesture motion by the hand shape image shown, user terminal gathers the certification hand shape image of user with palmmprint and transmits to service end; Service end extracts hand shape, palm print characteristics mating with preset hand shape, palm print characteristics; Service end is to transactional services end feedback matching result; Transactional services end performs respective transaction operation according to authentication result.Detailed description is launched to safety certifying method provided by the invention below.

Part I, introduces the step of registered user's characteristic information.

During registered user's characteristic information, need the preset palmprint image first gathering user, then extract palm print characteristics by palm grain identification method and save presets palm print characteristics.

As shown in Figure 3, during registered user's characteristic information, user terminal initiates registered user's feature request to service end; Smooth palm shape (such as, palm is stretched flat, the five fingers open) is positioned over the image acquisition region of user terminal by user, starts the preset palmprint image that image acquisition procedure gathers user; Transfer to service end after collecting preset palmprint image, service end utilizes palm grain identification method to extract the palm print characteristics of the user in preset palmprint image, preserves palm print characteristics after extraction.Since then, registered user's characterization step terminates.In the present embodiment, the step of registered user's characteristic information is identical with embodiment 1, please refer to concrete steps in embodiment 1 and embodiment in the concrete steps of registered user's characteristic information, is save length, has just repeated no longer one by one at this.

Part II, introduces the step of authenticated user identity.

As shown in figure 12, when user needs authenticating user identification in transaction scene, need by this method authenticated user identity.During authenticated user identity, user initiates transaction request by user terminal to transactional services end, and transactional services end initiates authentication request to service end; Service end generates palmmprint hand shape sequence and transfers to user terminal, sequentially shows hand shape image to user; User sequentially does corresponding gesture motion by the hand shape image shown, user terminal gathers the certification hand shape image of user with palmmprint and transfers to service end; Service end is done hand shape by the certification hand shape image of hand shape recognition method to band palmmprint and is identified and mate with preset hand-shaped characteristic; Service end is extracted palm print characteristics by palm grain identification method and is mated with preset palm print characteristics; Service end is to transactional services end feedback matching result; Transactional services end performs respective transaction operation according to authentication result.Detailed description is launched to the safety certifying method that the present embodiment provides below.

First, introduce user terminal and initiate transaction request to transactional services end, transactional services end initiates the step of authentication request to service end.When user needs to conclude the business, user terminal initiates transaction request to transactional services end; When transactional services end receives the transaction request of user, transactional services end initiates authentication request to safety certificate server.As shown in figure 12, when initiating authentication request, user terminal sends after authentication request information package.Preferably send in xml format.Authentication request information comprises ID, the Customs Assigned Number of user terminal.In this step, in order to ensure the fail safe of data verification further, after initiating authentication request, service end also needs the legitimacy of further authentication verification request, in authentication request information, therefore also comprise the random coded of generation.The data packet format example that user terminal encapsulates when initiating authentication request is as follows:

After transactional services termination receives the transaction request of user terminal, send authentication request to service end, after end feedback authentication result to be serviced, feedback result is sent to user terminal by transactional services end.It is as follows that transactional services end is sent to the data packet format example encapsulated when service end sends authentication request:

After service end receives the authentication request packet of transactional services end transmission, service end reads authentication request packet, by verifying the legitimacy of the mode authentication verification request of random coded.Service end examines Customs Assigned Number further by safety certificate server after the mode authentication verification request of checking random coded, if numbering is legal, then feeds back to transactional services end " permission certification ", otherwise feedback " identity is illegal ".Safety certificate server is to transactional services end feedback ID authentication request result.It is as follows that service end is sent to the data packet format example encapsulated when transactional services end sends authentication request result:

After service end receives the authentication request packet of user terminal transmission, service end reads authentication request packet, by verifying the legitimacy of the mode authentication verification request of random coded.If the verification passes, then start a certification thread and carry out follow-up certification; If checking is not passed through, then transmit result to user terminal in the form of a message, inform that user authentication request is illegal, certification terminates.Does user terminal, after the authentication request result receiving transactional services end feedback, initiatively access identify_url address (such as: https: // 10.77.12.9/identify.asp? id=459B63D3_1232_47cb_B568_43475715B79C).

As shown in figure 12, generate palmmprint hand shape sequence and sequentially show that the step of hand shape image is substantially the same manner as Example 1 to user, just having repeated no longer one by one at this.After safety certificate server receives the authentication request of client, a set of identification sequences can be generated, and feed back to user terminal.The hand shape information packet format sample of service end transmission is as follows:

As shown in figure 12, user sequentially does corresponding gesture motion by the hand shape image shown and to gather the step of the certification hand shape image of user with palmmprint substantially the same manner as Example 1, has just repeated no longer one by one at this.User terminal can draw out corresponding hand shape instruction legend at user terminal, and point out user after receiving the authentication requesting that server sends, and user terminal sends to safety certificate server at backstage timing acquiring user present image simultaneously.User terminal is as follows to the data packet format example encapsulated during the certification hand shape image of service end transmit band palmmprint:

3rd, introduce by the certification hand shape image of hand shape recognition method to band palmmprint do hand shape identify and with the step of preset hand-shaped characteristic comparison.

As shown in figure 12, do hand shape by the certification hand shape image of hand shape recognition method to band palmmprint and to identify and substantially the same manner as Example 1 with the step of preset hand-shaped characteristic comparison, just repeated no longer one by one at this.

4th, introduce and extract palm print characteristics and the step of mating with preset palm print characteristics by palm grain identification method.

