CN103207995A - PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method - Google Patents
PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method Download PDFInfo
- Publication number
- CN103207995A CN103207995A CN2013101744989A CN201310174498A CN103207995A CN 103207995 A CN103207995 A CN 103207995A CN 2013101744989 A CN2013101744989 A CN 2013101744989A CN 201310174498 A CN201310174498 A CN 201310174498A CN 103207995 A CN103207995 A CN 103207995A
- Authority
- CN
- China
- Prior art keywords
- facial image
- face
- pca
- people
- carried out
- 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 relates to a PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method. The method comprises the steps of: firstly creating a database including facial image samples; carrying out processing and training on the facial image samples in the database, processing facial images collected by a camera or a scanner; and finally comparing and identifying processed to-be-identified facial images with the processed and trained facial image samples, and outputting an identification result. According to the PCA-based 3D face identification method provided by the invention, an improved PCA algorithm is adopted to identify the images, and the identification efficiency is improved; and a 3D facial deformation model method is adopted to establish a 3D facial model method, and the facial identification precision is improved.
Description
Technical field
The present invention relates to a kind of face identification method, relate in particular to a kind of 3D face identification method based on PCA.
Background technology
Recognition of face is refered in particular to utilize to analyze and is compared the computer technology that people's face visual signature information is carried out the identity discriminating.Recognition of face is the computer technology research field of a hot topic, and the face tracking detecting is adjusted image automatically and amplified, and exposure intensity is adjusted in infrared detecting at night automatically; It belongs to biometrics identification technology, is the biological characteristic of biosome (generally refering in particular to the people) itself is distinguished the biosome individuality.In face recognition process, generally include the detection of people's face, face characteristic extraction and recognition of face step, feature extraction is an important step that influences recognition effect.The 2D face recognition technology is ripe at present, but because single 2D image can not provide identification required complete information, and the face characteristic of its feature extraction also is based on the face characteristic of 2D, therefore, accuracy of identification is difficult to further improve.
Summary of the invention
The present invention has overcome the deficiencies in the prior art, and a kind of 3D of extraction face characteristic, face identification method that accuracy of identification is high are provided.
For achieving the above object, the technical solution used in the present invention is: a kind of 3D face identification method based on PCA is characterized in that: comprise following steps:
1) creates the database with facial image sample;
2) the facial image sample in the database being carried out detection and location, whether in image have people face, if people's face is arranged, then make accurate in locating if judging;
3) the facial image sample after the detection and location is carried out pre-service;
4) pretreated facial image sample is carried out feature extraction;
5) utilize the BP neural network to the facial image sample training after feature extraction;
6) gather facial image by video camera or scanner;
7) facial image of gathering being carried out detection and location, whether in image have people face, if people's face is arranged, then make accurate in locating if judging;
8) facial image after the step 7) detection and location is carried out pre-service;
9) pretreated facial image in the step 8) is carried out feature extraction;
10) projection coefficient of facial image sample set in the projection coefficient of facial image to be identified in the step 9) and the database is compared the output recognition result.
In a preferred embodiment of the present invention, comprise that further the pretreated step of step 3) and step 8) is:
A) background in the deletion facial image is found out people's face in the position of image;
B) brightness of adjustment facial image.
In a preferred embodiment of the present invention, the pre-service that further comprises step 3) and step 8) is to set up the faceform with 3D shape and texture according to 3D people's shape of face varying model method.
In a preferred embodiment of the present invention, further comprise and adopt the PCA algorithm to extract face characteristic in step 4) and the step 9).
The invention solves the defective that exists in the background technology, the 3D face identification method based on PCA of the present invention utilizes improved PCA algorithm to image recognition, has improved recognition efficiency; Adopt 3D people's shape of face varying model method to set up 3D faceform's method, improved the accuracy of identification of people's face.
Description of drawings
The present invention is further described below in conjunction with drawings and Examples.
Fig. 1 is the operational flowchart of the 3D face identification method based on PCA of the present invention.
Embodiment
The present invention is further detailed explanation in conjunction with the accompanying drawings and embodiments now, and these accompanying drawings are the synoptic diagram of simplification, basic structure of the present invention only is described in a schematic way, so it only shows the formation relevant with the present invention.
