CN101246543B - Examiner identity identification method based on bionic and biological characteristic recognition - Google Patents
Examiner identity identification method based on bionic and biological characteristic recognition Download PDFInfo
- Publication number
- CN101246543B CN101246543B CN2008100198360A CN200810019836A CN101246543B CN 101246543 B CN101246543 B CN 101246543B CN 2008100198360 A CN2008100198360 A CN 2008100198360A CN 200810019836 A CN200810019836 A CN 200810019836A CN 101246543 B CN101246543 B CN 101246543B
- Authority
- CN
- China
- Prior art keywords
- examiner
- pen
- data
- signature
- sample set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 70
- 239000011664 nicotinic acid Substances 0.000 title abstract 2
- 238000000605 extraction Methods 0.000 claims abstract description 17
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 11
- 230000004927 fusion Effects 0.000 claims abstract description 11
- 230000001815 facial effect Effects 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 32
- 238000012545 processing Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 10
- 238000004891 communication Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 4
- 238000002611 posturography Methods 0.000 claims description 2
- 238000012795 verification Methods 0.000 abstract description 4
- 230000003044 adaptive effect Effects 0.000 abstract description 2
- 239000002245 particle Substances 0.000 description 15
- 230000006870 function Effects 0.000 description 13
- 238000003909 pattern recognition Methods 0.000 description 9
- 230000005540 biological transmission Effects 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 7
- 238000010606 normalization Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000015556 catabolic process Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000006731 degradation reaction Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 210000004709 eyebrow Anatomy 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000005021 gait Effects 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 238000002156 mixing Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- UPLPHRJJTCUQAY-WIRWPRASSA-N 2,3-thioepoxy madol Chemical compound C([C@@H]1CC2)[C@@H]3S[C@@H]3C[C@]1(C)[C@@H]1[C@@H]2[C@@H]2CC[C@](C)(O)[C@@]2(C)CC1 UPLPHRJJTCUQAY-WIRWPRASSA-N 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 210000000887 face Anatomy 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000011551 log transformation method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 238000011165 process development Methods 0.000 description 1
- 210000001747 pupil Anatomy 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 238000007789 sealing Methods 0.000 description 1
- 238000010561 standard procedure Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000013517 stratification Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Landscapes
- Collating Specific Patterns (AREA)
Abstract
The invention discloses an examiner identification system based on bionic and biometric identification, using synthetically various biometric identification methods based on high-dimension space geometric shape adaptive coverage theory to achieve identification of examiner identity. First of all, through an acquisition equipment, gripping-pen fingerprint is obtained, on-line signature and facial image, and then the data is mapped into high-dimension space observation point after feature extraction, last according to similar sample point continuity in the high-dimension space, through the relation between the observation point and sample set coverage area to obtain different biological characteristic network match degree, then through match degree fusion decision algorithm to identify identity of the examiner, and through automatic addition of new verification data to achieve sample set dynamic updating and trend forecast. Identification of the invention is fast, result is accurate, and the invention is not only suitable for examiner identification in existing examination mode, but also has broader application in the future machine- examination mode.
Description
Technical field
The present invention relates to a kind of identity identifying and authenticating system, be specifically related to a kind ofly merge multiple biological characteristic and based on the identity authentication system of bionical pattern-recognition.
Background technology
Examination is the concentrated inspection to examiner's acquisition of knowledge ability and psychological quality, is the only way which must be passed of each learner.Simultaneously, examination means that chance, image height examine so important examination meeting candidate's life is produced far-reaching influence.Yet in recent years, the examination orders of various society examinations have been subjected to severe challenge, impersonate phenomenon and emerge in an endless stream, havoc the seriousness and the fairness of examination.Tracing it to its cause, is because executive supervision is ineffective on the one hand, then is that technological means falls behind, and is difficult to the candidate is carried out title examination accurately because existing examination pattern is outmoded on the other hand.With the college entrance examination is example, and the invigilator personnel mainly come in person by the photo on the admission card for entrance examination relatively, photo on the I.D. and examinee and confirm examinee's identity, and when these certificates were forged, the invigilator personnel often were difficult to distinguish the true and false of certificate and make the test-taker get by under false pretences.
Along with the development of pattern recognition theory and perfect, obtained significant progress based on the biology of human body, the identity recognizing technology of behavioural characteristic, recognition of face, fingerprint recognition, iris recognition, signature identification, Gait Recognition etc. have all obtained corresponding achievement and progressively have been applied to produce actual in research field separately.These biological characteristics had both reflected still image information (people's face of people, fingerprint, iris), dynamic behaviour information (the on-line signature that has comprised the people again, gait), have good monopoly and exclusiveness, the accuracy rate of each biological characteristic authentication and identification, user's acceptance level and cost aspect are different, relative merits are respectively arranged, for the user, discerning and authenticate by people's face system is the most friendly mode, it is the most reliable that iris recognition and authentication then have been proved to be, stable and a kind of detection approach accurately, on-line signature and handwriting recognition system are gathered conveniently because of it and are simple to operate, are also accepted extensively by the user.But these recognition systems commonly used face many problems when using separately, such as face identification system to illumination, factor such as attitude and expression is very responsive, iris authentication system has very high requirement to the sample quality that collects, not easy to operate during collection, and in actual use probably because the client iris sample quality that collects is too poor or the user was lost efficacy suffering under the situation of ophthalmology disease etc., signature and person's handwriting identification, for same user, its signature and person's handwriting also can produce than big-difference under different times and different conditions, say nothing of the problem of its forgery that faces and personation.And, be used for the accurate evaluation of examiner's identity by the fusion between multiple living things feature recognition Verification System can be addressed the above problem effectively, prevented to adopt forged certificate deception invigilator personnel's the mode that impersonates from technological layer.
