CN103065156A - Distinguishing method for digital camera image and computer-generated image based on white balance - Google Patents
Distinguishing method for digital camera image and computer-generated image based on white balance Download PDFInfo
- Publication number
- CN103065156A CN103065156A CN2012105638931A CN201210563893A CN103065156A CN 103065156 A CN103065156 A CN 103065156A CN 2012105638931 A CN2012105638931 A CN 2012105638931A CN 201210563893 A CN201210563893 A CN 201210563893A CN 103065156 A CN103065156 A CN 103065156A
- Authority
- CN
- China
- Prior art keywords
- image
- white balance
- digital camera
- computer generated
- computer
- 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
Landscapes
- Image Processing (AREA)
Abstract
The invention relates to the technique of multimedia forensics, in particular to a distinguishing method for a digital camera image and a computer-generated image based on white balance. The distinguishing method for the digital camera image and the computer-generated image based on white balance aims to resolve the problem of the identification of the truth of digital image contents. By means of a theory that an image taken by a digital camera has the trace of being disposed by the white balance but an image generated by a computer does not need the disposal process, the distinguishment between a digital camera image and a computer-generated image can be achieved by detecting the presence or the absence of the trace of being disposed by the white balance in the images to be detected. The distinguishing method for the digital camera image and the computer-generated image based on white balance can be used for a formulation of a digital image authentication system, a distinguishment of an original cartoon and a pirated cartoon and the like, and gain an initiative in identification methods of the truth of digital image contents with better performance.
Description
Technical field
The invention belongs to multimedia evidence obtaining field, relate in particular to a kind of digital camera images based on white balance and computer generated image differentiating method.
Background technology
The universal arrival that indicates digital Age of PC and internet.Digital Age convenient self-evident, but consequent social concern is also following, for example ' electronics crime '.Along with constantly soaring electronics crime case, digital evidence obtaining (digitalforensics) technology is arisen at the historic moment.It is the relevant information that will comprise digital signal generation, storage, transmission etc. as the required evidence of crime survey and legal affairs, Email for example, office documents, sound, video, image, internet records, internal storage data, GPS information etc.In a plurality of directions of digital evidence obtaining subject, the popular direction of multimedia forensic technologies for just having risen in recent years.Whether authenticity and the reliability of multimedia evidence obtaining main checking numerical information (such as digital image, video, audio frequency etc.), main application are for the reliability at court's checking digital evidence, can accept and believe etc.Nowadays, digital picture because of it is popularized rapidly, wide-scale distribution and very easily the characteristic such as modification caused people's extensive concern.Digital picture is seen everywhere, and newpapers and periodicals, magazine, network etc. be its route of transmission all.In this simultaneously, even photo editing software from strength to strength even is forged photo so that the ordinary people still can edit and revise digital picture by PC.Along with the upgrading of image editing software, tampered image, computer generated image are more and more true to nature, even are difficult to the naked eye distinguish its authenticity.Yet except some specific Entertainment Scene, in the time of most of, people need to know definitely the digital picture true and false of seeing.Particularly in case involve economic interests, individual reputation, spin, crime survey etc., the information that the authenticity of digital image content, the reliability in source must be confirmed especially.Unfortunately, forge image and be deep into already our life, the numerous areas such as science, politics, commerce, law, media all have been subject to impact in various degree.And traditional verification method is difficult to deal with the digital picture " prestige crisis " of today.In order better to tackle the safety problem of digital picture, the passive forensic technologies of digital picture is used and is given birth to.The passive forensic technologies of digital picture belongs to the research frontier that occurs in recent years, and it all has in fields such as medical science, military affairs, court evidence, news report, insurance risk investigation, ecommerce dabbles, and uses very extensive.Research both at home and abroad still is in the junior stage, though the technical lot of challenges that is faced with does not affect market to its active demand at all, application prospect is good.
Digital camera images and computer generated image differentiation are the popular research of in the passive forensic technologies of digital picture branches.In today that graphics software spreads unchecked, computer generated image can be accomplished the vivid effect of " mixing the spurious with the genuine ".How the differentiate between images content is digital camera is taken the natural picture that obtains, and is record and the reflection to real world; Or the just embodiment of computer mapping technology, content is also untrue, or even to the distortion of reality, and can mislead the public.These all are the problems that must answer out in the face of non-amusement (such as criminal investigation, court's proof etc.) scene.It also is the emphasis of this patent research.
