CN105224917A - A kind of method and system utilizing color space to create skin color probability map - Google Patents
A kind of method and system utilizing color space to create skin color probability map Download PDFInfo
- Publication number
- CN105224917A CN105224917A CN201510570761.5A CN201510570761A CN105224917A CN 105224917 A CN105224917 A CN 105224917A CN 201510570761 A CN201510570761 A CN 201510570761A CN 105224917 A CN105224917 A CN 105224917A
- Authority
- CN
- China
- Prior art keywords
- skin
- probability
- color
- colour
- statistical form
- 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.)
- Granted
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
- G06V40/168—Feature extraction; Face representation
- G06V40/169—Holistic features and representations, i.e. based on the facial image taken as a whole
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of method and system utilizing color space to create skin color probability map, treat training coloured image and count, add up the number often organizing color in the number and the non-colour of skin often organizing color in the colour of skin, form training statistical form; Utilize statistical form to calculate often group color and belong to the probability of skin; To often organize color and belong to the probability normalization of skin, obtain skin color probability search table; Pre-service input picture; Judge whether face to be detected according to pretreated input picture, structure skin color probability map.It is more extensive that the present invention aims to provide a kind of scope of application, the skin color probability drawing generating method that processing speed is fast.
Description
Technical field
The invention belongs to image processing field, be specifically related to a kind of method and system utilizing color space to create skin color probability map.
Background technology
In recent years, along with camera is biometrics, mobile phone camera, whole people arrive in the epoch of photographing.Wherein a very important ingredient to the shooting of personage.Everybody all has sense of loving to make up and wearing beautiful clothes, and beautifying the portrait photo obtained also is one of focus.If know in picture, which position is skin, which position is not skin, and can be just subsequent treatment portrait photo, such as grind skin, whitening, makeups etc. provide important help.If manually smear out skin area, be comparatively loaded down with trivial details and complicated for common cellphone user.So the research utilizing image automatically to generate complexion model just seems particularly important.
Most is representational is in the world that MichaelJ.Jones proposes " StatisticalColorModelswithApplicationtoSkinDetection " for the algorithm of current acquisition complexion model.This method, by structure mixed Gauss model, calculates each pixel in figure and belongs to the probability of skin.But its algorithm effect is for unsatisfactory auto heterodyne image or darker image, and processing procedure is quite time-consuming.
Summary of the invention
In order to solve the problem, a kind of method and system utilizing color space to create skin color probability map of the present invention.It is more extensive that the present invention aims to provide a kind of scope of application, the skin color probability drawing generating method that processing speed is fast.
For achieving the above object, the technical solution used in the present invention is:
Utilize color space to create a method for skin color probability map, comprise following step:
(1) treat training coloured image to count, add up the number often organizing color in the number and the non-colour of skin often organizing color in the colour of skin, form training statistical form;
(2) utilize statistical form to calculate often group color and belong to the probability of skin;
(3) will often organize color and belong to the probability normalization of skin, obtain skin color probability search table;
(4) pre-service user input picture;
(5) judge whether face to be detected according to pretreated input picture, structure skin color probability map.
Further, described step (1) is specially, and comprises step:
To often opening coloured image to be trained, mark colour of skin position and non-colour of skin position;
And will often open coloured image to be trained, be YCbCrCg color space from RGB color space conversion;
Create the statistical form that six 256 row 256 arrange, each value in statistical form is used for counting the sum of all pixels belonging to corresponding color, six statistical forms are colour of skin Y-Cb statistical form respectively, colour of skin Y-Cr statistical form, colour of skin Y-Cg statistical form, non-colour of skin Y-Cb statistical form, non-colour of skin Y-Cr statistical form, non-colour of skin Y-Cg statistical form;
Each pixel of often opening in image to be trained is distinguished according to colour of skin position and non-colour of skin position and the difference of YCbCrCg color space, list corresponding statistical form in, form training statistical form;
Wherein, also need to add up all belong to the total number of pixel of the colour of skin and belong to the total number of pixel of the non-colour of skin.
Be further, colour of skin Y-Cb statistical form and non-colour of skin Y-Cb statistical form are one group, colour of skin Y-Cr statistical form and non-colour of skin Y-Cr statistical form are one group, colour of skin Y-Cg statistical form and non-colour of skin Y-Cg statistical form are one group, obtain Y-Cb, Y-Cr and Y-Cg tri-groups of statistical forms, always take up room as 256*256*3, far less than space 256*256*256*256 shared by YCbCrCg statistical form, also to take up room 256*256*256 far less than traditional RGB statistical form.
