CN106469289A - Facial image sex-screening method and system - Google Patents
Facial image sex-screening method and system Download PDFInfo
- Publication number
- CN106469289A CN106469289A CN201510501402.4A CN201510501402A CN106469289A CN 106469289 A CN106469289 A CN 106469289A CN 201510501402 A CN201510501402 A CN 201510501402A CN 106469289 A CN106469289 A CN 106469289A
- Authority
- CN
- China
- Prior art keywords
- facial image
- sex
- gabor
- gabor wavelet
- algorithm
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Abstract
The invention provides a kind of facial image sex-screening method and system, wherein, facial image sex-screening system includes:Image collection module, control process module and memory module.Extract the Gabor wavelet feature vector of original facial image using Gabor algorithm and Memetic algorithm, and one face images in facial image database Gabor wavelet feature vector, and set up a feature vector set, concentrate in described characteristic vector and find out the immediate characteristic vector of Gabor wavelet feature vector with described original facial image, and using the sex of the facial image corresponding to described immediate characteristic vector as described original facial image sex.Carry out dimensionality reduction without using means such as PCA, the identification data of more separating capacity being obtained in shorter process time, improving recognition performance, thus improve recognition accuracy.
Description
Technical field
The present invention relates to Computer Image Processing and area of pattern recognition, especially a kind of facial image sex inspection
Survey method and system.
Background technology
With social constantly progressive and each side, fast and effectively auto authentication is compeled highly necessary
Ask, biometrics identification technology has obtained development at full speed in recent decades.A kind of inherent as people belongs to
Property, and there is very strong self stability and individual difference, biological characteristic becomes auto authentication
The most preferable foundation.Current biometrics identification technology mainly includes:Fingerprint recognition, retina identification,
Iris identification, Gait Recognition, hand vein recognition, recognition of face etc..Compared with other recognition methodss, face is known
Not due to itself having direct, friendly, convenient feature, user to its no any mental maladjustment, therefore
It is easy to be accepted by user, thus widely being studied and applying.In addition additionally it is possible to know to face
Other result is for further analysis, obtains many extra enriching such as the sex about people, expression, age
Information, extends the application prospect of recognition of face.And face is one of important biological characteristic, facial image
On contained substantial amounts of information, such as sex, age ethnic group, identity etc..Wherein, the sex identification of face
Function is just attempt to give the facial image according to input for the computer, judges the ability of user sex.Man-machine
The development of interaction technique (HCI) makes computer vision, artificial intelligence, in monitoring, GUI Human Machine Interface
Etc. aspect play an increasingly important role.With the progress of technology, based on asking of facial image pattern recognition
Topic is increasingly becoming the focus of Recent study.Including Face datection, face identification, face character (property
Not, age, expression, race etc.) identification etc. all kinds of identification problems.Just it is so that based on the Gender Classification of face
Computer can judge the process of its sex according to the facial image of input.The sex identification seemingly people of face
" inherent " ability, but allow computer to be identified being not easy to, even if there being a large amount of being derived to calculate
The effort of the research worker of the every field such as machine vision, pattern recognition, artificial intelligence, psychology.
Face technology belongs to machine learning category, and technology and system are required for experiencing data training process, that is,
A large amount of facial images and corresponding mark are given to algorithm together as input, and algorithm can be according to these training datas
Automatically learn corresponding model thus being used for practical application.Problem is identified for sex, generally most common
Method is to describe face characteristic by Gabor wavelet, then carries out dimension-reduction treatment with PCA, finally adopts and divides
Known to class model, supporting vector machine model SVM carries out Gender Classification, but method recognition performance of today is simultaneously
It is not fine so that existing sex recognition accuracy is not high.
Content of the invention
It is an object of the invention to provide a kind of facial image sex-screening method and system, to solve existing property
The not high problem of other recognition accuracy.
In order to achieve the above object, the invention provides a kind of facial image sex-screening method and system, its
In, facial image sex-screening method comprises the following steps:
Obtain an original facial image;
Extract the Gabor wavelet feature of described original facial image using Gabor algorithm and Memetic algorithm
Vector;
Select a facial image database, extract the Gabor wavelet of the face images in described facial image database
Characteristic vector, and set up a feature vector set;
The Gabor wavelet feature vector found out with described original facial image is concentrated to connect most in described characteristic vector
Near characteristic vector, using the facial image sex corresponding to described immediate characteristic vector as described original
The sex of facial image.
