CN101599121A - The authenticating colorized face images system and method - Google Patents
The authenticating colorized face images system and method Download PDFInfo
- Publication number
- CN101599121A CN101599121A CNA2009101084338A CN200910108433A CN101599121A CN 101599121 A CN101599121 A CN 101599121A CN A2009101084338 A CNA2009101084338 A CN A2009101084338A CN 200910108433 A CN200910108433 A CN 200910108433A CN 101599121 A CN101599121 A CN 101599121A
- Authority
- CN
- China
- Prior art keywords
- face
- colorized
- face images
- people
- authenticating
- 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
Links
Images
Abstract
The invention discloses a kind of authenticating colorized face images system and method.Described authenticating colorized face images system comprises: network shooting module, facial image conversion module and face recognition module; Described authenticating colorized face images method comprises: the network shooting module of the first step, seizure colorized face images; Second the step, utilize the PCA method that colorized face images is transformed to new two-dimentional uncorrelated color space; The 3rd goes on foot, discerns the facial image in the new two-dimentional uncorrelated color space.Technical solution of the present invention authenticating colorized face images system and method is contactless to the image acquisition of people's face, recognition of face has intuitive, passivity and non-infringement, be a kind of more natural, directly, a kind of recognition method easily, the precision height of recognition result simultaneously.
Description
[technical field]
The present invention relates to authentication and computer realm, particularly a kind of authenticating colorized face images system and method.
[background technology]
Man face image acquiring is contactless, recognition of face has intuitive, passivity and non-infringement, people's face is the important channel that people distinguish different people simultaneously, it is topmost information source, therefore face authentication be a kind of more natural than identifications such as fingerprint, palmmprint, vein and irises, directly, a kind of recognition method easily, in numerous identifications field such as identity is differentiated, credit card identification, passport are checked, man-machine interaction is widely adopted with aspects such as, supervisory system, the system of registering, system's debarkation authentications.
In the recognition of face field, colorized face images can provide more useful information than gray level image.Proposed a lot of colour space transformation methods at present, but these methods are not considered the correlativity between each composition of color space after the conversion.For the new two-dimentional uncorrelated color space that some conversion obtain, recognition result awaits to improve.
[summary of the invention]
For the recognition result out of true of the identification system that overcomes prior art and the loaded down with trivial details property of computers log-on system, the invention provides a kind of authenticating colorized face images system and method.
The present invention solves the technical scheme that the prior art problem adopts: a kind of authenticating colorized face images system is provided.Described authenticating colorized face images system comprises: the network shooting module that is used to catch colorized face images; Be used to utilize PCA (Principle Component Analysis, principal component analysis) method colorized face images to be transformed to the facial image conversion module of new two-dimentional uncorrelated color space; And be used for discerning the face recognition module of the facial image of new two-dimentional uncorrelated color space; Described network shooting module, facial image conversion module and face recognition module connect successively.
According to authenticating colorized face images provided by the invention system one optimal technical scheme be: described authenticating colorized face images system further comprises: be used for detecting the people's face detection module whether image that the network shooting module catches has people's face.
According to authenticating colorized face images provided by the invention system one optimal technical scheme be: described authenticating colorized face images system further comprises: be used for the colorized face images that captures is cut apart and normalized pretreatment module.
According to authenticating colorized face images provided by the invention system one optimal technical scheme be: described face recognition module comprises: the characteristic extracting module that is used for the facial image of new two-dimentional uncorrelated color space is done LDA (LinearDiscriminant Analysis, linear discriminant analysis) feature extraction; And the face characteristic that is used for the face characteristic that will extract and face database people's face comparing module of comparing.
The present invention also provides a kind of authenticating colorized face images method.Described authenticating colorized face images method may further comprise the steps: the first step, seizure colorized face images; Second the step, utilize the PCA method that colorized face images is transformed to new two-dimentional uncorrelated color space; The 3rd goes on foot, discerns the facial image in the new two-dimentional uncorrelated color space.