Further, described step (2) is specially, and comprises step:
Calculate total number of all pixels and skin and noncutaneous probability;
To often organize each value in colour of skin statistical form, and respectively divided by the total number of skin pixel, obtain in skin, comprising the probability often organizing color; To each value in the non-colour of skin statistical form of respective sets, respectively divided by the total number of non-skin pixel, obtain in non-skin, comprising the probability often organizing color;
Show that often organizing color is skin probability respectively according to Bayes rule.
Further, utilize three groups of statistical forms to calculate Y-Cb respectively, in Y-Cr, Y-Cg, often organize the probability that color belongs to skin.
Further, described step (3) is specially, and comprises step:
Respectively calculating is carried out to 256*256 the skin probability value often organized in statistical form and obtain average and standard deviation; According to obtained average and standard deviation, to each value of skin probability in corresponding group of statistical form, carry out Gaussian normalization processing; Data acquisition after process is formed skin color probability search table.
Wherein, probability normalization span is [0,1].
Further, described step (4) is YCbCrCg image by input picture from RGB color space conversion; And Face datection is carried out to input picture, obtain face rectangle frame.
Further, described step (5) calculates according to face rectangle frame detected in step (4), comprises step:
If face do not detected, for each pixel in input picture, by the value of its Y-Cb, Y-Cr and Y-Cg, search in obtained skin color probability search table, and obtain skin color probability by calculating; If face detected, distinguish average and the variance of R, G and channel B in calculating input image; To Gauss's skin probability of each this point of calculating in image; For each pixel in input picture, by the value of its Y-Cb, Y-Cr and Y-Cg, search in obtained skin color probability search table, and obtain pre-skin color probability by calculating; By Gauss's skin probability and pre-skin color probability, obtain final skin color probability.
Wherein, when carrying out average and variance and calculating, sort from small to large after each pixel of the face detected is calculated, less explanation the closer to, then only select collating sequence to calculate compared with the pixel of small end.Advanced line ordering, then by the pixel computation of mean values come above, thus filter out in the colour of skin pixels such as the tooth that accounts for a small amount of number or black eye ball, improve accuracy rate.
On the other hand, present invention also offers a kind of system utilizing color space to create skin color probability map, comprising:
Statistical form generation module, to treating that training coloured image counts in a large number, adding up the number of often group group color in the number and non-skin often organizing color in skin, forming statistical form;
Skin probability computing module, often organizes the probability that color belongs to skin in counting statistics table;
Probability normalization module, is normalized the probability often organizing color and belong to skin;
Input picture pretreatment module, carries out pre-service to input picture;
Skin color probability map forms module, judges whether can face be detected in input picture, thus structure skin color probability map;
Wherein, statistical form generation module endpiece is connected with skin probability computing module inlet end, and skin probability computing module endpiece connects probability normalized mode block entrance end; Input picture pretreatment module endpiece and skin color probability map form module inlet end and are connected, and skin color probability map forms module inlet end and is also connected with probability normalization module.
Adopt the beneficial effect of the technical program: a kind of method and system utilizing color space to create skin color probability map proposed by the invention, traditional YCbCr color space improves, add Cg Color Channel, thus the colour of skin is added up in YCbCrCg space, create Y-Cb, Y-Cr, Y-Cg tri-groups of statistical forms, always take up room as 256*256*3, far less than space 256*256*256*256 shared by YCbCrCg statistical form, also to take up room 256*256*256 far less than traditional RGB statistical form; The advanced line ordering when asking Gaussian mean, then by the pixel computation of mean values come above, thus filtering out in the colour of skin tooth accounting for a small amount of number, the pixels such as black eye ball improve accuracy rate; The look-up table of YCbCrCg color will be used, and by R, G, the gaussian probability that channel B calculates combines, and makes the result of Face Detection more accurate.
Accompanying drawing explanation
Fig. 1 is the method main flow chart utilizing color space to create skin color probability map of the present invention;
Fig. 2 is the process flow diagram creating skin color probability search table in embodiments of the invention;
Fig. 3 is the method flow diagram constructing skin color probability map in embodiments of the invention;
Fig. 4 be the method for the invention by system architecture sketch.
Embodiment
In order to make the object, technical solutions and advantages of the present invention clearly, below in conjunction with accompanying drawing, the present invention is further elaborated.
A kind of method utilizing color space to create skin color probability map shown in Figure 1, comprises following step:
(1) treat training coloured image to count, add up the number often organizing color in the number and the non-colour of skin often organizing color in the colour of skin, form training statistical form, as shown in Figure 2.
Concrete steps are:
A) to often opening coloured image to be trained, artificial mark colour of skin position and non-colour of skin position.Such as colour of skin position is all labeled as white, and non-colour of skin position is all labeled as black.
