CN101162499A - Method for using human face formwork combination to contrast - Google Patents
Method for using human face formwork combination to contrast Download PDFInfo
- Publication number
- CN101162499A CN101162499A CNA2006101171417A CN200610117141A CN101162499A CN 101162499 A CN101162499 A CN 101162499A CN A2006101171417 A CNA2006101171417 A CN A2006101171417A CN 200610117141 A CN200610117141 A CN 200610117141A CN 101162499 A CN101162499 A CN 101162499A
- Authority
- CN
- China
- Prior art keywords
- template
- face
- fast
- compared
- people
- 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
Abstract
A comparison method by means of human face template combination includes the following steps: firstly, a template base for one or more human faces is created in advance, wherein, a fast template and a precise template are created for each human face in the template base; then, a fast template and a precise template are created for each human face to be compared; moreover, the fast template of the compared human face is compared with all the fast templates in the template base to obtain the similarity of the compared human face and each fast template in the template base; then, according to preset pick-up number N and the obtained similarity, the precise templates corresponding to the N fast templates most similar to the compared human faces are read from the templates base; moreover, the precise templates of the compared human faces and the read N precise templates are compared to obtain precise identification of the compared human faces, thereby increasing comparison rate and precision.
Description
Technical field
The present invention relates to a kind of method of utilizing the face template combination to compare.
Background technology
Continuous development along with recognition of face, the method of extracting face characteristic is more and more, these methods all respectively have relative merits, and the face characteristic that the algorithm that has extracts is simple, and comparison speed is also very fast, but discrimination is not high, this class template is referred to as fast Template, the face characteristic complexity that the algorithm that has extracts, and the precision of identification is very high, but the preceding relatively a kind of algorithm of comparison speed is slow, and this class template is referred to as accurate template.
And in large-scale comparison system, because it is very big to participate in the template number of comparison, compare with fast Template, then often the comparison result precision is not high, is difficult to discern people's face; And compare with accurate template, finishing then that once to compare institute's time spent long, the user is difficult to accept.
Therefore, how to solve the existing problem of large-scale comparison system and become the technical task that industry needs to be resolved hurrily in fact.
Summary of the invention
The object of the present invention is to provide a kind of method of utilizing the face template combination to compare, to realize the raising of comparison speed and comparison accuracy.
In order to achieve the above object, the invention provides a kind of method of utilizing the face template combination to compare, it mainly comprises step: 1) be pre-created the template base that comprises one or more people's faces, wherein, each people's face has all been created a fast Template and an accurate template in the described template base; 2) people's face to be compared is created fast Template and accurate template; 3) fast Template of people's face to be compared and all fast Template in the described template base are compared to obtain the people's face to be compared and the similarity of each fast Template in the template base; 4) according to default extraction number N and resulting similarity, in described template base, read N the fast Template pairing accurate template the most similar to people's face to be compared; 5) N that accurate template and step 4) read an accurate template of people's face to be compared compared with accurate identification people's face to be compared.
Wherein, fast Template in the described template base is provided with classification, described classification is a sex, age, one or more in occupation and the identity, in described step 2) and step 3) between also comprise step: (1) sets the classification of people's face fast Template to be compared, fast Template in the described template base in the described step 3) is its classification and the identical person of classification of people's face fast Template to be compared, described fast Template is extracted corresponding characteristic by the face characteristic data extraction method to whole people's face and is formed, described accurate template by the face characteristic data extraction method respectively to the corresponding characteristic of each extracted region of people's face and form, described face characteristic data extraction method is principal component analysis (PCA) algorithm (PCA) and linear discriminant analysis algorithm (LDA), and described step 4) comprises step: (1) is sorted the similarity that obtains according to descending; (2) intercepting top n similarity; (3) according to extracting a corresponding N accurate template in N the pairing fast Template self-template of similarity storehouse in the step (2).
In sum, of the present invention utilize method that face template combination compares by with fast Template be used for recognition of face after accurately template makes up, realized the comparison speed in the large scale system comparison and the raising of comparison accuracy.