According to authenticating colorized face images method one optimal technical scheme provided by the invention be: described authenticating colorized face images method further comprises: the step whether people's face is arranged in the image that detection network shooting module is caught.
According to authenticating colorized face images method one optimal technical scheme provided by the invention be: described authenticating colorized face images method further comprises: the colorized face images that captures is cut apart and normalized step.
According to authenticating colorized face images method one optimal technical scheme provided by the invention be: the described first step is specially: start the network shooting module, this module of initialization, set this module mode of operation, whether detect has people's face to occur in the viewfinder range, if nobody's face occurs, cycle detection then, up to there being people's face to occur, go out people's face with the rectangle circle, rectangle frame is according to people's little self size of adjusting automatically of being bold, according to default collection quantity k continuous shooting k time, the time interval between adjacent twice is s second, in the continuous shooting process, to confirm all whether people's face is arranged in the viewfinder range before taking each time, if people's face shifts out viewfinder range and causes detecting less than people's face then stop to take, keep the state of cycle detection, when retracting once more, continues people's face to take.
According to authenticating colorized face images method one optimal technical scheme provided by the invention be: described second step is specially: the colorized face images that comprises m pixel for a width of cloth, each pixel is regarded as a sample, the vector representation of 3*1 of this sample, total m*n sample of colorized face images that the n width of cloth is such, with the principal component analysis method to two main compositions of these sample extraction, obtain the vector that new sample is 2*1, between each dimension of the new sample that obtains is incoherent, the corresponding two-dimentional uncorrelated color space of the facial image of new composition of sample.
According to authenticating colorized face images method one optimal technical scheme provided by the invention be: the described step 3 step comprises following substep: at first, the facial image in the new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; Secondly, the face characteristic of extraction and the face characteristic in the face database are compared.
The present invention also provides a kind of computers log-on system.Described computers log-on system comprises: be used to register people's face Registering modules of face characteristic and be used for people's face authentication module that the face characteristic with current face characteristic and registration compares; Described people's face authentication module joins in the Gina dynamic link libraries.
According to computers log-on provided by the invention system one optimal technical scheme be: described people's face authentication module comprises: the network shooting module that is used to catch colorized face images; Be used to utilize the PCA method colorized face images to be transformed to the facial image conversion module of new two-dimentional uncorrelated color space; And be used for discerning the face recognition module of the facial image of new two-dimentional uncorrelated color space; Described network shooting module, facial image conversion module and face recognition module connect successively.
According to computers log-on provided by the invention system one optimal technical scheme be: described people's face authentication module further comprises: be used for detecting the people's face detection module whether image that the network shooting module catches has people's face.
According to computers log-on provided by the invention system one optimal technical scheme be: described people's face authentication module further comprises: be used for the colorized face images that captures is cut apart and normalized pretreatment module.
According to computers log-on provided by the invention system one optimal technical scheme be: described people's face authentication module comprises: the characteristic extracting module that is used for the facial image of new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; And the face characteristic that is used for the face characteristic that will extract and face database people's face comparing module of comparing.
According to computers log-on provided by the invention system one optimal technical scheme be: described people's face Registering modules comprises: the network shooting module that is used to catch colorized face images; Whether be used for detecting image that the network shooting module catches has people's face detection module of people's face; Be used for the colorized face images that captures is cut apart and normalized pretreatment module; Be used to utilize the PCA method colorized face images to be transformed to the facial image conversion module of new two-dimentional uncorrelated color space; Be used for the facial image of new two-dimentional uncorrelated color space is done the characteristic extracting module of linear discriminant analysis feature extraction and is used to store the face database module of facial image feature.
The present invention also provides a kind of computers log-on method.Described computers log-on method may further comprise the steps: the first step, registered user's face characteristic information; Second step, user's face characteristic of current face characteristic and registration is compared.
According to computers log-on method one optimal technical scheme provided by the invention be: described second step comprises following substep: at first, catch colorized face images; Secondly, utilize the PCA method that colorized face images is transformed to new two-dimentional uncorrelated color space; At last, the facial image in the new two-dimentional uncorrelated color space of identification.
