CN102368294A - Method for extracting sex characteristics under complex environment - Google Patents
Method for extracting sex characteristics under complex environment Download PDFInfo
- Publication number
- CN102368294A CN102368294A CN 201110264746 CN201110264746A CN102368294A CN 102368294 A CN102368294 A CN 102368294A CN 201110264746 CN201110264746 CN 201110264746 CN 201110264746 A CN201110264746 A CN 201110264746A CN 102368294 A CN102368294 A CN 102368294A
- Authority
- CN
- China
- Prior art keywords
- extracting
- complex environment
- human body
- characteristic
- sex
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to a method for extracting sex characteristics under a complex environment and comprises extraction of the sex characteristics. The method comprises the following steps: a. extracting the characteristics of a human face, a skin color, a hair style and a decoration of a human body; b. extracting the trained and detected human body characteristics through a linear discrimination method, and using a Fisher method to extract the characteristics of training samples and object images; c. classifying the object images through a dynamic clustering method and calculating an average recognition rate of a single classifier; d. training a support vector regression machine through the male and female training samples so as to obtain a group of parameter values; e. comparing classification accuracies of various kinds of fusion rules and the different fusion rules. In the method for extracting the sex characteristics under the complex environment, the dynamic clustering method is used; an average recognition rate can be calculated rapidly, conveniently and accurately; a convergence speed is fast.
Description
Technical field
The present invention relates to the field of biotechnology, especially the method for distilling of sex character under the complex environment.
Background technology
In June, 2010 is in international digital signage in second Shanghai and the exhibition of touch inquire technology; The Greater China manager Sui Dapeng of Samsung points out: the digital signage system of the said firm is incorporating more new technology; Accomplish collection and statistics through face tracking and recognition technology, and can adjust broadcast strategy in view of the above information such as audient crowd's number, sex, age intervals.Spectators' recognition system solution that Samsung is applied to catering industry can be the different different contents of audient colony broadcast, to accomplish the intelligence issue.
The extraction of the sex character that uses at present, method of operating is complicated, extracts loaded down with trivial detailsly, loses time.
Summary of the invention
The technical matters that the present invention will solve is: in order to overcome the above-mentioned middle problem that exists, a kind of method for distilling that extracts conveniently and calculate sex character under the simple complex environment is provided.
The technical solution adopted for the present invention to solve the technical problems is: the method for distilling of sex character under a kind of complex environment, comprise the extraction of sex character, and its concrete steps are following:
A. extract people's face, the colour of skin, the hair style of human body, the characteristic of decoration;
B. extract the characteristic and the characteristic of utilizing Fisher method extraction training sample and target image of the detected human body of training through the linear discriminant method; The Fisher linear discriminant analysis is as single classifier, and its recognition performance is all good on different databases, still; Utilize the Fisher linear approach when carrying out dimension-reduction treatment; Still can lose some Useful Informations, and these information maybe be very important for the step of back, this also is the weak point of this method;
C. through the dynamic clustering method target image is classified and calculate the average recognition rate of single classifier, the dynamic clustering algorithm, the C-mean algorithm is a kind of unsupervised learning method based on neighbour's rule commonly used; At first definite crowd's number c that needs chooses c representative point, uses these representative points as starting type; Again each sample X among the sample set H is found out at a distance of nearest representative point, X is grouped among the crowd at this nearest representative point place and goes, like this; Iteration just tentatively is divided into c crowd with neighbour's rule with H for the first time; Next iteration just on this basis with the mean vector of each group of last iteration gained as new representative point, by neighbour's rule H is divided into c crowd once more, up to hive off stable till; The speed of convergence of C-mean algorithm is than very fast, but its convergence result depends on choosing of initial cluster center;
D. through men and women's training sample support vector regression is trained and obtain one group of parameter value; Support vector regression (Support Vector Regression; Be called for short SVR), the support vector regression algorithm is the popularization of support vector machine method on regression problem.Through introducing insensitive loss function and kernel function, can be advantageously applied to nonlinear regression analysis, and the small sample set problem is had good estimated performance;
E. the nicety of grading with various fusion rules and different fusion rules compares.
The beneficial effect of the method for distilling of sex character is under the complex environment of the present invention: adopt the dynamic clustering method, can quick and conveniently calculate average recognition rate, its fast convergence rate exactly.
Embodiment
The method for distilling of sex character under the complex environment of the present invention comprises the extraction of sex character, and its concrete steps are following:
At first extract people's face, the colour of skin, the hair style of human body, the characteristic of decoration, face characteristic comprises global feature and local feature;
Secondly, extract the characteristic and the characteristic of utilizing Fisher method extraction training sample and target image of the detected human body of training through the linear discriminant method;
Again, through the dynamic clustering method target image is classified and calculate the average recognition rate of single classifier;
Then, through men and women's training sample support vector regression is trained to obtain one group of parameter value, utilize the SVR that trains that test sample book is classified;
At last, the nicety of grading with various fusion rules and different fusion rules compares.