Embodiment
The method of utilizing the face template combination to compare of the present invention is usually used in the comparison system on a large scale, and it mainly may further comprise the steps:
1) is pre-created the template base that comprises one or more people's faces, wherein, each people's face has all been created a fast Template and an accurate template in the described template base, wherein said fast Template is extracted corresponding characteristic by the face characteristic data extraction method to whole people's face and is formed, it has advantage on recognition speed, and each fast Template is provided with classification, described classification is a sex, age, one or more in occupation and the identity, for example, for public security system, it in use, the classification that can set fast Template is thief's class, plunder class, the class etc. of killing a person, and for common user, can be man or women as required with category setting, respectively to the corresponding characteristic of each extracted region of people's face and form, it has advantage to described accurate template on comparison accuracy by the face characteristic data extraction method, and described face characteristic data extraction method is principal component analysis (PCA) algorithm (PCA) and linear discriminant analysis algorithm (LDA).
2) people's face to be compared is created fast Template and accurate template, in the present embodiment, people's face to be compared is to obtain by the photo that the scene is imported, and extracts known to those skilled in the art the knowing of method of people's face from photo, is no longer given unnecessary details at this.
3) set the classification of people's face fast Template to be compared, for example, the classification of setting people's face fast Template to be compared is male.
4) fast Template of people's face to be compared and the fast Template in the described template base are compared to obtain the people's face to be compared and the similarity of each fast Template in the template base, because of the classification of people's face fast Template to be compared is the man, then the classification of the fast Template in its need and the template base is compared for man's fast Template, for example, through comparison, the similarity of No. 1 fast Template in people's face to be compared and the template base is 30, similarity 50 with No. 2 fast Template, similarity 40 with No. 3 fast Template, similarity 70 with No. 4 fast Template, be noted that, if do not set the classification of people's face fast Template to be compared, then all fast Template in itself and the template base can be compared.
5) according to default extraction number N and resulting similarity, read N the fast Template pairing accurate template the most similar to people's face to be compared in described template base, in the present embodiment, this step 4) may further comprise the steps again:
(1) similarity that obtains is sorted according to descending, can be got 70,50,40,30 after the ordering;
(2) intercepting top n similarity for example, intercepts preceding 2 similarities, and promptly 70,50;
(3) according to extracting corresponding 2 accurate templates in 2 the pairing fast Template self-template of similarity storehouses in the step (2), similarity 70 pairing fast Template are No. 4, similarity 50 pairing fast Template are No. 2, therefore, will read out in No. 4 pairing No. 4 accurate templates of fast Template and No. 2 pairing No. 2 accurate template self-template storehouses of fast Template.
It should be noted that and also can not adopt sortord to select corresponding accurately template, only need to pick out maximum 2 and get final product from the similarity that obtains.
6) N that accurate template and step 4) read an accurate template of people's face to be compared compared with accurate identification people's face to be compared, in the present embodiment, the accurate template of people's face to be compared and No. 4 and No. 2 accurate templates are compared obtain corresponding similarity, size according to similarity can accurately be discerned people's face to be compared, if the people's face to be compared and the similarity of No. 4 accurate templates are 60, with the similarity of No. 2 accurate templates be 90, then think people's face and No. 2 corresponding human face similarities of accurate template to be compared.
In addition, the user also can select different templates to be used for comparison as required, for example, only selects for use fast Template to compare, and also can only select for use accurate template to compare.
In sum, of the present invention utilize method that face template combination compares by with fast Template and accurately template make up, make extensive comparison system in comparison process, can take into account comparison speed and can take into account comparison accuracy again, realized the raising of comparison speed and comparison accuracy.
Claims (9)
1. one kind is utilized face template to make up the method for comparing, and it is characterized in that comprising step:
1) be pre-created the template base that comprises one or more people's faces, wherein, each people's face has all been created a fast Template and an accurate template in the described template base;
2) people's face to be compared is created fast Template and accurate template;
3) fast Template of people's face to be compared and the fast Template in the described template base are compared to obtain the people's face to be compared and the similarity of each fast Template in the template base;
4) according to default extraction number N and resulting similarity, in described template base, read N the fast Template pairing accurate template the most similar to people's face to be compared;
5) N that accurate template and step 4) read an accurate template of people's face to be compared compared with accurate identification people's face to be compared.