Identity identifying method about various biological characteristics fusions, more existing Study on forming achievements have obtained license, as publication number is that the Chinese patent of CN1304114A discloses a kind of identity identification method based on multi-biological characteristic, and this technology belongs to area of pattern recognition.It utilizes people's biological characteristic, as: face picture, iris, fingerprint, person's handwriting etc., the people is carried out identity authentication, and qualification result normalized to same scope with the standard method for normalizing with the output of whole features, adopt methods such as self-organizing feature map neural network and fuzzy neural network technology to merge again respectively; Publication number is that the Chinese patent of CN1794266A discloses identification and authentication method that biological characteristic merges, it is characterized in that, at first obtain people's face, iris, on-line signature and each biological characteristic of off line person's handwriting of user by various collecting devices, next these biological characteristics are sent into corresponding identification authentication sub module respectively and carried out feature extraction and template matches, and the mark that obtains after the coupling separately of output.These marks perhaps are admitted to the identification Fusion Module through after normalization, get to the end recognition result by steps such as degree of confidence are integrated; Perhaps be admitted to the authentication Fusion Module, be mapped to hyperspace and by after the sorter classification authentication result to the end; Perhaps identification authenticates fusion, the final recognition result after obtaining authenticating after merging once more.Through after merging, no matter verify or discern that total error rate has all obtained reduction than single creature feature identification Verification System.These two patents all are from method research angle, adopt various collecting devices to gather all biological informations, do not consider in the practical application applicability and to user's invasive, and main adopt remain traditional mode identification method, lay particular emphasis on the fusion of method.Publication number is that the Chinese patent of CN1464478A discloses the non-hypersphere geometrical body covering method in a kind of pattern-recognition, comprise the steps: (1) initialization sample space, sample space is divided into known sample subspace and unknown sample subspace two big classes; (2) beginning is at the training of certain type of sample; (3) construct mutual relationship between the sample of the same type according to rule, construct this sample subspace; (4) adopt non-hyperspherical geometrical body that every type of sample subspace is covered; (5) form the sample subspace of sealing.This invention proposes the best covering thought that similar sample distributes in feature space, solved effectively in the traditional mode identification and divided theoretical existing problem, but the feature space choice criteria in this method still awaits further research with the choosing method that covers with geometrical body.
Summary of the invention
The object of the invention provides a kind of based on bionical and examiner identity appraising system living things feature recognition, improvement by method, realization is identified fast, accurately to examiner's identity, both to be applicable to the examiner's identity authentication under the existing paper examination pattern, the machine in the future that is applicable to is again examined the identity authentication under the pattern.
For achieving the above object, the technical solution used in the present invention is: a kind of based on bionical and examiner identity appraising system living things feature recognition, comprise and utilize collector to gather multiple biological characteristic, the data of gathering are handled, authenticated evaluation according to training sample set, specifically may further comprise the steps:
(1) gather examiner's multinomial biological characteristic simultaneously, the data that collect are transferred to communication front-end equipment after by hardware encipher, described biological characteristic comprises on-line signature at least, fingerprint pattern and facial image hold a pen;
(2) collect the pre-service of data, comprise data are carried out filtering and regularization;
(3) from pretreated data, carry out feature extraction and characteristics combination, pass through the data soft encryption again, be transmitted through the network to back-end processing equipment then;
(4) back-end analysis of data is handled: the packet in that the back-end processing device decrypts receives from network, adopt the identification that realizes examiner's identity based on the living things feature recognition method of higher dimensional space geometrical body self-adaptation covering theory;
(5) sample set dynamically updates: the data that checking is passed through are added in the sample set, dynamically generate template corresponding, make new template reflect the trend of sample changed.
In the technique scheme, the acquisition method of described step (1) is, at first gather examiner's on-line signature by database, utilize people's metastable characteristics of attitude when signature to adopt camera to obtain people's face and signature posturography picture comparatively clearly, the fingerprint image that holds a pen of the signature pen collection that is provided with Multifunction Sensor by front end simultaneously, the described fingerprint image that holds a pen behave finger when holding a pen with between whole contact informations, comprise the fingerprint texture segment and the relative position of all contact positions.
Wherein, gather the pressure characteristic that holds a pen simultaneously at signature, its method is, in each contact position of described signature pen pressure transducer is set, gather hold a pen pressure in signature process variation and carry out record corresponding to the respective regions of the fingerprint image that holds a pen.
The synchronous acquisition mode of above-mentioned steps (1) can be called the mode that " one-stop " gathers biological characteristic, it has realized collecting static state as much as possible (people's face, fingerprint) and dynamic (signature) biological information in examiner's once signed writing process, have Noninvasive and accessibility preferably, especially it is everyone peculiar biological characteristic that the pressure that holds a pen in fingerprint image and when signature of holding a pen changes, have monopoly and exclusiveness, and can improve the accuracy rate of identification significantly after other biological characteristic combines.
Because the singularity of biological information, it is particularly outstanding that the data security problem seems.Native system at first carries out Data Format Transform with the data that collect by encryption chip, and then it is transferred to communication front-end equipment, has eliminated the potential safety hazard of data in transmission course.
Step in the technique scheme (2) is carried out pre-service to the data that collect, its fundamental purpose is the irrelevant information in the removal of images, recover useful real information, strengthen for information about detectability and reduced data to greatest extent, thus raising subsequent treatment result's reliability.In native system, mainly used following preprocess method: normalization, level and smooth, restore and strengthen.
Step in the technique scheme (3) comprises feature extraction, the detection of people's face and the feature extraction of on-line signature and the feature extraction of the fingerprint that holds a pen to carrying out feature extraction through pretreated data; Through after the feature extraction phases, how from numerous and complicated various feature, to filter out useful feature, and it is organically blended to form corresponding proper vector be the difficult point of area of pattern recognition always.The technical program is from bionical angle, adopt the thought of multiscale analysis to realize organically blending between the linked character collection, at first the feature after extracting being carried out yardstick divides, it is divided into Global Information and detailed information, and then further divide according to the relevance between feature, the feature that relevance is strong is divided into same linked character collection, carries out Feature Fusion at last, and linked character is synthesized proper vector.
After facial image, on-line signature and the fingerprint image that holds a pen had been carried out feature extraction, communication front-end equipment need be analyzed identification by Network Transmission to back-end processing equipment with merging good proper vector.Proper vector comprises examiner's biological information, and is the significant data source that back-end analysis identification and training sample set upgrade, and its importance is self-evident.The present invention adopts cryptographic algorithm, and a kind of multistep cryptographic algorithm is encrypted data, changes the data layout of proper vector, and the form with packet arrives back-end processing equipment with data by Network Transmission then, has protected examiner's personal information security effectively.