According to computers log-on provided by the invention system one optimal technical scheme be: described second step further comprises: detect the step whether people's face is arranged in the image that the network shooting module catches.
According to computers log-on provided by the invention system one optimal technical scheme be: described second step further comprises: the colorized face images that captures is cut apart and normalized step.
According to computers log-on provided by the invention system one optimal technical scheme be: the step of the facial image in the new two-dimentional uncorrelated color space of described identification comprises following substep: at first, the facial image in the new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; Secondly, the face characteristic of extraction and the face characteristic in the face database are compared.
According to computers log-on provided by the invention system one optimal technical scheme be: the described first step further comprises following substep: at first, catch colorized face images; Secondly. in the image that detection network shooting module is caught whether people's face is arranged; Once more, the colorized face images that captures is cut apart and normalization; Then, utilize the PCA method that colorized face images is transformed to new two-dimentional uncorrelated color space; Then, the facial image in the new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; At last, deposit the facial image feature in face database.
Beneficial technical effects of the present invention is: the man face image acquiring of authenticating colorized face images system and method and computers log-on system and method is contactless, recognition of face has intuitive, passivity and non-infringement, be a kind of more natural, directly, a kind of recognition method easily, the precision height of recognition result simultaneously.
[description of drawings]
Fig. 1 is a colorized face images Verification System structured flowchart in the embodiment of the invention;
Fig. 2 is a colorized face images authentication method process flow diagram in the embodiment of the invention;
Fig. 3 is the system chart of the colorized face images deriving means of computer login system and method in the embodiment of the invention.
Fig. 4 is the process flow diagram that lands of computer login system in the embodiment of the invention and method.
Fig. 5 is the login authentication sectional drawing of computer login system and method in the embodiment of the invention.
Fig. 6 is the face recognition process figure of the face authentication system of computer login system and method in the embodiment of the invention.
Fig. 7 is cut apart and the normalization synoptic diagram for people's face in the recognition of face of computer login system in the embodiment of the invention and method.
Fig. 8 is the colour space transformation synoptic diagram of computer login system and method in the embodiment of the invention.
Fig. 9 is the face authentication method synoptic diagram that the matching score layer of computer login system and method in the embodiment of the invention merges.
[embodiment]
The present invention is described in detail below in conjunction with drawings and Examples.
Please refer to Fig. 1, Fig. 1 is a colorized face images Verification System structured flowchart in the embodiment of the invention.Authenticating colorized face images system in the present embodiment comprises: the network shooting module that is used to catch colorized face images; Whether be used for detecting image that the network shooting module catches has people's face detection module of people's face; Be used for the colorized face images that captures is cut apart and normalized pretreatment module; Be used to utilize PCA (PrincipleComponent Analysis, principal component analysis) method colorized face images to be transformed to the facial image conversion module of new two-dimentional uncorrelated color space; And be used for discerning the face recognition module of the facial image of new two-dimentional uncorrelated color space; Described network shooting module, people's face detection module, pretreatment module, facial image conversion module and face recognition module connect successively.Described face recognition module comprises: the characteristic extracting module that is used for the facial image of new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; And the face characteristic that is used for the face characteristic that will extract and face database people's face comparing module of comparing.
Please refer to Fig. 2, Fig. 2 is a colorized face images authentication method process flow diagram in the embodiment of the invention.Authenticating colorized face images method in the present embodiment may further comprise the steps: the first step, seizure colorized face images; In the image that second step, detection network shooting module are caught whether people's face is arranged,, then carried out for the 3rd step,, then return the first step if do not have if having; The 3rd goes on foot, the colorized face images that captures is cut apart and normalization; The 5th the step, utilize the PCA method that colorized face images is transformed to new two-dimentional uncorrelated color space; Facial image in the 6th step, the new two-dimentional uncorrelated color space of identification: at first, the facial image in the new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; Secondly, the face characteristic of extraction and the face characteristic in the face database are compared.
The optimal technical scheme of present embodiment is: authenticating colorized face images system and method for the present invention is applied in the computers log-on system and method.