2. the method for utilizing the face template combination to compare as claimed in claim 1, it is characterized in that: the fast Template in the described template base is provided with classification.
3. the method for utilizing face template combination to compare as claimed in claim 2 is characterized in that: described classification is one or more in sex, age, occupation and the identity.
4. as claim 2 or the 3 described methods of utilizing face templates combinations to compare, it is characterized in that in described step 2) and step 3) between also comprise step: (1) sets the classification of people's face fast Template to be compared.
5. the method for utilizing the face template combination to compare as claimed in claim 4 is characterized in that: the fast Template in the described template base in the described step 3) is the classification identical person of its classification with people's face fast Template to be compared.
6. the method for utilizing the face template combination to compare as claimed in claim 1, it is characterized in that: described fast Template is extracted corresponding characteristic by the face characteristic data extraction method to whole people's face and is formed.
7. the method for utilizing face template combination to compare as claimed in claim 1 is characterized in that: described accurate template forms the corresponding characteristic of each extracted region of people's face respectively by the face characteristic data extraction method.
8. as claim 6 or the 7 described methods of utilizing the face template combination to compare, it is characterized in that: described face characteristic data extraction method is principal component analysis (PCA) algorithm and linear discriminant analysis algorithm.
9. the method for utilizing the face template combination to compare as claimed in claim 1 is characterized in that described step 4) comprises step:
(1) similarity that obtains is sorted according to descending;
(2) intercepting top n similarity;
(3) according to extracting a corresponding N accurate template in N the pairing fast Template self-template of similarity storehouse in the step (2).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101171417A CN101162499A (en) | 2006-10-13 | 2006-10-13 | Method for using human face formwork combination to contrast |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101171417A CN101162499A (en) | 2006-10-13 | 2006-10-13 | Method for using human face formwork combination to contrast |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101162499A true CN101162499A (en) | 2008-04-16 |
Family
ID=39297420
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006101171417A Pending CN101162499A (en) | 2006-10-13 | 2006-10-13 | Method for using human face formwork combination to contrast |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101162499A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101667248B (en) * | 2008-09-04 | 2013-03-27 | 索尼株式会社 | Image processing apparatus, imaging apparatus, image processing method, and program |
| CN103971106A (en) * | 2014-05-27 | 2014-08-06 | 深圳市赛为智能股份有限公司 | Multi-view human facial image gender identification method and device |
| CN104143083A (en) * | 2014-07-11 | 2014-11-12 | 北京神州智联科技有限公司 | Face recognition system based on process management |
| WO2014180108A1 (en) * | 2013-05-09 | 2014-11-13 | Tencent Technology (Shenzhen) Company Limited | Systems and methods for matching face shapes |
| CN104573611A (en) * | 2013-10-11 | 2015-04-29 | 杭州海康威视数字技术股份有限公司 | Distributed face recognition trunking system |
| CN106815566A (en) * | 2016-12-29 | 2017-06-09 | 天津中科智能识别产业技术研究院有限公司 | A kind of face retrieval method based on multitask convolutional neural networks |
| CN109344703A (en) * | 2018-08-24 | 2019-02-15 | 深圳市商汤科技有限公司 | Method for checking object and device, electronic equipment and storage medium |
| CN109446893A (en) * | 2018-09-14 | 2019-03-08 | 百度在线网络技术（北京）有限公司 | Face identification method, device, computer equipment and storage medium |
| CN109472183A (en) * | 2017-09-08 | 2019-03-15 | 上海银晨智能识别科技有限公司 | Image-recognizing method and device, system of deploying to ensure effective monitoring and control of illegal activities, computer readable storage medium |
| CN112232206A (en) * | 2020-10-16 | 2021-01-15 | 陈晶晶 | Face recognition method and face recognition platform based on big data and artificial intelligence |
