CN101414351A - Fingerprint recognition system and control method - Google Patents

Fingerprint recognition system and control method Download PDF

Info

Publication number
CN101414351A
CN101414351A CNA2008100464373A CN200810046437A CN101414351A CN 101414351 A CN101414351 A CN 101414351A CN A2008100464373 A CNA2008100464373 A CN A2008100464373A CN 200810046437 A CN200810046437 A CN 200810046437A CN 101414351 A CN101414351 A CN 101414351A
Authority
CN
China
Prior art keywords
fingerprint
module
client
prime
template
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.)
Pending
Application number
CNA2008100464373A
Other languages
Chinese (zh)
Inventor
章毅
纪禄平
蒲晓蓉
刘贵松
杨成福
Original Assignee
章毅
纪禄平
蒲晓蓉
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 章毅, 纪禄平, 蒲晓蓉 filed Critical 章毅
Priority to CNA2008100464373A priority Critical patent/CN101414351A/en
Publication of CN101414351A publication Critical patent/CN101414351A/en
Pending legal-status Critical Current

Links

Abstract

The invention discloses a fingerprint identifying system which comprises three large parts of a fingerprint database, a central server and a client terminal; the fingerprint database is mainly used for memorizing the fingerprint characteristic information after digital coding; the central server is mainly responsible for validating the validity of the client terminal, receiving the fingerprint characteristic data, comparing the characteristics and returning the fingerprint identifying result; the client terminal is mainly responsible for collecting the fingerprint, extracting and uploading the fingerprint characteristic data. The system overcomes the defects of the prior art, can provide an online/offline fingerprint identifying system the basic algorithm library of which can be seamlessly expanded and the functions of which can be selected and matched. The fingerprint identifying system can lead a user to obtain the ID identification and authentication services with high quality under the situation of not needing a great amount of capital investment.