In the embodiment of the invention, by revising the Windows login interface, self-defined dll file is realized the computer login system based on authenticating colorized face images.
The user logins mainly and is managed by Winlogo and assembly Gina dynamic link libraries thereof in Windows, it is responsible for the login of process user and cancellation, start user SHELL (shell, the Windows acquiescence is used Explorer.exe), input password, change password, locking and release computing machine etc.In order to realize the process of people's face login, we mainly realize by revising Gina dynamic link library adding face recognition module.In the login safety certification, adopt the facial image login to substitute traditional Windows login, the user who does not carry out facial image registration can't use this computing machine, and the system lock meeting has only the registered user to pass through after the Information Authentication of people's face could release and use this computing machine.
Please refer to Fig. 3, Fig. 4, in example of the present invention, computer login system land flow process, whole process is broadly divided into 5 major parts.
After system start-up was execution in step 21, system process logon execution in step 22 was written into self-defining GINA, and it is that people's face authentication module is finished that the process of whole login is given self-defined assembly GINA at this moment.People's face proof procedure mainly is made up of four steps, comprise the initialization camera, be written into the step 231 that people's face detects sorter, the step 232 that camera image is obtained, the step 233 that human face region is cut apart and with the storehouse in the people's face data contrasts step 234 of carrying out the data verification of people's face.
The step 231 of module initialization mainly is some initial work of carrying out when module is written into by system at first, and it mainly comprises the startup IP Camera, sets the camera mode of operation so that the work of next step image acquisition; Should be written in addition and use the grouped data that detects with people's face to be used for cutting apart of step 233 human face region.
The step 232 that camera image is obtained is that camera collection is become the discernible picture format of system to data-switching, does corresponding image pre-service work then, carries out image gray processing at last and cuts apart with the human face region that is used for step 233.
After obtaining that the image of people's face is arranged, enter the step 233 that human face region is cut apart, people's face that utilization has trained detects sorter and calibrates human face region, cuts out human face region to be used for the identification of step 234 people face on the one hand; Be presented on the window of login on the one hand.
Please refer to Fig. 5, after starting login, authentication interface has been replaced traditional password login frame, replacement be the facial image that detects people's face.
After obtaining human face region, enter with the storehouse in the people's face data contrasts step 234 of carrying out the data verification of people's face, the face recognition algorithms of mentioning among utilization the present invention is carried out recognition of face, if people's face data and system user people face Data Matching then successfully change the login success status over to.
System state becomes after the successful step 241 of login, system shows desktop, the user can operate accordingly according to the authority of oneself, if the user selects to nullify computing machine (step 243) or the user does not carry out operating system locking (step 242) to computing machine for a long time, then system should enter and nullify or lock-out state, at this moment, if the user wants to reuse computing machine, people's face login component forwards the initialization of step 231 face recognition module again to, restarts the process of people's face login authentication.
Please refer to Fig. 6, in the embodiment of the invention, the face recognition process of face authentication system is general face recognition process.Whole implement process can be divided into two parts, comprises the step 41 of people's face registration and the step 42 of recognition of face.
First is the step 41 of people's face registration, and the registration of people's face comprises: the step 411 that people's face detects, and whether be used to detect has people's face and locatees people's face in the camera capture scope; The step 412 of gathering is used for photographic images; Pretreated step 413 is used for that people's face is cut apart and the normalization of people's face; The step 414 of feature extraction is used to extract the information that comprises in the facial image, and the people's face information that obtains and corresponding people's name (or numbering) and user class (or classification) is saved in the face database 40.
The step 41 of people's face registration judges in the camera capture scope by the step 411 that people's face detects whether people's face is arranged, standby rectangle frame people from location face when people's face occurs, if user's this moment asks to gather, then system carries out acquisition step 412 images acquired, the image that obtains through pretreated step 413 be people's face cut apart with the normalization of people's face after carry out feature extraction step 414, name (or numbering) and the user class (or classification) with the feature extracted and corresponding people is saved in the face database 40 at last.