As shown in figure 12, extract palm print characteristics the step of mating with preset palm print characteristics is substantially the same manner as Example 1 by palm grain identification method, just repeated no longer one by one at this.Safety certificate server, after the image receiving user terminal transmission, can detect and send corresponding task to user terminal.The service end data packet format example that safety certificate server sends message encapsulation to user terminal when the certification hand shape image of the band palmmprint of comparison collection makes hand shape, palmmprint is as follows:

As shown in figure 12, in order to ensure the fail safe of verification process, before pressing palm grain identification method extraction palm print characteristics, when gathering the certification hand shape image of user with palmmprint, the palm confirming pickup area is needed to be continued presence, if the palm of pickup area is discontinuous, point out user authentication failure (such as in authentication procedures, user is when performing palmmprint hand shape image sequence, hand is shifted out the image acquisition region of user terminal by user, then think that this ID authentication request is illegal, the failure of prompting user authentication request, needs authenticating identity again).In technical scheme of the present invention, confirm that the palm of pickup area is that the object of continued presence is, prevent the forgery terminal of malice from gathering the behavior of image, thus improve the fail safe of authentication further.

3) tracing process adopts MeanShift (average drifting) method.

As shown in figure 12, in order to ensure the fail safe of verification process, in one embodiment of the present of invention, in the step of authenticated user identity, the present invention also provides and puts and the step verified by specified angle user's gesture.Technical scheme of the present invention is also prefixed multiple angles degree AI, gets 0 °≤AI≤90 °.

In the present invention, the granularity of hand angles is 1 °, so for each hand shape, all there are 91 kinds of angles., there are 637 (=7 × 91) and plant possibility in all hand shapes for each the unknown; For the hand shape sequence that a length is N, then have 637 ⁿplant combination.In the present invention, NN5, namely for each group hand shape sequence, at least exists 104881082626957 (=637 ⁵) plant combination.

As 508 in Fig. 8, several angles are wherein exemplified below respectively:

8) 0 degree: i.e. reference position, palm is flat to be stretched, and facies palmaris is parallel with user terminal screen; Finger tip vertically upward.

9) 15 degree: take reference position as original position, 15 degree are rotated counterclockwise.

10)-15 degree: be original position with reference position, turn clockwise 15 degree.

11) 30 degree: take reference position as original position, 30 degree are rotated counterclockwise.

12)-30 degree: be original position with reference position, turn clockwise 30 degree.

13) 45 degree: take reference position as original position, 45 degree are rotated counterclockwise.

14)-45 degree: be original position with reference position, turn clockwise 45 degree.

(a) logical operation 1: no matter current gesture is any, all shows appointment gesture preset in advance.

(b) logical operation 2: for whole (or part) gesture, show gesture in contrast.

(c) logical operation 3: for whole (or part) gesture, show another gesture that predefined is good.

Finally, introduce service end to transactional services end feedback matching result, transactional services end performs respective transaction operation according to authentication result.

Service end transmits authentication result to after subscriber authentication to transactional services end, and transactional services end is according to authentication result process transaction request.Pass through if authentication result is certification, then transactional services end performs corresponding transactional operation.If authentification failure, then feed back the information of certification again or Fail Transaction to user terminal.After certification completes, user terminal can receive the authentication result from safety certificate server, also can obtain corresponding transaction results by inquiry transactional services end simultaneously.No matter authentication success or failure, safety certificate server all can initiatively send result to service end of concluding the business after certification terminates.The data packet format example that service end encapsulates when transactional services end sends authentication result is as follows:

Embodiment 3

For embodying the superiority of inventing the safety certifying method provided further, the present invention also provides a kind of security certification system applying above-mentioned safety certifying method, and as shown in figure 13, this system comprises: service end and user terminal; Service end is connected with user terminal communication; User terminal gathers the preset palmprint image of user and is sent to service end, and service end is extracted palm print characteristics by palm grain identification method and saved presets palm print characteristics; User terminal initiates authentication request to service end, service end generates palmmprint hand shape sequence and transfers to user terminal, user terminal sequentially shows hand shape image, the hand shape image that user shows by user terminal sequentially does corresponding gesture motion, user terminal gathers the hand shape image of the band palmmprint of user and transfers to service end, service end is extracted palm print characteristics by palm grain identification method and is mated with preset palm print characteristics, and service end feeds back authentication result by matching result to user terminal.Detailed description is launched to security certification system provided by the invention below.

As shown in Figure 13, Figure 14, in security certification system provided by the invention, user terminal can be PC, can be PAD, also can be mobile phone etc. all can gather the electronic equipment of image.User terminal needs the image acquisition device user's characteristic information by self.Image collecting device is preferably camera.User terminal comprises data transmit-receive module, business logic modules image capture module.Interactive interface module, data transmit-receive module, image capture module communicate to connect with business logic modules respectively, and image capture module and data transmit-receive module communicate to connect.User initiates instruction by interactive interface module to business logic modules, business logic modules controls to send registration or authentication request by data transmit-receive module to service end, and business logic modules is controlled image capture module and gathers palmmprint hand shape image and transmitted to service end by data transmit-receive module.Business logic modules control data interactive module receives the palmmprint hand shape sequence of service end transmission and transfers to interactive interface module and sequentially show hand shape image to user.

As shown in figure 14, in order to ensure the fail safe of data further, user terminal also comprises data encryption module, and business logic modules, image capture module and data encryption module communicate to connect, and data encryption module is connected with data transmit-receive module.The image of image capture module collection is transmitted by data transmit-receive module after being encrypted by data encryption module.

As shown in figure 15, in order to ensure still can by user terminal registered user characteristic information or authenticated user identity when network condition is not good, user terminal also comprises image pre-processing module, and business logic modules, image capture module and image pre-processing module communicate to connect.Business logic modules controls image pre-processing module and the image information of image capture module collection is done preliminary treatment, is transmitted by pretreated image by data transceiving unit.It is identical with the pre-treatment step in above-mentioned safety certifying method that the image information (containing the preset palmprint image of preliminary treatment and certification hand shape image) that image pre-processing module gathers image capture module in security certification system provided by the invention makes the pretreated step of the preset palmprint image of preliminary treatment, just repeated no longer one by one at this.As shown in figure 15, further, the view data that image capture module gathers first by image pre-processing module process, then can be transmitted by data transmit-receive module after data encryption module encryption.