As shown in Figure 1, a kind of 3D face identification method based on PCA is characterized in that: comprise following steps:
S1, establishment have the database of facial image sample; If in actual applications, the facial image sample in the database may have only one or two, will cause BP train samples shortage in the subsequent step like this, so the database of abundant people's face object samples should be arranged in application.
S2, the facial image sample in the database being carried out detection and location, whether in image have people face, if people's face is arranged, then make accurate in locating if judging.
S3, the facial image sample after the detection and location is carried out pre-service.
S4, pretreated facial image sample is carried out feature extraction, adopt the PCA algorithm that the facial image sample is carried out feature extraction.
S5, utilize the BP neural network to the facial image sample training after feature extraction; The BP neural network is a kind of multilayer feedforward network of one way propagation.The BP learning algorithm is called " error Back-Propagation algorithm ", and basic thought is to adjust weight distribution by the minimal value of network error function to make neural network converge on steady state (SS), thereby makes network also can provide suitable output when accepting unknown input.
S6, gather facial image by video camera or scanner.
S7, the facial image of gathering being carried out detection and location, whether in image have people face, if people's face is arranged, then make accurate in locating if judging.
S8, the facial image after the step S7 detection and location is carried out pre-service.
S9, pretreated facial image among the step S8 is carried out feature extraction; Because the quantity of information of people's face picture that video camera or scanner are gathered is often very big, directly handle to produce huge calculated amount, so before recognition of face, will carry out feature extraction.In the dimension that reduces feature space, keep identifying information as much as possible, to reach effective classification.Feature extraction in this step adopts the PCA algorithm that the facial image sample is carried out feature extraction.
S10, the projection coefficient of facial image sample set in the projection coefficient of facial image to be identified among the step S9 and the database is compared the output recognition result.
Adopt the PCA algorithm that the facial image sample is carried out feature extraction among the present invention, this basic idea is: image vector is converted to low dimensional vector through after the Karhunen-Loeve transformation by high dimension vector, and formation low-dimensional linear vector space, it is proper subspace, then people's face is projected to this lower dimensional space, with the proper vector of resulting projection coefficient as identification.During identification people face, only need the projection coefficient of target sample collection in the projection coefficient of sample to be identified and the database is compared, to determine and which kind of is nearest.
The PCA algorithm, has identification accurately and the little advantage of calculated amount, but under following some situation, has defective preferably under the situation in picture quality to be identified, and first image to be identified and training image illuminance difference are bigger; It two is that people's face background difference is bigger.For improving the PCA algorithm, can do pre-service to image, in step S3 and S8, pre-service comprises two steps:
A, deletion background: the deletion background will be found out people's face in the position of image, to this, face identification method of the present invention is found out the position of people's face by the method for 3D deformation model modeling, and further find out eye position, calculate people's general scope of being bold according to the distance between two eyes then, suitable adjustment by to people's face scope makes its normalization.
The brightness of b, every facial image of adjustment: by setting a fixed value, adjust the gray-scale value of image pixel, make gradation of image mean value reach this fixed value, by homomorphic filtering, reduce uneven illumination.
Wherein, the pre-service of step S3 and step S8 is to set up the faceform with 3D shape and texture according to 3D people's shape of face varying model method, and the core concept of 3D people's shape of face varying model is to utilize the 3D faceform's of limited quantity linear combination to express shape and the texture of any one 3D people's face.
3D face identification method based on PCA of the present invention, the principal component analysis (PCA) that utilizes (principal compONent analysis, PCA) be a kind of method that is used for analyzing data in the multivariate statistical analysis, it is the method that with a kind of feature of lesser amt sample is described to reach reduction feature space dimension, the basis of method is the Karhunen-Loeve expansion, the great advantage of K-L conversion is that decorrelation is good, a large amount of irrelevant redundant informations in the image can be removed like this, reduce the structure complexity of BP neural network, also improved training effectiveness and the convergency factor of neural network simultaneously.
3D face identification method based on PCA of the present invention utilizes improved PCA algorithm to image recognition, has improved recognition efficiency; Adopt 3D people's shape of face varying model method to set up 3D faceform's method, improved the accuracy of identification of people's face.