Behind the packet that the back-end processing device decrypts receives from network it is authenticated evaluation, the present invention adopts diverse ways to carry out matching degree for global characteristics vector sum minutia vector and calculates.For the vectorial point that directly its mapping is become in the higher dimensional space of global characteristics, according to the continuity of similar sample point in higher dimensional space, training sample set should be covered by limited, the continuous zone of a slice, treats that by analysis the relation of observation station and training sample set overlay area can obtain the matching degree between this biometric sample and training sample set; Then adopt the method for dynamic programming to carry out template matches for minutia.Obtain a kind of matching degree of biological characteristic after then both result being multiplied each other, the decision making algorithm of employing fusion is at last handled the matching degree between multiple biometric sample and training sample set, obtain the whole matching degree after the matching degree weighted sum with multiple biological characteristic, the identification conclusion can be drawn with threshold ratio after, evaluation can be realized examiner's identity.
The priori that " continuity of similar sample point in higher dimensional space " distributes as sample point, think that certain biological characteristic (signature, fingerprint, the people's face) sample that comes from same individual should be continuous in higher dimensional space, be to undergo mutation between the sample, all there is a continually varying curve between any two sample points, makes one of them sample point carry out the transition to the another one sample point smoothly.Principle that Here it is " homology continuity " that is: is established among the feature space Rn all and is belonged to all of category-A things and be set A, if having any two element x and y in the set A, then for arbitrarily greater than 0 value ε, must have set B, make:
Corresponding to learning process, be exactly the distribution of training sample in feature space at things of the like description, select one or more suitable occluding surfaces, the continuous complicated geometirc physique that forms a higher dimensional space rationally covers training sample.And this complicated geometirc physique need carry out self-adaptation according to the distribution of training sample and chooses, and covers training sample set with reasonable manner.
When the user carries out after authentication passes through, new biometric sample information also automatically by system record.When limited storage space, training sample set must dynamically be deleted some stale informations when adding fresh information, and it is significant allowing training sample set embody its variation tendency in the mode that dynamically updates.
The update mode of training sample set can be by geometrical body in the higher dimensional space the selection of overlay area consider.The present invention has realized that the self-adaptation of overlay area dynamically generates, think that promptly training sample set after dynamically updating should realize the most reasonable dynamic covering to all samples in higher dimensional space, to verify that newly the biological characteristic that passes through is updated to training sample set, simultaneously with stale information filtering from the overlay area.
Because the technique scheme utilization, the present invention compared with prior art has following advantage:
1. on-line signature, people's face, signature posture have been gathered when gathering examiner's biological characteristic, also adopt the signature pen that has Multifunction Sensor, pressure variation when having gathered finger and the contacted whole fingerprints of pen and corresponding holding a pen, abundant biological information has monopoly and exclusiveness, in conjunction with after can improve the accuracy rate of identification significantly.
Data when collecting device is transferred to communication front-end equipment through hardware encryption---encryption chip is encrypted, when communication front-end equipment is transferred to back-end processing equipment through soft encryption---the multistep cryptographic algorithm is encrypted, and has guaranteed the safety of data.
3. total system has incorporated bionical thought, and the mode in the cognitive world of imitation people is carried out pattern-recognition, and its core concept is embodied in following three aspects:
(1) people's thinking can be divided into two kinds of logical thinking and thinkings in images, and it is based on the reasoning from logic process development by deduction that brain carries out logical thinking, carries out thinking in images and then be developing based on the summary of experience process of method of induction.Pattern-recognition is a kind of thinking in images problem more, and the variable quantity that it relates to is often a lot, and it is carried out strict reasoning from logic, and it is unpractical adopting the multi-variable function Modeling Calculation.Therefore, solve the describing mode that pattern recognition problem must find a kind of thinking in images, and the figure notion, geometrical body was described and just is a kind of effective describing mode of thinking in images during promptly the higher-dimension sky was asked.
(2) people has the dimensional variation notion when discerning.With signature, people's face and fingerprint is example, and the people not only observes its overall profile when identification, and pays close attention to its minutia.Therefore, when extracting signature, people's face and fingerprint characteristic, not only to consider global informations such as length breadth ratio, form, and will pay close attention to detailed information such as stroke intersection, canthus lines, identifying object be estimated exactly from different yardsticks.
(3) people brings in constant renewal in as the data base of identification.After the people has discerned certain object according to priori, the existing feature information of refining into of this object can be stored memory, so that accurate identification in the future.Therefore, sample to be tested and existing training sample set are compared and determine that it is true after, the new data that checking is passed through need be added in this sample set automatically, and again it is assessed, according to the distribution situation of each sample in higher dimensional space, reasonably remove outmoded sample, make new training sample set reflect the variation tendency of identifying object.
4. system of the present invention is from bionical thought, integrated use realizes evaluation to examiner's identity based on the various living things feature recognition methods of higher dimensional space geometrical body self-adaptation covering theory, identify quick, the result is accurate, be not only applicable to the examiner's identity authentication under the existing paper examination pattern, and examine at machine in the future and to have more wide application prospect under the pattern.
Description of drawings
Fig. 1 is the system flow synoptic diagram of embodiment one;
Fig. 2 is the fingerprint collecting synoptic diagram that holds a pen of embodiment one;
Fig. 3 is the on-line signature parameter synoptic diagram of embodiment one;
Fig. 4 is the homology continuity synoptic diagram of embodiment one.
Embodiment
Below in conjunction with drawings and Examples the present invention is further described:
Embodiment one:, a kind of referring to accompanying drawing 1 based on bionical and examiner identity appraising system living things feature recognition to shown in the accompanying drawing 4, elaborate by following steps:
The first step, the collection of on-line signature, hold a pen fingerprint pattern and facial image data.