As shown in Figure 13, Figure 17, service end comprises safety certificate server and the database of band identity authentication function, and database comprises palm print characteristics database and hand shape setting data storehouse.Palm print characteristics database and hand shape setting data storehouse and safety certificate server communicate to connect.Safety certificate server comprises data transmit-receive module, hand shape identification module, personal recognition module and dynamic hand shape generation module.Data transmit-receive module is connected with hand shape identification module, personal recognition module and dynamic hand shape generation module respectively, and hand shape identification module, dynamically hand shape generation module are connected with hand shape setting data storehouse respectively, personal recognition module and palm print characteristics DataBase combining; During registered user's characteristic information: the preset palmprint image received is transferred to personal recognition module and extracts palm print characteristics and be kept in palm print characteristics database by data transmit-receive module, and send feedback registering result to user terminal by data transmit-receive module; During authenticated user identity: dynamically hand shape generation module generates palmmprint hand shape sequence by preset hand shape image and transfers to user terminal by data transmit-receive module; The certification hand shape image of reception is transferred to hand shape identification module by data transmit-receive module, personal recognition module extracts hand shape, palm print characteristics mating with the preset hand shape in hand shape setting data storehouse, palm print characteristics database, palm print characteristics; Hand shape identification module, personal recognition module feed back authentication result by data transmit-receive module to user terminal according to matching result.It is identical with the step of the extraction palm print characteristics in above-mentioned safety certifying method that personal recognition module extracts the step of the user's palm print characteristics in preset palmprint image or in certification hand shape image, just repeated no longer one by one at this.In hand shape identification module identification certification hand shape image, the step of hand shape is identical with the hand shape identification step in above-mentioned safety certifying method, has just repeated no longer one by one at this.

As shown in Figure 16,17, in order to ensure the fail safe of data further, safety certificate server also comprises the data decryption module corresponding with user terminal, data transmit-receive module is connected with data decryption module respectively with personal recognition module, after safety certificate server receives the image of user terminal encryption, be sent to data decryption module by data transmit-receive module, be resent to personal recognition resume module by after data decryption module decryption processing.

As shown in figure 17, in order to ensure the fail safe of authentication further, service end also comprises palm tracking module: palm tracking module is connected with data transmit-receive module; Palm tracking module, when gathering the certification hand shape image of user with palmmprint, adopts hand shape area tracking algorithm opponent shape area tracking, determines that whether palmmprint hand shape image is continuous; If the discontinuous existence of the palm of pickup area, points out user authentication failure by data transmit-receive module; If the palm continued presence of pickup area, continue to gather other palmmprint hand shape images by data transmit-receive module prompting user.

As shown in figure 17, service end also comprises recognition decision module: recognition decision module respectively with hand shape identification module, personal recognition model calling; Service end is by after certification palmprint image identification matching treatment, and recognition decision module judges whether the palmmprint hand shape sequence generated gathers, identifies, mated; If do not complete palmmprint hand shape sequence, then recognition decision module is to the next hand shape image of user's transmission user, and user repeats above-mentioned steps until the palmmprint hand shape sequence generated all completes; If palmmprint hand shape sequence completes, then the palmprint information matching result of recognition decision module searches record, if there is the result that the match is successful, for user performs corresponding business operation, then by data transmit-receive module to the successful message of user terminal certified transmission; If without the palmprint information matching result that the match is successful, by data transmit-receive module to user terminal certified transmission failure.

Embodiment 4

For embodying the superiority of inventing the safety certifying method provided further, the present invention also provides a kind of security certification system applying above-mentioned safety certifying method, user, when transaction or other need authentication of users scene, needs by safety Verification System authenticated user identity.As shown in figure 18, this security certification system comprises service end and user terminal, server is connected with user terminal communication, also comprise the transactional services end be connected with server and user terminal communication, user terminal gathers the preset palmprint image of user and is sent to service end, and service end is extracted palm print characteristics by palm grain identification method and saved presets palm print characteristics, user initiates transaction authentication request by user terminal to transactional services end, transactional services end initiates identity energy authentication request to service end, service end generates palmmprint hand shape sequence and transfers to user terminal, user terminal sequentially shows hand shape image, the hand shape image that user shows by user terminal sequentially does corresponding gesture motion, user terminal gathers the hand shape image of the band palmmprint of user and transfers to service end, service end is extracted palm print characteristics by palm grain identification method and is mated with preset palm print characteristics, service end is to transactional services end user terminal feedback matching result, transactional services end starts transaction results information according to the matching result of feedback by user terminal.Detailed description is launched to security certification system provided by the invention below.

As shown in figure 14, user terminal comprises: data transmit-receive module, business logic modules image capture module; Interactive interface module, data transmit-receive module, image capture module communicate to connect with business logic modules respectively, and image capture module and data transmit-receive module communicate to connect; User initiates instruction by interactive interface module to business logic modules, business logic modules controls to send registration or authentication request by data transmit-receive module to service end, and business logic modules is controlled image capture module and gathers palmmprint hand shape image and transmitted to service end by data transmit-receive module; Business logic modules control data interactive module receives the palmmprint hand shape sequence of service end transmission and transfers to interactive interface module and sequentially show hand shape image to user.

As shown in figure 15, user terminal also comprises data encryption module, and business logic modules, image capture module and data encryption module communicate to connect, and data encryption module is connected with data transmit-receive module; The image of image capture module collection is transmitted by data transmit-receive module after being encrypted by data encryption module.

As shown in figure 15, user terminal also comprises image pre-processing module; Business logic modules, image capture module and image pre-processing module communicate to connect; Business logic modules controls image pre-processing module and the image information of image capture module collection is done preliminary treatment, is transmitted by pretreated image by data transceiving unit.It is identical with the pre-treatment step in above-mentioned safety certifying method that the image information (containing the preset palmprint image of preliminary treatment and certification hand shape image) that image pre-processing module gathers image capture module in security certification system provided by the invention makes the pretreated step of the preset palmprint image of preliminary treatment, just repeated no longer one by one at this.As shown in figure 15, further, the view data that image capture module gathers first by image pre-processing module process, then can be transmitted by data transmit-receive module after data encryption module encryption.