-
2006
- 2006-10-13 CN CNA2006101171417A patent/CN101162499A/en active Pending
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101667248B (en) * | 2008-09-04 | 2013-03-27 | 索尼株式会社 | Image processing apparatus, imaging apparatus, image processing method, and program |
| WO2014180108A1 (en) * | 2013-05-09 | 2014-11-13 | Tencent Technology (Shenzhen) Company Limited | Systems and methods for matching face shapes |
| CN104573611A (en) * | 2013-10-11 | 2015-04-29 | 杭州海康威视数字技术股份有限公司 | Distributed face recognition trunking system |
| CN104573611B (en) * | 2013-10-11 | 2018-03-20 | 杭州海康威视数字技术股份有限公司 | A kind of distributed recognition of face group system |
| CN103971106A (en) * | 2014-05-27 | 2014-08-06 | 深圳市赛为智能股份有限公司 | Multi-view human facial image gender identification method and device |
| CN103971106B (en) * | 2014-05-27 | 2018-07-27 | 深圳市赛为智能股份有限公司 | Various visual angles facial image gender identification method and device |
| CN104143083B (en) * | 2014-07-11 | 2018-03-02 | 北京神州智联科技有限公司 | A kind of face identification system of Kernel-based methods management |
| CN104143083A (en) * | 2014-07-11 | 2014-11-12 | 北京神州智联科技有限公司 | Face recognition system based on process management |
| CN106815566A (en) * | 2016-12-29 | 2017-06-09 | 天津中科智能识别产业技术研究院有限公司 | A kind of face retrieval method based on multitask convolutional neural networks |
| CN109472183A (en) * | 2017-09-08 | 2019-03-15 | 上海银晨智能识别科技有限公司 | Image-recognizing method and device, system of deploying to ensure effective monitoring and control of illegal activities, computer readable storage medium |
| CN109344703A (en) * | 2018-08-24 | 2019-02-15 | 深圳市商汤科技有限公司 | Method for checking object and device, electronic equipment and storage medium |
| CN109446893A (en) * | 2018-09-14 | 2019-03-08 | 百度在线网络技术（北京）有限公司 | Face identification method, device, computer equipment and storage medium |
| CN112232206A (en) * | 2020-10-16 | 2021-01-15 | 陈晶晶 | Face recognition method and face recognition platform based on big data and artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101162499A (en) | Method for using human face formwork combination to contrast | |
| CN103914686B (en) | A kind of face alignment authentication method and system shone based on certificate photo with collection | |
| CN100356388C (en) | Biocharacteristics fusioned identity distinguishing and identification method | |
| US11256902B2 (en) | People-credentials comparison authentication method, system and camera | |
| CN105825176A (en) | Identification method based on multi-mode non-contact identity characteristics | |
| CN106295672B (en) | A kind of face identification method and device | |
| JP6472184B2 (en) | Object identification device, object identification method, and program | |
| US20120288148A1 (en) | Image recognition apparatus, method of controlling image recognition apparatus, and storage medium | |
| KR20150079583A (en) | Biometric authentication in connection with camera-equipped devices | |
| US9990538B2 (en) | Face recognition apparatus and method using physiognomic feature information | |
| US9911027B2 (en) | Fingerprint authentication system, fingerprint authentication program and fingerprint authentication method | |
| Kim et al. | A survey and proposed framework on the soft biometrics technique for human identification in intelligent video surveillance system | |
| CN102945366A (en) | Method and device for face recognition | |
| US20140254892A1 (en) | Face recognition apparatus, system and method for managing users based on user grouping | |
| CN106056083B (en) | A kind of information processing method and terminal | |
| CN106709417A (en) | Multimodal biological recognition system and use method thereof | |
| CN103020607A (en) | Face recognition method and face recognition device | |
| CN103258190A (en) | Face recognition method used for mobile terminal | |
| KR20160148863A (en) | Apparatus for user verification | |
| CN106991448A (en) | A kind of portrait compares processing method | |
| CN101324919A (en) | Photograph video contrast method | |
| CN110197098A (en) | Identity identifying method based on face and iris feature | |
| CN201698456U (en) | Real-time face detecting and recognizing device | |
| CN110458094A (en) | Device class method based on fingerprint similarity | |
| US8879805B2 (en) | Automated image identification method |
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 |
Open date: 20080416 |