CN1704966A - Method for detecting pornographic images - Google Patents
Method for detecting pornographic images Download PDFInfo
- Publication number
- CN1704966A CN1704966A CN 200410042877 CN200410042877A CN1704966A CN 1704966 A CN1704966 A CN 1704966A CN 200410042877 CN200410042877 CN 200410042877 CN 200410042877 A CN200410042877 A CN 200410042877A CN 1704966 A CN1704966 A CN 1704966A
- Authority
- CN
- China
- Prior art keywords
- image
- face region
- skin
- human face
- area
- 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
- 238000000034 method Methods 0.000 title claims description 17
- 230000000007 visual effect Effects 0.000 claims abstract description 13
- 238000001514 detection method Methods 0.000 claims description 36
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 claims description 6
- 238000004040 coloring Methods 0.000 claims description 3
- 230000003068 static effect Effects 0.000 claims description 3
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 claims description 2
- YTAHJIFKAKIKAV-XNMGPUDCSA-N [(1R)-3-morpholin-4-yl-1-phenylpropyl] N-[(3S)-2-oxo-5-phenyl-1,3-dihydro-1,4-benzodiazepin-3-yl]carbamate Chemical compound O=C1[C@H](N=C(C2=C(N1)C=CC=C2)C1=CC=CC=C1)NC(O[C@H](CCN1CCOCC1)C1=CC=CC=C1)=O YTAHJIFKAKIKAV-XNMGPUDCSA-N 0.000 claims 1
- 238000010998 test method Methods 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000001815 facial effect Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
This invention discloses a test method for sexy images characterizing in removing images with few skin color pixels, carrying out man-face detect to judge if the image contains man-faces, setting up a body skin color model by the detected color distribution information of the man-face zone, detecting the body skin zone in the image, extracting the visual characters with high level semantic contents related to man-faces, finally sorting the images based on the extracted image visual characters to judge if the image is sexy.
Description
Technical field
The present invention relates to the method that a kind of pornographic image detects, specifically, it is based on the pornographic image detection method of the high-rise visual signature relevant with people's face information.
Background technology
The illegal propagation of interconnected network pornography image has brought a series of social concern and legal issue.How to protect the netizen not to be subjected to the harassing and wrecking of pornographic image, and prevent that the teenager from becoming a problem demanding prompt solution to the visit of pornographic image.A kind of solution route is that computer system judged in the past at display image whether picture material is Pornograph, and this just need be based on the pornographic image detection method of image vision content.
Existing based on the pornographic image detection method of image vision content in (for example, disclosed in the 198th page to the 203rd page of Asia computer vision meeting (ACCV 2004) collection of thesis in 2004, the visual signature of the token image vision content that is adopted only has rudimentary semantic content, these features comprise color, texture and shape etc.These features are limited to the descriptive power of picture material, and its described picture material and people differ greatly to the understanding of picture material.For example there are similar color characteristic in people's face close-up image and a pornographic image in two images, textural characteristics and shape facility probably.To these two images, utilize existing technology to be difficult to distinguish out which opens is pornographic image.
How be applicable to the senior semantic feature of describing the pornographic image content in the abstract image and utilize these features to detect pornographic image and become a difficult problem based on the pornographic image detection method of image vision content.
Summary of the invention
The technical problem to be solved in the present invention provides a kind of pornographic image detection method, overcomes the deficiency of the pornographic detection method of existing visual signature based on rudimentary semantic content in the image, improves the accuracy rate that pornographic image detects.