B) will often open coloured image to be trained, be YCbCrCg color space from RGB color space conversion, computing formula is as follows:
Y=0.299*R+0.587*G+0.114*B
Cb=-0.148*R-0.291*G+0.439*B+128
Cr=0.439*R-0.368*G-0.071*B+128
Cg=-0.318*R+0.439*G-0.121*B+128
Above-mentioned R, G, B are the value of three passages in RGB color space respectively, and Y, Cb, Cr, Cg are the values of YCbCrCg color space, and wherein the span of R, G, B, Y, Cb, Cr, Cg is [0,255].
C) create the statistical form that six 256 row 256 arrange, each value in statistical form is used for counting the sum of all pixels belonging to corresponding color.
Wherein, six statistical forms are colour of skin Y-Cb statistical form respectively, colour of skin Y-Cr statistical form, colour of skin Y-Cg statistical form, non-colour of skin Y-Cb statistical form, non-colour of skin Y-Cr statistical form, non-colour of skin Y-Cg statistical form; Wherein, colour of skin Y-Cb statistical form and non-colour of skin Y-Cb statistical form are one group, and colour of skin Y-Cr statistical form and non-colour of skin Y-Cr statistical form are one group, and colour of skin Y-Cg statistical form and non-colour of skin Y-Cg statistical form are one group.Create Y-Cb, Y-Cr, Y-Cg tri-groups of statistical forms, always take up room as 256*256*3, far less than space 256*256*256*256 shared by YCbCrCg statistical form, also to take up room 256*256*256 far less than traditional RGB statistical form.
Suppose Y [100] Cb [200]=30000 in colour of skin Y-Cb table, so represent all Y values in training process and be 100 and Cb value is 200, and the total number of the pixel belonging to skin area is 30000.
D) each pixel of often opening in image to be trained to be distinguished according to colour of skin position and non-colour of skin position and the difference of YCbCrCg color space, list corresponding statistical form in.
If it is colour of skin point according to artificial mark figure, so Y-Cb of colour of skin statistics point, the corresponding count value of Y-Cr, Y-Cg adds 1 respectively; If be not colour of skin point, the Y-Cb of so non-colour of skin statistics point, the corresponding count value of Y-Cr, Y-Cg adds 1 respectively.Add up all total number Skincount of pixel belonging to the colour of skin, number notSkincount. total with the pixel belonging to the non-colour of skin simultaneously
(2) utilize statistical form to calculate often group color and belong to the probability of skin, as shown in Figure 2.
Calculate total number of all pixels, skin and noncutaneous probability; To each value in one group of colour of skin statistical form, divided by the total number of skin pixel, obtain in skin, comprising the probability often organizing color; In like manner to each value in non-colour of skin statistical form, divided by the total number of non-skin pixel, obtain in non-skin, comprising the probability often organizing color; Show that often organizing color is skin probability respectively according to Bayes rule.
Calculate total number Totalcount=Skincount+notSkincount of all pixels, calculate the probability P skin=Skincount/Totalcount of skin afterwards.
In like manner calculate noncutaneous probability P notskin=notSkincount/Totalcount.
Above in 2 formulas, Pskin and Pnotskin span is [0,1].
To each value in colour of skin statistical form Y-Cb, divided by the total number Skincount of skin pixel, obtain the probability P 1 comprising certain group color Y, Cb in skin; In like manner to each value in non-colour of skin statistical form Y-Cb, divided by the total number notSkincount of non-skin pixel, obtain the probability P 2 comprising certain group color YCbCrCg in non-skin.
The span of P1 and P2 is [0,1], according to Bayes rule: P1*Pskin/ (P1*Pskin+P2*Pnotskin) obtains the probability that wherein a group of color Y, Cb are skin; In like manner utilize the method to calculate the probability of all 256*256 group colors, wherein Pbayes1 [Y] [Cb], the span of Y and Cb is: [0,255].
In like manner calculate 256*256 group color Y to colour of skin table Y-Cr and non-colour of skin table Y-Cr, Cr is probability P bayes2 [Y] [Cr] of skin, the span of Y and Cr: [0,255].
In like manner calculate 256*256 group color Y to colour of skin table Y-Cg and non-colour of skin table Y-Cg, Cg is probability P bayes3 [Y] [Cg] of skin, the span of a and b: [0,255].
The span of Pbayes1, Pbayes2, Pbayes3 is all 0 to just infinite.
(3) will often organize color and belong to the probability normalization of skin, obtain skin color probability search table, as shown in Figure 2.