Description

A kind of fingerprint recognition system and control method thereof
Technical field
The present invention relates to the biometrics identification technology field, be specifically related to fingerprint recognition system and control method thereof.
Background technology
Fingerprint recognition, everyone comprises that the skin lines of fingerprint has nothing in common with each other, and presents uniqueness and constant throughout one's life on pattern, breakpoint and point of crossing.Just can be mapped a people with his fingerprint in view of the above, by with his fingerprint and the finger print data of preserving in advance compare, just can verify its true identity, Here it is fingerprint identification technology.Fingerprint recognition is mainly carried out identity authentication according to information such as the lines of somatic fingerprint, minutias to the operation or the person of being operated, have benefited from integrated manufacturing technology of hyundai electronics and rapid and reliable algorithm research, begun to enter into our daily life, it is the most deep to become in the present biological detection research, most widely used general, develop proven technique.
The feature of the fingerprint that can be used for discerning is divided into two big class, i.e. general characteristic and local features.General characteristic is meant that mainly those directly just can observed feature with human eye, as basic grain pattern, pattern district, core point, trigpoint, style line, streakline quantity etc.Local feature then is the node on the fingerprint, as destination node, bifurcation, isolated point extremely direction and position etc.Fingerprint identification technology relates generally to four functions: read fingerprint image, extract feature, preserve data and comparison.At the beginning, read the image of somatic fingerprint by the fingerprint fetch equipment, get after the fingerprint image, carry out preliminary processing to original image, make it more clear.Next, identification of fingerprint software is set up the numeral of fingerprint---characteristic.At last,, the template of two fingerprints is compared, calculate their similarity degree, thereby obtain matching result by the fuzzy comparative approach of computing machine.
Fingerprint recognition has caused the extensive concern of domestic and international academia and industry member as a technological means of utilizing the intrinsic biological information of human body to discern individual true identity, and becomes a research focus in the area of computer aided identification field gradually.Simultaneously, it makes the software developer constantly develop and develop fingerprint recognition product and the technology that makes new advances in some successful Application aspect commercial.Because domestic scientific research institutions and enterprise's research is in this respect started late, the basis is also relatively weaker, can't determine the input risk of fingerprint recognition system project and the income of expection, add present many technological difficulties of also not resolved that in fingerprint recognition system, still have, and every profession and trade is not enough to the dynamics of investment of disposing fingerprint recognition system, and it is very slow that these reasons all make domestic fingerprint recognition use progress.The core technology of present fingerprint recognition Related product all is to be provided by external famous colleges and universities and large enterprise mostly, as Michigan State University, Microsoft and the IBM etc. of the U.S.; Domestic also have only minority enterprise to develop the independent intellectual property right algorithm for recognizing fingerprint, and buy and use these core technologies to need user effort great amount of manpower and material resources.These fingerprint recognition systems are generally used and are designed towards single function, as fingerprint attendance system, fingerprint door control system and electronic fingerprint lock etc., also do not occur a kind of fingerprint recognition system software based on many application integration on the market.This makes enterprise that the fingerprint recognition product is had a demand choosing, can reduce service efficiency unavoidably during deployment system, increasing client's gross investment, therefore usually can bring inconvenience to the client.
Positive fingerprint series of products in the ZKFinger of control science and technology, the Hangzhou in the Fingerpass embedded fingerprint recognition system that the present comparatively famous fingerprint recognition system of domestic industry circle is the Chinese Academy of Sciences, Shenzhen, and the product of company such as IBM, SUN, IDTECK, Sony, Compaq, these products have all obtained promoting preferably on market, but certain defective are also arranged and need to continue perfect place.The not enough aspect of these fingerprint recognition products mainly shows: 1, all be at the feature and function exploitation, lack other function expansion, single purchase is difficult to tackle multiple applicable cases; 2, software systems are only supported specific fingerprint collecting instrument, and hardware device is not had versatility; 3, algorithm upgrading difficulty is difficult to the seamless loading of implementation algorithm module; 4, interface is non-public, is difficult to realize flexible secondary development.
Summary of the invention
Technical matters to be solved by this invention is a kind of fingerprint recognition system and control method thereof, this system overcomes the defective of prior art, provide Zai Xian off-line, the basic algorithm storehouse can fingerprint recognition system service seamless expansion, that function can be matched, the user is dropped under the situation of substantial contribution not needing, obtain high-quality identification and authentication service.
Technical matters proposed by the invention is to solve like this: a kind of fingerprint recognition system is provided, comprise fingerprint database, central server and client three parts, it is characterized in that, fingerprint database is mainly used to store the fingerprint characteristic information through after the numerical coding, central server mainly is responsible for the checking client legitimacy, is received fingerprint characteristic data, aspect ratio to loopback fingerprint recognition result, client mainly is responsible for gathering fingerprint, extracting and upload fingerprint characteristic data; Central server comprises legitimate verification module, system management module, feature comparing module, second development interface module, algorithm load-on module, flow custom module, monitoring module and identification service module, client comprises finger print acquisition module, fingerprint pretreatment module, characteristic extracting module and pattern classification module, data communication module connects central server and client, wherein:
The legitimate verification module: server carries out the rights of using checking to the client of request service, prevents that unauthorized user from proposing the fingerprint recognition services request to central server.
Finger print acquisition module: client off-line is gathered fingerprint image or online acquisition vital fingerprint image, wherein off-line is gathered the picture that fingerprint is mainly gathered BMP form and JPG form, the JPG formatted file carries out data decomposition to it earlier, decompose file from the isolated code table of packed data and quantization table, obtain view data in that it is carried out inverse discrete cosine transform;
The fingerprint pretreatment module: the fingerprint that client is gathered strengthens, and comprises the field of direction device of fingerprint effective coverage segmenting device and fingerprint and the device that utilizes the M-PCNN network that the later image of pre-service is carried out filtering;
Characteristic extracting module: global characteristics and the minutia of extracting fingerprint that client is gathered, the mode type that comprises fingerprint, client's end points, bifurcation position and coordinate figure thereof comprise the Fourier spectrum feature deriving means and the minutia extraction element that extract global characteristics;
The pattern classification module: main being responsible for classified to the pattern of fingerprint, and fingerprint pattern is divided into six classes: whirlpool, left side ring, right ring, dicyclo, arch form and cusped arch;
Feature comparing module: the fingerprint characteristic of storing in the fingerprint characteristic of client upload and the database is compared, calculate characteristic similarity between the two, comprise fingerprint alignment means and similarity calculation element;
Second development interface module: provide second development interface, for the user constructs the new business logic;
The algorithm load-on module: to the new algorithm module of finishing load, the sign and the interface message of management algorithm module;
Flow custom module: safeguard existing flow process, newly-increased flow process, be each stage placement algorithm of flow process;
Monitoring module: supervise that each is online, the services request of offline client and ruuning situation;
Identification service module: handle the service of client-requested, mainly comprise fingerprint register, fingerprint recognition, fingerprint authentication and connection request;
System management module: the rudimentary algorithm apolegamy of the loading of management rudimentary algorithm, fingerprint recognition, the management of fingerprint client authorization, the assembling of fingerprint recognition application system and the assessment of algorithm operation result;
Data communication module: between client and server, establish a communications link, send and receive fingerprint characteristic data and fingerprint recognition object information.
According to fingerprint recognition system provided by the present invention, it is characterized in that described second development interface module mainly is that the function of the second development interface in the following table will be provided:
Numbering The interface function function The interface function prototype
1 System initialization API_INI_SYS
2 Read finger print data API_READ_DATA
3 Fingerprint classification API_FIG_CLASSIFY
4 The calculated direction field API_GET_ORIENTATION
5 Fingerprint recognition API_RECOGNITION
6 Fingerprint authentication API_VERIFY
These interface functions are encapsulated in the user and can programme in the dynamic link library of the FIG_NN_API by name that calls.
A kind of method for controlling fingerprint identification is characterized in that, comprises step:
(1) client connects central server by communication interface, after the legitimate verification of client identity process, obtains corresponding services request mandate;
(2) gather finger print data to be identified, carry out pattern classification and rough handling;
(3) finger print data with the rough handling of step (2) process carries out the fingerprint pre-service, comprises following little step:
1. the fingerprint effective coverage is cut apart: fingerprint image is divided into the image block of a series of 16 * 16 non-intersections, and each piece is labeled as B (1,1) respectively, B (1,2) ..., B (i, j), utilize then following formula v (i, j)
v ( i , j ) = ( x 1 - x ‾ ) 2 + ( x 2 - x ‾ ) 2 + . . . + ( x n - x ‾ ) 2 N
Calculate the grey scale pixel value variance of each image block, wherein x nWith x represent respectively this figure fast in the gray-scale value of pixel, N represents the pixel quantity that comprises in the segment segmentation threshold v to be set θ=11.5, respectively with the variance yields and the v of each segment θCompare, if v (i, j)〉v θ, then (i j) is judged as effective finger-print region to this segment B, otherwise is judged as the fingerprint background;
2. the field of direction of fingerprint is calculated: calculate respectively segment B (i, j) in each pixel gradient G in the x and y direction xAnd G y, utilize formula d (i, j)
d ( i , j ) = 1 2 arctan ( Σ i b ′ = 1 w Σ j b ′ = 1 w 2 G x ( i b ′ , j b ′ ) G y ( i b ′ , j b ′ ) Σ i b ′ = 1 w Σ j b ′ = 1 w G x 2 ( i b ′ , j b ′ ) - G y 2 ( i b ′ , j b ′ ) )
(i, local direction j) is in the formula to calculate segment B The coordinate of pixel in the expression segment, w represents the pixel wide of segment, value like the segment local direction value that calculates is only got in four component values 0, π/4, π/3 and 3 π/4 recently, local direction d (the i that segment is final, j) ∈ { 0,4 π/4, π, 3 π/4}, the consistance feature correction field of direction result of calculation of direction of passage field, (i is in 5 * 5 neighborhood D scope j) at segment B, calculate its consistance value C (i, j)
C ( i , j ) = 1 24 Σ ( i ′ , j ′ ) ∈ D | d ( i , j ) - d ( i ′ , j ′ ) | 2
Have in this formula
Here d=mod (d (i ', j ')-d (i, j)+2 π)
If C (i, j)<0.35, then with segment B (i, local direction j) are adjusted into the most significant direction of local direction in the neighborhood D, otherwise segment B (i, local direction j) remains unchanged;
3. use the M-PCNN network that the later image of pre-service is carried out filtering;
(4) after treatment fingerprint image in the step (3) is carried out feature extraction, comprises that global characteristics extracts and the minutia extraction:
1. global characteristics is extracted as the Fourier spectrum feature: at first the fingerprint image with input is divided into 32 * 32 image block, and segment is done two dimensional discrete Fourier transform, and formula is
G ( m , n ) = 1 N Σ i = 0 N - 1 Σ K = 0 N - 1 ( g ( i , k ) exp ( - j 2 π ( mi N + kn N ) ) )
In the formula, (m n) is the codomain coordinate of pixel, and (i k) is the frequency domain coordinate of pixel correspondence, and each subgraph piece has just obtained the fourier spectrum figure of full figure through after the Fourier transform, and it is quantized into the global characteristics vector of fingerprint by rule;
2. minutia is extracted, and comprises core point, bifurcation and end points;
(5) send the characteristic of the fingerprint image that is extracted through step (4) to central server;
(6) central server receives and decomposes the characteristic of uploading, and is stored in characteristic formation to be matched;
(7) the fingerprint characteristic comparing module is according to the matching algorithm searching database of the current apolegamy of system, and the similarity data that obtain compared in record at every turn;
(8) central server is determined current recognition result, and the loopback clients corresponding;
(9) client receives result, and carries out subsequent action in view of the above;
Wherein in the step (1), the treatment step of the services request of central server customer in response end is as follows:
1. communication module receives the fingerprint recognition services request from client, and verifies its legitimacy; 2. central server sends request by communication module to client and allows signal, illustrates that current request is is checked and approved, and can send fingerprint characteristic data.
Control method according to fingerprint recognition provided by the present invention is characterized in that, the middle M-PCNN network of step (3) carries out filtering to the later image of pre-service and may further comprise the steps:
1. during initialization, according to size of images structure PCNN network, and the PCNN neuron is corresponding one by one with the image slices vegetarian refreshments;
2. under the cooperation of the field of direction and segmentation result, allow the PCNN operation, and constantly revise the neuronic load signal strength of igniting, when neuronic activity during greater than given threshold value, neuron is lighted a fire, and the gray-scale value of pixel is upgraded;
3. set the neuron operation rule, allow PCNN constantly move, the neuron of lighting a fire in the network triggers grey scale pixel value is made amendment, and when the PCNN network stabilization, has just obtained the later image of filtered enhancing from each neuronic load signal value.
Control method according to fingerprint recognition provided by the present invention is characterized in that, in the minutia extraction step, it is on the basis of fingerprint direction of fingerprint field that core point is mentioned in the step (4), utilizes the poincare method to extract this value Wherein Δ (k) meets
&Delta; ( k ) = &delta; ( k ) , if | &delta; ( k ) | < &pi; 2 &pi; + &delta; ( k ) , if &delta; ( k ) &le; - &pi; 2 &pi; - &delta; ( k ) , otherwise
δ(k)=O′(Ψ x(i′),Ψ y(i′))-O′(Ψ x(i),Ψ y(i)),
i′=(i+1)mod?N Ψ
After calculate finishing, judge again p (i, value j) is if (i j)=0.5, then is a core point to P; The extraction of bifurcation and end points detects by the sleiding form method and obtains, and at first defines the dot structure template of bifurcation and end points, respectively with this template by from left to right, from top to bottom order traversal full figure.The every slip of template once, just the matching degree of calculation template and image corresponding region, when meeting value greater than 0.7 the time, the minutia corresponding with using template with regard to the current pixel of process decision chart picture is consistent, if promptly meet the bifurcated template, then this pixel is a bifurcation, if meet the end points template, then this pixel is exactly an end points.
Control method according to fingerprint recognition provided by the present invention is characterized in that, in the step (7), the fingerprint characteristic contrast may further comprise the steps:
1. fingerprint alignment: input fingerprint feature point set Vi={v1, v2 ..., Vn}, wherein vi=(x, y, θ), Tj={t1, t2 ... tm} is the unique point set of template fingerprint, and at first to the core point alignment, after the alignment, the coordinate of Vi point set is done as down conversion:
X ' i=x i+ (x To-x Io) and y ' i=y i+ (y To-y Io),
(x wherein To, y To) and (x Io, y Io) difference template fingerprint and the core point of importing fingerprint; And then the rotation alignment, the rotation alignment is then by asking a reflection transformation t to obtain:
t ( v ) = cos a - sin a &Delta;x sin a cos a &Delta;y 0 0 1 x y 1
In above-mentioned conversion, the core point alignment, Δ x and Δ y determine that a is by making expression formula | v t-t (v i) | 2<θ sets up, and asks Vi and Vt match point logarithm maximum to obtain, and wherein θ is a preset threshold, is set to 0.05;
2. similarity is calculated: mate for global characteristics, the feature of its input fingerprint and template fingerprint is the fourier spectrum feature, proper vector is exactly spectrogram pixel serialization result, similarity between the two is not wait the Euclidean distance between the long vector to realize by calculating, and similarity is wherein calculated by following formula:
s = ( 1 - dis max ( v i , v t ) ) &times; 100 %
In the formula, dis is both Euclidean distance values, max (v i, v t) expression gets in input vector and the template vector maximum vector length; For the minutia coupling, be to be benchmark with the template, will import point set Vi by (Δ x, Δ y, a) conversion projects to after template point set Vt goes up, calculate distance error less than 5 location of pixels in, all match points determine that to quantity the similarity value of this moment is calculated by following formula:
s = n couple max ( v i , v t ) &times; 100 %
In the formula, n CoupleRepresent the successfully quantity of the point of pairing, max (v i, v t) still expression input point set and template point are concentrated the minutiae point number of maximum;
3. fingerprint recognition: calculate after the similarity data S between input fingerprint and the template fingerprint, again with S and default recognition threshold T sRelatively, the T here sAdjusted according to different security requirements, if S 〉=T sJudge that then two fingerprints successfully mate, promptly come from same finger, otherwise judge that two fingerprints are not to come from same finger.
Function of the present invention by the processing stage reside at client and central server respectively, the basic algorithm storehouse can constantly load new fingerprint Processing Algorithm module, the fingerprint recognition process can be matched different Processing Algorithm unit, to adapt to the applied environment of different performance requirement, system with central server to each client centralized control, can effectively prevent to insert this system and initiate illegal services request without the client of system's use authority, it provides complete second development interface to the third party, make things convenient for the user to make up new application systems software in view of the above based on fingerprint recognition, system develops based on neural network model, as Pulse Coupled Neural Network (PCNN), support vector machine (SVM) etc., have good concurrent ability and fuzzy diagnosis ability, the built-in adaptive middleware module of fingerprint equipment of system makes the fingerprint collecting instrument of most of manufacturer production on this compatible market of platform energy.
The present invention uses fingerprint recognition and the test of heuristics assessment is integrated in one, and both can be deployed as conventional fingerprint recognition application software, also can be configured as the experimental test Evaluation Platform of fingerprint rudimentary algorithm module newly developed.The identification and the authentication service that can the concentrated area all fingerprint client users be facilitated only need to buy the client right to use and can obtain rapidly and efficiently fingerprint recognition service, are beneficial to the user and reduce investment outlay.Communication in the system between the client and server is all undertaken by the internet that extensively distributes at present, makes full use of existing resource to save communications cost.Client can be matched fingerprint enhancement algorithms, thinning algorithm, feature extraction algorithm etc., and server also can be matched the fingerprint database searching algorithm, and the Processing Algorithm on basis such as fingerprint characteristic matching algorithm also can be adjusted the similarity decision rule in real time.The user can be configured to the single application systems software of function with it after blocking some function, also can be configured to the test of heuristics Evaluation Platform.
The present invention can support 500 clients at most, make the small user not need to drop under the situation of substantial contribution purchase and deployment whole system, can obtain high-quality fingerprint recognition service, thereby provide strong backing for the authentication demand of company and enterprise.
The present invention provides unified fingerprint recognition service for the user on the internet: the various Processing Algorithm modules (as fingerprint image enhancing, field of direction calculating, pattern classification etc.) that the user can using system provides after by registration.The user also can be within the scope of authority, at the suitable Processing Algorithm of application characteristic apolegamy separately, as selecting Gabor filtering enhancement algorithms or RNN enhancement algorithms.Simultaneously, for the protection of user's data safety and data-privacy is considered that total system only writes down fingerprint characteristic data, any original fingerprint image of arriving of record acquisition does not help realizing the protection of individual privacy.
Fingerprint recognition system integrated platform software function provided by the present invention is very powerful, has contained the most ripe algorithm in current fingerprint Recognition field; System adopts distributed structure/architecture, can make the application deployment of system break through geological restraint and eliminate bottleneck and realize load balancing preferably, thereby improve the overall system handling capacity.
With the capable exploitation of C# language, total system adopts the three-layer architecture model on the Net platform in the present invention.Wherein the fingerprint client tier mainly comprises fingerprint collecting equipment, the adaptive middleware of tool interface system, and fingerprint strengthens objects such as module, characteristic extracting module, communication module; Central server mainly comprises database retrieval module, feature comparing module, system management module etc.; Database mainly is made of MS SQL SERVER 2000.Though native system logically is based on the distributed system exploitation, whole system also can be deployed on same the computing machine physically, saves hardware investment.
The performance history of fingerprint recognition system integrated platform software is carried out according to national software engineering standard fully, and adopt OO distributed component development scheme, taken into full account extensibility and the transplantability and the data security of system in the design phase, thereby guaranteed that these software systems have had preferably extendability, data security and adaptability widely.
Principal feature of the present invention is summarized as follows:
1, system adopts distributed three-layer architecture model, develop according to the soft project standard, client and application server logical separation, thereby the operating load of balanced client process machine and server have improved the upper limit that system can support client terminal quantity greatly.
2, the fingerprint algorithm storehouse content of the system integration is very abundant, the fingerprint Processing Algorithm that has almost contained present all main flows, the user can also develop new Processing Algorithm and be increased to corresponding rudimentary algorithm storehouse according to actual conditions, rationally matches, uses according to application characteristic for the user.
3, the fingerprint collecting equipment interface middleware of the system integration make this system can the support the market on most fingerprint collecting equipment, this middleware module can also constantly be expanded, make system can support more equipment, improved the adaptive faculty of system hardware environment.
4, the function of system realizes based on distributed, multithreading thought.The source of the fingerprint characteristic data of handling is distribution and transparent, the user need not to know the concrete deployed position of application server and database server in the system, whole system promptly can be deployed on same the computing machine, also can be deployed on the computing machines different in the network.In addition, application server is corresponding with it at service thread of client unlatching that each proposes services request, has improved the concurrent processing ability of system.
5, system adopts the client identity authentication policy; can guarantee of the illegal use of unauthorized client end to native system; and Server Transport and processing only is the fingerprint characteristic, thereby plays the protection user resources, guarantees data security and protect the purpose of individual privacy.
6, system provides collocation strategy flexibly, after the user looks concrete condition blockade redundancy feature, native system can be configured to single fingerprint identification application system, also can be configured to experiment test, the Performance Evaluation platform of newly-increased algorithm.
Along with network the popularizing gradually of China, former application ever-changing, distributed, the versatility of the fingerprint recognition system software forward of function singleness, integrated-type comprehensive platform software direction develop at present.Can predict that following fingerprint identification software developing direction is towards the internet, towards distributed structure/architecture, feature richness, freedom and flexibility.Native system is that the user provides complete fingerprint recognition service with online and offline mode, and its algorithms library can freely be expanded, and can adapt to multiple fingerprint collecting instrument, can respond the services request of a plurality of clients concomitantly; Simultaneously native system is by carrying out strict use authority and authentication to client, and only handles the characteristic of extracting, and can prevent that so illegal use from also having guaranteed the security of system, also can effectively protect individual privacy simultaneously.
Description of drawings
Fig. 1 is a system assumption diagram of the present invention;
Fig. 2 is a functional hierarchy distribution plan of the present invention;
Fig. 3 is the functional block diagram of client of the present invention;
Fig. 4 is the functional block diagram of central server of the present invention;
Fig. 5 is a control flow chart of the present invention;
Fig. 6 is an algorithm load-on module control flow chart of the present invention;
Fig. 7 is a fingerprint recognition flow process apolegamy control flow chart of the present invention;
Fig. 8 is a legitimate verification process flow diagram of the present invention;
Fig. 9 is a JPG algorithm flow chart of the present invention;
Figure 10 is that online acquisition living body finger print function of the present invention realizes schematic diagram;
Figure 11 is the neuronal structure figure of PCNN network of the present invention.
Embodiment
The present invention is further described below in conjunction with embodiment.
Fingerprint recognition system integrated platform software provided by the present invention strictly comes each functional module of design system according to standard fingerprint recognition workflow, extensibility based on system is considered, when design, make each module independent, nuclear interface standardizing, the variation of each inside modules can not cause the change that other module is big like this, such module independent helps the fast updating and the expansion upgrading of system, to adapt to the develop rapidly of fingerprint identification technology.