In order to solve the problems of the technologies described above, the invention provides a kind of pornographic image detection method, may further comprise the steps:
A) system initialization is set various threshold values according to user's request, sets up the complexion model of the overall situation;
B) complexion model of the imagery exploitation overall situation to input carries out Face Detection to image;
C) do you judge that whether the ratio of the skin pixel in the image is above preset threshold? if, execution in step d), if not, execution in step p);
D) people's face detects, judge that image comprises people's face? if, execution in step f), if not, execution in step e);
E) according to the lower-level vision feature of image, comprise color, shape and textural characteristics judge whether image is pornographic image, and testing process finishes;
F) utilize the colouring information of detected human face region to set up the human body complexion model, promptly characterize the color distribution of human body with the color distribution of human face region;
G) according to each human body complexion model in the enterprising pedestrian's body of entire image detection of skin regions;
H) result according to detection of skin regions generates colour of skin mask image;
I) according to the colour of skin mask image that generates, have the visual signature of senior semantic content in the abstract image, the visual signature of described senior semantic content comprises: the number of people's face in the image; The relative size of human face region in the image, this characterizing definition are the ratio of sum of all pixels in number of pixels in the human face region and the image; The relative distance of human face region and picture centre, this characterizing definition behaviour face regional center point and the distance of image center and the ratio of picture traverse or height; The ratio of whole area of skin color areas and human face region area in the image, this characterizing definition is the number of whole skin pixels in the image and the ratio of the number of pixels in the human face region; The maximum area of skin color area and the ratio of human face region area in the image, this characterizing definition are the ratio of the number of pixels in the maximum area of skin color and the number of pixels in the human face region in the image;
J) do you judge whether people's face number surpasses preset threshold? if, execution in step p); If not, execution in step k);
K) do you judge that whether the relative size of human face region is above preset threshold? if, execution in step p); If not, execution in step l);
L) do you judge that whether the relative distance of human face region position and picture centre is above preset threshold? if, execution in step m), if not, execution in step p);
M) do you judge in the image that whether the ratio of all area of skin color areas and human face region area is above preset threshold? if, execution in step n), if not, execution in step p);
N) do you judge that whether the ratio of area of skin color area maximum in the image and human face region area is above preset threshold? if, execution in step o), if not, execution in step p);
O) think that input picture is a pornographic image, testing process finishes;
P) think that input picture is a normal picture, testing process finishes.
In such scheme, step a) and b) in, described complexion model is the static model that off-line is set up.
In such scheme, in the step f), described complexion model is according to the dynamic model of the online foundation of color distribution information of detected human face region in the step d).
In such scheme, a human body complexion model in the step d) in the corresponding step f) of detected each human face region.
In such scheme, step h) colour of skin mask image in is the summation of each testing result in the step g).
As from the foregoing, the present invention detects by utilizing the high-rise visual signature relevant with people's face information to carry out pornographic image, can effectively improve accuracy rate, reduces error rate, especially reduces the generation of facial image being thought by mistake the situation of pornographic image.
Description of drawings
Fig. 1 is a pornographic image detection method process flow diagram of the present invention.
Embodiment
The present invention at first judges whether there is large-area area of skin color in the image by Face Detection, if there is not large-area area of skin color in image, is normal picture with spectral discrimination directly.Otherwise, in the image that contains the large tracts of land area of skin color, carry out people's face and detect, if detect the appearance of people's face, then contain the people in the key diagram picture, whether the people that subsequent step is used for authentication image exposes the excess skin zone.
As shown in Figure 1, the pornographic image detection method based on the high-rise visual signature relevant with people's face information involved in the present invention mainly may further comprise the steps:
Step 100 utilizes the complexion model of the overall situation that image is carried out Face Detection; In step 100, overall complexion model is the static model of setting up in the system initialization;
Does step 110 judge whether the ratio of the skin pixels in the image surpasses preset threshold? if, execution in step 120, if not, execution in step 230;
In the step 110, the skin pixel ratio is defined as the ratio of sum of all pixels in the number of skin pixel in the image and the image.
In step 120, people's face detects the number that the information of exporting comprises people's face in the image, the absolute position of human face region and the absolute size of human face region; The absolute position coordinate of personnel selection face regional center point in the plane of delineation of human face region characterizes, and the absolute size of human face region characterizes with the sum of all pixels that comprises in the human face region;
Step 140 utilizes the colouring information of detected human face region to set up the human body complexion model;
In step 140, characterize the distribution of human body complexion with the color distribution of human face region.