Second portion is recognition of face, and recognition of face comprises: the step 421 that people's face detects, and purposes is with step 411; The step 422 of gathering, purposes is with step 412; Pretreated step 423, purposes is with step 413; The step 424 of feature extraction, purposes is with step 414; The step 425 of people's face comparison is used for calculating the distance between test face characteristic and all face characteristics of face database and obtains arest neighbors, exports recognition result 426 at last.
The step 42 of recognition of face judges in the camera capture scope by the step 421 that people's face detects whether people's face is arranged, standby rectangle frame people from location face when people's face occurs, if user's this moment asks to gather, then system carries out the step 422 of gathering, take all scenery in the camera capture scope, the image that obtains through pretreated step 423 be people's face cut apart with the normalization of people's face after carry out feature extraction step 424, carry out step 425 at last eigenface and the eigenface in the face database extracted are carried out the comparison of people's face, find the eigenface of arest neighbors, output recognition result 426; When with the distance of the eigenface of arest neighbors during less than certain threshold value, the name (or numbering) and the user class (or classification) of output arest neighbors eigenface during not less than certain threshold value, be can be regarded as the people's face authentication authorization and accounting that not have to mate and are failed.
Please refer to Fig. 7,51 is the image that collects; 52 for cutting apart the facial image in the rectangle frame that obtains; 53 are normalized to image after the 56*52 size for picture shape.
The image 51 that collects goes out the interior facial image 52 of rectangle frame through segmented extraction, and the normalization of process shape obtains the image 53 of 56*52 size again; Can remove unnecessary background information through over-segmentation, normalization can be transformed to same size with face images through shape, helps the contrast of subsequent step.
The step 414 of feature extraction was divided into for two steps among Fig. 6: the first step is a colour space transformation, as shown in Figure 8; Second step is for the linear discriminant analysis feature extraction, shown in the step 91 among Fig. 9.
In the colour space transformation process, the vector of at first color-values of each pixel in the RGB image all being regarded as a 3*1, each pixel is regarded a sample as, with the color space dimensionality reduction of PCA method to pixel, promptly extract two main compositions of sample, can obtain the vector of 2*1, corresponding image is two-dimentional uncorrelated color space, through all uncorrelated between the every twenty percent branch that pixel comprised after the PCA processing, concrete conversion process is as follows:
Suppose to have M width of cloth coloured image A
1, A
2..., A
MTraditional rgb color space representation is represented color with three compositions, and each pixel is all by the vector representation of a 3*1.
We realize the uncorrelated color space from original colour space transformation to two dimension in the following method.At first, for a figure coloured image A
jWhole N pixels use a respectively
J1, a
J2..., a
JNThe vector of 3*1 is represented, A
1, A
2..., A
MBe training sample, the covariance matrix that can define a 3*3 is as follows:
Wherein a is all a
Ji(j=1,2 ..., M, i=1,2 ..., average N).Solve G
tAll eigenwerts and proper vector.Suppose v
1And v
2Be the proper vector of respectively corresponding two eigenvalue of maximum, for each pixel a
Ji(j=1,2 ..., M, i=1,2 ..., N) can calculate first and second hue components by following two conversion formulas:
In a new two-dimentional uncorrelated color space, represent original three-dimensional color image with the result who obtains above.In new two-dimentional uncorrelated color space, image A originally
jUse B
jBe expressed as follows:
B
j＝(z
j1，z
j2，...z
jN)，
Z wherein
Ji, i=1,2 ... N is the vector of 2*1
For the sake of simplicity, below with two vector representation original image A
jExpression B in the two-dimensional space that newly obtains
j:
Obviously, X and Y are the vector of two N*1.
Please refer to Fig. 8, Fig. 8 is a recognition of face feature extraction first step colour space transformation in the embodiment of the invention, and wherein 53 is that width of cloth size is the coloured image of 56*52; 61 is that pixel is the vector of a 3*1 in the expression in original color space in the coloured image; 62 is the expression in the two-dimentional uncorrelated color space of a pixel behind colour space transformation.