As shown in Figure 13, Figure 17, service end comprises: the safety certificate server of band identity authentication function and database, and database comprises palm print characteristics database and hand shape setting data storehouse; Palm print characteristics database and hand shape setting data storehouse and safety certificate server communicate to connect; Safety certificate server comprises data transmit-receive module, hand shape identification module, personal recognition module and dynamic hand shape generation module; Data transmit-receive module is connected with hand shape identification module, personal recognition module and dynamic hand shape generation module respectively, and hand shape identification module, dynamically hand shape generation module are connected with hand shape setting data storehouse respectively, personal recognition module and palm print characteristics DataBase combining; During registered user's characteristic information: the preset palmprint image received is transferred to personal recognition module and extracts palm print characteristics and be kept in palm print characteristics database by data transmit-receive module, and send feedback registering result to user terminal by data transmit-receive module; During authenticated user identity: dynamically hand shape generation module generates palmmprint hand shape sequence by preset hand shape image and transfers to user terminal by data transmit-receive module; The certification hand shape image of reception is transferred to hand shape identification module by data transmit-receive module, personal recognition module extracts hand shape, palm print characteristics mating with the preset hand shape in hand shape setting data storehouse, palm print characteristics database, palm print characteristics; Hand shape identification module, personal recognition module feed back authentication result by data transmit-receive module to user terminal according to matching result.It is identical with the step of the extraction palm print characteristics in above-mentioned safety certifying method that personal recognition module extracts the step of the user's palm print characteristics in preset palmprint image or in certification hand shape image, just repeated no longer one by one at this.In hand shape identification module identification certification hand shape image, the step of hand shape is identical with the hand shape identification step in above-mentioned safety certifying method, has just repeated no longer one by one at this.

As shown in Figure 16, Figure 17, service end also comprises the data decryption module corresponding with user terminal; Data transmit-receive module is connected with data decryption module respectively with personal recognition module; Safety certificate server is sent to data decryption module by data transmit-receive module, by reprocessing after data decryption module decryption processing after receiving the image of user terminal encryption.

As shown in figure 17, service end also comprises palm tracking module: palm tracking module is connected with data transmit-receive module; Palm tracking module, when gathering the certification hand shape image of user with palmmprint, adopts hand shape area tracking algorithm opponent shape area tracking, determines that whether palmmprint hand shape image is continuous; If the discontinuous existence of the palm of pickup area, points out user authentication failure by data transmit-receive module; If the palm continued presence of pickup area, continue to gather other palmmprint hand shape images by data transmit-receive module prompting user.

Embodiment 5

As shown in figure 13, the invention provides a kind of safety certificate server based on gesture identification, safety certificate server is used for the authority of authenticated user in security certification system; Safety certificate server is connected with user terminal communication; Safety certificate server receives the preset palmprint image that user terminal gathers user, and safety certificate server extracts palm print characteristics by palm grain identification method and saves presets palm print characteristics; Safety certificate server receives user terminal and initiates authentication request to safety certificate server, safety certificate server generates palmmprint hand shape sequence and transfers to user terminal, user terminal sequentially shows hand shape image, the hand shape image that user shows by user terminal sequentially does corresponding gesture motion, user terminal gathers the hand shape image of the band palmmprint of user and transfers to safety certificate server, safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, and safety certificate server feeds back authentication result by matching result to user terminal.

As shown in figure 14, user terminal comprises: data transmit-receive module, business logic modules image capture module; Interactive interface module, data transmit-receive module, image capture module communicate to connect with business logic modules respectively, and image capture module and data transmit-receive module communicate to connect; User initiates instruction by interactive interface module to business logic modules, business logic modules controls to send registration or authentication request by data transmit-receive module to safety certificate server, and business logic modules is controlled image capture module and gathers palmmprint hand shape image and transmitted to safety certificate server by data transmit-receive module; The palmmprint hand shape sequence of business logic modules control data interactive module reception server transmission also transfers to interactive interface module and sequentially shows hand shape image to user.

As shown in Figure 13, Figure 17, server comprises: the safety certificate server of band identity authentication function and database, and database comprises palm print characteristics database and hand shape setting data storehouse; Palm print characteristics database and hand shape setting data storehouse and safety certificate server communicate to connect; Safety certificate server comprises data transmit-receive module, hand shape identification module, personal recognition module and dynamic hand shape generation module; Data transmit-receive module is connected with hand shape identification module, personal recognition module and dynamic hand shape generation module respectively, and hand shape identification module, dynamically hand shape generation module are connected with hand shape setting data storehouse respectively, personal recognition module and palm print characteristics DataBase combining; During registered user's characteristic information: the preset palmprint image received is transferred to personal recognition module and extracts palm print characteristics and be kept in palm print characteristics database by data transmit-receive module, and send feedback registering result to user terminal by data transmit-receive module; During authenticated user identity: dynamically hand shape generation module generates palmmprint hand shape sequence by preset hand shape image and transfers to user terminal by data transmit-receive module; The certification hand shape image of reception is transferred to hand shape identification module by data transmit-receive module, personal recognition module extracts hand shape, palm print characteristics mating with the preset hand shape in hand shape setting data storehouse, palm print characteristics database, palm print characteristics; Hand shape identification module, personal recognition module feed back authentication result by data transmit-receive module to user terminal according to matching result.It is identical with the step of the extraction palm print characteristics in above-mentioned safety certifying method that personal recognition module extracts the step of the user's palm print characteristics in preset palmprint image or in certification hand shape image, just repeated no longer one by one at this.In hand shape identification module identification certification hand shape image, the step of hand shape is identical with the hand shape identification step in above-mentioned safety certifying method, has just repeated no longer one by one at this.