In step 150, utilize in the step 140 each the human body complexion model that generates all in image, to carry out one time Face Detection, suppose that human face detection tech detects 2 people's faces are arranged in the image, generates 2 human body complexion models of correspondence with it so in the step 150.Utilize each human body complexion model in image, to carry out Face Detection, judge whether the color value of each pixel in the image satisfies this human body complexion model,, judge that this pixel is a skin pixel if satisfy, otherwise, judge that this pixel is non-skin pixel.At last, each pixel in the image, as long as its color value satisfies in 2 human body complexion models any one, just this pixel of mark is a skin pixel in colour of skin mask image.
Step 160 generates colour of skin mask image according to the result of detection of skin regions; In step 160, colour of skin mask image is exactly that the skin pixel in the image is represented with different marks with non-skin pixel.A kind of mark of skin pixel, but not the another kind of mark of skin pixel is represented.Colour of skin mask image is a bianry image.
Step 170 according to the colour of skin mask image that generates, has the visual signature of senior semantic content in the abstract image;
In step 170, the visual signature of senior semantic content comprises the number of people's face in the image; The relative size of human face region in the image; The relative distance of human face region and picture centre; The ratio of whole area of skin color areas and human face region area in the image; The area of skin color area of maximum and the ratio of human face region area in the image.People's face number, these three information of the absolute position of human face region and the absolute size of human face region are provided by step 120.People's face number is exactly the people's face number that comprises in the detected image of people's face detection algorithm; The relative size of human face region is defined as the ratio of sum of all pixels in number of pixels in the human face region and the image in the image; The relative distance of human face region and picture centre is defined as the distance of human face region central point and image center and the ratio of picture traverse or height; The ratio of whole area of skin color areas and human face region area is defined as the sum of the pixel that is judged to be the colour of skin in the colour of skin mask image and the ratio of human face region colour of skin sum in the image; The long-pending ratio of maximum area of skin color area and people's face is defined as the ratio of the sum of all pixels human face region colour of skin sum in the area of skin color maximum in the colour of skin mask image in the image.
Does step 180 judge whether people's face number surpasses preset threshold? if, execution in step 230; If not, execution in step 190;
In step 180, the threshold value of predefined people's face number must be greater than 2.
Does step 190 judge whether the size of human face region surpasses preset threshold? if, execution in step 230; If not, carry out next step;
Whether step 220, the ratio of judging in the image area of maximum area of skin color and human face region area less than preset threshold, if, execution in step 230, if not, execution in step 240;
More than in each step used each threshold value require to pre-establish according to difference by the user to the performance of of the present invention and method, different threshold values may cause of the present invention and method has different accuracys rate.
Step 230 thinks that input picture is a normal picture;
Step 240 thinks that input picture is a pornographic image.
Be an instantiation below:
At first, collect the image that contains human body skin in a large number, statistics various YCbCr color values in these images appear as the probability of the colour of skin and the probability of the non-colour of skin respectively, according to Bayes rule, calculate the skin color probability value of every kind of YCbCr color value correspondence.Choose a colour of skin threshold value t,,, otherwise judge this non-skin pixel in pixel position if the skin color probability of the color value of pixel, judges that this pixel is a skin pixel greater than colour of skin threshold value t for each pixel in the image.The choosing of colour of skin threshold value t make the ratio maximum of just inspection rate of the colour of skin and false drop rate.Just inspection rate of the colour of skin is defined as the ratio of detected true skin pixel number and all skin pixel numbers.Colour of skin false drop rate is defined as the number of pixels that error-detecting is the colour of skin and the ratio of all non-skin pixel numbers.The skin color probability of shades of colour value correspondence and colour of skin threshold value t have constituted overall complexion model in the YCbCr color space.Preestablish following threshold value: the threshold value of people's face number, the threshold value of the relative size of human face region in the image; The threshold value of the relative distance of human face region and picture centre; The threshold value of the ratio of whole area of skin color areas and human face region area in the image; The threshold value of the ratio of the area of skin color area of maximum and human face region area in the image.More than each threshold value require to pre-establish according to difference by the user to the performance of of the present invention and method, different threshold values may cause of the present invention and method has different accuracys rate.