The color-values of each pixel of one width of cloth coloured image 53 all can be regarded the vector 61 of a 3*1 as, and the original color space can obtain two-dimentional uncorrelated color space behind the PCA colour space transformation, and corresponding pixel color values is expressed as the vector 62 of 2*1.
Please refer to Fig. 9, Fig. 9 is the score layer fusion recognition method in the embodiment of the invention, and in the embodiment of the invention, for the image of the uncorrelated color space of two dimension, each width of cloth image can represent with two matrixes respectively that matrix X7111 and matrix Y7112 respectively represent one dimension; At this method for expressing of the uncorrelated color space image of two dimension, different face authentication methods is proposed, be embodied in the step 425 that second step (linear discriminant analysis feature extraction) and the people's face of the step 424 of feature extraction among Fig. 4 compare.The feature extraction mode that different face authentication methods is corresponding different and people's face comparison mode, the face authentication method of using in the embodiment of the invention is a score layer fusion recognition method, process was divided into for two steps: second step 91 of feature extraction, be used to handle colour space transformation result 70, extract useful information; The step 425 of people's face comparison, be used for calculating the distance between test face characteristic and the face database face characteristic, output recognition result 426, when with the distance of the eigenface of arest neighbors during less than certain threshold value, the name (or numbering) and the user class (or classification) of output arest neighbors eigenface, during not less than certain threshold value, can be regarded as the people's face authentication authorization and accounting that not have coupling and fail.
Second step of feature extraction, 91 processes were the step 712 of linear discriminant analysis feature extraction, be used to extract the feature of matrix, the matrix source is two matrix inputs, be respectively matrix X7111 and matrix Y7112, bidimensional in the corresponding two-dimentional uncorrelated color space of these two matrixes, the result of linear discriminant analysis feature extraction enters the step 425 of people's face comparison respectively.
In the step 425 of people's face comparison, two parts feature that feature extraction goes out respectively with face database in character pair calculate matching score 4251, here the matching score available range or the similarity of indication are represented, suppose to obtain the matching score result and be respectively s1 and s2, the total matching score of people's face is the two weighted sum a*s1+b*s2 in test person face and the face database so, to this test person face classification 4253, draw recognition result 426 according to the arest neighbors principle of classification; When with the distance of the eigenface of arest neighbors during less than certain threshold value, the name (or numbering) and the user class (or classification) of output arest neighbors eigenface during not less than certain threshold value, be can be regarded as the people's face authentication authorization and accounting that not have to mate and are failed; Wherein, weights a and b obtain according to the training to training sample, in general, because first dimension in the new two-dimentional uncorrelated color space that the corresponding PCA conversion of matrix X7111 obtains, in the highest flight, and second dimension in the new two-dimentional uncorrelated color space that the corresponding PCA conversion of matrix Y7112 obtains accounts for back burner, so the value of a is greater than the value of b.
Above content be in conjunction with concrete optimal technical scheme to further describing that the present invention did, can not assert that concrete enforcement of the present invention is confined to these explanations.For the general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, can also make some simple deduction or replace, all should be considered as belonging to protection scope of the present invention.
Claims (10)
1. authenticating colorized face images system is characterized in that: described authenticating colorized face images system comprises: the network shooting module that is used to catch colorized face images; Be used to utilize the principal component analysis method colorized face images to be transformed to the facial image conversion module of new two-dimentional uncorrelated color space; And be used for discerning the face recognition module of the facial image of new two-dimentional uncorrelated color space; Described network shooting module, facial image conversion module and face recognition module connect successively.
2. authenticating colorized face images according to claim 1 system is characterized in that: described authenticating colorized face images system further comprises: be used for detecting the people's face detection module whether image that the network shooting module catches has people's face.
3. authenticating colorized face images according to claim 1 system is characterized in that: described authenticating colorized face images system further comprises: be used for the colorized face images that captures is cut apart and normalized pretreatment module.
4. authenticating colorized face images according to claim 1 system, it is characterized in that: described face recognition module comprises: the characteristic extracting module that is used for the facial image of new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction; And the face characteristic that is used for the face characteristic that will extract and face database people's face comparing module of comparing.