To 256*256 value in the skin probability of one group of statistical form, carry out calculating and obtain average and standard deviation; Utilize institute's average that obtains and standard deviation, to each value of the skin probability of this group statistical form, carry out Gaussian normalization processing; Data after process are put into skin color probability search table.
To all 256*256 values in Pbayes1 [Y] [Cb], add up the number count1 of wherein non-zero value, all values are added up, obtain probability and ProSum1.
Average mean1=ProSum1/count1 afterwards.
Afterwards to each Pbayes1 [Y] [Cb], carry out calculating squre1, computing formula is: squre1=(Pbayes1 [Y] [Cb]-mean1) * (Pbayes1 [Y] [Cb]-mean1).
Again to all 256*256 values, cumulative square1, obtains squreSum.
With squreSum/cont1, obtain the variance sigma1 of Pbayes1 [Y] [Cb], afterwards standard deviation SDeviation1:SDeviation1=sqrt (sigma1) is obtained to sigma1 evolution.
Last again to each value in Pbayes1 [Y] [Cb], carry out Gaussian normalization processing:
Pbayes1[Y][Cb]=((Pbayes1[Y][Cb]-mean1)/(3*SDeviation1)+1)/2
Pbayes1 [Y] [Cb] after such process, if it is greater than 1 remaining, makes it equal 1 99% all in [0,1]; If be less than 0, it is made to equal 0.Finally make the span [0,1] of Pbayes1 [Y] [Cb].
In like manner, same normalized is carried out to Pbayes2 [Y] [Cr] and Pbayes3 [Y] [Cg].
Wherein, probability normalization span is [0,1].
(4) pre-service user input picture, as shown in Figure 3.
Be YCbCrCg image by input picture from RGB color space conversion; Face datection is carried out to input picture, obtains face rectangle frame.
(5) judge whether face to be detected according to pretreated input picture, structure skin color probability map, as shown in Figure 3.
If face do not detected, for each pixel in input picture, by the value of its Y-Cb, Y-Cr and Y-Cg, search in obtained skin color probability search table, and obtain skin color probability by calculating; If face detected, calculate average and the variance of R, G and channel B respectively; To Gauss's skin probability of each this point of calculating in image; For each pixel in input picture, by the value of its Y-Cb, Y-Cr and Y-Cg, search in obtained skin color probability search table, and obtain pre-skin color probability by calculating; By Gauss's skin probability and pre-skin color probability, obtain final skin color probability.
Wherein, when carrying out average and variance and calculating, sort from small to large after each pixel of the face detected is calculated, less explanation the closer to, then only select the preceding pixel of collating sequence to calculate.Advanced line ordering, then by the pixel computation of mean values come above, thus filter out in the colour of skin pixels such as the tooth that accounts for a small amount of number or black eye ball, improve accuracy rate.
A) were it not for and face detected, so for each pixel in image, by its Y, Cb, Cr, the value of Cg, search Pbayes1 [Y] [Cb], Pbayes2 [Y] [Cr], the value of Pbayes3 [Y] [Cg], last skin color probability is the quadratic sum evolution again of these three values, and computing formula is as follows:
Pout=sqrt(Pbayes1[Y][Cb]*Pbayes1[Y][Cb]+
Pbayes2[Y][Cr]*Pbayes2[Y][Cr]+
Pbayes3[Y][Cg]*Pbayes3[Y][Cg])
Make again Pout be greater than 1 equal 1.The span [0,1] of Pout.
If b) face detected, to all pixels in face rectangle frame, carry out pixel at R Color Channel and add up, obtain R passage and Rsum, and number of pixels Rectcount in rectangle frame.Obtain the average of R passage divided by Rectcount with Rsum, computing formula is as follows:
Ravg=Rsum/Rectcount
Wherein the span of Ravg is [0,255].
Afterwards again to each pixel in face rectangle frame, carry out a square calculating, computing formula is as follows:
Rsqure=(R-Ravg)*(R-Ravg)
Then Rsqure is sorted from small to large, less explanation is the closer to Ravg, then only select collating sequence first three/mono-(or 1/2nd) pixel, recalculate the Ravg ' at these positions, in like manner calculating acquisition Rsqure '=(R-Ravg ') and * (R-Ravg '), each Rsqure ' in face rectangle frame is added up and obtains Rsquresum.Then calculate the variance Rsigma of R passage, computing formula is as follows:
Rsigma=Rsquresum/Rectcount
In like manner obtain average and variance Gavg, the Gsigma of G passage, the average of channel B and variance Bavg, Bsigma.
To point each in image, calculate Pout according to method when face not detected.