Image to input if the color space of image is not the YCbCr color space, then is converted to view data with the YCbCr color space and represents.Utilize overall complexion model to carry out Face Detection, the number of skin pixel in the statistical picture is if the ratio of skin pixel sum and total number of image pixels is normal picture with spectral discrimination directly less than preset threshold.The skin pixel sum is called the skin pixel ratio with the ratio of total number of image pixels.
Carrying out people's face for the skin pixel ratio greater than the image of preset threshold detects, people's face detection algorithm can adopt prior art, for example calendar year 2001 computer vision and pattern-recognition meeting (CVPR 2001) the collection of thesis first volume the 511st to the 518th page disclosed.The color of a people's facial skin is suitable substantially with his color at other position of (she) health, and therefore the colour of skin distribution situation of available face characterizes the colour of skin distribution of its health.The color of calculating all pixels in the detected human face region is at Y, Cb, and average and variance on three color components of Cr comprise: (μ
y, δ
y 2), (μ
Cb, δ
Cb 2), (μ
Cr, δ
Cr 2).Human body complexion is at Y, Cb, and each represents that with a normal distribution parameter assignment that characterizes normal distribution on each color component is average and the variance on the aforementioned human face region corresponding color component distribution situation of three color components of Cr.
For each outer pixel of human face region, if its color value (y, cb cr) satisfy condition | y-μ
y|＜τ
yδ
yAnd | cr-μ
Cr|＜τ
Crδ
CrAnd | cb-μ
Cd|＜τ
Cdδ
Cb, determine that it is skin pixels.τ
y, τ
Cr, τ
CbBe respectively three colour of skin threshold values on the color component.τ
y, τ
Cr, τ
CbBe initially set to 2.5,2,2 respectively.At first utilize initial colour of skin threshold value to carry out Face Detection, judge then whether each colour of skin among the initial detecting result has smooth texture together with the zone.If certain zone has smooth texture, think that then it is a skin area, otherwise each colour of skin threshold value all multiply by 0.9, carries out Face Detection with new threshold value in this district then, this process repeats, till the zone has smooth texture.
Detect and the result of Face Detection based on people's face, can be from the following feature of image contract: the number of people's face the image; The relative size of human face region in the image; The relative distance of human face region and picture centre; The ratio of whole area of skin color areas and human face region area in the image; The area of skin color area of maximum and the ratio of human face region area in the image.People's face number is exactly the people's face number that comprises in the detected image of people's face detection algorithm; The relative size of human face region is defined as the ratio of sum of all pixels in number of pixels in the human face region and the image in the image; The relative distance of human face region and picture centre is defined as the distance of human face region central point and image center and the ratio of picture traverse or height; The ratio of whole area of skin color areas and human face region area is defined as the sum of the pixel that is judged to be the colour of skin in the colour of skin mask image and the ratio of human face region colour of skin sum in the image; The long-pending ratio of maximum area of skin color area and people's face is defined as the ratio of the sum of all pixels human face region colour of skin sum in the area of skin color maximum in the colour of skin mask image in the image.
Utilize above-mentioned five features to judge whether image is pornographic image.At first, if the people's face in the image outnumbers preset threshold, image is a normal picture; Otherwise, then judge the relative size of human face region, if the relative size of human face region surpasses preset threshold, it is normal picture that process decision chart looks like; Otherwise, then judge the relative distance of human face region and picture centre again, if this relative distance less than preset threshold, the judgement image is a normal picture; Otherwise whether the ratio of then judging in the image all area of skin color areas and human face region area is less than preset threshold, if it is normal picture that process decision chart looks like; Otherwise, then judge in the image the maximum area of skin color area and the ratio of human face region area, if the ratio of the area of skin color area of maximum and human face region area less than preset threshold, if, the judgement image is a normal picture, otherwise for judging that image is a pornographic image.
It should be noted last that: above embodiment is the unrestricted technical scheme of the present invention in order to explanation only, although the present invention is had been described in detail with reference to the foregoing description, those of ordinary skill in the art is to be understood that: still can make amendment or be equal to replacement the present invention, and not breaking away from any modification or partial replacement of the spirit and scope of the present invention, it all should be encompassed in claim scope of the present invention.