Adopt the mode shown in the accompanying drawing 1, utilize database collection examiner's on-line signature, utilize the camera collection facial image simultaneously and the fingerprint image that holds a pen, and the Multifunction Sensor collection by the dedicated signatures anterior end of pen hold a pen fingerprint image and the pressure information that holds a pen when writing, this sensor not only can be gathered the hold a pen fingerprint image of people when signing, and the pressure that holds a pen in can the perception signature process changes, the described fingerprint image that holds a pen is different from the single piece of complete finger print image that the conventional fingerprint collecting device collects, it write down when the people holds a pen finger with between whole contact informations, comprise the fingerprint texture segment and the relative position of all contact positions; Simultaneously, the variation of the pressure of each contact position in signature process is also by sensor acquisition and corresponding to the respective regions of the fingerprint image that holds a pen.
Subsequently, native system carries out Data Format Transform with the data that collect by special encryption chip, again it is transferred to communication front-end equipment, has guaranteed the data security of user biological characteristic information effectively.
In second step, the data that collect are carried out pre-service.
Pretreated fundamental purpose is the irrelevant information in the removal of images, recovers useful real information, strengthens for information about detectability and reduced data to greatest extent, thus raising subsequent treatment result's reliability.In native system, mainly used following preprocess method:
(1) normalization
Make some feature of image under given conversion, have a kind of graphics standard form of invariance.Some character of image, for example the area of object and girth just had constant character originally for rotation of coordinate.In the ordinary course of things, the influence of some factor or some character of transfer pair image can be eliminated or weakens by normalized, thereby can be selected as the foundation of measurement image.Gray scale normalization, geometrical normalization and transform normalization are three kinds of method for normalizing that obtain the image invariance.
(2) level and smooth
The technology of random noise in the removal of images.To the basic demand of smoothing technique is that in the cancellation noise image outline or lines to be thickened unclear.Smoothing method commonly used has median method, part to ask the method for average and the k neighbour method of average.The regional area size can be fixed, and also can be that pointwise is with the gray-scale value size variation.In addition, application space frequency field band-pass filtering method sometimes.
(3) restore
Proofread and correct the image degradation that a variety of causes caused, make reconstruction or estimate that the image that obtains approaches the desirable image field of degenerating that do not have as much as possible.Usually the image degradation phenomenon can take place in actual applications, the aberration of optical system for example, the relative motion of camera and object all can make facial image degenerate.Basic recovery technique is that (x, (x is y) with ideal image f (x, convolution y) y) to regard degenrate function h as the degraded image g that obtains.Their Fourier transform exist concern G (u, v)=H (u, v) F (u, v).After determining degenrate function according to degradation mechanism, just from then on relational expression obtain F (u, v), again with Fourier inversion obtain f (x, y).Usually M (u, v)=(u v) is called inverse filter to 1/H.During practical application, (u v) increases with the distance of leaving uv plane initial point and descends rapidly, and u is worked as in the reinforcement of noise in the high-frequency range because H
2+ v
2During greater than a certain boundary value W0, (u v) equals 1 to make M.The selection of W0 should make H, and (u is v) at u
2+ v
2In≤W0 the scope zero point can not appear.The algebraic method of image restoration is based on the least square method optimum criterion.Seek an estimated value, make goodness criterion function value minimum.This method is fairly simple, can derive the least square method S filter.When not having noise, S filter becomes desirable inverse filter.
(4) strengthen
Information in the image is strengthened selectively and suppressed, improving the visual effect of image, or change image into be more suitable for form, so that data pick-up or identification in machine processing.For example an Image Intensified System can be given prominence to the outline line of image by Hi-pass filter, thereby makes shape and girth that machine can the measuring wheel profile.Image enhancement technique has several different methods, and contrast broadening, log-transformation, density stratification and histogram equalization etc. all can be used for changing image tone and outstanding details.Often to use diverse ways during practical application, test repeatedly just and can reach satisfied effect
The 3rd step, pretreated data are carried out feature extraction, and be combined into proper vector, behind the data soft encryption, arrive back-end processing equipment then by Network Transmission.
(1) feature extraction of on-line signature
Comprised abundant characteristic information in the on-line signature, on time-spatial information (si) (as accompanying drawing 3) and transform domain information division can be with the following table sorts of various features.
In addition, some global characteristics in the on-line signature also merit attention, and pen is gone up line time and accounted for ratio of T.T. or the like in the length breadth ratio of the deadline of for example signing, signature image, the signature process.
For time dependent features such as speed, inclination angle, pressure, the present invention at first is merged into it proper vector, and then is arranged in the proper vector chain, adopts the method for dynamic programming to carry out template matches; And for global characteristics, the present invention is arranged in the observation station for the treatment of that is mapped to after the proper vector in the higher dimensional space with it, utilizes the relation for the treatment of in the higher dimensional space between observation station and the sample overlay area to determine its matching degree.
(2) people's face detects and feature extraction
The present invention is based on complexion model, directly adopt YCbCr color space and passing threshold that area of skin color is adjudicated, to each point in the image (x, y), we with f (x, y) represent that whether this point belongs to skin pixel, obtains following formula:
Adopt particle swarm optimization algorithm that area of skin color is cut apart then.Particle swarm optimization algorithm is based on the evolution algorithmic of colony, and its thought source is in artificial life and EVOLUTIONARY COMPUTATION theory.During PSO solving-optimizing problem, problem separate corresponding to one " particle " in the search volume (particle) or " main body " (agent).Each particle all has position and the speed (direction of decision flight and distance) of oneself, also has an adaptive value by optimised function decision.Each particle is remembered, is followed current optimal particle, searches in solution space.The process of each iteration is not a completely random, if find better solutions, will seek the next one on this basis and separate.
Make PSO be initialized as a group random particles (RANDOM SOLUTION), in iteration each time, particle upgrades oneself by following the tracks of two " extreme values ": first is exactly preferably separating of being found of particle itself, be called extreme point (representing its position) with pbest, another extreme point among the PSO is preferably separating of finding at present of whole population, is called global extremum point (representing its position with gbest).
After finding two optimal values, particle upgrades speed and the position of oneself by following formula.
V
i＝ω×V
i+c
1×rand()×(pbest
i-x
i)+c
2×rand()×(gbest
i-x
i) (2)
x
i＝x
i+V
i (3)
In above-mentioned two formulas:
I=1,2 ..., M, M are the sums of particle in this colony;
V
iIt is particle's velocity;
Rand () is the random number between (0,1);
x
iIt is the current location of particle;
c
1And c
2It is the study factor;
ω is non-negative, is called inertial factor.