Preferably, in above-mentioned facial image sex-screening method, using Gabor algorithm and Memetic
Algorithm extracts the Gabor wavelet wave filter that the Gabor wavelet feature vector of described original facial image is used
It is defined as follows:
Wherein, x '=xcos θ+ysin θ,
Y '=- xsin θ+ycos θ,
X, y are each pixel coordinate in matrix, and λ is wavelength, and θ is the anglec of rotation,For phase place, γ is output
The transverse and longitudinal ratio of wavelength, b is bandwidth.
Preferably, in above-mentioned facial image sex-screening method, using Gabor algorithm and Memetic
Before algorithm extracts the Gabor wavelet feature vector of described original facial image, first to described original face figure
Obtain face gray level image as carrying out gray processing process, and described face gray level image is normalized
Obtain normalization facial image, extract described normalization face figure using Gabor algorithm and Memetic algorithm
The Gabor wavelet feature vector of picture.
Preferably, in above-mentioned facial image sex-screening method, described face gray level image is returned
One change is processed, and the step obtaining normalization facial image includes:
Using the AdaBoost algorithm based on Harr feature, described face gray level image is detected, work as detection
To carrying out next step during face information;
Obtain left and right eye coordinates using AdaBoost algorithm;
Rotate described face gray level image so that right and left eyes are the level of state, obtain one second face gray-scale maps
Picture;
Postrotational face gray level image is carried out with cutting making it wide and high is in predetermined ratio;
Face gray level image after cutting is zoomed in and out and obtains normalization facial image, described normalization face
The pixel of image is an intended pixel.
Preferably, in above-mentioned facial image sex-screening method, the scope of described predetermined ratio is
0.8: 1~1: 1.2.
Preferably, in above-mentioned facial image sex-screening method, the scope of described intended pixel is
80 × 80~500 × 500.
Preferably, in above-mentioned facial image sex-screening method, using Gabor algorithm and Memetic
The step that algorithm extracts the Gabor wavelet feature vector of described normalization facial image includes:
Using one group S Gabor wavelet wave filter, described normalization facial image is processed, each institute
State Gabor wavelet wave filter and obtain an eigenvalue;
S described eigenvalue is formed the Gabor wavelet feature vector of original facial image, described primitive man
The dimension of the Gabor wavelet feature vector of face image is S, and S is less than or equal to 100.
Preferably, in above-mentioned facial image sex-screening method, filtered using one group of S Gabor wavelet
Ripple device is processed to described normalization facial image, and each described Gabor wavelet wave filter obtains a spy
The step of value indicative includes:
Each Gabor wavelet wave filter is carried out dot product with described normalization face diagram, each Gabor is little
The dot-product operation of wave filter obtains an eigenvalue.
Preferably, in above-mentioned facial image sex-screening method, described characteristic vector concentrate find out with
The immediate characteristic vector of Gabor wavelet feature vector of described original facial image, and will be described closest
The step of the sex as described original facial image for the facial image sex corresponding to characteristic vector include:
Calculate the Gabor wavelet of each of described feature vector set characteristic vector and described original facial image
Mahalanobis distance between characteristic vector;
Described characteristic vector is found out by K- nearest neighbor algorithm and concentrates the Gabor wavelet with described original facial image
The immediate characteristic vector of characteristic vector, by the facial image sex corresponding to described immediate characteristic vector
Sex as described original facial image.
Present invention also offers a kind of facial image sex-screening system, including:
Image collection module, for obtaining an original facial image;
Control process module, extracts described original facial image using Gabor algorithm and Memetic algorithm
Gabor wavelet feature vector, and extract the Gabor wavelet feature of face images in a facial image database
Vector, and set up a feature vector set;
Memory module, for store described original facial image, the Gabor wavelet of described original facial image
Facial image in characteristic vector, described facial image database and described feature vector set.
In the facial image sex-screening method and system that the present invention provides, using Gabor algorithm and
Memetic algorithm extracts the Gabor wavelet feature vector of original facial image, and institute in a facial image database
There is the Gabor wavelet feature vector of facial image, and set up a feature vector set, in described feature vector set
In find out the immediate characteristic vector of Gabor wavelet feature vector with described original facial image, and by institute
State the sex as described original facial image for the sex of facial image corresponding to immediate characteristic vector.
Carry out dimensionality reduction without using means such as PCA, the identification of more separating capacity can be obtained in shorter process time
Data, improves recognition performance, thus improve recognition accuracy.