This system global structure as shown in Figure 1, its deployment comprises four parts, is respectively: fingerprint database, central server, communication network and client.Wherein the system platform of fingerprint database is windows server2003, and database platform is MS SQL server 2005 (Enterprise); The system platform of central server (identified server) is windows Server 2003; Communication network is common internet or intranet environment; The operation platform of client is windows 2000/XP, " second wood " XYZ-1 type fingerprint collecting instrument of the accurate configuration of affix., the core identification module of native system operates on the center identification server, and other processing module then runs on the client.Whole network is the center with the center identification server, the plurality of client end is Star Schema with it and connects, in the normal operation of system, be responsible for off-line or gather fingerprint image online by client, and the fingerprint character code that processing obtains sent to central server by network, and discern (1:N) or checking (1:1) by the core processing module on the central server, central server finishes the back to current fingerprint processing and returns result to workstation.
Fingerprint recognition system functional hierarchy provided by the present invention distributes as shown in Figure 2, orlop is software and hardware operation platform (computer hardware equipment, operating system, a fingerprint instrument), on lower floor the integrated platform software systems, some processing procedures that fingerprint recognition is required have been comprised, such as processing procedures such as the management of platform, fingerprint collecting, pre-service, feature extraction, fingerprint comparisons.
Fig. 3 and Fig. 4 have shown two cores of this platform software system from software composition function aspect, i.e. the functional module that comprises of central server and workstation (client) subsystem software.As shown in Figure 3, the functional module that client comprises is respectively: communication module, finger print acquisition module, fingerprint pretreatment module, pattern classification module, characteristic extracting module, display module.Wherein communication module main with " central server " on the respective modules swap data, send client identity information, upload fingerprint characteristic data, receive institute and ask the result of serving; Finger print acquisition module is mainly finished the online or off-line input of fingerprint, and function such as simple pre-service; Pretreatment module is mainly finished functions such as fingerprint image enhancing, calculated direction field, refinement; The pattern classification module mainly is responsible for the pattern of fingerprint is classified (divide six classes: whirlpool, left side ring, right ring, dicyclo, arch form and cusped arch); Characteristic extracting module is mainly finished the global characteristics of fingerprint or minutiae feature is extracted, and presses the string format coding, waits for that communication module sends to server; The duty of the main display workstation of display module, mode of operation, connection status, registration or identification (containing checking) result etc.The functional module that " central server " subsystem comprises is respectively: communication module, system management module, algorithm load-on module, flow custom module, monitoring module, identification service module.Wherein communication module mainly be responsible for to receive legitimate verification information that client sends, also receives service request type (registration or identification) and the corresponding fingerprint characteristic data of uploading, and to client loopback result; Basic data, client station management and the platform security control etc. on configuration-system operational factor, management backstage mainly are responsible in system management; The algorithm load-on module finish to new algoritic module load, sign of management algorithm module, interface message etc.; The flow custom module is responsible for safeguarding existing flow process, newly-increased flow process, is each stage placement algorithm of flow process etc.; Condition monitoring mainly is responsible for the services request and the ruuning situation of each online (off-line) client of supervision; The identification service module is responsible for handling the service of client-requested, mainly comprises kernel service contents such as fingerprint register, fingerprint recognition (1:N), fingerprint authentication (1:1), connection request.Below several emphasis modules are described:
The legitimate verification module: server carries out the rights of using checking to the client of request service, prevents that unauthorized user from proposing the fingerprint recognition services request to central server.The realization principle of this part here in order to narrate conveniently, is designated as IPc with the IP address of client as shown in Figure 8, and the IP address of central server is designated as IPs, and port numbers is n_port.When central server starts, the service routine that to create a port numbers automatically be n_port.Client need be before central server request legitimate verification, needs connect with central server earlier, and client proposes connection request by socket (socket) in this module, parameter be (IPs, n_port).After server has responded the connection request of client (connect confirm) as figure,, client communicates to connect just having set up with server.Communicate to connect after the foundation, client is by data structure format (Term_code, IPc, Mac_addr, User_name, User_PSW, Authorized) assembling identity information, here Term_code refers to client code, and IPc and Mac_addr are respectively the IP address and the physical addresss of client, and User_name and User_PSW are respectively operator's account number and password.After the identity information assembling was finished, client sent this identity information via communicating to connect to server.After server is received this identity information, compare with the legitimate client client information of preserving in the background data base, if there is no the client of this identity judges that then active client is the unauthorized client end, and disconnects it at once and communicate to connect.If judge that active client is legal, then produce a group length at random and be 8 character string STR_USER and return to client as the client identification code.In follow-up fingerprint recognition, checking request service, it oneself is legal user's identity of process server authorizes that client relies on this identification code str_user to show.
Finger print acquisition module: client off-line is gathered fingerprint image or online acquisition vital fingerprint image.The submodule of this module constitutes situation, and it is as shown in the table:
The picture of BMP, JPG form is supported in the collection of this module off-line, and for the JPG fingerprint image that off-line is gathered, module is carried out the JPG decoding and converted data bitmap to it.The JPG decoding algorithm flow process that adopts is shown in 9, and wherein, the JPG packed data obtains by decomposing file, and and then according to its obtain decoding needed code table and quantization table, IDCT carries out inverse discrete cosine transform to image.After the JPG decoding, obtain the data of YcbCr color system, and it is carried out gradation conversion, obtain importing 256 rank gradation datas of JPG fingerprint image.If the off-line input is BMP form fingerprint image, then, directly obtain the DIB view data by decomposing the BMP bitmap file.Also do not pass through greyscale transformation if not this DIB view data, then also will carry out greyscale transformation to it, finally obtain 256 rank gradation datas, these gradation datas will be as the benchmark input data of Fingerprint Processing Module.
If pass through the living body finger print of fingerprint collecting equipment online acquisition, then the output data of equipment directly is exactly 256 rank gradation datas, reads in the core buffer that therefore only needs the slave unit interface to indicate to get final product, and does not need to carry out other processing again.These software systems are integrated at present scientific and technological U.R.U 2000 fingerprint capturers of middle control and second Ochnaceae skill XYZ-1 fingerprint acquisition instrument, but all fingerprint collecting equipment on the support the market in theory.This part function of online acquisition living body finger print realizes that principle as shown in figure 10, detect the fingerprint equipment of installing on the client by the detection sub-module in the module among the figure, and read corresponding allocation list according to testing result, from allocation list, obtain the driving interface information (INT_FUNCTION_NAME) of this equipment, determine to gather action triggers signal (CAP_FLAG) according to these information again.When gathering fingerprint, whether module then has the fingerprint collecting action to take place according to this CAP_FLAG signal determining, if having then the finger print data buffer zone first address and the data volume that transmit according to interface read the finger print data that collects.Realize that by the aftermentioned process fingerprint equipment expands in the module: at first load the driver of fingerprint equipment, create the configuration list item for this new equipment then.The elementary field that the configuration file list item comprises is: device numbering (DEV_ID), device name (DEV_NAME), interface function name (INT_FUNCTION_NAME), path (INT_PATH) etc.
The fingerprint pretreatment module: client strengthens the fingerprint image of being gathered.This module adopts two steps that low-quality fingerprint image is strengthened, and first step is carried out pre-service to the original fingerprint image, and second step is with the M-PCNN network the later image of pre-service to be carried out filtering.Wherein the first step is divided into two processes again, is respectively that the fingerprint effective coverage is cut apart with the field of direction of fingerprint and calculated, and the original fingerprint image is carried out in the pretreated step, the image block that earlier the original fingerprint image of input is divided into a series of 16 * 16 non-intersections, each piece is labeled as B (1,1) respectively, and B (1,2), ..., B (i, j), utilize then following formula v (i, j)
v ( i , j ) = ( x 1 - x &OverBar; ) 2 + ( x 2 - x &OverBar; ) 2 + . . . + ( x n - x &OverBar; ) 2 N
Calculate the grey scale pixel value variance of each image block.X in the formula nWith x represent respectively this figure fast in the gray-scale value of pixel, N represents the pixel quantity that comprises in the segment.Segmentation threshold v is set θ=11.5, respectively with the variance yields and the v of each segment θCompare, if v (i, j)〉v θ, then (i j) is judged as effective finger-print region to this segment B, otherwise is judged as the fingerprint background.After this processing process finished, the fingerprint effective coverage had just split from background.
And then with the field of direction of the later fingerprint effective coverage correspondence of gray scale descent method computed segmentation.The specific algorithm flow process that adopts in the module is as follows:
(1) calculate respectively segment B (i, j) in each pixel gradient G in the x and y direction xAnd G y
(2) utilize formula d (i, j)
d ( i , j ) = 1 2 arctan ( &Sigma; i b &prime; = 1 w &Sigma; j b &prime; = 1 w 2 G x ( i b &prime; , j b &prime; ) G y ( i b &prime; , j b &prime; ) &Sigma; i b &prime; = 1 w &Sigma; j b &prime; = 1 w G x 2 ( i b &prime; , j b &prime; ) - G y 2 ( i b &prime; , j b &prime; ) )
(i, local direction j) is in the formula to calculate segment B The coordinate of pixel in the expression segment, w represents the pixel wide of segment.Value like the segment local direction value of utilizing formula to calculate is only got in four component values 0, π/4, π/3 and 3 π/4 recently, so final local direction d (i, j) ∈ { 0,4 π/4, π, 3 π/4} of segment;
(3) the consistance feature correction field of direction result of calculation of direction of passage field.Segment B (i, in 5 * 5 neighborhood D scope j), calculate its consistance value C (i, j):
C ( i , j ) = 1 24 &Sigma; ( i &prime; , j &prime; ) &Element; D | d ( i , j ) - d ( i &prime; , j &prime; ) | 2
Have in this formula
Here d=mod (d (i ', j ')-d (i, j)+2 π)
If C (i, j)<0.35, then with segment B (i, local direction j) are adjusted into the most significant direction of local direction in the neighborhood D, otherwise segment B (i, local direction j) remains unchanged.
In fingerprint strengthens, use be the PCNN filtering method.The neuronal structure of PCNN network is shown in Figure 11, and neuronic active U (m ', n ') obeys following rule:
u m &prime; n &prime; ( t + 1 ) = M m &prime; n &prime; &CenterDot; x m &prime; n &prime; ( t ) + &Sigma; ab &NotEqual; m &prime; n &prime; ab &Element; L m &prime; n &prime; w ab &CenterDot; &phi; ( N ab , O m &prime; n &prime; ) , y m &prime; n &prime; ( t + 1 ) = f ( u m &prime; n &prime; ( t + 1 ) - &theta; m &prime; n &prime; ( t + 1 ) ) .
Initialized the time, according to size of images structure PCNN network, and the PCNN neuron is corresponding one by one with the image slices vegetarian refreshments.In fact the enhancing of image is exactly under the cooperation of the field of direction and segmentation result, allows PCNN move, and constantly revises the neuronic load signal strength of igniting.For four matrixes of PCNN filter design as follows:
K 0 = 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 . ? K 1 = 0 0 0 1 1 0 0 1 1 1 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 . ? K 2 = 0 1 1 1 0 0 1 1 1 0 0 1 1 1 0 0 1 1 1 0 0 1 1 1 0 , ? K 3 = 1 1 0 0 0 1 1 1 0 0 0 1 1 1 0 0 0 1 1 1 0 0 0 1 1
When neuronic activity during greater than given threshold value, neuron is lighted a fire, and then the gray-scale value of pixel upgrades by following formula:
x m &prime; n &prime; ( t ) = x m &prime; n &prime; ( t - 1 ) + &delta; N &Sigma; ab &NotEqual; m &prime; n &prime; ab &Element; L m &prime; n &prime; | x ab ( t - 1 ) - x m &prime; n &prime; ( t - 1 ) | K ab O m &prime; n &prime;
The coefficient δ that wherein has determines by following rule:
By above-mentioned rule, PCNN constantly moves, and the neuron of lighting a fire in the network triggers grey scale pixel value is made amendment.When the PCNN network stabilization, just obtained the later image of filtered enhancing from each neuronic load signal value.