In each dimension, particle all has a maximum constraints speed V
MaxIf the speed of certain one dimension surpasses the V that sets
Max, the speed of this one dimension just is restricted to V so
Max(V
Max＞0).
(2) in the formula, the ω value is bigger, and then global optimizing ability is strong, a little less than the local optimal searching ability; Otherwise the ω value is less.When initial, ω is taken as constant, and experiment afterwards finds that dynamically ω can obtain than the better optimizing result of fixed value.Dynamically ω can change at PSO search procedure neutral line, also can dynamically change according to certain measure function of PSO performance.What the present invention adopted is linear decrease weights strategies.
ω
(t)＝(ω
ini-ω
end)(G
k-g)/G
k+ω
end (4)
G
kBe maximum evolutionary generation, ω
IniBe initial inertia weights, ω
EndBe iteration inertia weights when the maximum algebraically.
After area of skin color was determined, the zone of being cut apart might comprise some non-face zones, therefore needed to adopt fuzzy membership function that people's face is differentiated.Fuzzy membership functions is defined as follows:
Native system is explained people's face with 3 features.Be respectively the height of people's face, ratio and people's face vertical offset angle of height and the width, adopt fuzzy membership functions these parameter fuzzy to be divided the fuzzy membership functions μ that does not obtain three features
1(x), μ
2(x) and μ
3(x), whether be total subordinate function of people's face with its product as this zone of differentiation, that is:
μ(x)＝μ
1(x)·μ
2(x)·μ
3(x) (6)
If the membership function value in a certain zone, thinks then that this zone is exactly a human face region greater than a certain threshold value, use for the single face of native system, then the zone of membership function value maximum is a human face region.
Need to extract face characteristic after detecting human face region, face characteristic mainly comprises the variation of position, size and the shape of eyes, eyebrow, face.At first utilize principal component analysis to combine with the human face ratio characteristic, after obtaining the preliminary region of eyes, use technology such as rim detection, can determine the accurate position of pupil rapidly and accurately, utilize the color characteristic of the ratio feature of people's face and eyebrow, face then, can extract eyebrow and face very accurately.The identification of people's face will be can be used for after remarkable characteristic in these facial characteristics or the attributes extraction.
(3) the hold a pen feature extraction of fingerprint
As shown in Figure 2, the fingerprint that holds a pen that is collected among the present invention is different from single piece of traditional complete finger print, the finger that is the user when signature of its record with between whole contact informations, comprise the fingerprint texture segment and the relative position of all contact positions.Therefore, classic method and new method have been merged in its characteristic extraction procedure.
Fingerprint generally has two kinds of structures: Henry fingerprint pattern and Galton feature.The Henry classification is that standard designs qualitatively, characterizes the one-piece construction of crest pattern, is generally used for dividing finger print data.The Galton General Definition four characteristics: the starting point and the terminal point of (1) crest; (2) branch; (3) isolated area; (4) disturb.In order from the digital finger-print image, to extract feature, at first selected to comprise eight features of four Galton characteristics, and it has been divided into two classes: original and synthetic.Primitive character is respectively a little, crest terminal point and branch.On this basis, defined composite character again: isolated zone, burr intersects, bridge and short peak.In addition, fingerprint image can also extract following feature for holding a pen: fingerprint number of slices, relative position etc. between each segment.
(4) proper vector is synthetic
Through after the feature extraction phases, how from numerous and complicated various feature, to filter out useful feature, and it is organically blended to form corresponding proper vector be the difficult point of area of pattern recognition always.The present invention adopts the thought of multiscale analysis to realize organically blending between the linked character collection from bionical angle.
At first, each feature is carried out yardstick divide, it is divided into Global Information and detailed information.And then further divide according to the relevance between feature, pressure information, the writing speed of signature and the variation at stroke direction angle of writing down on the dynamics variation of the center of effort that holds a pen during such as signature, the numerical digit plate have stronger correlativity, can be divided into same linked character collection.Carry out Feature Fusion at last, linked character is synthesized proper vector.
Discussion with the front is an example, makes d
i(i is a center of effort quantity) expression dynamics that holds a pen, p represents that the pressure that writes down on the numerical digit plate, v represents the writing speed of signing, θ represents deflection, then can obtain a linked character vector φ (d
i, p, v, θ).
(5) data soft encryption and Network Transmission
Because the singularity of biological information, it is particularly outstanding that the data security problem seems.Consider the information security issue in the Network Transmission, native system has adopted a kind of multistep cryptographic algorithm that the transmission data are encrypted.
This algorithm uses a series of numeral (such as 128 keys), produces one repeatably but the pseudorandom Serial No. of height randomization.Once use 256 list items, use random number sequence to produce password commentaries on classics table, promptly 256 random numbers are placed in the matrix, then it is sorted, produce an arbitrarily table of ordering according to initial position, the numeral in the table is between 0 to 255.Produced the table of 256 concrete bytes like this.Allow this tandom number generator follow to produce remaining number in this table, to such an extent as to each table is different.Next, use " Shot Gun Technique " technology to produce decoding table.Basically, if a is mapped to b, b necessarily can be mapped to a so, so b[a[n]]=n (n is a number between 0 to 255).Assignment in a circulation makes the black list of the decoding table of one 256 byte corresponding to 256 bytes that produce previously.
By this method, can produce a such table, the order of table is at random, so what produce that the random number of these 256 bytes uses is the secondary pseudorandom, has used two 16 extra passwords.Now, two conversion tables have been arranged, basic encrypting and decrypting process is as described below: previous byte ciphertext is the index of the table of this 256 byte, perhaps, in order to improve cipher round results, can use unnecessary 8 value, even use verification and or the crc algorithm produce the index byte.Suppose that this table is 256 * 256 array, it satisfies so:
encrypt1＝a[encrypt0][value]
Variable ' encrypt1 ' is a data encrypted, and ' encrypt0 ' is previous enciphered data (or functional value of the several enciphered datas in front).Very natural, first data need one " seed ", and this " seed " must be remembered.