Brief description
Fig. 1 is the flow chart of facial image sex-screening method in the embodiment of the present invention;
Fig. 2 is the flow chart of step S2 in the embodiment of the present invention;
Fig. 3 is the flow chart of step S3 in the embodiment of the present invention;
Fig. 4 is the structural representation of facial image sex-screening system in the embodiment of the present invention.
Specific embodiment
Below in conjunction with schematic diagram, the specific embodiment of the present invention is described in more detail.According to following
Description and claims, advantages and features of the invention will become apparent from.It should be noted that, accompanying drawing all adopts
The very form of simplification and all using non-accurately ratio, only in order to convenient, lucidly aid in illustrating the present invention
The purpose of embodiment.
Embodiments provide a kind of facial image sex-screening method, specifically, as shown in figure 1, bag
Include following steps:
S1:Obtain an original facial image.
Wherein, described original facial image is will be for the other facial image of identification, described original face
Image is generally coloured image, the abundant information being comprised, if directly processed to this coloured image,
The speed processing is slower, and during the sex to described original facial image judges, is not required to
Want all information included in coloured image, therefore before carrying out next step, need to this cromogram
As being processed, reduce the color information that it is comprised, to improve speed and the efficiency of image procossing.Typically
, it is that gray proces are carried out to this coloured image, gray level image will be converted into by this coloured image, therefore need
Described original facial image is carried out gray processing process, thus obtaining a gray level image, i.e. described face ash
Degree image.
S2:Described face gray level image is normalized, obtains normalization facial image.
Specifically, as shown in Fig. 2 comprising the following steps:
S21:Using the AdaBoost algorithm based on Harr feature, described face gray level image is detected,
Carry out next step when face information is detected.
When not finding face information after described face gray level image is detected, then can report an error.
When face information is detected, then carry out next step.
S22:Obtain left and right eye coordinates using AdaBoost algorithm.
Detect the binocular information in described face gray level image, and obtain right and left eyes using AdaBoost algorithm and sit
Mark.
Modern left eye coordinates are:P1=(xi, yi);Formula (1)
Right eye coordinate is:P2=(xj, yj).Formula (2)
S23:Rotate described face gray level image so that right and left eyes are the level of state.
Rotate described face gray level image, the angle between adjustment eyes line and horizontal line, until eyes connect
Line is parallel with horizontal line, that is, so that eyes is the level of state.
In setting right and left eyes, distance in the heart is d, and the eyes line of centres is ω with the wedge angle of horizontal axis, as
Described face gray level image needs the angle value of rotation, and co-ordinate-type (1) and formula (2) using right and left eyes can
?:
Formula (3)
Described face gray level image then can be calculated needs the angle value ω of rotation as follows:
ω=arctan ((yi-yj)/(xi-xj));Formula (4)
Assume face center in described face gray level image coordinate be C (xe, ye), then can calculate:
Formula (5)
Formula (6)
Centered on face C, face rotation is carried out for the anglec of rotation with θ, after rotation, the coordinate at face center is (x ', y '),
Correlation computations formula is as follows:
X '=xe+cosθ(x-xe)+sinθ(y-ye);Formula (7)
Y '=ye+cosθ(y-ye)+sinθ(x-xe).Formula (8)
S24:Postrotational face gray level image is carried out with cutting making it wide and high is in predetermined ratio.
Postrotational face gray level image is carried out after cutting so as to wide and high is in predetermined ratio.
The scope of described predetermined ratio is 0.8: 1~1: 1.2.Preferably, described predetermined ratio is 1: 1.Namely
Say preferably so that the wide and high ratio of face gray level image after cutting is 1: 1.
S25:Face gray level image after cutting is zoomed in and out and obtains normalization facial image, described normalization
The pixel of facial image is an intended pixel.
Face gray level image after cutting is enlarged or reduces process, obtain normalization facial image,
The pixel making described normalization facial image is an intended pixel.The scope of described intended pixel is
80 × 80~500 × 500.Preferably, described intended pixel is 100 × 100.Return described in so can not only keeping
One change facial image on human face image information integrity moreover it is possible to reduce feature extraction after dimension, thus
Both ensure that the accuracy of judged result, improve processing speed again, improve efficiency.
During scaling, the scaling in its horizontal direction is:
Formula (9)
Scaling on vertical direction is:
β=100/2d;Formula (10)
Wherein, d is distance in the heart in two.
Then obtaining scaled matrix is:
Formula (11)
S3:The Gabor extracting described normalization facial image using Gabor algorithm and Memetic algorithm is little
Wave characteristic vector.