Characteristic extracting module: global characteristics that client takes the fingerprint and minutia comprise the mode type of fingerprint, client's end points, bifurcation position and coordinate figure thereof.The global characteristics that uses in the native system comprises core point, bifurcation and end points as Fourier spectrum feature, the minutia of use, divides two steps feature that takes the fingerprint in this module:
The Fourier spectrum Feature Extraction: at first the fingerprint image with input is divided into 32 * 32 image block, and segment is done two dimensional discrete Fourier transform, and formula is
G ( m , n ) = 1 N &Sigma; i = 0 N - 1 &Sigma; K = 0 N - 1 ( g ( i , k ) exp ( - j 2 &pi; ( mi N + kn N ) ) )
In the formula, (m n) is the codomain coordinate of pixel, and (i k) is the frequency domain coordinate of pixel correspondence.Each subgraph piece has just obtained the fourier spectrum figure of full figure after Fourier transform, this spectrogram is quantized into the global characteristics vector of fingerprint by rule.
The extraction of minutia: the detail characteristics of fingerprints that relates in this module is core point, bifurcation and end points, and wherein core point is on the basis of direction of fingerprint field, utilizes the poincare method to extract this value Wherein Δ (k) meets
&Delta; ( k ) = &delta; ( k ) , if | &delta; ( k ) | < &pi; 2 &pi; + &delta; ( k ) , if&delta; ( k ) &le; - &pi; 2 &pi; - &delta; ( k ) , otherwise
δ(k)=O′(Ψ x(i′),Ψ y(i′))-O′(Ψ x(i),Ψ y(i)),
      i′=(i+1)mod?N Ψ
After calculating end, judge that again (i, value j) is if (i j)=0.5, then is a core point to P to p.Bifurcation in the fingerprint and end points detect by the sleiding form method and obtain, and are respectively the definition of bifurcation and end points and use following dot structure template:
Detect the template of the template detection end points of bifurcated
Defined after these two feature templates, respectively with this template by from left to right, from top to bottom order traversal full figure.The every slip of template once, just the matching degree of calculation template and image corresponding region, when meeting value greater than 0.7 the time, the minutia corresponding with using template with regard to the current pixel of process decision chart picture is consistent, if promptly meet the bifurcated template, then this pixel is a bifurcation, if meet the end points template, then this pixel is exactly an end points.
The spectrum signature that Fourier filtering obtains, directly with the gray-scale value of pixel in the spectrogram by from left to right, from top to bottom order carries out serial code, obtain proper vector, detected minutia is by data structure (unique point sequence number, minutiae point type, the X coordinate, the Y coordinate, deflection) encode, special to the minutia data.
Data communication module: be responsible between client and central server, establishing a communications link, sending and receive fingerprint characteristic data and fingerprint recognition object information.This module realizes by the SOCKET technology, details as after.At server end, use canonical function
Socket (AF_INET, SOCK_STREAM, IPPROTO_TCP); Create a service SOCKET, with the bind function itself and server ip address are bound again, and on formulation port (n_port), monitor.In client, utilizing SOCKET to create one is the SOCKET object of parameter with client ip address and Service-Port n_port, and this object just communicates with server on designated port.
System management module: this module functions is to be responsible for the loading in rudimentary algorithm storehouse, each rudimentary algorithm apolegamy of fingerprint recognition, client end empowerment management, the assembling of fingerprint recognition application system, the assessment of algorithm operation result,
(1) algorithm loads submodule.
Earlier algorithms of different is packaged into independent DLL (dynamic link library), an algorithm is packaged into a DLL, and unique name.Copy to the algorithms library file with the COPY function then, implementation algorithm DLL uploads to system.For DLL creates the configuration list item, the structure of this configuration list item is as follows then:
Structure?Algorithm_Config
{
AL_TYPE: algorithm classification
AL_CODE: algorithm numbering
AL_NAME: algorithm title
AL_FUNC: the entrance function of the corresponding DLL of algorithm
FUNC_IN: the input parameter explanation of algorithm entrance function
FUNC_OUT: the output parameter explanation of algorithm entrance function
}
The classification of algorithm comprises four big classes, is respectively: fingerprint strengthens (classification sign indicating number 0), feature extraction (classification sign indicating number 1), fingerprint matching (classification sign indicating number 2), fingerprint classification (classification sign indicating number 3).The data layout of algorithm numbering is: classification sign indicating number+sequence number.
(2) client-side management submodule.
Client-side management comprises maintenance of information and client authorization management, wherein maintenance of information mainly be to client device information and client user information increase, operation such as deletion, modification, the client device information structure is (Term_code, IPc, Mac_addr, User_name, User_PSW, Authorized), here Term_code refers to client code, IPc and Mac_addr are respectively the IP address and the physical addresss of client, and User_name and User_PSW are responsible for operator's account number and password, and Authorized is an authorization flag.Client authorization is to utilize the UPDATE operation of SQL, for the Authorized field in the client-side information table is provided with the TRUE value, shows that this client has obtained server authorizes, can propose services request to server.
(3) application system assembling submodule.
Comprise the two large divisions, the one, the custom-built system flow process, the 2nd, algorithm parameter is set.This module customizes its treatment scheme at the fingerprint recognition system of different qualities, selects different enhancement algorithms, feature extraction algorithm, matching algorithm, fingerprint classification algorithm respectively by the fingerprint treatment scheme, and selection result is preserved.It mainly is after each algorithm is selected to determine on stream, to be respectively each algorithm relevant parameters is set, as the size that strengthens filter window, the threshold value θ of matching algorithm, categorical measure of fingerprint classification or the like that algorithm parameter is set.
(4) system performance assessment submodule.
In this module, mainly be to evaluate and test the results of property that given fingerprint recognition system obtains at given sample fingerprint from three aspects, be respectively FAR (false acceptance rate) FRR (false rejection rate), ROC resultant curve, response time properties T.
Feature comparing module: the fingerprint characteristic of storing in the fingerprint characteristic of client upload and the database is compared, calculate characteristic similarity between the two.This part function realizes by following three the processing stage:
(1) fingerprint alignment
For matching algorithm is had rotation and translation stability, in order to narrate convenient hypothesis point set Vi={v1, v2 ..., Vn} is the unique point set of input fingerprint, wherein vi=(x, y, θ), Tj={t1, t2 ... tm} is that the unique point of template fingerprint is gathered.In order to realize the integral body alignment of Vi and Tj, at first to the core point alignment, after the alignment, the coordinate of Vi point set is done as down conversion:
X ' i=x i+ (x To-x Io) and y ' i=y i+ (y To-y Io), (x wherein To, y To) and (x Io, y Io) difference template fingerprint and the core point of importing fingerprint.And then the rotation alignment, the rotation alignment is then by asking a reflection transformation t to obtain:
t ( v ) = cos a - sin a &Delta;x sin a cos a &Delta;y 0 0 1 x y 1
In above-mentioned conversion, because carried out the core point alignment, so Δ x and Δ y determine, in fact only need to determine that again anglec of rotation a gets final product, and a is by making expression formula
| v t-t (v i) | 2<θ sets up, and asks Vi and Vt match point logarithm maximum to obtain, and wherein θ is a preset threshold, is set to 0.05.
(2) similarity is calculated
Mate for global characteristics, the feature of its input fingerprint and template fingerprint is the fourier spectrum feature, proper vector is exactly spectrogram pixel serialization result, this moment, similarity between the two was not wait the Euclidean distance between the long vector to realize by calculating, and this moment, similarity was calculated by following formula
s = ( 1 - dis max ( v i , v t ) ) &times; 100 %
In the formula, dis is both Euclidean distance values, max (v i, v t) expression gets in input vector and the template vector maximum vector length.
For the similarity between the minutia point set, be to be benchmark with the template, will import point set Vi by (a) conversion projects to after template point set Vt goes up for Δ x, Δ y, calculate distance error less than 5 location of pixels in,
All match points determine that to quantity the similarity value of this moment is calculated by following formula:
s = n couple max ( v i , v t ) &times; 100 %
In the formula, n CoupleRepresent the successfully quantity of the point of pairing, max (v i, v t) still expression input point set and template point are concentrated the minutiae point number of maximum.
(3) fingerprint recognition result
Calculate after the similarity data S between input fingerprint and the template fingerprint, again with S and default recognition threshold T sRelatively, the T here sCan be adjusted according to different security requirements, if S 〉=T sJudge that then two fingerprints successfully mate, promptly come from same finger, otherwise judge that two fingerprints are not to come from same finger.
Second development interface module: the merit second development interface is provided, makes things convenient for the user to construct the new business logic.In this module, mainly provide following second development interface, these interface functions all are encapsulated in the dynamic link library of FIG_NN_API by name, and the user can programme and call.
Numbering The interface function function The interface function prototype
1 System initialization API_INI_SYS
2 Read finger print data API_READ_DATA
3 Fingerprint classification API_FIG_CLASSIFY
4 The calculated direction field API_GET_ORIENTATION
5 Fingerprint recognition API_RECOGNITION
Fingerprint authentication API_VERIFY
Its basic functions of fingerprint recognition system integrated platform software provided by the present invention provides reliable fingerprint recognition service, in this software be the process design handled of fingerprint recognition as shown in Figure 5.Generally, the single fingerprint recognition is handled and is divided into four relatively independent unit, is respectively: fingerprint collecting unit, fingerprint register unit, fingerprint matching unit, fingerprint storage unit.Fingerprint collecting unit: during the online acquisition living body finger print, this part function is realized by the fingerprint collecting equipment of special use, this kind equipment is many at present, the output of fingerprint collecting equipment generally is 256 grades of gray scale fingerprint images that resolution is not less than 500DPI (Dot Per Inch), and picture quality can be slightly variant with the difference of fingerprint equipment sometimes; And when off-line is gathered fingerprint, input then be the static fingerprint gray scale pictures of selecting by artificial supplementary mode.The fingerprint register unit: this part is identical with " fingerprint matching unit " some function.No matter be the living body finger print that online acquisition arrives, or the static fingerprint image of off-line input, all to pass through fingerprint pretreatment operation such as active zone is cut apart, crestal line enhancing, calculated direction field, refinement earlier, carry out fingerprint pattern classification, extraction feature, feature coding preservation again, thereby finish once complete fingerprint register process.The fingerprint matching unit: also there is a part of identical functions in this part with " fingerprint register unit ".For fingerprint image to be identified from the fingerprint capturer input, just as registering unit, still to carry out operations such as pre-service, binaryzation, refinement, classify, extract feature, feature coding then, in registered fingerprint template storehouse, search for comparison at last, and provide final fingerprint matching conclusion by fingerprint pattern.The fingerprint storage unit: be convenient management and assurance safety of data, the related data of native system all adopt the database mode storage.The function that mainly comprises is: basic data manipulation function such as data write, data retrieval, data are read, Data Update, data deletion.
Fingerprint recognition system integrated platform software provided by the present invention, algorithm wherein load the operation control flow and the fingerprint recognition flow process of corn module and match control flow shown in accompanying drawing 6 and 7.When loading algorithm, want earlier dynamic link libraries (DLL) file of selection algorithm correspondence, and then the attribute information of input algorithm title, DLL title, entrance function title, input, function return value scheduling algorithm, and then these information are preserved as database.When apolegamy fingerprint recognition flow process, determine process name earlier, the specific algorithm the processing stage of matching each for this flow process again strengthens stage, feature extraction phases, matching stage etc. as fingerprint.
Fingerprint recognition system integrated platform software provided by the present invention, integrated domestic and international famous fingerprint algorithm, algorithms library has comprised five big classes generally, be respectively " selection of fingerprint instrument ", " fingerprint image enhancing ", " field of direction calculating ", " crestal line refinement ", " fingerprint pattern classification ", " feature extraction ", " fingerprint comparison " etc., the following expression of all kinds of algorithms that comprise:
Fingerprint recognition system integrated platform software provided by the present invention both can block redundancy feature, and system configuration is become function single fingerprint identification software, also can be configured as the test experiments platform of newly-increased algorithm.In test experiments, the main performance of coming the newly-increased algoritic module of comprehensive assessment from the index of " feature extraction time ", " minutiae point false drop rate ", " false acceptance rate (FAR; False Accept Rate) ", " false rejection rate (FRR; False Reject Rate) ", " etc. error rate (EER; Equal Error Rate) " and " the runnability curve of acquisition (ROC, Receive OperatingCharacteristic Curve) " some quantifications like this.