The pseudo-random sequence that is produced during encryption is very random, any sequence that can be designed to want.About the detailed information of this random series, decrypting ciphertext is not unpractical.For example: the sequence of some ASCII character, may be converted to some mess codes at random as " processing " without any meaning, each byte all depends on the ciphertext of its previous byte, rather than actual value.For this conversion of any single character, hidden the effective real length of enciphered data.
In the 4th step, adopt based on the living things feature recognition method of bionical and higher dimensional space geometrical body self-adaptation covering theory and identify examiner's identity.
The present invention adopts diverse ways to carry out matching degree for global characteristics vector sum minutia vector and calculates.Directly its mapping is become point in the higher dimensional space for global characteristics vector, calculate the matching degree between itself and training sample set; Then adopt the method for dynamic programming to carry out template matches for minutia.Obtain a kind of matching degree of biological characteristic after then both result being multiplied each other, obtain the whole matching degree after the matching degree weighted sum with multiple biological characteristic at last, can draw the identification conclusion after with threshold ratio.
Be identified as example with on-line signature, at first the vector of each linked character in the minutia adopted the method for dynamic programming to solve the matching degree of itself and training sample set, the matching degree weighted sum that obtains is obtained the minutia matching degree.Then the global characteristics DUAL PROBLEMS OF VECTOR MAPPING is become a point in the higher dimensional space, obtain corresponding global characteristics matching degree according to the relation of itself and training sample set overlay area.At last details and global characteristics matching degree are multiplied each other and obtain final on-line signature whole matching degree.At last with after the matching degree weighted sum of on-line signature, people's face, the fingerprint that holds a pen with threshold ratio, finally realize the checking of examiner's identity.
In the 5th step, dynamically update training sample set.
When the user carries out after authentication passes through, new biometric sample information also automatically by system record.When limited storage space, training sample set must dynamically be deleted some stale informations when adding fresh information.This is reasonably in actual applications, so people's signature custom may be along with the time changes, and it is significant allowing training sample set embody its variation tendency in the mode that dynamically updates.
The update mode of training sample set can be by geometrical body in the higher dimensional space the selection of overlay area consider.The present invention has realized that the self-adaptation of overlay area dynamically generates, think that promptly training sample set after dynamically updating should realize the most reasonable dynamic covering to all samples in higher dimensional space, to verify that newly the biological characteristic that passes through is updated to training sample set, simultaneously with stale information filtering from the overlay area.
Claims (5)
1. one kind based on bionical and examiner's identity identifying method living things feature recognition, comprises and utilizes collector to gather multiple biological characteristic, the data of gathering are handled, authenticated evaluation according to training sample set, it is characterized in that specifically may further comprise the steps:
(1) gather examiner's multinomial biological characteristic simultaneously, the data that collect are transferred to communication front-end equipment after by hardware encipher, described biological characteristic comprises on-line signature at least, fingerprint pattern and facial image hold a pen;
(2) collect the pre-service of data, comprise data are carried out filtering and regularization;
(3) from pretreated data, carry out feature extraction and characteristics combination, pass through the data soft encryption again, be transmitted through the network to back-end processing equipment then, wherein, during to the biological characteristic combination after extracting, at first the feature after extracting being carried out yardstick divides, it is divided into Global Information and detailed information, and then further divides according to the relevance between feature, the feature that relevance is strong is divided into same linked character collection, carry out Feature Fusion at last, linked character is synthesized proper vector;
(4) back-end analysis of data is handled: the packet that receives from network in the back-end processing device decrypts, employing realizes the identification of examiner's identity based on the living things feature recognition method of higher dimensional space geometrical body self-adaptation covering theory, concrete recognition methods is, directly its mapping is become point in the higher dimensional space for global characteristics, calculate the matching degree between itself and the training sample set; Adopt the method for dynamic programming to carry out template matches for minutia, obtain a kind of matching degree of biological characteristic after then both result being multiplied each other, obtain the whole matching degree after the matching degree weighted sum with multiple biological characteristic at last, draw the identification conclusion after with threshold ratio;
(5) sample set dynamically updates: the data that checking is passed through are added in the sample set, dynamically generate template corresponding, make new template reflect the trend of sample changed.
2. according to claim 1 based on bionical and examiner's identity identifying method living things feature recognition, it is characterized in that: the acquisition method of described step (1) is, at first gather examiner's on-line signature by database, utilize people's metastable characteristics of attitude when signature to adopt camera to obtain people's face and signature posturography picture comparatively clearly, the fingerprint image that holds a pen of the signature pen collection that is provided with Multifunction Sensor by front end simultaneously, the described fingerprint image that holds a pen behave finger when holding a pen with between whole contact informations, comprise the fingerprint texture segment and the relative position of all contact positions.
3. according to claim 2 based on bionical and examiner's identity identifying method living things feature recognition, it is characterized in that: gather the pressure characteristic that holds a pen simultaneously at signature, its method is, each contact position at described signature pen is provided with pressure transducer, gather hold a pen pressure in signature process variation and carry out record corresponding to the respective regions of the fingerprint image that holds a pen.
4. according to claim 1 based on bionical and examiner's identity identifying method living things feature recognition, it is characterized in that in the step (3) the The data multistep cryptographic algorithm after the characteristics combination being carried out soft encryption.