Traditional method that is used for digital camera image and computer generated image differentiation is based on the differentiating method of Wavelet Detection.These class methods are often by carrying out multistage decomposition to small echo, and ask for a plurality of statistics to form distinguishing characteristic at each subband.In existing these class methods, common way is the progression of to distinguish precision and constantly increasing wavelet decomposition in order to improve, and ask for statistical nature at more subbands, be that cost exchanges the raising of distinguishing precision for being on the increase assumed condition, increasing the distinguishing characteristic dimension, this will inevitably cause the statistical nature dimension bigger than normal; Meanwhile, wavelet transformation can not well excavate the marginal information of image, and one of important difference between the artificial image that the captured real world images that obtains of digital camera and computing machine generate different with edge details just, this has also caused the differentiation precision still to remain the deficiency that improves.
Summary of the invention
The object of the invention is to overcome above-mentioned weak point, thereby solve in digital camera images and the computer generated image differentiation, the statistical nature dimension is excessive, the not high problem of differentiation precision.
The invention provides a kind of digital camera images based on white balance and computer generated image differentiating method, comprise following steps: step 1, set up image data base, comprise some digital camera images and some computer generated images in the image data base;
Step 2 is carried out statistical nature to each width of cloth image in the image data base and is extracted, and obtains set of eigenvectors;
Step 3 is sent into the set of eigenvectors of gained in the step 2 in the sorter and to be trained, and obtains for the differentiation model of distinguishing digital camera images and computer generated image;
Step 4 is carried out statistical nature to image to be measured and is extracted, and the gained proper vector is sent into sorter, distinguishes model by step 3 gained and judges.
And in step 2 and the step 4, the implementation of arbitrary image being carried out the statistical nature extraction comprises following substep,
Step 2.1 is selected n kind white balance algorithm, image is carried out the light source of n kind white balance algorithm and estimates;
Step 2.2 is estimated for step 2.1 gained n kind light source, calculates respectively R, G, and the gain coefficient on three passages of B is enumerated this 3 * n gain coefficient
Wherein c ∈ r, g, b}, i={1,2 ..., n};
Step 2.3 according to the 3 * n that tries to achieve in the step 2.2 gain factor, is passed through formula
Calculate 3 other * n statistical nature;
Step 2.4, the gain coefficient that step 2.2 is tried to achieve
As the 1st to the 3rd * n feature, the statistical nature that step 2.3 is tried to achieve
As the 3rd * n+1 to the 6 * n feature, be combined into final proper vector.
And, get n=5.
The present invention pursues the more excellent digital picture discrimination method of performance, be used for judging in the digital picture authentication whether image is that digital camera is taken the true picture that obtains, or the artificial image by the graphics software generation, and then for whether picture material truly provides reliable basis for estimation, be beneficial to the development of China's technique of criminal investigation research, the formulation of propelling relevant law.
Embodiment
The present invention is by analyzing the principle of digital camera imaging and the imaging of computerized mapping software, pointed out that digital camera takes pictures in the imaging process, must process this step through white balance, computer generated image then needn't be like this, then take this difference as starting point, design a kind of digital camera images based on white balance and computer generated image differentiating method.
The ultimate principle of the method design is as follows:
The basic skills that most white balances are processed is followed Von Kries model, 1. image light source is investigated the colour temperature of evaluate image that is:; 2. according to the result of color temperature estimation, set the color modulation gain factor of each color layers on the image; 3. according to the color modulation gain factor of setting, the modulation image color, the image of gained is the image that white balance was processed.
For with digital camera images, its image has passed through white balance and has processed, the colour temperature of light source is identical or be comparable to the standard sources colour temperature that white balance algorithm sets on the image, and computer generated image is quite different.Therefore, if the logarithmic code camera image carries out the white balance processing, be equivalent to for the second time white balance processing, its color modulation is changed not quite inevitable or even is not needed modulation.Process and computer generated image is carried out white balance, be equivalent to for the first time white balance processing this moment, its color modulation is changed relatively large.In other words, the white balance processing has or not and can judge by the assessment to color temp in the image generative process.Whether be to have passed through the problems such as which kind of white balance Processing Algorithm in the true picture taken of camera, the imaging process all in the unknown situation at image, estimate by known image being carried out multiple WHITE TONE, and extract statistical nature; The method that then adopts sorter to train the differentiation model realizes the differentiation of digital camera images and computer generated image.
Embodiment describes technical solution of the present invention in detail.
A kind of digital camera images and computer generated image differentiating method based on white balance provided by the invention, be used for judging in the digital picture authentication whether image is that digital camera is taken the true picture that obtains, or the artificial image that generates by graphics software, and then truly whether provide reliable basis for estimation for picture material.Technical solution of the present invention can adopt computer software technology to realize automatically operation.Describe by the following examples technical solution of the present invention in detail.