To point each in image, utilize R, G, channel B value and the Ravg obtained, Rsigma, Gavg, Gsigma, Bavg, Bsigma calculate Gauss's skin probability of this point, and computing formula is as follows:
Pgaus=exp(-(R-Ravg)*(R-Ravg)/Rsigma
-(G-Gavg)*(G-Gavg)/Gsigma
-(B-Bavg)*(B-Bavg)/Bsigma)
Wherein, the span of Pgaus is [0,1].
By (c), two probability multiplications of (d) step, obtain last skin color probability:
Presult=Pout*Pgaus。Presult span [0,1].
Based on identical inventive concept, as shown in Figure 4, present invention also offers a kind of system utilizing color space to create skin color probability map, comprising:
Statistical form generation module, to treating that training coloured image counts in a large number, adding up the number often organizing color in the number and the non-colour of skin often organizing color in the colour of skin, forming statistical form;
Skin probability computing module, often organizes the probability that color belongs to skin in counting statistics table;
Probability normalization module, is normalized the probability often organizing color and belong to skin;
Input picture pretreatment module, carries out pre-service to input picture;
Skin color probability map forms module, judges whether can face be detected in input picture, thus structure skin color probability map;
Wherein, statistical form generation module endpiece is connected with skin probability computing module inlet end, and skin probability computing module endpiece connects probability normalized mode block entrance end; Input picture pretreatment module endpiece and skin color probability map form module inlet end and are connected, and skin color probability map forms module inlet end and is also connected with probability normalization module.
More than show and describe ultimate principle of the present invention and principal character and advantage of the present invention.The technician of the industry should understand; the present invention is not restricted to the described embodiments; what describe in above-described embodiment and instructions just illustrates principle of the present invention; without departing from the spirit and scope of the present invention; the present invention also has various changes and modifications, and these changes and improvements all fall in the claimed scope of the invention.The claimed scope of this reality invention is defined by appending claims and equivalent thereof.
Claims (9)
1. utilize color space to create a method for skin color probability map, it is characterized in that, comprise following step:
(1) treat training coloured image to count, add up the number often organizing color in the number and the non-colour of skin often organizing color in the colour of skin, form training statistical form;
(2) utilize statistical form to calculate often group color and belong to the probability of skin;
(3) will often organize color and belong to the probability normalization of skin, obtain skin color probability search table;
(4) pre-service user input picture;
(5) judge whether face to be detected according to pretreated input picture, structure skin color probability map.
2. a kind of method utilizing color space to create skin color probability map according to claim 1, it is characterized in that, described step (1) comprises step:
To often opening coloured image to be trained, mark colour of skin position and non-colour of skin position;
And will often open coloured image to be trained, be YCbCrCg color space from RGB color space conversion;
Create the statistical form that six 256 row 256 arrange, each value in statistical form is used for counting the sum of all pixels belonging to corresponding color, six statistical forms are colour of skin Y-Cb statistical form respectively, colour of skin Y-Cr statistical form, colour of skin Y-Cg statistical form, non-colour of skin Y-Cb statistical form, non-colour of skin Y-Cr statistical form, non-colour of skin Y-Cg statistical form;
Each pixel of often opening in image to be trained is distinguished according to colour of skin position and non-colour of skin position and the difference of YCbCrCg color space, list corresponding statistical form in, form training statistical form;
Wherein, also need to add up all belong to the total number of pixel of the colour of skin and belong to the total number of pixel of the non-colour of skin.
3. a kind of method utilizing color space to create skin color probability map according to claim 2, it is characterized in that, colour of skin Y-Cb statistical form and non-colour of skin Y-Cb statistical form are one group, colour of skin Y-Cr statistical form and non-colour of skin Y-Cr statistical form are one group, colour of skin Y-Cg statistical form and non-colour of skin Y-Cg statistical form are one group, obtain Y-Cb, Y-Cr and Y-Cg tri-groups of statistical forms.
4. a kind of method utilizing color space to create skin color probability map according to claim 3, it is characterized in that, described step (2) comprises step:
Calculate total number of all pixels and skin and noncutaneous probability;
To often organize each value in colour of skin statistical form, and respectively divided by the total number of skin pixel, obtain in skin, comprising the probability often organizing color; To each value in the non-colour of skin statistical form of respective sets, respectively divided by the total number of non-skin pixel, obtain in non-skin, comprising the probability often organizing color;
Show that often organizing color is skin probability respectively according to Bayes rule.
5. a kind of method utilizing color space to create skin color probability map according to claim 5, is characterized in that, utilize three groups of statistical forms to calculate Y-Cb respectively, often organize the probability that color belongs to skin in Y-Cr, Y-Cg.