Above foundation desirable embodiment of the present invention is enlightenment, and by above-mentioned description, the related personnel can carry out various change and modification fully in the scope that does not depart from this invention technological thought.The technical scope of this invention is not limited to the content on the instructions, must determine technical scope according to the claim scope.
Claims (4)
1. 3D face identification method based on PCA is characterized in that: comprise following steps:
1) creates the database with facial image sample;
2) the facial image sample in the database being carried out detection and location, whether in image have people face, if people's face is arranged, then make accurate in locating if judging;
3) the facial image sample after the detection and location is carried out pre-service;
4) pretreated facial image sample is carried out feature extraction;
5) utilize the BP neural network to the facial image sample training after feature extraction;
6) gather facial image by video camera or scanner;
7) facial image of gathering being carried out detection and location, whether in image have people face, if people's face is arranged, then make accurate in locating if judging;
8) facial image after the step 7) detection and location is carried out pre-service;
9) pretreated facial image in the step 8) is carried out feature extraction;
10) projection coefficient of facial image sample set in the projection coefficient of facial image to be identified in the step 9) and the database is compared the output recognition result.
2. the 3D face identification method based on PCA according to claim 1, it is characterized in that: the pretreated step of step 3) and step 8) is:
A) background in the deletion facial image is found out people's face in the position of image;
B) brightness of adjustment facial image.
3. the 3D face identification method based on PCA according to claim 2 is characterized in that: the pre-service of step 3) and step 8) is to set up the faceform with 3D shape and texture according to 3D people's shape of face varying model method.
4. the 3D face identification method based on PCA according to claim 1 is characterized in that: adopt the PCA algorithm to extract face characteristic in step 4) and the step 9).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013101744989A CN103207995A (en) | 2013-05-13 | 2013-05-13 | PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013101744989A CN103207995A (en) | 2013-05-13 | 2013-05-13 | PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103207995A true CN103207995A (en) | 2013-07-17 |
Family
ID=48755212
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2013101744989A Pending CN103207995A (en) | 2013-05-13 | 2013-05-13 | PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103207995A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104573641A (en) * | 2014-12-26 | 2015-04-29 | 苏州福丰科技有限公司 | Face recognition method under illumination change condition |
CN105117692A (en) * | 2015-08-05 | 2015-12-02 | 福州瑞芯微电子股份有限公司 | Real-time face identification method and system based on deep learning |
CN106841212A (en) * | 2016-12-30 | 2017-06-13 | 湖南大学 | A kind of bottle mouth defect detection method based on local PCA and BP neural network |
CN107423678A (en) * | 2017-05-27 | 2017-12-01 | 电子科技大学 | A kind of training method and face identification method of the convolutional neural networks for extracting feature |
CN109993125A (en) * | 2019-04-03 | 2019-07-09 | 腾讯科技(深圳)有限公司 | Model training method, face identification method, device, equipment and storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009099125A (en) * | 2007-09-27 | 2009-05-07 | Aisin Aw Co Ltd | Image recognition device, image recognition program, and point information collection device and navigation device using them |
WO2011019192A2 (en) * | 2009-08-11 | 2011-02-17 | 주식회사 크라스아이디 | System and method for recognizing face using ir lighting |
CN102982322A (en) * | 2012-12-07 | 2013-03-20 | 大连大学 | Face recognition method based on PCA (principal component analysis) image reconstruction and LDA (linear discriminant analysis) |
-
2013
- 2013-05-13 CN CN2013101744989A patent/CN103207995A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009099125A (en) * | 2007-09-27 | 2009-05-07 | Aisin Aw Co Ltd | Image recognition device, image recognition program, and point information collection device and navigation device using them |