Claims (6)
1, a kind of pornographic image detection method is characterized in that may further comprise the steps:
A) system initialization is set various threshold values according to user's request, sets up the complexion model of the overall situation;
B) complexion model of the imagery exploitation overall situation to input carries out Face Detection to image;
C) do you judge that whether the ratio of the skin pixel in the image is above preset threshold? if, carry out next step, if not, execution in step p);
D) people's face detects, judge that image comprises people's face? if, execution in step f), if not, carry out next step;
E) according to the lower-level vision feature of image, comprise color, shape and textural characteristics judge whether image is pornographic image, and testing process finishes;
F) utilize the colouring information of detected human face region to set up the human body complexion model, promptly characterize the color distribution of human body with the color distribution of human face region;
G) according to each human body complexion model in the enterprising pedestrian's body of entire image detection of skin regions;
H) result according to detection of skin regions generates colour of skin mask image;
I) according to the colour of skin mask image that generates, have the visual signature of senior semantic content in the abstract image, the visual signature of described senior semantic content comprises: the number of people's face in the image; The relative size of human face region in the image, this characterizing definition are the ratio of sum of all pixels in number of pixels in the human face region and the image; The relative distance of human face region and picture centre, this characterizing definition behaviour face regional center point and the distance of image center and the ratio of picture traverse or height; The ratio of whole area of skin color areas and human face region area in the image, this characterizing definition is the number of whole skin pixels in the image and the ratio of the number of pixels in the human face region; The maximum area of skin color area and the ratio of human face region area in the image, this characterizing definition are the ratio of the number of pixels in the maximum area of skin color and the number of pixels in the human face region in the image;
J) do you judge whether people's face number surpasses preset threshold? if, execution in step p); If not, carry out next step;
K) do you judge that whether the relative size of human face region is above preset threshold? if, execution in step p); If not, carry out next step;
L) do you judge that whether the relative distance of human face region position and picture centre is above preset threshold? if, carry out next step, if not, execution in step p);
M) do you judge in the image that whether the ratio of all area of skin color areas and human face region area is above preset threshold? if, carry out next step, if not, execution in step p);
N) do you judge that whether the ratio of area of skin color area maximum in the image and human face region area is above preset threshold? if, carry out next step, if not, execution in step p);
O) think that input picture is a pornographic image, testing process finishes;
P) think that input picture is a normal picture, testing process finishes.
2, a kind of pornographic image detection method as claimed in claim 1 is characterized in that, at step a) and b) in, described complexion model is the static model that off-line is set up.
3, a kind of pornographic image detection method as claimed in claim 1 is characterized in that, in step f), described complexion model is according to the dynamic model of the online foundation of color distribution information of detected human face region in the step d).
4, a kind of pornographic image detection method as claimed in claim 1 is characterized in that, a human body complexion model in the step d) in the corresponding step f) of detected each human face region.
5, a kind of pornographic image detection method as claimed in claim 1 is characterized in that step h) in colour of skin mask image be the summation of each testing result in the step g).