6. a kind of method utilizing color space to create skin color probability map according to claim 6, it is characterized in that, described step (3) comprises step:
Respectively calculating is carried out to 256*256 the skin probability value often organized in statistical form and obtain average and standard deviation;
According to obtained average and standard deviation, to each value of skin probability in corresponding group of statistical form, carry out Gaussian normalization processing;
Data acquisition after process is formed skin color probability search table;
Wherein, probability normalization span is [0,1].
7. a kind of color space that utilizes according to claim 7 creates the method for skin color probability map, it is characterized in that, described step (4) is YCbCrCg image by input picture from RGB color space conversion; And Face datection is carried out to input picture, obtain face rectangle frame.
8. a kind of method utilizing color space to create skin color probability map according to claim 8, it is characterized in that, described step (5) calculates according to face rectangle frame detected in step (4), comprises step:
If face do not detected, for each pixel in input picture, by the value of its Y-Cb, Y-Cr and Y-Cg, search in obtained skin color probability search table, and obtain skin color probability by calculating;
If face detected, distinguish average and the variance of R, G and channel B in calculating input image; To Gauss's skin probability of each this point of calculating in image; For each pixel in input picture, by the value of its Y-Cb, Y-Cr and Y-Cg, search in obtained skin color probability search table, and obtain pre-skin color probability by calculating; By Gauss's skin probability and pre-skin color probability, obtain final skin color probability;
Wherein, when carrying out average and variance and calculating, sort from small to large after each pixel of the face detected is calculated, less explanation the closer to, then only select collating sequence to calculate compared with the pixel of small end.
9. utilize color space to create a system for skin color probability map, it is characterized in that, comprising:
Statistical form generation module, to treating that training coloured image counts in a large number, adding up the number of often group group color in the number and the non-colour of skin often organizing color in the colour of skin, forming statistical form;
Skin probability computing module, often organizes the probability that color belongs to skin in counting statistics table;
Probability normalization module, is normalized the probability often organizing color and belong to skin;
Input picture pretreatment module, carries out pre-service to input picture;
Skin color probability map forms module, judges whether can face be detected in input picture, thus structure skin color probability map;
Wherein, statistical form generation module endpiece is connected with skin probability computing module inlet end, and skin probability computing module endpiece connects probability normalized mode block entrance end; Input picture pretreatment module endpiece and skin color probability map form module inlet end and are connected, and skin color probability map forms module inlet end and is also connected with probability normalization module.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510570761.5A CN105224917B (en) | 2015-09-10 | 2015-09-10 | A kind of method and system using color space creation skin color probability map |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510570761.5A CN105224917B (en) | 2015-09-10 | 2015-09-10 | A kind of method and system using color space creation skin color probability map |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105224917A true CN105224917A (en) | 2016-01-06 |
CN105224917B CN105224917B (en) | 2019-06-21 |
Family
ID=54993877
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510570761.5A Active CN105224917B (en) | 2015-09-10 | 2015-09-10 | A kind of method and system using color space creation skin color probability map |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105224917B (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105701474A (en) * | 2016-01-15 | 2016-06-22 | 西安交通大学 | Method for identifying video smog in combination with color and appearance characteristics |
CN105894480A (en) * | 2016-05-03 | 2016-08-24 | 成都索贝数码科技股份有限公司 | High-efficiency facial beautification device easy for parallel realization |
CN105913400A (en) * | 2016-05-03 | 2016-08-31 | 成都索贝数码科技股份有限公司 | Device for obtaining high-quality and real-time beautiful image |
CN106570472A (en) * | 2016-11-02 | 2017-04-19 | 华为技术有限公司 | Skin color detecting method and device and terminal |