With above-mentioned foundation desirable embodiment of the present invention is enlightenment, and through above-mentioned description, the related work 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 confirm its technical scope according to the claim scope.
Claims (1)
1. the method for distilling of sex character under the complex environment comprises the extraction of sex character, and it is characterized in that: its concrete steps are following:
A. extract people's face, the colour of skin, the hair style of human body, the characteristic of decoration;
B. extract the characteristic and the characteristic of utilizing Fisher method extraction training sample and target image of the detected human body of training through the linear discriminant method;
C. through the dynamic clustering method target image is classified and calculate the average recognition rate of single classifier; D. through men and women's training sample support vector regression is trained and obtain one group of parameter value;
E. the nicety of grading with various fusion rules and different fusion rules compares.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110264746 CN102368294A (en) | 2011-09-07 | 2011-09-07 | Method for extracting sex characteristics under complex environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110264746 CN102368294A (en) | 2011-09-07 | 2011-09-07 | Method for extracting sex characteristics under complex environment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102368294A true CN102368294A (en) | 2012-03-07 |
Family
ID=45760857
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201110264746 Pending CN102368294A (en) | 2011-09-07 | 2011-09-07 | Method for extracting sex characteristics under complex environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102368294A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102930454A (en) * | 2012-10-07 | 2013-02-13 | 乐配(天津)科技有限公司 | Intelligent 3D (Three Dimensional) advertisement recommendation method based on multiple perception technologies |
CN103514454A (en) * | 2012-09-27 | 2014-01-15 | Tcl集团股份有限公司 | Support vector machine gender classification method based on online learning |
CN104915000A (en) * | 2015-05-27 | 2015-09-16 | 天津科技大学 | Multisensory biological recognition interaction method for naked eye 3D advertisement |
CN114493229A (en) * | 2022-01-20 | 2022-05-13 | 广东电网有限责任公司电力调度控制中心 | Regulation and control business arrangement agent method and system based on unsupervised learning technology |
-
2011
- 2011-09-07 CN CN 201110264746 patent/CN102368294A/en active Pending
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103514454A (en) * | 2012-09-27 | 2014-01-15 | Tcl集团股份有限公司 | Support vector machine gender classification method based on online learning |
CN103514454B (en) * | 2012-09-27 | 2016-12-21 | Tcl集团股份有限公司 | Method based on on-line study support vector machine Gender Classification |
CN102930454A (en) * | 2012-10-07 | 2013-02-13 | 乐配(天津)科技有限公司 | Intelligent 3D (Three Dimensional) advertisement recommendation method based on multiple perception technologies |
CN104915000A (en) * | 2015-05-27 | 2015-09-16 | 天津科技大学 | Multisensory biological recognition interaction method for naked eye 3D advertisement |
CN114493229A (en) * | 2022-01-20 | 2022-05-13 | 广东电网有限责任公司电力调度控制中心 | Regulation and control business arrangement agent method and system based on unsupervised learning technology |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105528794B (en) | Moving target detecting method based on mixed Gauss model and super-pixel segmentation | |
CN105160317B (en) | One kind being based on area dividing pedestrian gender identification method | |
CN106296695B (en) | Adaptive threshold natural target image segmentation extraction algorithm based on conspicuousness | |
Kamlapurkar | Detection of plant leaf disease using image processing approach | |
CN104036247A (en) | Facial feature based face racial classification method | |
CN103226835B (en) | Based on method for tracking target and the system of online initialization gradient enhancement regression tree | |
CN106295584A (en) | Depth migration study is in the recognition methods of crowd's attribute | |
CN107103326A (en) | The collaboration conspicuousness detection method clustered based on super-pixel | |
CN102298709A (en) | Energy-saving intelligent identification digital signage fused with multiple characteristics in complicated environment | |
CN101526994B (en) | Fingerprint image segmentation method irrelevant to collecting device | |
CN104134071A (en) | Deformable part model object detection method based on color description | |
CN104361313A (en) | Gesture recognition method based on multi-kernel learning heterogeneous feature fusion | |
Hassanat et al. | Colour-based lips segmentation method using artificial neural networks | |
CN102368294A (en) | Method for extracting sex characteristics under complex environment | |
CN104301585A (en) | Method for detecting specific kind objective in movement scene in real time | |
CN108960586A (en) | A kind of non-invasive load recognition methods adapting to scene changes | |
CN105718885A (en) | Human face characteristic point tracking method | |
CN103985130A (en) | Image significance analysis method for complex texture images | |
CN103678552A (en) | Remote-sensing image retrieving method and system based on salient regional features | |
CN106844785A (en) | Saliency segmentation-based content-based image retrieval method | |
CN103150562B (en) | A kind of ant group many cells tracker based on premise of image background extraction | |
CN106250818B (en) | A kind of total order keeps the face age estimation method of projection | |
Borkatulla et al. | Bangladeshi medicinal plant dataset | |
CN103903017A (en) | Face recognition method based on self-adaption soft histogram local binary patterns | |
CN110276350B (en) | Target detection method for marine ship |
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 |
Application publication date: 20120307 |