5. authenticating colorized face images method, it is characterized in that: described authenticating colorized face images method may further comprise the steps:
A. catch colorized face images;
B. utilize the principal component analysis method that colorized face images is transformed to new two-dimentional uncorrelated color space;
C. discern the facial image in the new two-dimentional uncorrelated color space.
6. authenticating colorized face images method according to claim 5 is characterized in that: described authenticating colorized face images method further comprises: the step whether people's face is arranged in the image that detection network shooting module is caught.
7. authenticating colorized face images method according to claim 5 is characterized in that: described authenticating colorized face images method further comprises: the colorized face images that captures is cut apart and normalized step.
8. authenticating colorized face images method according to claim 5, it is characterized in that: described steps A is specially: start the network shooting module, this module of initialization, set this module mode of operation, whether detect has people's face to occur in the viewfinder range, if nobody's face occurs, cycle detection then, up to there being people's face to occur, go out people's face with the rectangle circle, rectangle frame is according to people's little self size of adjusting automatically of being bold, according to default collection quantity k continuous shooting k time, the time interval between adjacent twice is s second, in the continuous shooting process, to confirm all whether people's face is arranged in the viewfinder range before taking each time, cause detecting less than people's face then stop to take if people's face shifts out viewfinder range, the state that keeps cycle detection continues to take when people's face is retracted once more.
9. authenticating colorized face images method according to claim 5, it is characterized in that: described step B is specially: with principal component analysis method two main compositions of m*n sample extraction to the colorized face images of m pixel of the n width of cloth, obtain the vector of new sample 2*1, uncorrelated between each dimension of described new sample, the facial image correspondence of new composition of sample, two-dimentional uncorrelated color space; Wherein, each pixel is regarded a sample as, and with the vector representation of 3*1, then the colorized face images of a n width of cloth m pixel has m*n sample.
10. authenticating colorized face images method according to claim 5 is characterized in that: described step C comprises following substep:
C1. the facial image in the new two-dimentional uncorrelated color space is done the linear discriminant analysis feature extraction;
C2. the face characteristic of extraction and the face characteristic in the face database are compared.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009101084338A CN101599121A (en) | 2009-06-30 | 2009-06-30 | The authenticating colorized face images system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009101084338A CN101599121A (en) | 2009-06-30 | 2009-06-30 | The authenticating colorized face images system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101599121A true CN101599121A (en) | 2009-12-09 |
Family
ID=41420562
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2009101084338A Pending CN101599121A (en) | 2009-06-30 | 2009-06-30 | The authenticating colorized face images system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101599121A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102163279A (en) * | 2011-04-08 | 2011-08-24 | 南京邮电大学 | Color human face identification method based on nearest feature classifier |
| CN102346846A (en) * | 2011-09-16 | 2012-02-08 | 由田信息技术(上海)有限公司 | Face snap-shooting and contour analysis system |
| CN106485186A (en) * | 2015-08-26 | 2017-03-08 | 阿里巴巴集团控股有限公司 | Image characteristic extracting method, device, terminal device and system |
| CN108205620A (en) * | 2016-12-20 | 2018-06-26 | 航天信息股份有限公司 | The method of automated log on tax administration self-service terminating machine and tax administration self-service terminating machine |
| CN109376717A (en) * | 2018-12-14 | 2019-02-22 | 中科软科技股份有限公司 | Personal identification method, device, electronic equipment and the storage medium of face comparison |
| CN110267038A (en) * | 2019-06-28 | 2019-09-20 | 广东中星微电子有限公司 | Coding method and device, coding/decoding method and device |