CN106981066A (en) * | 2017-03-06 | 2017-07-25 | 武汉嫦娥医学抗衰机器人股份有限公司 | A kind of interior face image dividing method based on the colour of skin |
CN106991379A (en) * | 2017-03-09 | 2017-07-28 | 广东欧珀移动通信有限公司 | Human body skin recognition methods and device and electronic installation with reference to depth information |
CN107274452A (en) * | 2017-05-31 | 2017-10-20 | 成都品果科技有限公司 | A kind of small pox automatic testing method |
CN107424115A (en) * | 2017-05-31 | 2017-12-01 | 成都品果科技有限公司 | A kind of colour of skin correction algorithm based on face key point |
CN108389234A (en) * | 2018-02-05 | 2018-08-10 | 深圳市恒晨电器有限公司 | A kind of heterochromatic detection algorithm of tablet LCD, storage device and mobile terminal |
CN111047619A (en) * | 2018-10-11 | 2020-04-21 | 展讯通信(上海)有限公司 | Face image processing method and device and readable storage medium |
CN111145086A (en) * | 2019-12-27 | 2020-05-12 | 北京奇艺世纪科技有限公司 | Image processing method and device and electronic equipment |
CN113656627A (en) * | 2021-08-20 | 2021-11-16 | 北京达佳互联信息技术有限公司 | Skin color segmentation method and device, electronic equipment and storage medium |
CN113657401A (en) * | 2021-08-24 | 2021-11-16 | 凌云光技术股份有限公司 | Probability map visualization method and device for defect detection |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040017938A1 (en) * | 2002-04-08 | 2004-01-29 | James Cooper | Method and apparatus for detecting and/or tracking one or more colour regions in an image or sequence of images |
CN1700238A (en) * | 2005-06-23 | 2005-11-23 | 复旦大学 | Method for dividing human body skin area from color digital images and video graphs |
CN1704966A (en) * | 2004-05-28 | 2005-12-07 | 中国科学院计算技术研究所 | Method for detecting pornographic images |
CN1763765A (en) * | 2004-10-21 | 2006-04-26 | 佳能株式会社 | Method, device and storage medium for detecting face complexion area in image |
CN101251898A (en) * | 2008-03-25 | 2008-08-27 | 腾讯科技(深圳)有限公司 | Skin color detection method and apparatus |
CN102236786A (en) * | 2011-07-04 | 2011-11-09 | 北京交通大学 | Light adaptation human skin colour detection method |
CN102324025A (en) * | 2011-09-06 | 2012-01-18 | 北京航空航天大学 | Human face detection and tracking method based on Gaussian skin color model and feature analysis |
-
2015
- 2015-09-10 CN CN201510570761.5A patent/CN105224917B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040017938A1 (en) * | 2002-04-08 | 2004-01-29 | James Cooper | Method and apparatus for detecting and/or tracking one or more colour regions in an image or sequence of images |
CN1704966A (en) * | 2004-05-28 | 2005-12-07 | 中国科学院计算技术研究所 | Method for detecting pornographic images |
CN1763765A (en) * | 2004-10-21 | 2006-04-26 | 佳能株式会社 | Method, device and storage medium for detecting face complexion area in image |
CN1700238A (en) * | 2005-06-23 | 2005-11-23 | 复旦大学 | Method for dividing human body skin area from color digital images and video graphs |
CN101251898A (en) * | 2008-03-25 | 2008-08-27 | 腾讯科技(深圳)有限公司 | Skin color detection method and apparatus |
CN102236786A (en) * | 2011-07-04 | 2011-11-09 | 北京交通大学 | Light adaptation human skin colour detection method |
CN102324025A (en) * | 2011-09-06 | 2012-01-18 | 北京航空航天大学 | Human face detection and tracking method based on Gaussian skin color model and feature analysis |
Non-Patent Citations (2)
Title |
---|
MICHAEL J.JONES等: "Statistical Color Models with Application to Skin Detection", 《INTERNATIONAL JOURNAL OF COMPUTER VISION》 * |
黄秀常: "基于改进YCbCr空间的肤色检测模式分析", 《计算机仿真》 * |
Cited By (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105701474B (en) * | 2016-01-15 | 2019-02-05 | 西安交通大学 | A kind of video smoke recognition methods of color combining and external physical characteristic |
CN105701474A (en) * | 2016-01-15 | 2016-06-22 | 西安交通大学 | Method for identifying video smog in combination with color and appearance characteristics |
CN105894480A (en) * | 2016-05-03 | 2016-08-24 | 成都索贝数码科技股份有限公司 | High-efficiency facial beautification device easy for parallel realization |
CN105913400A (en) * | 2016-05-03 | 2016-08-31 | 成都索贝数码科技股份有限公司 | Device for obtaining high-quality and real-time beautiful image |
CN106570472A (en) * | 2016-11-02 | 2017-04-19 | 华为技术有限公司 | Skin color detecting method and device and terminal |
CN106570472B (en) * | 2016-11-02 | 2019-11-05 | 华为技术有限公司 | A kind of skin color detection method, device and terminal |
CN106981066A (en) * | 2017-03-06 | 2017-07-25 | 武汉嫦娥医学抗衰机器人股份有限公司 | A kind of interior face image dividing method based on the colour of skin |
CN106981066B (en) * | 2017-03-06 | 2019-07-12 | 武汉嫦娥医学抗衰机器人股份有限公司 | A kind of interior face image dividing method based on the colour of skin |
CN106991379A (en) * | 2017-03-09 | 2017-07-28 | 广东欧珀移动通信有限公司 | Human body skin recognition methods and device and electronic installation with reference to depth information |