Introduce Memetic algorithm on the basis of Gabor algorithm Gabor wavelet wave filter is optimized.
Gabor wavelet wave filter after optimization is defined as follows:
Formula (12)
Wherein, x '=xcos θ+ysin θ, formula (13)
Y '=- xsin θ+ycos θ, formula (14)
X, y are each pixel coordinate in matrix, and λ is wavelength, and θ is the anglec of rotation,For phase place, γ is output
The transverse and longitudinal ratio of wavelength, b is bandwidth.
Specifically, as shown in figure 3, comprising the following steps:
S31:Using one group S Gabor wavelet wave filter, described normalization facial image is processed, often
Individual described Gabor wavelet wave filter obtains an eigenvalue.
Each Gabor filter matrix and described normalization facial image are carried out dot product, each Gabor is little
The dot-product operation of wave filter obtains an eigenvalue.
Specifically, described normalization facial image is divided into multiple regions, thus each Gabor is filtered
Device matrix carries out local dot product, thus each Gabor is little with each region in described normalization facial image
The dot-product operation of wave filter obtains an eigenvalue.
S32:S described eigenvalue is formed the Gabor wavelet feature vector of original facial image, described former
The dimension of the Gabor wavelet feature vector of beginning facial image is S.
Each Gabor wavelet wave filter obtains an eigenvalue, and S Gabor wavelet wave filter just has S
Individual eigenvalue, this S eigenvalue is arranged according to the order obtaining, thus constituting with characteristic vector,
It is the Gabor wavelet feature vector of original facial image.The Gabor wavelet of described original facial image is special
The dimension levying vector is identical with the number of Gabor wavelet wave filter, as S.Wherein, S is less than or equal to
100.
S4:Select a facial image database, extract the Gabor of the face images in described facial image database
Wavelet character vector, and set up a feature vector set.
Select a facial image database in disclosed facial image database, to each of this facial image database
Facial image all repeat the above steps S1, S2 and S3, and by each the face figure in this facial image database
The characteristic vector of picture and the sex shown by this facial image carry out corresponding.In fact, here it is one to this
The process of bright the proposed study of facial image sex-screening method, thus set up a feature vector set.
S5:Concentrate the Gabor wavelet feature vector found out with described original facial image in described characteristic vector
Immediate characteristic vector, and using the facial image sex corresponding to described immediate characteristic vector as institute
State the sex of original facial image.
First, calculate the Gabor of each of described feature vector set characteristic vector and described original facial image
Mahalanobis distance between wavelet character vector.
Then, find out described characteristic vector by K- nearest neighbor algorithm (kNN, k-NearestNeighbor) to concentrate
With the immediate characteristic vector of Gabor wavelet feature vector of described original facial image, will be described closest
The facial image sex corresponding to characteristic vector as described original facial image sex.
The embodiment of the present invention additionally provides a kind of facial image sex-screening system, specifically, as shown in figure 4,
Including:Control process module 101, memory module 102 and image collection module 103, described image obtains
Module 103 is used for obtaining an original facial image.Described control process module 101 utilize Gabor algorithm and
Memetic algorithm extracts the Gabor wavelet feature vector of described original facial image, and extracts a face figure
As the Gabor wavelet feature vector of face images in storehouse, and set up a feature vector set.Described storage
Module 102 is used for storing the Gabor wavelet feature arrow of described original facial image, described original facial image
Facial image in amount, described facial image database and described feature vector set.
To sum up, in facial image sex-screening method and system provided in an embodiment of the present invention, using Gabor
Algorithm and the Gabor wavelet feature vector of the Memetic algorithm original facial image of extraction, and a face figure
As the Gabor wavelet feature vector of face images in storehouse, and set up a feature vector set, in described spy
Levy the immediate characteristic vector of Gabor wavelet feature vector found out in vector set with described original facial image,
And using the sex of the facial image corresponding to described immediate characteristic vector as described original facial image
Sex.Carry out dimensionality reduction without using means such as PCA, more separating capacity can be obtained in shorter process time
Identification data, improve recognition performance, thus improve recognition accuracy.
Above are only the preferred embodiments of the present invention, the present invention is not played with any restriction effect.Appoint
What person of ordinary skill in the field, in the range of without departing from technical scheme, to the present invention
Disclose technical scheme and technology contents make any type of equivalent or modification etc. change, all belong to without departing from
The content of technical scheme, still falls within protection scope of the present invention.