5. according to claim 1 based on bionical and examiner's identity identifying method living things feature recognition, the update mode that it is characterized in that training sample set in the step (5) is: the training sample set after dynamically updating should be realized the most reasonable dynamic covering to all samples in higher dimensional space, the self-adaptation that is the overlay area dynamically generates, to verify that newly the biological characteristic that passes through is updated to training sample set, simultaneously with stale information filtering from the overlay area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008100198360A CN101246543B (en) | 2008-03-18 | 2008-03-18 | Examiner identity identification method based on bionic and biological characteristic recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008100198360A CN101246543B (en) | 2008-03-18 | 2008-03-18 | Examiner identity identification method based on bionic and biological characteristic recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101246543A CN101246543A (en) | 2008-08-20 |
| CN101246543B true CN101246543B (en) | 2010-06-02 |
Family
ID=39946983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008100198360A Expired - Fee Related CN101246543B (en) | 2008-03-18 | 2008-03-18 | Examiner identity identification method based on bionic and biological characteristic recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101246543B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103178961A (en) * | 2013-02-04 | 2013-06-26 | 快车科技有限公司 | Safe information interaction method and related device |
| CN113011318A (en) * | 2021-03-16 | 2021-06-22 | 南京新知艺测科技有限公司 | Identity recognition system in on-computer evaluation |
Families Citing this family (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8635211B2 (en) | 2009-06-11 | 2014-01-21 | Dolby Laboratories Licensing Corporation | Trend analysis in content identification based on fingerprinting |
| CN102236793A (en) * | 2010-04-29 | 2011-11-09 | 比亚迪股份有限公司 | Method for rapidly detecting skin color |
| WO2012102110A1 (en) * | 2011-01-27 | 2012-08-02 | 株式会社エヌ・ティ・ティ・ドコモ | Mobile information terminal, grip characteristic learning method, and grip characteristic authentication method |
| WO2012114881A1 (en) * | 2011-02-21 | 2012-08-30 | 株式会社エヌ・ティ・ティ・ドコモ | Gripping characteristics learning authentication system and gripping characteristics learning authentication method |
| CN102739944A (en) * | 2012-05-10 | 2012-10-17 | 蒋耘晨 | Interactive type expandable camera |
| CN102708716B (en) * | 2012-05-16 | 2015-03-04 | 深圳市海云天科技股份有限公司 | Answering device and examination data processing method thereof |
| CN102722284B (en) * | 2012-06-15 | 2013-07-03 | 重庆傲雄信息技术有限公司 | Touch-panel-based handwritten signing recording method |
| CN102970510B (en) * | 2012-11-23 | 2015-04-15 | 清华大学 | Method for transmitting human face video |
| CN103235957B (en) * | 2013-04-18 | 2016-05-11 | 武汉汉德瑞庭科技有限公司 | A kind of online handwriting authentication method and system based on palmar side surface information |
| CN103236044B (en) * | 2013-05-02 | 2016-01-20 | 北京理工大学 | Based on the micro scanning image reconstructing method of Taylor expansion integration degradation model |
| SG11201600050SA (en) * | 2013-07-05 | 2016-02-26 | Dbs Bank Ltd | System and method for providing bank transactions with a remote bank device |
| CN104463074B (en) * | 2013-09-12 | 2017-10-27 | 金佶科技股份有限公司 | The discrimination method and device for identifying of true and false fingerprint |
| CN104579948A (en) * | 2013-10-29 | 2015-04-29 | 国家计算机网络与信息安全管理中心 | Method and device for fragmenting message |
| CN103927351B (en) * | 2014-04-10 | 2017-06-06 | 清华大学 | Fingerprint retrieval method and system based on attitude updating |
| CN104102916B (en) * | 2014-07-03 | 2017-12-01 | 宁波摩米创新工场电子科技有限公司 | Face Image Recognition System based on man face geometric size normalization |
| CN104794665A (en) * | 2015-04-07 | 2015-07-22 | 宋阳 | Examination and evaluation system and control method thereof |
| CN105303632B (en) * | 2015-10-22 | 2018-04-03 | 苏州工业园区服务外包职业学院 | A kind of mobile monitor is registered system and method for work |
| KR102466995B1 (en) * | 2015-12-21 | 2022-11-14 | 삼성전자주식회사 | Method and device for authenticating user |
| CN105740683B (en) * | 2016-01-20 | 2018-10-12 | 北京信安盟科技有限公司 | Based on multifactor, multi engine, the man-machine auth method being combined and system |
| CN105631272B (en) * | 2016-02-02 | 2018-05-11 | 云南大学 | A kind of identity identifying method of multiple security |
| CN107045618B (en) * | 2016-02-05 | 2020-07-03 | 北京陌上花科技有限公司 | Facial expression recognition method and device |
| CN106203029A (en) * | 2016-06-28 | 2016-12-07 | 联想(北京)有限公司 | A kind of verification method and electronic equipment |
| CN106203329B (en) * | 2016-07-08 | 2019-02-22 | 安徽工业大学 | A method of identity template is established based on eyebrow and carries out identification |
| CN107169474A (en) * | 2017-06-16 | 2017-09-15 | 郑州云海信息技术有限公司 | A kind of crime method for early warning based on intelligent Computation Technology |
| CN107346381A (en) * | 2017-06-30 | 2017-11-14 | 方亚南 | A kind of pen and method for identifying writing people's identity |
| CN107657241B (en) * | 2017-10-09 | 2021-03-26 | 河海大学常州校区 | Signature pen-oriented signature authenticity identification system |
| CN107909687A (en) * | 2017-11-07 | 2018-04-13 | 惠州市瑞时智控科技有限公司 | A kind of Intelligent door control system and its control method based on machine learning |
| CN107657822A (en) * | 2017-11-17 | 2018-02-02 | 嘉兴四维智城信息科技有限公司 | Highway moisture film early warning system and method for traffic video automatic identification |
| CN109171670B (en) * | 2018-06-25 | 2021-02-05 | 天津海仁医疗技术有限公司 | 3D blood vessel imaging algorithm based on reverse principal component analysis method |
| CN109492560A (en) * | 2018-10-26 | 2019-03-19 | 深圳力维智联技术有限公司 | Facial image Feature fusion, device and storage medium based on time scale |
| CN109816567A (en) * | 2018-11-08 | 2019-05-28 | 深圳壹账通智能科技有限公司 | A kind of online testing method, apparatus, equipment and storage medium |
| CN109376723A (en) * | 2018-12-19 | 2019-02-22 | 赵太方 | A kind of portable identification Verification System based on face recognition technology |