CN106991379B (en) * | 2017-03-09 | 2020-07-10 | Oppo广东移动通信有限公司 | Human skin recognition method and device combined with depth information and electronic device |
CN107274452B (en) * | 2017-05-31 | 2020-07-24 | 成都品果科技有限公司 | Automatic detection method for acne |
CN107424115A (en) * | 2017-05-31 | 2017-12-01 | 成都品果科技有限公司 | A kind of colour of skin correction algorithm based on face key point |
CN107274452A (en) * | 2017-05-31 | 2017-10-20 | 成都品果科技有限公司 | A kind of small pox automatic testing method |
CN108389234A (en) * | 2018-02-05 | 2018-08-10 | 深圳市恒晨电器有限公司 | A kind of heterochromatic detection algorithm of tablet LCD, storage device and mobile terminal |
CN108389234B (en) * | 2018-02-05 | 2021-10-08 | 深圳市恒晨电器有限公司 | Flat LCD (liquid Crystal display) heterochromatic detection method, storage device and mobile terminal |
CN111047619A (en) * | 2018-10-11 | 2020-04-21 | 展讯通信(上海)有限公司 | Face image processing method and device and readable storage medium |
CN111047619B (en) * | 2018-10-11 | 2022-09-30 | 展讯通信(上海)有限公司 | Face image processing method and device and readable storage medium |
CN111145086A (en) * | 2019-12-27 | 2020-05-12 | 北京奇艺世纪科技有限公司 | Image processing method and device and electronic equipment |
CN113656627A (en) * | 2021-08-20 | 2021-11-16 | 北京达佳互联信息技术有限公司 | Skin color segmentation method and device, electronic equipment and storage medium |
CN113656627B (en) * | 2021-08-20 | 2024-04-19 | 北京达佳互联信息技术有限公司 | Skin color segmentation method and device, electronic equipment and storage medium |
CN113657401A (en) * | 2021-08-24 | 2021-11-16 | 凌云光技术股份有限公司 | Probability map visualization method and device for defect detection |
CN113657401B (en) * | 2021-08-24 | 2024-02-06 | 凌云光技术股份有限公司 | Probability map visualization method and device for defect detection |
Also Published As
Publication number | Publication date |
---|---|
CN105224917B (en) | 2019-06-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105224917A (en) | A kind of method and system utilizing color space to create skin color probability map | |
CN109785258B (en) | Face image restoration method based on multi-discriminator generated countermeasure network | |
CN104036474B (en) | A kind of Automatic adjustment method of brightness of image and contrast | |
CN104200470B (en) | Blue screen image-matting method | |
CN108805839A (en) | Combined estimator image defogging method based on convolutional neural networks | |
CN103854261B (en) | The bearing calibration of colour cast image | |
CN109376596A (en) | Face matching process, device, equipment and storage medium | |
CN106326939A (en) | Parameter optimization method and system of convolutional neural network | |
CN109344742A (en) | Characteristic point positioning method, device, storage medium and computer equipment | |
CN102436668A (en) | Automatic Beijing Opera facial mask making-up method | |
CN108388905B (en) | A kind of Illuminant estimation method based on convolutional neural networks and neighbourhood context | |
CN102867295B (en) | A kind of color correction method for color image | |
CN105913051A (en) | Device and method for updating template library for face image recognition | |
CN103426194B (en) | A kind of preparation method of full animation expression | |
CN107657625A (en) | Merge the unsupervised methods of video segmentation that space-time multiple features represent | |
CN106303354A (en) | A kind of face specially good effect recommends method and electronic equipment | |
CN104636759A (en) | Method for obtaining picture recommending filter information and picture filter information recommending system | |
CN110188670A (en) | Face image processing process, device in a kind of iris recognition and calculate equipment | |
CN104392425B (en) | A kind of image enchancing method of the adjust automatically contrast based on face | |
CN106530309A (en) | Video matting method and system based on mobile platform | |
CN109587466A (en) | The method and apparatus of colored shadow correction | |
CN106097354A (en) | A kind of combining adaptive Gauss Face Detection and the hand images dividing method of region growing | |
CN104599297A (en) | Image processing method for automatically blushing human face | |
CN106097261B (en) | Image processing method, device, storage medium and terminal device | |
CN104504716A (en) | Solid wood floor paint automatic color blending method based on BP neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CP02 | Change in the address of a patent holder | ||
CP02 | Change in the address of a patent holder |
Address after: 610041 the 13 floor of No. 1 middle Tianfu Avenue, No. 1268, high-tech zone, Chengdu, Sichuan. Patentee after: Chengdu PinGuo Digital Entertainment Ltd. Address before: 610041 No. 216 South City Road, Chengdu hi tech Zone, Sichuan Patentee before: Chengdu PinGuo Digital Entertainment Ltd. |