Claims (10)
1. a kind of facial image sex-screening method is it is characterised in that comprise the following steps:
Obtain an original facial image;
Extract the Gabor wavelet feature of described original facial image using Gabor algorithm and Memetic algorithm
Vector;
Extract the Gabor wavelet feature vector of face images in a facial image database, and set up a feature
Vector set;
The Gabor wavelet feature vector found out with described original facial image is concentrated to connect most in described characteristic vector
Near characteristic vector, and using the sex of the facial image corresponding to described immediate characteristic vector as described
The sex of original facial image.
2. facial image sex-screening method as claimed in claim 1 is it is characterised in that utilize Gabor
The Gabor wavelet feature vector of algorithm and the Memetic algorithm described original facial image of extraction is used
Gabor wavelet wave filter is defined as follows:
Wherein, x'=xcos θ+ysin θ,
Y'=-xsin θ+ycos θ,
X, y are each pixel coordinate in matrix, and λ is wavelength, and θ is the anglec of rotation,For phase place, γ is output
The transverse and longitudinal ratio of wavelength, b is bandwidth.
3. facial image sex-screening method as claimed in claim 1 is it is characterised in that utilize Gabor
Before the Gabor wavelet feature vector of algorithm and the Memetic algorithm described original facial image of extraction, first right
Described original facial image carries out gray processing process and obtains face gray level image, and to described face gray level image
It is normalized and obtains normalization facial image, extract institute using Gabor algorithm and Memetic algorithm
State the Gabor wavelet feature vector of normalization facial image.
4. facial image sex-screening method as claimed in claim 3 is it is characterised in that to described face
Gray level image is normalized, and the step obtaining normalization facial image includes:
Using the AdaBoost algorithm based on Harr feature, described face gray level image is detected, work as detection
To carrying out next step during face information;
Obtain left and right eye coordinates using AdaBoost algorithm;
Rotate described face gray level image so that right and left eyes are the level of state;
Postrotational face gray level image is carried out with cutting making it wide and high is in predetermined ratio;
Face gray level image after cutting is zoomed in and out and obtains normalization facial image, described normalization face
The pixel of image is an intended pixel.
5. facial image sex-screening method as claimed in claim 4 is it is characterised in that described pre- definite proportion
The scope of example is 0.8:1~1:1.2.
6. facial image sex-screening method as claimed in claim 4 is it is characterised in that described pre- fixation
The scope of element is 80 × 80~500 × 500.
7. facial image sex-screening method as claimed in claim 3 is it is characterised in that utilize Gabor
The step bag of the Gabor wavelet feature vector of algorithm and the Memetic algorithm described normalization facial image of extraction
Include:
Using one group S Gabor wavelet wave filter, described normalization facial image is processed, each institute
State Gabor wavelet wave filter and obtain an eigenvalue;
S described eigenvalue is formed the Gabor wavelet feature vector of described original facial image, described former
The dimension of the Gabor wavelet feature vector of beginning facial image is S, and S is less than or equal to 100.
8. facial image sex-screening method as claimed in claim 7 is it is characterised in that utilize one group of S
Individual Gabor wavelet wave filter is processed to described normalization facial image, each described Gabor wavelet filter
The step that ripple device obtains an eigenvalue includes:
Each Gabor wavelet wave filter is carried out dot product with described normalization face diagram, each Gabor is little
The dot-product operation of wave filter obtains an eigenvalue.
9. facial image sex-screening method as claimed in claim 1 is it is characterised in that in described feature
The immediate characteristic vector of Gabor wavelet feature vector with described original facial image is found out in vector set,
And using the facial image sex corresponding to described immediate characteristic vector as described original facial image property
Other step includes:
Calculate the Gabor wavelet of each of described feature vector set characteristic vector and described original facial image
Mahalanobis distance between characteristic vector;
Described characteristic vector is found out by K- nearest neighbor algorithm and concentrates the Gabor wavelet with described original facial image
The immediate characteristic vector of characteristic vector, by the facial image sex corresponding to described immediate characteristic vector
Sex as described original facial image.