As shown in Figure 16, Figure 17, safety certificate server also comprises the data decryption module corresponding with user terminal; Data transmit-receive module is connected with data decryption module respectively with personal recognition module; Safety certificate server is sent to data decryption module by data transmit-receive module, by reprocessing after data decryption module decryption processing after receiving the image of user terminal encryption.

As shown in figure 17, safety certificate server also comprises palm tracking module: palm tracking module is connected with data transmit-receive module; Palm tracking module, when gathering the certification hand shape image of user with palmmprint, adopts hand shape area tracking algorithm opponent shape area tracking, determines that whether palmmprint hand shape image is continuous; If the discontinuous existence of the palm of pickup area, points out user authentication failure by data transmit-receive module; If the palm continued presence of pickup area, continue to gather other palmmprint hand shape images by data transmit-receive module prompting user.

As shown in figure 17, safety certificate server also comprises recognition decision module: recognition decision module respectively with hand shape identification module, personal recognition model calling; Safety certificate server is by after certification palmprint image identification matching treatment, and recognition decision module judges whether the palmmprint hand shape sequence generated gathers, identifies, mated; If do not complete palmmprint hand shape sequence, then recognition decision module is to the next hand shape image of user's transmission user, and user repeats above-mentioned steps until the palmmprint hand shape sequence generated all completes; If palmmprint hand shape sequence completes, then the palmprint information matching result of recognition decision module searches record, if there is the result that the match is successful, for user performs corresponding business operation, then by data transmit-receive module to the successful message of user terminal certified transmission; If without the palmprint information matching result that the match is successful, by data transmit-receive module to user terminal certified transmission failure.

Embodiment 6

As shown in figure 18, the invention provides a kind of safety certificate server based on gesture identification, safety certificate server is used for the authority of authenticated user in security certification system, safety certificate server and user terminal and trading server communicate to connect, safety certificate server receives the preset palmprint image that user terminal gathers user, and safety certificate server extracts palm print characteristics by palm grain identification method and saves presets palm print characteristics, user initiates transaction authentication request by user terminal to trading server, safety certificate server receives trading server and initiates authentication request to safety certificate server, safety certificate server generates palmmprint hand shape sequence and transfers to user terminal, user terminal sequentially shows hand shape image, the hand shape image that user shows by user terminal sequentially does corresponding gesture motion, user terminal gathers the hand shape image of the band palmmprint of user and transfers to safety certificate server, safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, server is to trading server user terminal feedback matching result, trading server starts transaction results information according to the matching result of feedback by user terminal.

As shown in Figure 13, Figure 17, safety certificate server comprises: the safety certificate server of band identity authentication function and database, and database comprises palm print characteristics database and hand shape setting data storehouse; Palm print characteristics database and hand shape setting data storehouse and safety certificate server communicate to connect; Safety certificate server comprises data transmit-receive module, hand shape identification module, personal recognition module and dynamic hand shape generation module; Data transmit-receive module is connected with hand shape identification module, personal recognition module and dynamic hand shape generation module respectively, and hand shape identification module, dynamically hand shape generation module are connected with hand shape setting data storehouse respectively, personal recognition module and palm print characteristics DataBase combining; During registered user's characteristic information: the preset palmprint image received is transferred to personal recognition module and extracts palm print characteristics and be kept in palm print characteristics database by data transmit-receive module, and send feedback registering result to user terminal by data transmit-receive module; During authenticated user identity: dynamically hand shape generation module generates palmmprint hand shape sequence by preset hand shape image and transfers to user terminal by data transmit-receive module; The certification hand shape image of reception is transferred to hand shape identification module by data transmit-receive module, personal recognition module extracts hand shape, palm print characteristics mating with the preset hand shape in hand shape setting data storehouse, palm print characteristics database, palm print characteristics; Hand shape identification module, personal recognition module feed back authentication result by data transmit-receive module to user terminal according to matching result.It is identical with the step of the extraction palm print characteristics in above-mentioned safety certifying method that personal recognition module extracts the step of the user's palm print characteristics in preset palmprint image or in certification hand shape image, just repeated no longer one by one at this.In hand shape identification module identification certification hand shape image, the step of hand shape is identical with the hand shape identification step in above-mentioned safety certifying method, has just repeated no longer one by one at this.

Embodiment 7

As shown in figure 13, the invention provides a kind of safety certification user terminal based on gesture identification, it is characterized in that: user terminal is used for the authority of authenticated user in security certification system; User terminal can be PC, can be PAD, also can be mobile phone etc. all can gather the electronic equipment of image.User terminal needs the image acquisition device user's characteristic information by self.Image collecting device is preferably camera.Safety certificate server is connected with user terminal communication; The preset palmprint image that user terminal gathers user transfers to safety certificate server, and safety certificate server extracts palm print characteristics by palm grain identification method and saves presets palm print characteristics; Safety certificate server receives user terminal and initiates authentication request to safety certificate server, safety certificate server generates palmmprint hand shape sequence and transfers to user terminal, user terminal sequentially shows hand shape image, the hand shape image that user shows by user terminal sequentially does corresponding gesture motion, user terminal gathers the hand shape image of the band palmmprint of user and transfers to safety certificate server, safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, and safety certificate server feeds back authentication result by matching result to user terminal.

Embodiment 8

As shown in figure 18, the invention provides a kind of safety certification user terminal based on gesture identification, user terminal is used for the authority of authenticated user in security certification system; User terminal can be PC, can be PAD, also can be mobile phone etc. all can gather the electronic equipment of image.User terminal needs the image acquisition device user's characteristic information by self.Image collecting device is preferably camera.User terminal and Security Authentication Service and trading server communicate to connect, the preset palmprint image that user terminal gathers user transfers to safety certificate server, and safety certificate server extracts palm print characteristics by palm grain identification method and saves presets palm print characteristics, user terminal initiates transaction authentication request to trading server, safety certificate server receives trading server and initiates authentication request to safety certificate server, safety certificate server generates palmmprint hand shape sequence and transfers to user terminal, user terminal sequentially shows hand shape image, the hand shape image that user shows by user terminal sequentially does corresponding gesture motion, user terminal gathers the hand shape image of the band palmmprint of user and transfers to safety certificate server, safety certificate server extracts palm print characteristics by palm grain identification method and mates with preset palm print characteristics, server is to trading server user terminal feedback matching result, trading server starts transaction results information according to the matching result of feedback by user terminal.