Claims (6)

1, a kind of fingerprint recognition system, comprise fingerprint database, central server and client three parts, it is characterized in that, fingerprint database is mainly used to store the fingerprint characteristic information through after the numerical coding, central server mainly is responsible for the checking client legitimacy, is received fingerprint characteristic data, aspect ratio to loopback fingerprint recognition result, client mainly is responsible for gathering fingerprint, extracting and upload fingerprint characteristic data; Central server comprises legitimate verification module, system management module, feature comparing module, second development interface module, algorithm load-on module, flow custom module, monitoring module and identification service module, client comprises finger print acquisition module, fingerprint pretreatment module, characteristic extracting module and pattern classification module, data communication module connects central server and client, wherein:
The legitimate verification module: server carries out the rights of using checking to the client of request service, prevents that unauthorized user from proposing the fingerprint recognition services request to central server.
Finger print acquisition module: client off-line is gathered fingerprint image or online acquisition vital fingerprint image, wherein off-line is gathered the picture that fingerprint is mainly gathered BMP form and JPG form, the JPG formatted file carries out data decomposition to it earlier, decompose file from the isolated code table of packed data and quantization table, obtain view data in that it is carried out inverse discrete cosine transform;
The fingerprint pretreatment module: the fingerprint that client is gathered strengthens, and comprises the field of direction device of fingerprint effective coverage segmenting device and fingerprint and the device that utilizes the M-PCNN network that the later image of pre-service is carried out filtering;
Characteristic extracting module: global characteristics and the minutia of extracting fingerprint that client is gathered, the mode type that comprises fingerprint, client's end points, bifurcation position and coordinate figure thereof comprise the Fourier spectrum feature deriving means and the minutia extraction element that extract global characteristics;
The pattern classification module: main being responsible for classified to the pattern of fingerprint, and fingerprint pattern is divided into six classes: whirlpool, left side ring, right ring, dicyclo, arch form and cusped arch;
Feature comparing module: the fingerprint characteristic of storing in the fingerprint characteristic of client upload and the database is compared, calculate characteristic similarity between the two, comprise fingerprint alignment means and similarity calculation element;
Second development interface module: provide second development interface, for the user constructs the new business logic;
Algorithm load-on module: finish that new algoritic module loads, the sign of management algorithm module and interface message;
Flow custom module: safeguard existing flow process, newly-increased flow process, be each stage placement algorithm of flow process;
Monitoring module: supervise that each is online, the services request of offline client and ruuning situation;
Identification service module: handle the service of client-requested, mainly comprise fingerprint register, fingerprint recognition, fingerprint authentication and connection request;
System management module: the rudimentary algorithm apolegamy of the loading of management rudimentary algorithm, fingerprint recognition, the management of fingerprint client authorization, the assembling of fingerprint recognition application system and the assessment of algorithm operation result;
Data communication module: between client and server, establish a communications link, send and receive fingerprint characteristic data and fingerprint recognition object information.
2, fingerprint recognition system according to claim 1 is characterized in that, described second development interface module mainly is that the function of the second development interface in the following table will be provided:
Numbering The interface function function The interface function prototype 1 System initialization API_INI_SYS 2 Read finger print data API_READ_DATA 3 Fingerprint classification API_FIG_CLASSIFY 4 The calculated direction field API_GET_ORIENTATION 5 Fingerprint recognition API_RECOGNITION 6 Fingerprint authentication API_VERIFY
These interface functions are encapsulated in the user and can programme in the dynamic link library of the FIG_NN_API by name that calls.
3, a kind of method for controlling fingerprint identification is characterized in that, may further comprise the steps:
(1) client connects central server by communication interface, after the legitimate verification of client identity process, obtains corresponding services request mandate;
(2) gather finger print data to be identified, carry out pattern classification and rough handling;
(3) finger print data with the rough handling of step (2) process carries out the fingerprint pre-service, comprises following little step:
1. the fingerprint effective coverage is cut apart: fingerprint image is divided into the image block of a series of 16 * 16 non-intersections, and each piece is labeled as B (1,1) respectively, B (1,2) ..., B (i, j), utilize then following formula v (i, j)
v ( i , j ) = ( x 1 - x &OverBar; ) 2 + ( x 2 - x &OverBar; ) 2 + . . . + ( x n - x &OverBar; ) 2 N
Calculate the grey scale pixel value variance of each image block, wherein x nWith x represent respectively this figure fast in the gray-scale value of pixel, N represents the pixel quantity that comprises in the segment segmentation threshold v to be set θ=11.5, respectively with the variance yields and the v of each segment θCompare, if v (i, j)〉v θ, then (i j) is judged as effective finger-print region to this segment B, otherwise is judged as the fingerprint background;
2. the field of direction of fingerprint is calculated: calculate respectively segment B (i, j) in each pixel gradient G in the x and y direction xAnd G y, utilize formula d (i, j)
d ( i , j ) = 1 2 arctan ( &Sigma; i b &prime; = 1 w &Sigma; j b &prime; = 1 w 2 G x ( i b &prime; , j b &prime; ) G y ( i b &prime; , j b &prime; ) &Sigma; i b &prime; = 1 w &Sigma; j b &prime; = 1 w G x 2 ( i b &prime; , j b &prime; ) - G y 2 ( i b &prime; , j b &prime; ) )
(i, local direction j) is in the formula to calculate segment B The coordinate of pixel in the expression segment, w represents the pixel wide of segment, value like the segment local direction value that calculates is only got in four component values 0, π/4, π/3 and 3 π/4 recently, local direction d (the i that segment is final, j) ∈ { 0,4 π/4, π, 3 π/4}, the consistance feature correction field of direction result of calculation of direction of passage field, (i is in 5 * 5 neighborhood D scope j) at segment B, calculate its consistance value C (i, j)
C ( i , j ) = 1 24 &Sigma; ( i &prime; , j &prime; ) &Element; D | d ( i , j ) - d ( i &prime; , j &prime; ) | 2
Have in this formula
Here d=mod (d (i ', j ')-d (i, j)+2 π) if C (i, j)<0.35, then with segment B (i, local direction j) are adjusted into the most significant direction of local direction in the neighborhood D, otherwise segment B (i, local direction j) remains unchanged;
3. use the M-PCNN network that the later image of pre-service is carried out filtering;
(4) after treatment fingerprint image in the step (3) is carried out feature extraction, comprises that global characteristics extracts and the minutia extraction:
1. global characteristics is extracted as the Fourier spectrum feature: at first the fingerprint image with input is divided into 32 * 32 image block, and segment is done two dimensional discrete Fourier transform, and formula is
G ( m , n ) = 1 N &Sigma; i = 0 N - 1 &Sigma; K = 0 N - 1 ( g ( i , k ) exp ( - j 2 &pi; ( mi N + kn N ) ) )
In the formula, (m n) is the codomain coordinate of pixel, and (i k) is the frequency domain coordinate of pixel correspondence, and each subgraph piece has just obtained the fourier spectrum figure of full figure after Fourier transform, and it is quantized into the global characteristics vector of fingerprint by rule;
2. minutia is extracted, and comprises core point, bifurcation and end points;
(5) send the characteristic of the fingerprint image that is extracted through step (4) to central server;
(6) central server receives and decomposes the characteristic of uploading, and is stored in characteristic formation to be matched;
(7) the fingerprint characteristic comparing module is according to the matching algorithm searching database of the current apolegamy of system, and the similarity data that obtain compared in record at every turn;
(8) central server is determined current recognition result, and the loopback clients corresponding;
(9) client receives result, and carries out subsequent action in view of the above;
Wherein in the step (1), the treatment step of the services request of central server customer in response end is as follows:
1. communication module receives the fingerprint recognition services request from client, and verifies its legitimacy; 2. central server sends request by communication module to client and allows signal, illustrates that current request is is checked and approved, and can send fingerprint characteristic data.
4, the control method of fingerprint recognition according to claim 3 is characterized in that, the middle M-PCNN network of step (3) carries out filtering to the later image of pre-service and may further comprise the steps:
1. during initialization, according to size of images structure PCNN network, and the PCNN neuron is corresponding one by one with the image slices vegetarian refreshments;
2. under the cooperation of the field of direction and segmentation result, allow the PCNN operation, and constantly revise the neuronic load signal strength of igniting, when neuronic activity during greater than given threshold value, neuron is lighted a fire, and the gray-scale value of pixel is upgraded;
3. set the neuron operation rule, allow PCNN constantly move, the neuron of lighting a fire in the network triggers grey scale pixel value is made amendment, and when the PCNN network stabilization, has just obtained the later image of filtered enhancing from each neuronic load signal value.
5, the control method of fingerprint recognition according to claim 3 is characterized in that, in the minutia extraction step, its core point is on the basis of direction of fingerprint field in the step (4), utilizes the poincare method to extract this value Wherein Δ (k) meets
&Delta; ( k ) = &delta; ( k ) if | &delta; ( k ) | < &pi; 2 &pi; + &delta; ( k ) , if&delta; ( k ) &le; - &pi; 2 &pi; - &delta; ( k ) , otherwise ,
δ(k)=O′(Ψ x(i′),Ψ y(i′))-O′(Ψ x(i),Ψ y(i)),
i′=(i+1)modN Ψ
After calculate finishing, judge again p (i, value j) is if (i j)=0.5, then is a core point to P; The extraction of bifurcation and end points detects by the sleiding form method and obtains, and at first defines the dot structure template of bifurcation and end points, respectively with this template by from left to right, from top to bottom order traversal full figure.The every slip of template once, just the matching degree of calculation template and image corresponding region, when meeting value greater than 0.7 the time, the minutia corresponding with using template with regard to the current pixel of process decision chart picture is consistent, if promptly meet the bifurcated template, then this pixel is a bifurcation, if meet the end points template, then this pixel is exactly an end points.
6, the control method of fingerprint recognition according to claim 3 is characterized in that, in the step (7), the fingerprint characteristic contrast may further comprise the steps:
1. fingerprint alignment: input fingerprint feature point set Vi={v1, v2 ..., Vn}, wherein vi=(x, y, θ), Tj={t1, t2 ... tm} is the unique point set of template fingerprint, and at first to the core point alignment, after the alignment, the coordinate of Vi point set is done as down conversion:
X ' i=x i+ (x To-x Io) and y ' i=y i+ (y To-y Io),
(x wherein To, y To) and (x Io, y Io) difference template fingerprint and the core point of importing fingerprint; And then the rotation alignment, the rotation alignment is then by asking a reflection transformation t to obtain:
t ( v ) = cos a - sin a &Delta;x sin a cos a &Delta;y 0 0 1 x y 1
In above-mentioned conversion, the core point alignment, Δ x and Δ y determine that a is by making expression formula | v t-t (v i) | 2<θ sets up, and asks Vi and Vt match point logarithm maximum to obtain, and wherein θ is a preset threshold, is set to 0.05;
2. similarity is calculated: mate for global characteristics, the feature of its input fingerprint and template fingerprint is the fourier spectrum feature, proper vector is exactly spectrogram pixel serialization result, similarity between the two is not wait the Euclidean distance between the long vector to realize by calculating, and wherein similarity is calculated by following formula:
s = ( 1 - dis max ( v i , v t ) ) &times; 100 %
In the formula, dis is both Euclidean distance values, max (v i, v t) expression gets in input vector and the template vector maximum vector length; For the minutia coupling, be to be benchmark with the template, will import point set Vi by (Δ x, Δ y, a) conversion projects to after template point set Vt goes up, calculate distance error less than 5 location of pixels in, all match points determine that to quantity the similarity value of this moment is calculated by following formula:
s = n couple max ( v i , v t ) &times; 100 %
In the formula, n CoupleRepresent the successfully quantity of the point of pairing, max (v i, v t) still expression input point set and template point are concentrated the minutiae point number of maximum;
3. fingerprint recognition: calculate after the similarity data S between input fingerprint and the template fingerprint, again with S and default recognition threshold T sRelatively, the T here sAdjusted according to different security requirements, if S 〉=T sJudge that then two fingerprints successfully mate, promptly come from same finger, otherwise judge that two fingerprints are not to come from same finger.
CNA2008100464373A 2008-11-03 2008-11-03 Fingerprint recognition system and control method Pending CN101414351A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNA2008100464373A CN101414351A (en) 2008-11-03 2008-11-03 Fingerprint recognition system and control method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNA2008100464373A CN101414351A (en) 2008-11-03 2008-11-03 Fingerprint recognition system and control method