The embodiment realization flow comprises following steps:
Step 1, set up image data base:
Can set up the proprietary image data base for training digital camera images and computer generated image differentiation model during implementation, include digital camera images, computer generated image is some (can collect voluntarily, the database that also can adopt other people to set up is such as the DVMM database of Columbia University).
Step 2, statistical nature extracts:
Each width of cloth image in the image data base is all extracted proper vector according to the statistical nature extracting method, obtain the set of eigenvectors that all Characteristic of Image vectors consist of in the image data base.
Step 3, the training area sub-model:
The set of eigenvectors of gained in the step 2 sent in the sorter train, obtain for the differentiation model of distinguishing digital camera images and computer generated image;
Step 4, testing image is differentiated:
When judging such as needs whether image is computer generated image, extract the feature of testing image by statistical nature extracting method consistent in the step 2, the gained proper vector is sent into sorter, judges by the differentiation model that has trained.
For ease of implementing reference, the embodiment of the invention further provides the concrete solution procedure of the statistical nature extracting method of mentioning in the above-mentioned differentiating method as follows:
Step 2.1, image light source is estimated:
Embodiment gets n=5, image is carried out the first step operation of 5 kinds of white balance algorithm, be that light source estimates (4 kinds of white balance algorithm are arranged generally, Gray World, MaxRGB, Shades of Grey, and Grey Edge, but Grey Edge is divided into 2 kinds of situations according to the parameter difference, altogether 5 kinds of light source estimations technique).
Step 2.2, the extraction of feature set one:
5 kinds of light sources for step 2.1 are estimated, calculate respectively R, G, and the gain coefficient on three passages of B is enumerated this 15 gain coefficients
Wherein c ∈ r, g, b}, i={1,2 ..., 5}, with these 15 statistical natures as the 1st to the 15th feature;
Step 2.3, the extraction of feature set two:
15 gain factors according to trying to achieve in the step 2.2 pass through formula
Calculate 15 other statistical natures, with these 15 statistical natures as the 16th to the 30th feature;
Wherein, abs (.) expression is the operation that takes absolute value, and English is absolute, is the function representation that takes absolute value in all kinds of computereses (comprising the programming languages commonly used such as C language, VB language, Pascal language, Matlab language).
Step 2.4, the formation of final feature:
The feature that step 2.2 is tried to achieve is as the 1st to the 15th feature; The feature that step 2.3 is tried to achieve is as the 16th to the 30th feature, merges to be combined into final proper vector.
Specific embodiment described herein only is to the explanation for example of the present invention's spirit.Those skilled in the art can make various modifications or replenish or adopt similar mode to substitute described specific embodiment, but can't depart from spirit of the present invention or surmount the defined scope of appended claims.
Claims (3)
1. digital camera images and computer generated image differentiating method based on a white balance is characterized in that: comprise following steps,
Step 1 is set up image data base, comprises some digital camera images and some computer generated images in the image data base;
Step 2 is carried out statistical nature to each width of cloth image in the image data base and is extracted, and obtains set of eigenvectors;
Step 3 is sent into the set of eigenvectors of gained in the step 2 in the sorter and to be trained, and obtains for the differentiation model of distinguishing digital camera images and computer generated image;
Step 4 is carried out statistical nature to image to be measured and is extracted, and the gained proper vector is sent into sorter, distinguishes model by step 3 gained and judges.
2. as claimed in claim 1 based on digital camera images and the computer generated image differentiating method of white balance, it is characterized in that:
In step 2 and the step 4, the implementation of arbitrary image being carried out the statistical nature extraction comprises following substep,
Step 2.1 is selected n kind white balance algorithm, image is carried out the light source of n kind white balance algorithm and estimates;
Step 2.2 is estimated for step 2.1 gained n kind light source, calculates respectively R, G, and the gain coefficient on three passages of B is enumerated this 3 * n gain coefficient
Wherein c ∈ r, g, b}, i={1,2 ..., n};
Step 2.3 according to the 3 * n that tries to achieve in the step 2.2 gain factor, is passed through formula
Calculate 3 other * n statistical nature;
Step 2.4, the gain coefficient that step 2.2 is tried to achieve
As the 1st to the 3rd * n feature, the statistical nature that step 2.3 is tried to achieve
As the 3rd * n+1 to the 6 * n feature, be combined into final proper vector.