The safety certifying method based on gesture identification, terminal, server and system that inventing provides not only can be applied to the authentication mode in above-described embodiment, can also by other equipment with being used in mobile payment, long-range enabling, unblanking of safe deposit box, the gate inhibition of armament depot and authentication.In frontier inspection clearance, police and judicial, financial instrument, ecommerce, social security welfare, the public safety fields such as information network and gate inhibition, work attendance, school, hospital, venue, the civil areas such as supermarket can be applied.Can be able to realize doing in above-described embodiment 1-8 suitable modification in above-mentioned applied environment.Concrete enforcement distortion, with reference to above-described embodiment 1-8, does not remake detailed having repeated at this.

To sum up, safety certifying method based on gesture identification provided by the invention, terminal, server and system, dynamic gesture sequence and the identification of palm feature are combined, neither need user's keeping and carry extra identity keepsake, reduce again the possibility that identity information is held as a hostage to a great extent, improve the fail safe of authentication.1) loss of certification keepsake is prevented.Owing to present invention uses the feature of staff, therefore natural possess anti-lost feature.2) fail safe of long-distance identity-certifying is substantially increased.First, present invention uses the track algorithm based on palm feature, to guarantee in the process of authentication, the certification terminal of being held as a hostage is difficult to adopt the mode switched fast to forge staff image; Then, in the process that palm is followed the tracks of, the present invention employs the dynamic sequence based on Random-Rotation angle and preset hand shape produced by server simultaneously, to guarantee that the certification terminal of being held as a hostage cannot by playing the palm video recorded or artificial fingerprint carry out fake user; Finally, while safety certificate server verifies dynamic hand shape sequence, also detect the palm print characteristics of user to identify the identity of user further.Based on the synchronous execution of above-mentioned three kinds of methods, the present invention effectively can improve the fail safe of long-distance identity-certifying under certification terminal is held as a hostage situation.3)

By image acquisition device existing on mobile terminal or camera, without the need to additionally increasing special capture setting, the cost of further reduction system, user-friendly.

Those skilled in the art should understand, embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the present invention can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disc store, CD-ROM, optical memory etc.) of computer usable program code.

The present invention describes with reference to according to the flow chart of the method for the embodiment of the present invention, equipment (system) and computer program and/or block diagram.Should understand can by the combination of the flow process in each flow process in computer program instructions realization flow figure and/or block diagram and/or square frame and flow chart and/or block diagram and/or square frame.These computer program instructions can being provided to the processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing device to produce a machine, making the instruction performed by the processor of computer or other programmable data processing device produce device for realizing the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.

These computer program instructions also can be stored in can in the computer-readable memory that works in a specific way of vectoring computer or other programmable data processing device, the instruction making to be stored in this computer-readable memory produces the manufacture comprising command device, and this command device realizes the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.

These computer program instructions also can be loaded in computer or other programmable data processing device, make on computer or other programmable devices, to perform sequence of operations step to produce computer implemented process, thus the instruction performed on computer or other programmable devices is provided for the step realizing the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.

Although describe the preferred embodiments of the present invention, those skilled in the art once obtain the basic creative concept of cicada, then can make other change and amendment to these embodiments.So claims are intended to be interpreted as comprising preferred embodiment and falling into all changes and the amendment of the scope of the invention.

Obviously, those skilled in the art can carry out various change and modification to the present invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.

Above execution mode is only for illustration of the present invention; and be not limitation of the present invention; the those of ordinary skill of relevant technical field; without departing from the spirit and scope of the present invention; various change and modification can also be made; therefore all equivalent technical schemes also belong to category of the present invention, and scope of patent protection of the present invention should be defined by the claims.

Claims

1. based on a safety certifying method for gesture identification, it is characterized in that: comprise the step of registered user's characteristic information and the step of authenticated user identity;

The step of described registered user's characteristic information comprises:

Gather the preset palmprint image of user;

The step of described authenticated user identity comprises:

User initiates authentication request;

Feedback matching result.

2. safety certifying method as claimed in claim 1, is characterized in that: the step of the preset palmprint image of described collection user specifically comprises:

Smooth palm facies palmaris is positioned over image capture position by user;

Gather user's palm facies palmaris image;

To palm facies palmaris Image semantic classification.

3. safety certifying method as claimed in claim 2, is characterized in that: the described step to palm facies palmaris Image semantic classification specifically comprises:

To the palm facies palmaris image coarse segmentation process gathered;

To the palm facies palmaris image median filter process after equilibrium;

4. safety certifying method as claimed in claim 3, is characterized in that: the step of the described palm facies palmaris image coarse segmentation process to gathering specifically comprises:

Utilize the area of skin color of H component extraction hand.

5. safety certifying method as claimed in claim 3, is characterized in that: the described palm facies palmaris display foreground to collection and background are distinguished and the step of equilibrium treatment specifically comprises:

6. safety certifying method as claimed in claim 1, is characterized in that: described preset palmprint image or certification hand shape image are by camera collection.

7. safety certifying method as claimed in claim 1, is characterized in that: the step of described extraction palm print characteristics specifically comprises:

Locate the facies palmaris region in preset palmprint image;

Cut the facies palmaris region in preset palmprint image;

Extract the palm print characteristics in preset palmprint image.