| CN109784083B (en) * | 2019-02-22 | 2022-09-02 | 吉林大学 | Bionic encryption system based on fusion of grip strength information and dorsal hand vein information |
| CN110599639B (en) * | 2019-08-13 | 2021-05-07 | 深圳市天彦通信股份有限公司 | Identity verification method and related product |
| CN110855720A (en) * | 2019-12-26 | 2020-02-28 | 北京易通云行科技有限公司 | Self-service signature machine of examining drives with face recognition function |
| CN112788026A (en) * | 2020-08-20 | 2021-05-11 | 王红根 | Information management method and management system based on mobile internet and biological authentication |
| CN112014804B (en) * | 2020-10-30 | 2021-01-26 | 西南交通大学 | Radar signal sorting method based on bionic pattern recognition algorithm of ball covering |
| CN112983406B (en) * | 2021-03-15 | 2022-03-25 | 西南石油大学 | Natural gas hydrate reservoir parameter index evaluation method |
| CN114550316B (en) * | 2022-04-27 | 2022-08-05 | 广州商景网络科技有限公司 | One-stop credible biological characteristic data acquisition terminal equipment and acquisition and sharing method |
| CN115619410B (en) * | 2022-10-19 | 2024-01-26 | 闫雪 | Self-adaptive financial payment platform |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1304114A (en) * | 1999-12-13 | 2001-07-18 | 中国科学院自动化研究所 | Identity identification method based on multiple biological characteristics |
| CN1464478A (en) * | 2002-06-20 | 2003-12-31 | 中国科学院半导体研究所 | A non-hypersphere geometrical body covering method for pattern recognition |
| CN1974266A (en) * | 2006-12-08 | 2007-06-06 | 青特集团有限公司 | Revolving mechanism with front mounted cylinder for damper |
-
2008
- 2008-03-18 CN CN2008100198360A patent/CN101246543B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1304114A (en) * | 1999-12-13 | 2001-07-18 | 中国科学院自动化研究所 | Identity identification method based on multiple biological characteristics |
| CN1464478A (en) * | 2002-06-20 | 2003-12-31 | 中国科学院半导体研究所 | A non-hypersphere geometrical body covering method for pattern recognition |
| CN1974266A (en) * | 2006-12-08 | 2007-06-06 | 青特集团有限公司 | Revolving mechanism with front mounted cylinder for damper |
Non-Patent Citations (6)
| Title |
|---|
| 方志刚,叶伟中.多生物特征识别技术综述.计算机工程29 9.2003,29(9),140-142. |
| 方志刚,叶伟中.多生物特征识别技术综述.计算机工程29 9.2003,29(9),140-142. * |
| 王守觉.仿生模式识别(拓扑模式识别)－－一种模式识别新模型的理论与应用.电子学报30 10.2002,30(10),1417-1420. |
| 王守觉.仿生模式识别(拓扑模式识别)——一种模式识别新模型的理论与应用.电子学报30 10.2002,30(10),1417-1420. * |
| 金涌 等.改善手写签名动态特征稳定性的局部相关分析.华中理工大学学报26 12.1998,26(12),71-73. |
| 金涌等.改善手写签名动态特征稳定性的局部相关分析.华中理工大学学报26 12.1998,26(12),71-73. * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103178961A (en) * | 2013-02-04 | 2013-06-26 | 快车科技有限公司 | Safe information interaction method and related device |
| CN103178961B (en) * | 2013-02-04 | 2017-05-17 | 快车科技有限公司 | Safe information interaction method and related device |
| CN113011318A (en) * | 2021-03-16 | 2021-06-22 | 南京新知艺测科技有限公司 | Identity recognition system in on-computer evaluation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101246543A (en) | 2008-08-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101246543B (en) | Examiner identity identification method based on bionic and biological characteristic recognition | |
| Valdes-Ramirez et al. | A Review of Fingerprint Feature Representations and Their Applications for Latent Fingerprint Identification: Trends and Evaluation. | |
| Yuan et al. | Fingerprint liveness detection using an improved CNN with image scale equalization | |
| Dai et al. | Multifeature-based high-resolution palmprint recognition | |
| CN100492400C (en) | Matching identification method by extracting characters of vein from finger | |
| CN102521575B (en) | Iris identification method based on multidirectional Gabor and Adaboost | |
| CN101604376B (en) | Method for identifying human faces based on HMM-SVM hybrid model | |
| CN105469076B (en) | Face alignment verification method based on multi-instance learning | |
| CN103336960A (en) | Human face identification method based on manifold learning | |
| CN105184254B (en) | A kind of identity identifying method and system | |
| CN107169479A (en) | Intelligent mobile equipment sensitive data means of defence based on fingerprint authentication | |
| CN102254165A (en) | Hand back vein identification method based on fusion of structural coding features and texture coding features | |
| CN101819629B (en) | Supervising tensor manifold learning-based palmprint identification system and method | |
| Maltoni et al. | Fingerprint recognition | |
| CN103942545A (en) | Method and device for identifying faces based on bidirectional compressed data space dimension reduction | |
| CN109359607A (en) | A kind of palm print and palm vein fusion identification method based on texture | |
| Kanjan et al. | A comparative study of fingerprint matching algorithms | |
| Jain et al. | A filterbank-based representation for classification and matching of fingerprints | |
| Chen | Extended feature set and touchless imaging for fingerprint matching | |
| Shariatmadar et al. | An efficient method for finger-knuckle-print recognition by using the information fusion at different levels | |
| Chinnappan et al. | Fingerprint recognition technology using deep learning: a review | |
| Jain | Automatic Fingerprint Matching Using Extended Feature Set | |
| Li et al. | The evolution of biometrics | |
| Shi et al. | A novel fingerprint matching algorithm based on minutiae and global statistical features | |
| Palanikumar et al. | Advanced palmprint recognition using unsharp masking and histogram equalization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| ASS | Succession or assignment of patent right |
Owner name: SUZHOU INSTITUTE OF NANO-TECH AND NANO-BIONICS (SI Free format text: FORMER OWNER: SUZHOU NANO TECHNIQUE + NANO BIONIC RESEARCH INST. Effective date: 20100928 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20100928 Address after: 215125 Suzhou Industrial Park, Jiangsu Province, if the waterway No. 398, No. Patentee after: Suzhou Institute of Nano-Tech and Bionics (SINANO), Chinese Academy of Sciences Address before: 215125, Suzhou, Jiangsu province Suzhou Industrial Park alone villa lake high Parish, if the waterway 398 Patentee before: Suzhou Nano Technique & Nano Bionic Research Inst. |
|
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100602 Termination date: 20140318 |