10. a kind of facial image sex-screening system is it is characterised in that include:
Image collection module, for obtaining an original facial image;
Control process module, extracts described original facial image using Gabor algorithm and Memetic algorithm
Gabor wavelet feature vector, and extract the Gabor wavelet feature of face images in a facial image database
Vector, and set up a feature vector set;
Memory module, for storing the Gabor wavelet of described original facial image, described original facial image
Facial image in characteristic vector, described facial image database and described feature vector set.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510501402.4A CN106469289A (en) | 2015-08-16 | 2015-08-16 | Facial image sex-screening method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510501402.4A CN106469289A (en) | 2015-08-16 | 2015-08-16 | Facial image sex-screening method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106469289A true CN106469289A (en) | 2017-03-01 |
Family
ID=58214859
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510501402.4A Pending CN106469289A (en) | 2015-08-16 | 2015-08-16 | Facial image sex-screening method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106469289A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1794264A (en) * | 2005-12-31 | 2006-06-28 | 北京中星微电子有限公司 | Method and system of real time detecting and continuous tracing human face in video frequency sequence |
CN104050457A (en) * | 2014-06-26 | 2014-09-17 | 浙江大学 | Human face gender identification method based on small sample training library |
CN104463243A (en) * | 2014-12-01 | 2015-03-25 | 中科创达软件股份有限公司 | Sex detection method based on average face features |
CN104463142A (en) * | 2014-12-26 | 2015-03-25 | 中科创达软件股份有限公司 | Gender identification method and device based on facial images |
CN104598888A (en) * | 2015-01-28 | 2015-05-06 | 广州远信网络科技发展有限公司 | Human face gender recognition method |
-
2015
- 2015-08-16 CN CN201510501402.4A patent/CN106469289A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1794264A (en) * | 2005-12-31 | 2006-06-28 | 北京中星微电子有限公司 | Method and system of real time detecting and continuous tracing human face in video frequency sequence |
CN104050457A (en) * | 2014-06-26 | 2014-09-17 | 浙江大学 | Human face gender identification method based on small sample training library |
CN104463243A (en) * | 2014-12-01 | 2015-03-25 | 中科创达软件股份有限公司 | Sex detection method based on average face features |
CN104463142A (en) * | 2014-12-26 | 2015-03-25 | 中科创达软件股份有限公司 | Gender identification method and device based on facial images |
CN104598888A (en) * | 2015-01-28 | 2015-05-06 | 广州远信网络科技发展有限公司 | Human face gender recognition method |
Non-Patent Citations (1)
Title |
---|
周家锐等: "基于Gabor小波与Memetic算法的人脸识别方法", 《电子学报》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Cament et al. | Fusion of local normalization and Gabor entropy weighted features for face identification | |
CN102542281B (en) | Non-contact biometric feature identification method and system | |
CN102194131B (en) | Fast human face recognition method based on geometric proportion characteristic of five sense organs | |
WO2019145578A1 (en) | Neural network image processing apparatus | |
CN109033940B (en) | A kind of image-recognizing method, calculates equipment and storage medium at device | |
CN102938065B (en) | Face feature extraction method and face identification method based on large-scale image data | |
Vijayan et al. | Twins 3D face recognition challenge | |
Khan et al. | A comparative analysis of gender classification techniques | |
CN104239769B (en) | Personal identification method and system based on finger vein features | |
CN111062292B (en) | Fatigue driving detection device and method | |
CN104915656B (en) | A kind of fast human face recognition based on Binocular vision photogrammetry technology | |
CN106203375A (en) | A kind of based on face in facial image with the pupil positioning method of human eye detection | |
Yu et al. | An eye detection method based on convolutional neural networks and support vector machines | |
CN111914616A (en) | Finger vein identification and anti-counterfeiting integrated method, device, storage medium and equipment | |
Hsu et al. | A novel eye center localization method for multiview faces | |
CN105893916A (en) | New method for detection of face pretreatment, feature extraction and dimensionality reduction description | |
CN104021384A (en) | Face recognition method and device | |
CN109977909B (en) | Finger vein identification method and system based on minutia area matching | |
Hu et al. | Fusion of region-based representations for gender identification | |
Long et al. | Near infrared face image quality assessment system of video sequences | |
Kim et al. | Facial age estimation via extended curvature Gabor filter | |
CN106469289A (en) | Facial image sex-screening method and system | |
Wu et al. | Spontaneous versus posed smile recognition via region-specific texture descriptor and geometric facial dynamics | |
Xia et al. | An accurate eye localization approach for smart embedded system | |
Bhowmik et al. | Polar fusion technique analysis for evaluating the performances of image fusion of thermal and visual images for human face recognition |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
EE01 | Entry into force of recordation of patent licensing contract | ||
EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20170301 Assignee: Shanghai Li Ke Semiconductor Technology Co., Ltd. Assignor: Leadcore Technology Co., Ltd. Contract record no.: 2018990000159 Denomination of invention: Face image gender detection method and system License type: Common License Record date: 20180615 |
|
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170301 |