8. safety certifying method as claimed in claim 7, it is characterized in that, the step in the facies palmaris region in the preset palmprint image in described location specifically comprises:

9. safety certifying method as claimed in claim 8, it is characterized in that, the step in the facies palmaris region in the preset palmprint image of described cutting specifically comprises:

Find out the mid point b that key point a and key point c forms line segment;

With b point for initial point, make the normal f of line segment ac;

Extracting P × P region is facies palmaris region.

10. safety certifying method as claimed in claim 1, is characterized in that: described palmmprint hand shape sequence is from hand shape candidate collection, select the hand shape image of multiple different hand shapes to generate.

11. safety certifying methods as claimed in claim 1, is characterized in that: described palmmprint hand shape sequence is stochastic generation.

12. safety certifying methods as claimed in claim 1, is characterized in that: described to user show hand shape image with hand shape instruction illustrated form show at user terminal.

13. safety certifying methods as claimed in claim 1, is characterized in that: the step of the certification hand shape image of described collection user with palmmprint also comprises:

14. safety certifying methods as claimed in claim 1, is characterized in that: the step of the preset palmprint image of described collection user also comprises the step verified preset hand shape image, specifically comprises:

Preset hand shape image is showed user successively;

Save presets the result of hand shape image checking.

15. safety certifying methods as claimed in claim 1, is characterized in that: also comprise the pretreated step of certification hand shape image after the step of the certification hand shape image of described collection user with palmmprint, specifically comprise:

To the certification hand shape image coarse segmentation process gathered;

16. safety certifying methods as claimed in claim 1, is characterized in that:

The step of the authentication image encryption to band palmmprint is also comprised after the step of the certification hand shape image of described collection user with palmmprint;

17. safety certifying methods as claimed in claim 1, is characterized in that: the step of described extraction hand-shaped characteristic comprises:

18. safety certifying methods as claimed in claim 17, is characterized in that, the step that described all hand shape masks with pre-setting compare as matching degree specifically comprises:

Fill result and determine corresponding matching degree.

19. safety certifying methods as claimed in claim 18, is characterized in that, the step that the described hand shape image by extraction fills hand shape mask corresponding to preset hand shape image also comprises:

Corresponding matching degree is determined according to filling result.

20. safety certifying methods as claimed in claim 1, is characterized in that, also comprise before described feedback matching result:

21. 1 kinds, based on the safety certifying method of gesture identification, is characterized in that: comprise the step of registered user's characteristic information and the step of authenticated user identity;

The step of described registered user's characteristic information comprises:

The step of described authenticated user identity comprises:

User terminal initiates authentication request to service end;

22. safety certifying methods as claimed in claim 21, is characterized in that: the step of the preset palmprint image of described collection user specifically comprises:

User terminal gathers user's palm facies palmaris image;

User terminal is to palm facies palmaris Image semantic classification.

23. safety certifying methods as claimed in claim 22, is characterized in that: the described step to palm facies palmaris Image semantic classification specifically comprises:

24. safety certifying methods as claimed in claim 23, is characterized in that: the step of the described palm facies palmaris image coarse segmentation process to gathering specifically comprises:

User terminal utilizes the area of skin color of H component extraction hand.

25. safety certifying methods as claimed in claim 23, is characterized in that: the described palm facies palmaris display foreground to collection and background are distinguished and the step of equilibrium treatment specifically comprises:

26. safety certifying methods as claimed in claim 21, is characterized in that: described preset palmprint image or certification hand shape image are by camera collection.

27. safety certifying methods as claimed in claim 21, is characterized in that: the step of described extraction palm print characteristics specifically comprises:

Service end locates the facies palmaris region in preset palmprint image;

Service end cuts the facies palmaris region in preset palmprint image;

Service end extracts the palm print characteristics in preset palmprint image.

28. safety certifying methods as claimed in claim 27, it is characterized in that, the step in the facies palmaris region in the preset palmprint image in described location specifically comprises:

29. safety certifying methods as claimed in claim 28, it is characterized in that, the step in the facies palmaris region in the preset palmprint image of described cutting specifically comprises:

It is facies palmaris region that service end extracts P × P region.

30. safety certifying methods as claimed in claim 21, is characterized in that: described palmmprint hand shape sequence is that service end selects the hand shape image of multiple different hand shapes to generate from hand shape candidate collection.

31. safety certifying methods as claimed in claim 21, is characterized in that: described palmmprint hand shape sequence is stochastic generation.

32. safety certifying methods as claimed in claim 21, is characterized in that: the described hand shape image shown to user is shown at user terminal with hand shape instruction illustrated form.

33. safety certifying methods as claimed in claim 21, is characterized in that: the step of the certification hand shape image of described collection user with palmmprint also comprises:

34. safety certifying methods as claimed in claim 21, is characterized in that: the step of the preset palmprint image of described collection user also comprises the step to the preset hand shape image checking of service end, specifically comprises:

35. safety certifying methods as claimed in claim 21, is characterized in that: also comprise the pretreated step of certification hand shape image after the step of the certification hand shape image of described collection user with palmmprint, specifically comprise:

36. safety certifying methods as claimed in claim 21, is characterized in that:

The step of user terminal to the authentication image encryption of band palmmprint is also comprised after the step of the certification hand shape image of described collection user with palmmprint;

37. safety certifying methods as claimed in claim 21, is characterized in that: the step of described extraction hand-shaped characteristic comprises:

38. safety certifying methods as claimed in claim 27, is characterized in that, the described step compared as matching degree with all hand shape masks that service end pre-sets specifically comprises:

Fill result and determine corresponding matching degree.

39. safety certifying methods as claimed in claim 28, is characterized in that, the step that the described hand shape image by extraction fills hand shape mask corresponding to preset hand shape image also comprises:

40. safety certifying methods as claimed in claim 21, is characterized in that, also comprise before described feedback matching result:

41. 1 kinds, based on the safety certifying method of gesture identification, is characterized in that: comprise the step of registered user's characteristic information and the step of authenticated user identity;

The step of described registered user's characteristic information comprises:

The step of described authenticated user identity comprises:

User terminal initiates transaction request to transactional services end;

Transactional services end sends authentication request to service end;

42. safety certifying methods as claimed in claim 41, is characterized in that: the step of the preset palmprint image of described collection user specifically comprises:

User terminal gathers user's palm facies palmaris image;

User terminal is to palm facies palmaris Image semantic classification.