Publications (1)

Publication Number Publication Date
CN101414351A true CN101414351A (en) 2009-04-22

Family

ID=40594882

Family Applications (1)

Application Number Title Priority Date Filing Date
CNA2008100464373A Pending CN101414351A (en) 2008-11-03 2008-11-03 Fingerprint recognition system and control method

Country Status (1)

Country Link
CN (1) CN101414351A (en)

Cited By (54)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101958792A (en) * 2009-07-17 2011-01-26 华为技术有限公司 Method and device for authenticating finger print of user
CN102045162A (en) * 2009-10-16 2011-05-04 电子科技大学 Personal identification system of permittee with tri-modal biometric characteristic and control method thereof
CN102354367A (en) * 2011-09-23 2012-02-15 杭州晟元芯片技术有限公司 Fingerprint identification card and fingerprint identification method running on card
CN101719272B (en) * 2009-11-26 2012-07-04 上海大学 Three-dimensional image segmentation method based on three-dimensional improved pulse coupled neural network
CN102708364A (en) * 2012-05-31 2012-10-03 西安电子科技大学 Cascade-classifier-based fingerprint image classification method
CN102737231A (en) * 2012-05-30 2012-10-17 深圳光启创新技术有限公司 Visible light communication-based fingerprint verification system and method
CN103220265A (en) * 2012-01-19 2013-07-24 西门子公司 Industrial automation system and method for safeguarding the system
WO2013139221A1 (en) * 2012-03-19 2013-09-26 腾讯科技(深圳)有限公司 Authentication method, device and system based on biological characteristics
CN103368745A (en) * 2013-07-19 2013-10-23 江南大学 User identity strong authentication method for educational information resource guarantee
CN103473492A (en) * 2013-09-05 2013-12-25 北京百纳威尔科技有限公司 Method and user terminal for recognizing permission
CN103617384A (en) * 2013-11-27 2014-03-05 广州御银科技股份有限公司 Method and system for verifying starting permissions of bill sorting machine
CN103701774A (en) * 2013-12-03 2014-04-02 乐视网信息技术(北京)股份有限公司 Login method and device
CN103930893A (en) * 2012-07-13 2014-07-16 Befs有限公司 Portable storage device using fingerprint recognition and method for controlling same
CN103955496A (en) * 2014-04-18 2014-07-30 大连恒锐科技股份有限公司 Fast field tire trace pattern retrieval algorithm
CN103995889A (en) * 2014-06-03 2014-08-20 广东欧珀移动通信有限公司 Method and device for classifying pictures
WO2014166069A1 (en) * 2013-04-09 2014-10-16 Jian Lian Informatized operation recording system self-contained in urban rail transit system device
CN104113869A (en) * 2014-06-20 2014-10-22 北京拓明科技有限公司 Signaling data-based prediction method and system for potential complaint user
CN104282058A (en) * 2014-08-08 2015-01-14 厦门华数电力科技有限公司 Unlocking method of Bluetooth-based safety intelligent lock system with video monitoring function
TWI473024B (en) * 2011-12-23 2015-02-11 Nat Inst Chung Shan Science & Technology Life fingerprint authenticity verification technology
CN104573462A (en) * 2015-01-30 2015-04-29 北京得意音通技术有限责任公司 Fingerprint and voiceprint dual-authentication method and fingerprint and voiceprint collector
CN104679413A (en) * 2015-03-09 2015-06-03 广东欧珀移动通信有限公司 Control method and control device for playing music
CN104881667A (en) * 2014-02-28 2015-09-02 阿里巴巴集团控股有限公司 Characteristic information extraction method and apparatus
CN105205442A (en) * 2015-08-07 2015-12-30 北京天诚盛业科技有限公司 Fingerprint acquisition method and device
CN105488375A (en) * 2015-11-27 2016-04-13 胡金钱 Fingerprint verification seal and seal fingerprint verification system and method
CN105654026A (en) * 2015-07-16 2016-06-08 宇龙计算机通信科技(深圳)有限公司 Fingerprint storage method and apparatus thereof, fingerprint identification method and apparatus thereof
US9367728B2 (en) 2013-12-05 2016-06-14 Egis Technology Inc. Fingerprint recognition method and device thereof
CN105678059A (en) * 2015-12-18 2016-06-15 延边大学 Numerical calculation system
CN105814904A (en) * 2014-11-07 2016-07-27 华为技术有限公司 Media resource sharing method and device
CN105894619A (en) * 2015-08-11 2016-08-24 乐视体育文化产业发展(北京)有限公司 Bicycle unlocking method and device and bicycle
CN106055636A (en) * 2016-05-30 2016-10-26 长江岩土工程总公司(武汉) Portable intelligent identification method for rock
CN106295483A (en) * 2015-06-11 2017-01-04 联想(北京)有限公司 A kind of information processing method and electronic equipment
CN106529614A (en) * 2016-09-08 2017-03-22 河源市公安局 Biological feature collection method and terminal
WO2017054285A1 (en) * 2015-09-30 2017-04-06 宇龙计算机通信科技(深圳)有限公司 Fingerprint recognition method and mobile terminal
CN106709450A (en) * 2016-12-23 2017-05-24 上海斐讯数据通信技术有限公司 Recognition method and system for fingerprint images
CN106843913A (en) * 2017-03-29 2017-06-13 百帝安(北京)科技有限公司 Small area fingerprint recognition flush bonding module interface
CN106888207A (en) * 2017-02-21 2017-06-23 中国联合网络通信集团有限公司 Authentication method, system and SIM
CN107357553A (en) * 2017-07-11 2017-11-17 郑州云海信息技术有限公司 A kind of data capture management method and device
CN107392113A (en) * 2017-06-28 2017-11-24 深圳森阳环保材料科技有限公司 A kind of fingerprint recognition device and storage device
CN107563842A (en) * 2017-08-08 2018-01-09 上海斐讯数据通信技术有限公司 A kind of ticket-booking system and booking method based on fingerprint recognition
CN107615299A (en) * 2016-03-16 2018-01-19 指纹卡有限公司 For assessing the method and system of fingerprint template
CN108197611A (en) * 2018-02-05 2018-06-22 广东职业技术学院 A kind of humanoid robot pattern recognition system
CN108255946A (en) * 2017-12-13 2018-07-06 东莞市奥海科技股份有限公司 Coding management method and device, equipment and the storage medium of a kind of laser carving equipment
CN108304483A (en) * 2017-12-29 2018-07-20 东软集团股份有限公司 A kind of Web page classification method, device and equipment
CN108681284A (en) * 2018-05-23 2018-10-19 文丹 A kind of intelligent home control system based on fingerprint recognition
CN108921126A (en) * 2018-07-20 2018-11-30 北京开普云信息科技有限公司 A kind of automatic identification signature stamp or the method and device of handwritten signature
CN109711546A (en) * 2018-12-21 2019-05-03 深圳市商汤科技有限公司 Neural network training method and device, electronic equipment and storage medium
CN109766738A (en) * 2017-11-09 2019-05-17 比亚迪股份有限公司 A kind of fingerprint identification method and device, computer readable storage medium
CN109936584A (en) * 2017-12-15 2019-06-25 天津铂创国茂电子科技发展有限公司 The method of fingerprint recognition and speech recognition based on cloud branch server
CN109948430A (en) * 2019-01-28 2019-06-28 北京旷视科技有限公司 Characteristic sharing method, device, system, processing end and storage medium
CN109977909A (en) * 2019-04-04 2019-07-05 山东财经大学 Finger vein identification method and system based on minutiae point Region Matching
CN110427876A (en) * 2019-07-31 2019-11-08 中国工商银行股份有限公司 Refer to vein identification method, device and computer system
CN111123327A (en) * 2019-12-31 2020-05-08 广东轻工职业技术学院 Positioning and tracking system based on Beidou satellite navigation
CN112019977A (en) * 2020-09-04 2020-12-01 广州郝舜科技有限公司 Audio acquisition device for big data acquisition
TWI730589B (en) 2020-01-14 2021-06-11 大陸商北京集創北方科技股份有限公司 Optical living body feature detection method and biological feature acquisition device and information processing device using the same