| CN110851457A (en) * | 2019-10-21 | 2020-02-28 | 广东优世联合控股集团股份有限公司 | Method for searching login information |
| CN113160444A (en) * | 2021-04-20 | 2021-07-23 | 范传进 | Attendance machine based on end pipe cloud architecture and use method thereof |
-
2009
- 2009-06-30 CN CNA2009101084338A patent/CN101599121A/en active Pending
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102163279A (en) * | 2011-04-08 | 2011-08-24 | 南京邮电大学 | Color human face identification method based on nearest feature classifier |
| CN102163279B (en) * | 2011-04-08 | 2012-11-28 | 南京邮电大学 | Color human face identification method based on nearest feature classifier |
| CN102346846A (en) * | 2011-09-16 | 2012-02-08 | 由田信息技术(上海)有限公司 | Face snap-shooting and contour analysis system |
| CN106485186A (en) * | 2015-08-26 | 2017-03-08 | 阿里巴巴集团控股有限公司 | Image characteristic extracting method, device, terminal device and system |
| CN106485186B (en) * | 2015-08-26 | 2020-02-18 | 阿里巴巴集团控股有限公司 | Image feature extraction method and device, terminal equipment and system |
| CN108205620A (en) * | 2016-12-20 | 2018-06-26 | 航天信息股份有限公司 | The method of automated log on tax administration self-service terminating machine and tax administration self-service terminating machine |
| CN108205620B (en) * | 2016-12-20 | 2021-07-02 | 航天信息股份有限公司 | Method for automatically logging in tax self-service terminal and tax self-service terminal |
| CN109376717A (en) * | 2018-12-14 | 2019-02-22 | 中科软科技股份有限公司 | Personal identification method, device, electronic equipment and the storage medium of face comparison |
| CN110267038A (en) * | 2019-06-28 | 2019-09-20 | 广东中星微电子有限公司 | Coding method and device, coding/decoding method and device |
| CN110267038B (en) * | 2019-06-28 | 2022-07-29 | 广东中星微电子有限公司 | Encoding method and device, and decoding method and device |
| CN110851457A (en) * | 2019-10-21 | 2020-02-28 | 广东优世联合控股集团股份有限公司 | Method for searching login information |
| CN113160444A (en) * | 2021-04-20 | 2021-07-23 | 范传进 | Attendance machine based on end pipe cloud architecture and use method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101599121A (en) | The authenticating colorized face images system and method | |
| Cui et al. | An iris image synthesis method based on PCA and super-resolution | |
| CN101587485B (en) | Face information automatic login method based on face recognition technology | |
| CN103886283A (en) | Method for fusing multi-biometric image information for mobile user and application thereof | |
| CN104951940A (en) | Mobile payment verification method based on palmprint recognition | |
| Yang et al. | A novel finger-vein recognition method with feature combination | |
| CN107122700A (en) | A kind of palm print and palm vein combined registering and recognition methods based on video | |
| Pravallika et al. | SVM classification for fake biometric detection using image quality assessment: Application to iris, face and palm print | |
| Khan et al. | Low dimensional representation of dorsal hand vein features using principle component analysis (PCA) | |
| US20200302039A1 (en) | Authentication verification using soft biometric traits | |
| CN109815780A (en) | A kind of high-precision fingerprint identification method and system based on image procossing | |
| CN103345599A (en) | Virtual desktop login method based on face recognition technology | |
| Seal et al. | Minutiae based thermal face recognition using blood perfusion data | |
| Lakshmi et al. | Off-line signature verification using Neural Networks | |
| Wang et al. | Finger multimodal features fusion and recognition based on CNN | |
| CN114241584A (en) | Face optimization identification method and terminal | |
| Vijayalakshmi et al. | Finger and palm print based multibiometric authentication system with GUI interface | |
| Krichen et al. | Specific texture analysis for iris recognition | |
| Shang et al. | Face and lip-reading authentication system based on android smart phones | |
| Kanjan et al. | A comparative study of fingerprint matching algorithms | |
| Wang et al. | Quality-based score level fusion for continuous authentication with motion sensor and face | |
| CN103700151A (en) | Morning run check-in method | |
| Sanches et al. | A single sensor hand biometric multimodal system | |
| Khan et al. | Dorsal hand vein biometric using Independent Component Analysis (ICA) | |
| Ma | Finger vein identification based on 2DPCA |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20091209 |