6, a kind of pornographic image detection method as claimed in claim 1 is characterized in that step j) in the threshold value of predefined people's face number greater than 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100428773A CN1323370C (en) | 2004-05-28 | 2004-05-28 | Method for detecting pornographic images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100428773A CN1323370C (en) | 2004-05-28 | 2004-05-28 | Method for detecting pornographic images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1704966A true CN1704966A (en) | 2005-12-07 |
| CN1323370C CN1323370C (en) | 2007-06-27 |
Family
ID=35577322
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2004100428773A Expired - Lifetime CN1323370C (en) | 2004-05-28 | 2004-05-28 | Method for detecting pornographic images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1323370C (en) |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100447808C (en) * | 2007-01-12 | 2008-12-31 | 郑文明 | Method for classification human facial expression and semantics judgement quantization method |
| CN101441717B (en) * | 2007-11-21 | 2010-12-08 | 中国科学院计算技术研究所 | Method and system for detecting eroticism video |
| CN101923652A (en) * | 2010-07-23 | 2010-12-22 | 华中师范大学 | Pornographic picture identification method based on joint detection of skin colors and featured body parts |
| CN102034107A (en) * | 2010-12-02 | 2011-04-27 | 西安电子科技大学 | Unhealthy image differentiating method based on robust visual attention feature and sparse representation |
| CN101360174B (en) * | 2007-08-03 | 2011-05-25 | 广达电脑股份有限公司 | Digital image classifying method and apparatus |
| CN102117413A (en) * | 2011-03-01 | 2011-07-06 | 金华就约我吧网络科技有限公司 | Method for automatically filtering defective image based on multilayer feature |
| CN102129575A (en) * | 2011-03-24 | 2011-07-20 | 成都四方信息技术有限公司 | Pornographic image analysis system based on color space skin color model |
| CN102163286A (en) * | 2010-02-24 | 2011-08-24 | 中国科学院自动化研究所 | Pornographic image evaluating method |
| US8014567B2 (en) * | 2006-07-19 | 2011-09-06 | Electronics And Telecommunications Research Institute | Method and apparatus for recognizing gesture in image processing system |
| CN101477703B (en) * | 2008-10-11 | 2011-09-14 | 大连大学 | Human body animation process directly driven by movement capturing data based on semantic model |
| CN102236796A (en) * | 2011-07-13 | 2011-11-09 | Tcl集团股份有限公司 | Method and system for sorting defective contents of digital video |
| CN102324036A (en) * | 2011-09-02 | 2012-01-18 | 北京新媒传信科技有限公司 | Obtain the method and apparatus of face complexion area in the image |
| CN102360435A (en) * | 2011-10-26 | 2012-02-22 | 西安电子科技大学 | Undesirable image detecting method based on connotative theme analysis |
| CN102521610A (en) * | 2011-12-08 | 2012-06-27 | 北京新媒传信科技有限公司 | Image filtering method and device |
| CN102547794A (en) * | 2012-01-12 | 2012-07-04 | 郑州金惠计算机***工程有限公司 | Identification and supervision platform for pornographic images and videos and inappropriate contents on wireless application protocol (WAP)-based mobile media |
| CN102542304A (en) * | 2012-01-12 | 2012-07-04 | 郑州金惠计算机***工程有限公司 | Region segmentation skin-color algorithm for identifying WAP (Wireless Application Protocol) mobile porn image |
| CN102567738A (en) * | 2012-01-06 | 2012-07-11 | 华南理工大学 | Rapid detection method for pornographic videos based on Gaussian distribution |
| CN103312770A (en) * | 2013-04-19 | 2013-09-18 | 无锡成电科大科技发展有限公司 | Method for auditing resources of cloud platform |
| US8611644B2 (en) | 2008-09-26 | 2013-12-17 | Tencent Technology (Shenzhen) Company Limited | Method and apparatus for training classifier, method and apparatus for image recognition |
| CN105224917A (en) * | 2015-09-10 | 2016-01-06 | 成都品果科技有限公司 | A kind of method and system utilizing color space to create skin color probability map |
| CN106446803A (en) * | 2016-09-07 | 2017-02-22 | 北京小米移动软件有限公司 | Live content recognition processing method, device and equipment |
| CN106454492A (en) * | 2016-10-12 | 2017-02-22 | 武汉斗鱼网络科技有限公司 | Live pornographic content audit system and method based on delayed transmission |
| CN108229353A (en) * | 2017-12-21 | 2018-06-29 | 深圳市商汤科技有限公司 | Sorting technique and device, electronic equipment, storage medium, the program of human body image |