Cited By (76)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101958792A (en) * 2009-07-17 2011-01-26 华为技术有限公司 Method and device for authenticating finger print of user
CN102045162A (en) * 2009-10-16 2011-05-04 电子科技大学 Personal identification system of permittee with tri-modal biometric characteristic and control method thereof
CN101719272B (en) * 2009-11-26 2012-07-04 上海大学 Three-dimensional image segmentation method based on three-dimensional improved pulse coupled neural network
CN102354367A (en) * 2011-09-23 2012-02-15 杭州晟元芯片技术有限公司 Fingerprint identification card and fingerprint identification method running on card
TWI473024B (en) * 2011-12-23 2015-02-11 Nat Inst Chung Shan Science & Technology Life fingerprint authenticity verification technology
CN103220265A (en) * 2012-01-19 2013-07-24 西门子公司 Industrial automation system and method for safeguarding the system
CN103220265B (en) * 2012-01-19 2018-03-30 西门子公司 Industrial automation system and the method protected to it
US10664581B2 (en) 2012-03-19 2020-05-26 Tencent Technology (Shenzhen) Company Limited Biometric-based authentication method, apparatus and system
WO2013139221A1 (en) * 2012-03-19 2013-09-26 腾讯科技(深圳)有限公司 Authentication method, device and system based on biological characteristics
US10108792B2 (en) 2012-03-19 2018-10-23 Tencent Technology (Shenzhen) Company Limited Biometric-based authentication method, apparatus and system
CN102737231B (en) * 2012-05-30 2016-04-27 深圳光启智能光子技术有限公司 Based on fingerprint verification system and the method for visible light communication
CN102737231A (en) * 2012-05-30 2012-10-17 深圳光启创新技术有限公司 Visible light communication-based fingerprint verification system and method
CN102708364A (en) * 2012-05-31 2012-10-03 西安电子科技大学 Cascade-classifier-based fingerprint image classification method
CN103930893A (en) * 2012-07-13 2014-07-16 Befs有限公司 Portable storage device using fingerprint recognition and method for controlling same
WO2014166069A1 (en) * 2013-04-09 2014-10-16 Jian Lian Informatized operation recording system self-contained in urban rail transit system device
CN104871178A (en) * 2013-04-09 2015-08-26 简炼 Informatized operation recording system self-contained in urban rail transit system device
CN103368745A (en) * 2013-07-19 2013-10-23 江南大学 User identity strong authentication method for educational information resource guarantee
CN103473492B (en) * 2013-09-05 2016-11-02 北京百纳威尔科技有限公司 Authority recognition method and user terminal
CN103473492A (en) * 2013-09-05 2013-12-25 北京百纳威尔科技有限公司 Method and user terminal for recognizing permission
CN103617384A (en) * 2013-11-27 2014-03-05 广州御银科技股份有限公司 Method and system for verifying starting permissions of bill sorting machine
CN103617384B (en) * 2013-11-27 2017-01-04 广州御银科技股份有限公司 A kind of cleaning-sorting machine start method for verifying authority and system
CN103701774A (en) * 2013-12-03 2014-04-02 乐视网信息技术(北京)股份有限公司 Login method and device
US9367728B2 (en) 2013-12-05 2016-06-14 Egis Technology Inc. Fingerprint recognition method and device thereof
CN104881667A (en) * 2014-02-28 2015-09-02 阿里巴巴集团控股有限公司 Characteristic information extraction method and apparatus
CN103955496A (en) * 2014-04-18 2014-07-30 大连恒锐科技股份有限公司 Fast field tire trace pattern retrieval algorithm
CN103995889B (en) * 2014-06-03 2017-11-03 广东欧珀移动通信有限公司 Picture classification method and device
CN103995889A (en) * 2014-06-03 2014-08-20 广东欧珀移动通信有限公司 Method and device for classifying pictures
CN104113869B (en) * 2014-06-20 2017-12-22 北京拓明科技有限公司 A kind of potential report user's Forecasting Methodology and system based on signaling data
CN104113869A (en) * 2014-06-20 2014-10-22 北京拓明科技有限公司 Signaling data-based prediction method and system for potential complaint user
CN104282058B (en) * 2014-08-08 2016-08-24 厦门华数电力科技有限公司 The method for unlocking of the safety intelligent lock system of band video monitoring based on bluetooth
CN104282058A (en) * 2014-08-08 2015-01-14 厦门华数电力科技有限公司 Unlocking method of Bluetooth-based safety intelligent lock system with video monitoring function
CN105814904A (en) * 2014-11-07 2016-07-27 华为技术有限公司 Media resource sharing method and device
CN105814904B (en) * 2014-11-07 2019-07-23 华为技术有限公司 Media resource sharing method and equipment
CN104573462B (en) * 2015-01-30 2017-05-03 北京得意音通技术有限责任公司 Fingerprint and voiceprint dual-authentication method
CN104573462A (en) * 2015-01-30 2015-04-29 北京得意音通技术有限责任公司 Fingerprint and voiceprint dual-authentication method and fingerprint and voiceprint collector
CN104679413A (en) * 2015-03-09 2015-06-03 广东欧珀移动通信有限公司 Control method and control device for playing music
CN106295483A (en) * 2015-06-11 2017-01-04 联想(北京)有限公司 A kind of information processing method and electronic equipment
CN106295483B (en) * 2015-06-11 2020-02-21 联想(北京)有限公司 Information processing method and electronic equipment
CN105654026A (en) * 2015-07-16 2016-06-08 宇龙计算机通信科技(深圳)有限公司 Fingerprint storage method and apparatus thereof, fingerprint identification method and apparatus thereof
WO2017008350A1 (en) * 2015-07-16 2017-01-19 宇龙计算机通信科技(深圳)有限公司 Fingerprint storage method and device, and fingerprint recognition method and device
CN105205442A (en) * 2015-08-07 2015-12-30 北京天诚盛业科技有限公司 Fingerprint acquisition method and device
CN105894619A (en) * 2015-08-11 2016-08-24 乐视体育文化产业发展(北京)有限公司 Bicycle unlocking method and device and bicycle
WO2017054285A1 (en) * 2015-09-30 2017-04-06 宇龙计算机通信科技(深圳)有限公司 Fingerprint recognition method and mobile terminal
CN105488375A (en) * 2015-11-27 2016-04-13 胡金钱 Fingerprint verification seal and seal fingerprint verification system and method
CN105678059A (en) * 2015-12-18 2016-06-15 延边大学 Numerical calculation system
CN107615299B (en) * 2016-03-16 2019-04-02 指纹卡有限公司 For assessing the method and system of fingerprint template
CN107615299A (en) * 2016-03-16 2018-01-19 指纹卡有限公司 For assessing the method and system of fingerprint template
CN106055636B (en) * 2016-05-30 2019-12-13 长江岩土工程总公司(武汉) portable intelligent rock identification method
CN106055636A (en) * 2016-05-30 2016-10-26 长江岩土工程总公司(武汉) Portable intelligent identification method for rock
CN106529614A (en) * 2016-09-08 2017-03-22 河源市公安局 Biological feature collection method and terminal
CN106709450A (en) * 2016-12-23 2017-05-24 上海斐讯数据通信技术有限公司 Recognition method and system for fingerprint images
CN106888207B (en) * 2017-02-21 2020-02-21 中国联合网络通信集团有限公司 Authentication method, system and SIM card
CN106888207A (en) * 2017-02-21 2017-06-23 中国联合网络通信集团有限公司 Authentication method, system and SIM
CN106843913A (en) * 2017-03-29 2017-06-13 百帝安(北京)科技有限公司 Small area fingerprint recognition flush bonding module interface
CN107392113B (en) * 2017-06-28 2021-01-22 南京鼎峯云信息科技有限公司 Fingerprint identification device and storage device
CN107392113A (en) * 2017-06-28 2017-11-24 深圳森阳环保材料科技有限公司 A kind of fingerprint recognition device and storage device
CN107357553A (en) * 2017-07-11 2017-11-17 郑州云海信息技术有限公司 A kind of data capture management method and device
CN107563842A (en) * 2017-08-08 2018-01-09 上海斐讯数据通信技术有限公司 A kind of ticket-booking system and booking method based on fingerprint recognition
CN109766738A (en) * 2017-11-09 2019-05-17 比亚迪股份有限公司 A kind of fingerprint identification method and device, computer readable storage medium
CN108255946A (en) * 2017-12-13 2018-07-06 东莞市奥海科技股份有限公司 Coding management method and device, equipment and the storage medium of a kind of laser carving equipment
CN109936584A (en) * 2017-12-15 2019-06-25 天津铂创国茂电子科技发展有限公司 The method of fingerprint recognition and speech recognition based on cloud branch server
CN108304483B (en) * 2017-12-29 2021-01-19 东软集团股份有限公司 Webpage classification method, device and equipment
CN108304483A (en) * 2017-12-29 2018-07-20 东软集团股份有限公司 A kind of Web page classification method, device and equipment
CN108197611A (en) * 2018-02-05 2018-06-22 广东职业技术学院 A kind of humanoid robot pattern recognition system
CN108681284A (en) * 2018-05-23 2018-10-19 文丹 A kind of intelligent home control system based on fingerprint recognition
CN108921126A (en) * 2018-07-20 2018-11-30 北京开普云信息科技有限公司 A kind of automatic identification signature stamp or the method and device of handwritten signature
CN108921126B (en) * 2018-07-20 2019-07-23 北京开普云信息科技有限公司 A kind of automatic identification signature stamp or the method and device of handwritten signature
CN109711546A (en) * 2018-12-21 2019-05-03 深圳市商汤科技有限公司 Neural network training method and device, electronic equipment and storage medium
CN109711546B (en) * 2018-12-21 2021-04-06 深圳市商汤科技有限公司 Neural network training method and device, electronic equipment and storage medium
CN109948430A (en) * 2019-01-28 2019-06-28 北京旷视科技有限公司 Characteristic sharing method, device, system, processing end and storage medium
CN109977909A (en) * 2019-04-04 2019-07-05 山东财经大学 Finger vein identification method and system based on minutiae point Region Matching
CN110427876A (en) * 2019-07-31 2019-11-08 中国工商银行股份有限公司 Refer to vein identification method, device and computer system
CN111123327B (en) * 2019-12-31 2021-04-30 广东轻工职业技术学院 Positioning and tracking system based on Beidou satellite navigation
CN111123327A (en) * 2019-12-31 2020-05-08 广东轻工职业技术学院 Positioning and tracking system based on Beidou satellite navigation
TWI730589B (en) 2020-01-14 2021-06-11 大陸商北京集創北方科技股份有限公司 Optical living body feature detection method and biological feature acquisition device and information processing device using the same
CN112019977A (en) * 2020-09-04 2020-12-01 广州郝舜科技有限公司 Audio acquisition device for big data acquisition

Similar Documents

Publication Publication Date Title
CN101414351A (en) Fingerprint recognition system and control method
AU2009246750B2 (en) Fingerprint representation using gradient histograms
KR101897709B1 (en) System and method for user access of dispensing unit
US7519558B2 (en) Biometrically enabled private secure information repository
CN105654027B (en) A kind of method and device of fingerprint recognition
CN101809582B (en) Smart identity system
CN101571920A (en) Biometric authentication system, authentication client terminal, and biometric authentication method
US20190354736A1 (en) Identification verification using a device with embedded radio-frequency identification functionality
CN105160302B (en) Multimodal Biometrics general-purpose platform and identity identifying method
CN105069622A (en) Face identification payment system facing mobile terminal and face identification payment method facing mobile terminal
CN103886283A (en) Method for fusing multi-biometric image information for mobile user and application thereof
CN108875341A (en) A kind of face unlocking method, device, system and computer storage medium
CN106228133A (en) User authentication method and device
Omri et al. Cloud-based mobile system for biometrics authentication
Zhang et al. Android-based multimodal biometric identification system using feature level fusion
CN109670788A (en) Medical insurance checking method, device, equipment and storage medium based on data analysis
US20200293760A1 (en) Multi-modal identity recognition
CN107742140B (en) Intelligent identity information identification method based on RFID technology
CN105184236A (en) Robot-based face identification system
WO2003025718A2 (en) Private secure information repository with biometric authentication
CN107086985A (en) Fingerprint authority checking method and system
Dalwadi et al. Face recognition-based attendance system using real-time computer vision algorithms
CN107615299A (en) For assessing the method and system of fingerprint template
Siddiqui et al. Development of Secure Multimodal Biometric System for Person Identification Using Feature Level Fusion: Fingerprint and Iris
CN208506767U (en) A kind of intelligent cabinet fingerprint recognition system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Open date: 20090422