WO2011019192A2 (en) * | 2009-08-11 | 2011-02-17 | 주식회사 크라스아이디 | System and method for recognizing face using ir lighting |
CN102982322A (en) * | 2012-12-07 | 2013-03-20 | 大连大学 | Face recognition method based on PCA (principal component analysis) image reconstruction and LDA (linear discriminant analysis) |
Non-Patent Citations (5)
Title |
---|
吴通理等: "三维人脸识别及其关键技术", 《福建电脑》, no. 2, 1 February 2007 (2007-02-01) * |
朱巍峰: "3D人脸识别研究探索", 《现代电子技术》, vol. 34, no. 11, 1 June 2011 (2011-06-01) * |
柳杨: "三维人脸识别算法综述", 《系统仿真学报》, vol. 18, 30 December 2006 (2006-12-30) * |
王跃明等: "三维人脸识别研究综述", 《计算机辅助设计与图形学学报》, vol. 20, no. 7, 15 July 2008 (2008-07-15) * |
郑明伟: "基于PCA和BP神经网络的人脸识别方法", 《潍坊高等职业教育》, vol. 7, no. 3, 15 September 2011 (2011-09-15) * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104573641A (en) * | 2014-12-26 | 2015-04-29 | 苏州福丰科技有限公司 | Face recognition method under illumination change condition |
CN105117692A (en) * | 2015-08-05 | 2015-12-02 | 福州瑞芯微电子股份有限公司 | Real-time face identification method and system based on deep learning |
CN106841212A (en) * | 2016-12-30 | 2017-06-13 | 湖南大学 | A kind of bottle mouth defect detection method based on local PCA and BP neural network |
CN106841212B (en) * | 2016-12-30 | 2019-06-21 | 湖南大学 | A kind of bottle mouth defect detection method based on local PCA and BP neural network |
CN107423678A (en) * | 2017-05-27 | 2017-12-01 | 电子科技大学 | A kind of training method and face identification method of the convolutional neural networks for extracting feature |
CN109993125A (en) * | 2019-04-03 | 2019-07-09 | 腾讯科技(深圳)有限公司 | Model training method, face identification method, device, equipment and storage medium |
CN109993125B (en) * | 2019-04-03 | 2022-12-23 | 腾讯科技(深圳)有限公司 | Model training method, face recognition device, face recognition equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104463117B (en) | A kind of recognition of face sample collection method and system based on video mode | |
CN104537336B (en) | A kind of face identification method and system for possessing self-learning function | |
CN110956094A (en) | RGB-D multi-mode fusion personnel detection method based on asymmetric double-current network | |
CN101587485B (en) | Face information automatic login method based on face recognition technology | |
US20210034840A1 (en) | Method for Recognzing Face from Monitoring Video Data | |
Sugiarto et al. | Wood identification based on histogram of oriented gradient (HOG) feature and support vector machine (SVM) classifier | |
CN106610969A (en) | Multimodal information-based video content auditing system and method | |
CN106874929B (en) | Pearl classification method based on deep learning | |
CN103207995A (en) | PCA (Principal Component Analysis)-based 3D (three dimensional) face identification method | |
CN103886589A (en) | Goal-oriented automatic high-precision edge extraction method | |
CN110390308B (en) | Video behavior identification method based on space-time confrontation generation network | |
CN103440035A (en) | Gesture recognition system in three-dimensional space and recognition method thereof | |
CN105718866A (en) | Visual target detection and identification method | |
CN101853397A (en) | Bionic human face detection method based on human visual characteristics | |
CN105335691A (en) | Smiling face identification and encouragement system | |
Mady et al. | Efficient real time attendance system based on face detection case study “MEDIU staff” | |
CN103218612A (en) | 3D (Three-Dimensional) face recognition method | |
CN106097456A (en) | Oblique photograph outdoor scene three dimensional monolithic model method based on self-adapting cluster algorithm | |
CN111382601A (en) | Illumination face image recognition preprocessing system and method for generating confrontation network model | |
CN110991554B (en) | Improved PCA (principal component analysis) -based deep network image classification method | |
Diyasa et al. | Multi-face Recognition for the Detection of Prisoners in Jail using a Modified Cascade Classifier and CNN | |
CN109446964A (en) | Face detection analysis method and device based on end-to-end single-stage multiple scale detecting device | |
Anggraini | Face recognition using principal component analysis and self organizing maps | |
CN108664886A (en) | A kind of fast face recognition method adapting to substation's disengaging monitoring demand | |
CN103235943A (en) | Principal component analysis-based (PCA-based) three-dimensional (3D) face 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 | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130717 |