| CN111199233A (en) * | 2019-12-30 | 2020-05-26 | 四川大学 | Improved deep learning pornographic image identification method |
| CN112651321A (en) * | 2020-12-21 | 2021-04-13 | 浙江商汤科技开发有限公司 | File processing method and device and server |
-
2004
- 2004-05-28 CN CNB2004100428773A patent/CN1323370C/en not_active Expired - Lifetime
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8014567B2 (en) * | 2006-07-19 | 2011-09-06 | Electronics And Telecommunications Research Institute | Method and apparatus for recognizing gesture in image processing system |
| CN100447808C (en) * | 2007-01-12 | 2008-12-31 | 郑文明 | Method for classification human facial expression and semantics judgement quantization method |
| CN101360174B (en) * | 2007-08-03 | 2011-05-25 | 广达电脑股份有限公司 | Digital image classifying method and apparatus |
| CN101441717B (en) * | 2007-11-21 | 2010-12-08 | 中国科学院计算技术研究所 | Method and system for detecting eroticism video |
| US8611644B2 (en) | 2008-09-26 | 2013-12-17 | Tencent Technology (Shenzhen) Company Limited | Method and apparatus for training classifier, method and apparatus for image recognition |
| CN101477703B (en) * | 2008-10-11 | 2011-09-14 | 大连大学 | Human body animation process directly driven by movement capturing data based on semantic model |
| CN102163286A (en) * | 2010-02-24 | 2011-08-24 | 中国科学院自动化研究所 | Pornographic image evaluating method |
| CN102163286B (en) * | 2010-02-24 | 2013-03-20 | 中国科学院自动化研究所 | Pornographic image evaluating method |
| CN101923652A (en) * | 2010-07-23 | 2010-12-22 | 华中师范大学 | Pornographic picture identification method based on joint detection of skin colors and featured body parts |
| CN102034107A (en) * | 2010-12-02 | 2011-04-27 | 西安电子科技大学 | Unhealthy image differentiating method based on robust visual attention feature and sparse representation |
| CN102034107B (en) * | 2010-12-02 | 2012-12-05 | 西安电子科技大学 | Unhealthy image differentiating method based on robust visual attention feature and sparse representation |
| CN102117413B (en) * | 2011-03-01 | 2012-11-14 | 金华就约我吧网络科技有限公司 | Method for automatically filtering defective image based on multilayer feature |
| CN102117413A (en) * | 2011-03-01 | 2011-07-06 | 金华就约我吧网络科技有限公司 | Method for automatically filtering defective image based on multilayer feature |
| CN102129575A (en) * | 2011-03-24 | 2011-07-20 | 成都四方信息技术有限公司 | Pornographic image analysis system based on color space skin color model |
| CN102236796B (en) * | 2011-07-13 | 2014-01-22 | Tcl集团股份有限公司 | Method and system for sorting defective contents of digital video |
| CN102236796A (en) * | 2011-07-13 | 2011-11-09 | Tcl集团股份有限公司 | Method and system for sorting defective contents of digital video |
| CN102324036B (en) * | 2011-09-02 | 2014-06-11 | 北京新媒传信科技有限公司 | Method and device for acquiring human face skin color region from image |
| CN102324036A (en) * | 2011-09-02 | 2012-01-18 | 北京新媒传信科技有限公司 | Obtain the method and apparatus of face complexion area in the image |
| CN102360435A (en) * | 2011-10-26 | 2012-02-22 | 西安电子科技大学 | Undesirable image detecting method based on connotative theme analysis |
| CN102521610B (en) * | 2011-12-08 | 2013-11-13 | 北京新媒传信科技有限公司 | Image filtering method and device |
| CN102521610A (en) * | 2011-12-08 | 2012-06-27 | 北京新媒传信科技有限公司 | Image filtering method and device |
| CN102567738B (en) * | 2012-01-06 | 2014-12-31 | 华南理工大学 | Rapid detection method for pornographic videos based on Gaussian distribution |
| CN102567738A (en) * | 2012-01-06 | 2012-07-11 | 华南理工大学 | Rapid detection method for pornographic videos based on Gaussian distribution |
| CN102547794A (en) * | 2012-01-12 | 2012-07-04 | 郑州金惠计算机***工程有限公司 | Identification and supervision platform for pornographic images and videos and inappropriate contents on wireless application protocol (WAP)-based mobile media |
| CN102542304A (en) * | 2012-01-12 | 2012-07-04 | 郑州金惠计算机***工程有限公司 | Region segmentation skin-color algorithm for identifying WAP (Wireless Application Protocol) mobile porn image |
| CN102547794B (en) * | 2012-01-12 | 2015-05-06 | 郑州金惠计算机***工程有限公司 | Identification and supervision platform for pornographic images and videos and inappropriate contents on wireless application protocol (WAP)-based mobile media |