43. safety certifying methods as claimed in claim 42, is characterized in that: the described step to palm facies palmaris Image semantic classification specifically comprises:

44. safety certifying methods as claimed in claim 43, is characterized in that: the step of the described palm facies palmaris image coarse segmentation process to gathering specifically comprises:

User terminal utilizes the area of skin color of H component extraction hand.

45. safety certifying methods as claimed in claim 43, is characterized in that: the described palm facies palmaris display foreground to collection and background are distinguished and the step of equilibrium treatment specifically comprises:

46. safety certifying methods as claimed in claim 41, is characterized in that: described preset palmprint image or certification hand shape image are by camera collection.

47. safety certifying methods as claimed in claim 41, is characterized in that: the step of described extraction palm print characteristics specifically comprises:

Service end locates the facies palmaris region in preset palmprint image;

Service end cuts the facies palmaris region in preset palmprint image;

Service end extracts the palm print characteristics in preset palmprint image.

48. safety certifying methods as claimed in claim 47, it is characterized in that, the step in the facies palmaris region in the preset palmprint image in described location specifically comprises:

49. safety certifying methods as claimed in claim 48, it is characterized in that, the step in the facies palmaris region in the preset palmprint image of described cutting specifically comprises:

It is facies palmaris region that service end extracts P × P region.

50. safety certifying methods as claimed in claim 41, is characterized in that: described palmmprint hand shape sequence is that service end selects the hand shape image of multiple different hand shapes to generate from hand shape candidate collection.

51. safety certifying methods as claimed in claim 41, is characterized in that: described palmmprint hand shape sequence is stochastic generation.

52. safety certifying methods as claimed in claim 41, is characterized in that: the described hand shape image shown to user is shown at user terminal with hand shape instruction illustrated form.

53. safety certifying methods as claimed in claim 41, is characterized in that: the step of the certification hand shape image of described collection user with palmmprint also comprises:

54. safety certifying methods as claimed in claim 41, is characterized in that: the step of the preset palmprint image of described collection user also comprises the step to the preset hand shape image checking of service end, specifically comprises:

55. safety certifying methods as claimed in claim 41, is characterized in that: also comprise the pretreated step of certification hand shape image after the step of the certification hand shape image of described collection user with palmmprint, specifically comprise:

56. safety certifying methods as claimed in claim 41, is characterized in that:

57. safety certifying methods as claimed in claim 41, is characterized in that: the step of described extraction hand-shaped characteristic comprises:

58. safety certifying methods as claimed in claim 47, is characterized in that, the described step compared as matching degree with all hand shape masks that service end pre-sets specifically comprises:

Fill result and determine corresponding matching degree.

59. safety certifying methods as claimed in claim 48, is characterized in that, the step that the described hand shape image by extraction fills hand shape mask corresponding to preset hand shape image also comprises:

60. safety certifying methods as claimed in claim 41, is characterized in that, also comprise before described feedback matching result:

61. safety certifying methods as claimed in claim 41, is characterized in that, described transactional services end performs transaction by authentication result, and comprises to the step of user terminal feedback trading result:

62. 1 kinds, based on the security certification system of gesture identification, is characterized in that: comprise service end and user terminal;

Described service end is connected with described user terminal communication;

63. security certification systems as claimed in claim 62, is characterized in that: described user terminal comprises: data transmit-receive module, business logic modules image capture module;

64. security certification systems as described in claim 63, it is characterized in that: described user terminal also comprises data encryption module, business logic modules, image capture module and data encryption module communicate to connect, and data encryption module is connected with data transmit-receive module;

65. security certification systems as described in claim 63, is characterized in that: described user terminal also comprises image pre-processing module;

66. security certification systems as claimed in claim 62, is characterized in that: described service end comprises:

During registered user's characteristic information:

During authenticated user identity:

67. security certification systems as described in claim 64, is characterized in that: described service end also comprises the data decryption module corresponding with user terminal;

68. security certification systems as described in claim 66, is characterized in that: described service end also comprises palm tracking module:

69. security certification systems as described in claim 66, is characterized in that: described service end also comprises recognition decision module:

70. 1 kinds, based on the security certification system of gesture identification, is characterized in that: comprise service end and user terminal;

Described server is connected with described user terminal communication;

71. security certification systems as described in claim 70, is characterized in that: described user terminal comprises: data transmit-receive module, business logic modules image capture module;

72. security certification systems as described in claim 71, it is characterized in that: described user terminal also comprises data encryption module, business logic modules, image capture module and data encryption module communicate to connect, and data encryption module is connected with data transmit-receive module;

73. security certification systems as described in claim 71, is characterized in that: described user terminal also comprises image pre-processing module;

74. security certification systems as described in claim 70, is characterized in that: described service end comprises:

During registered user's characteristic information:

During authenticated user identity:

75. security certification systems as described in claim 72, is characterized in that: described service end also comprises the data decryption module corresponding with user terminal;

76. security certification systems as described in claim 74, is characterized in that: described service end also comprises palm tracking module:

77. security certification systems as described in claim 74, is characterized in that: described service end also comprises recognition decision module:

78. 1 kinds, based on the safety certificate server of gesture identification, is characterized in that:

Described safety certificate server is connected with user terminal communication;

79. 1 kinds, based on the safety certificate server of gesture identification, is characterized in that:

Described safety certificate server and user terminal and trading server communicate to connect;

80. 1 kinds, based on the safety certification user terminal of gesture identification, is characterized in that:

Described user terminal and safety certificate server communicate to connect;

81. 1 kinds, based on the safety certification user terminal of gesture identification, is characterized in that:

Described user terminal and safe safety certificate server and trading server communicate to connect;