3. as claimed in claim 1 based on digital camera images and the computer generated image differentiating method of white balance, it is characterized in that: get n=5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210563893.1A CN103065156B (en) | 2012-12-21 | 2012-12-21 | Based on digital camera images and the computer generated image differentiating method of white balance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210563893.1A CN103065156B (en) | 2012-12-21 | 2012-12-21 | Based on digital camera images and the computer generated image differentiating method of white balance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103065156A true CN103065156A (en) | 2013-04-24 |
| CN103065156B CN103065156B (en) | 2016-01-13 |
Family
ID=48107779
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210563893.1A Active CN103065156B (en) | 2012-12-21 | 2012-12-21 | Based on digital camera images and the computer generated image differentiating method of white balance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103065156B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019085889A1 (en) * | 2017-10-30 | 2019-05-09 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | White balance processing method, electronic device and computer readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070258618A1 (en) * | 2006-05-05 | 2007-11-08 | Yun-Qing Shi | System and/or method for image tamper detection |
| CN101655912A (en) * | 2009-09-17 | 2010-02-24 | 上海交通大学 | Method for detecting computer generated image and natural image based on wavelet transformation |
-
2012
- 2012-12-21 CN CN201210563893.1A patent/CN103065156B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070258618A1 (en) * | 2006-05-05 | 2007-11-08 | Yun-Qing Shi | System and/or method for image tamper detection |
| CN101655912A (en) * | 2009-09-17 | 2010-02-24 | 上海交通大学 | Method for detecting computer generated image and natural image based on wavelet transformation |
Non-Patent Citations (1)
| Title |
|---|
| 王敏: "《中国优秀硕士学位论文全文数据库》", 15 August 2012, article "基于色温估计自动白平衡算法研究与实现", pages: 15-26 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019085889A1 (en) * | 2017-10-30 | 2019-05-09 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | White balance processing method, electronic device and computer readable storage medium |
| US10609352B2 (en) | 2017-10-30 | 2020-03-31 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | White balance processing method, electronic device and computer readable storage medium |
| US10812767B2 (en) | 2017-10-30 | 2020-10-20 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | White balance processing method, electronic device and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103065156B (en) | 2016-01-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Guo et al. | Fake colorized image detection | |
| Swaminathan et al. | Digital image forensics via intrinsic fingerprints | |
| CN108596141B (en) | Detection method and system for generating face image by deep network | |
| Gill et al. | A review paper on digital image forgery detection techniques | |
| CN104933721B (en) | Stitching image altering detecting method based on color filter array characteristic | |
| CN111444826B (en) | Video detection method, device, storage medium and computer equipment | |
| CN105184238A (en) | Human face recognition method and system | |
| US8503768B2 (en) | Shape description and modeling for image subscene recognition | |
| CN104408780A (en) | Face recognition attendance system | |
| Zhang et al. | No one can escape: A general approach to detect tampered and generated image | |
| Zhang et al. | Distinguishing photographic images and photorealistic computer graphics using visual vocabulary on local image edges | |
| CN101655912B (en) | Method for detecting computer generated image and natural image based on wavelet transformation | |
| CN114842524B (en) | Face false distinguishing method based on irregular significant pixel cluster | |
| Mani et al. | A survey on digital image forensics: Metadata and image forgeries | |
| Conotter et al. | Detecting photographic and computer generated composites | |
| CN104408736A (en) | Characteristic-similarity-based synthetic face image quality evaluation method | |
| CN103065156B (en) | Based on digital camera images and the computer generated image differentiating method of white balance | |
| CN117373136A (en) | Face counterfeiting detection method based on frequency mask and attention consistency | |
| Chen et al. | Identification of image global processing operator chain based on feature decoupling | |
| CN114329050A (en) | Visual media data deduplication processing method, device, equipment and storage medium | |
| KR20230072994A (en) | Method of processing image and apparatus thereof | |
| CN107424141A (en) | A kind of face-image method for evaluating quality based on probability block | |
| CN103065155B (en) | A kind of digital camera type identifier method based on CFA | |
| CN112434547A (en) | User identity auditing method and device | |
| CN114092743B (en) | Compliance detection method and device for sensitive picture, storage medium and equipment |
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: 20210707 Address after: 215000 unit 01, 5 / F, building a, 388 Xinping street, Suzhou Industrial Park, Suzhou City, Jiangsu Province Patentee after: BOOSLINK SUZHOU INFORMATION TECHNOLOGY Co.,Ltd. Address before: 430072 Hubei Province, Wuhan city Wuchang District of Wuhan University Luojiashan Patentee before: WUHAN University |
|
| TR01 | Transfer of patent right |