| CN103312770A (en) * | 2013-04-19 | 2013-09-18 | 无锡成电科大科技发展有限公司 | Method for auditing resources of cloud platform |
| CN105224917A (en) * | 2015-09-10 | 2016-01-06 | 成都品果科技有限公司 | A kind of method and system utilizing color space to create skin color probability map |
| CN105224917B (en) * | 2015-09-10 | 2019-06-21 | 成都品果科技有限公司 | A kind of method and system using color space creation skin color probability map |
| CN106446803A (en) * | 2016-09-07 | 2017-02-22 | 北京小米移动软件有限公司 | Live content recognition processing method, device and equipment |
| CN106454492A (en) * | 2016-10-12 | 2017-02-22 | 武汉斗鱼网络科技有限公司 | Live pornographic content audit system and method based on delayed transmission |
| CN108229353A (en) * | 2017-12-21 | 2018-06-29 | 深圳市商汤科技有限公司 | Sorting technique and device, electronic equipment, storage medium, the program of human body image |
| CN111199233A (en) * | 2019-12-30 | 2020-05-26 | 四川大学 | Improved deep learning pornographic image identification method |
| CN112651321A (en) * | 2020-12-21 | 2021-04-13 | 浙江商汤科技开发有限公司 | File processing method and device and server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1323370C (en) | 2007-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1704966A (en) | Method for detecting pornographic images | |
| CN1301489C (en) | Method for extracting tooth area from tooth image and personal recongition method and device | |
| CN108121991B (en) | Deep learning ship target detection method based on edge candidate region extraction | |
| CN108830294A (en) | A kind of augmentation method of image data | |
| CN106951869B (en) | A kind of living body verification method and equipment | |
| US20090310861A1 (en) | Image processing | |
| CN109948566B (en) | Double-flow face anti-fraud detection method based on weight fusion and feature selection | |
| CN101980242A (en) | Human face discrimination method and system and public safety system | |
| CN111898538B (en) | Certificate authentication method and device, electronic equipment and storage medium | |
| CN1945628A (en) | Video frequency content expressing method based on space-time remarkable unit | |
| CN1975759A (en) | Human face identifying method based on structural principal element analysis | |
| CN104463134B (en) | A kind of detection method of license plate and system | |
| CN113989858B (en) | Work clothes identification method and system | |
| WO2023165616A1 (en) | Method and system for detecting concealed backdoor of image model, storage medium, and terminal | |
| CN104361339B (en) | Slap shape Graph Extraction and recognition methods | |
| CN108830882A (en) | Video abnormal behaviour real-time detection method | |
| CN109447064B (en) | A kind of duplicate rows License Plate Segmentation method and system based on CNN | |
| CN106682669A (en) | Image processing method and mobile terminal | |
| CN112418334B (en) | Method for identifying deformation fault of skirtboard grating of railway bullet train | |
| CN113705469A (en) | Face recognition method and device, electronic equipment and computer readable storage medium | |
| Abdallah et al. | A new color image database for benchmarking of automatic face detection and human skin segmentation techniques | |
| CN108510483A (en) | A kind of calculating using VLAD codings and SVM generates color image tamper detection method | |
| CN110086860B (en) | Data anomaly detection method and device under Internet of things big data environment | |
| CN109800758A (en) | A kind of natural scene character detecting method of maximum region detection | |
| CN111860661B (en) | Data analysis method and device based on user behaviors, electronic equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20181218 Address after: 100190 488, 4th floor, No. 10 South Road, Zhongguancun Academy of Sciences, Haidian District, Beijing Patentee after: Beijing Skyvein Net Computer Co.,Ltd. Address before: 100080 No. 6 South Road, Zhongguancun Academy of Sciences, Beijing, Haidian District Patentee before: Institute of Computing Technology, Chinese Academy of Sciences |
|
| TR01 | Transfer of patent right | ||
| CX01 | Expiry of patent term |
Granted publication date: 20070627 |