CN103544692A - Blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on basis of statistical judgment - Google Patents
Blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on basis of statistical judgment Download PDFInfo
- Publication number
- CN103544692A CN103544692A CN201210243242.4A CN201210243242A CN103544692A CN 103544692 A CN103544692 A CN 103544692A CN 201210243242 A CN201210243242 A CN 201210243242A CN 103544692 A CN103544692 A CN 103544692A
- Authority
- CN
- China
- Prior art keywords
- sub
- block
- statistics
- detection method
- distorted
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Image Processing (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses a blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on the basis of statistical judgment, and relates to the field of image processing techniques and machine vision. The blind detection method includes Step 1), acquiring original data of pictures; Step 2), segmenting the JPED images; Step 3), computing DCT (discrete cosine transformation) coefficients; Step 4), extracting vectors and creating histograms; Step 5), judging tampered pixel blocks; Step 6), judging tampered areas; Step 7), eliminating interference noise; Step 8), positioning the tampered areas. Compared with certain existing detection method, the blind detection method has the advantages that shortcomings of high computational complexity, low computation speed and limited identification means of the traditional algorithm can be effectively overcome, and the authenticity of image data can be effectively confirmed.
Description
[technical field]
The present invention relates to image processing techniques and field of machine vision, in particular, the present invention relates to a kind of dual compression jpeg picture based on statistics judgement and distort blind checking method.
[background technology]
Along with the high speed development of computer technology, people utilize some image editing tools to revise, create become very simple of a width digital picture, are no lack of and have people to be not intended to or have a mind to, and even malice is transmitted the digital picture of forging through meticulously.A lot of fields are higher to the security requirement of digital picture, and any small change all may affect actual information to be expressed.Once the digital picture of distorting and forging is used to media, scientific discovery, military field, law court's material evidence etc., can society, science, politics, military each side be produced and be had a strong impact on undoubtedly.When recipient receives this class image information, often wish to screen judgement to image: whether this image is original image, and whether some pixel was tampered, if be tampered, can navigate to the pixel being tampered.
Traditionally, people, when differentiating the true and false of image, adopt the method for embed digital watermark or digital signature in digital picture in advance.Yet, these methods be easily subject to equipment cost and image source and recipient the factor such as separated impact and cannot popularize, so do not rely on any prior imformation to distort detection and location to a width digital picture, be that tool is of great significance.
Due to any, to distorting of digital picture, all can inevitably cause the especially variation of statistical nature of image internal data.As evidence of crime all can be left in any scene of a crime, the variation of these data characteristicses, has become again to detect digital picture evidence and foundation that whether process is distorted on the contrary.
[summary of the invention]
The object of the invention is to effectively overcome the defect that calculated amount is large, computing velocity is excessively slow, means of identification is limited of traditional algorithm, provide a kind of dual compression jpeg picture based on statistics judgement to distort blind checking method.
Technical scheme of the present invention is achieved in that it comprises the following steps:
Step1), obtain picture primary data information (pdi) I (M, N), I (M, N) is carried out to spatial alternation, be converted to brightness space Y(M, N), wherein, Y represents the Y component space in YCrCb space, MN is respectively row pixel count and the row pixel count of picture;
Step2), the cutting apart of jpeg image: by Y(M, N) be divided into nonoverlapping sub-block blk(i), i=1,2,3
Step3), calculate DCT coefficient: each sub-block blk (i) is carried out to DCT(discrete cosine transform), obtain the DCT coefficient of each sub-block;
Step4), extraction and the histogrammic foundation of vector: the DCT coefficient of each sub-block blk (i) is arranged to (getting a front n coefficient) by Z-type, obtain new sub-block sequence vecter (i), i=1,2,3 Extract the coefficient of each sub-block vecter (i) same sequence number, set up DCT coefficient histogram bar (i) model, i=1 wherein, 2,3
Step5), judgement tampered image: the periodic feature distributing according to correlative character between sub-block coefficient and histogram bar (i), the otherness that differentiation second-compressed and first compression coefficient histogram distribute; Join probability Principle of Statistics, calculates the probability that each sub-block may be distorted simultaneously;
Step6), judge tampered region: the judgement whether result that the threshold values obtaining by statistics rule is processed Step5 distorts;
Step7), eliminate interfering noise: with wave filter, abate the noise dry;
Step8), positioning tampering region: according to above-mentioned steps label tampered region;
In described Step2: big or small Wei8*8， unit sub-block blk(i) is pixel;
In described Step4: n is 21;
In described Step6, threshold values Threshold obtains in experiment according to statistics in advance, and the scope of threshold values Threshold is 0.51-0.60; Judge that the method whether block of pixels distorts is: lower than above-mentioned threshold values Threshold, think to distort piece, higher than threshold values Threshold, think the non-piece of distorting.
Beneficial effect of the present invention is: method of the present invention can be to given digital picture, judge whether it distorts through region, the means of distorting comprise and copy, paste, splice, scratch figure, synthetic etc., also can orient the region of distorting, thereby effectively confirm the true or false of view data if having.Compare with more existing detection methods, the present invention has effectively overcome the defect that traditional algorithm calculated amount is large, computing velocity is slow, means of identification is limited, provides a kind of dual compression jpeg picture based on statistics judgement to distort blind checking method.
[accompanying drawing explanation]
Fig. 1 is for to carry out after Y conversion being divided into mutual nonoverlapping 8*8 sub-block figure to original image;
Fig. 2 arranges by Z-type the DCT coefficient table of getting after dct transform;
Fig. 3 is the histogram that DCT coefficient that in Fig. 2, No. 2 blocks were processed without PS builds;
Fig. 4 is the histogram that DCT coefficient that in Fig. 2, No. 2 blocks were processed through PS builds;
Fig. 5 is the former figure that process is not distorted;
Fig. 6 is that the present invention tests final effect figure.
[embodiment]
Below in conjunction with drawings and Examples, the invention will be further described.
The dual compression jpeg picture of statistics judgement is distorted a blind checking method, and it comprises the following steps:
Step1), obtaining of data: read raw information, the former picture of cutting, obtains picture primary data information (pdi) I (M, N), making it pixel count M, N can be divided exactly by 8, cutting figure is carried out to spatial alternation and be converted to brightness space Y component, wherein, Y component is the Y component space in YCrCb space;
Step2), cutting apart of jpeg image: Y component is divided into mutual nonoverlapping 8*8 sub-block, and unit is pixel, as shown in Figure 1;
Step3), calculate DCT coefficient: each above-mentioned sub-block is carried out to discrete cosine transform, obtain the DCT coefficient of each sub-block;
Step4), extraction and the histogrammic foundation of vector: the DCT coefficient of each sub-block blk (i) is pressed to Z-type and arrange, get wherein front n coefficient, as shown in Figure 2, in the present embodiment, n gets 21, thereby obtains new sub-block sequence vecter (i), extract the coefficient of each sub-block vecter (i) same sequence number, set up DCT coefficient histogram bar (i) model (as Fig. 3, shown in 4), in bar (i), i=1,2,3,, n;
Step5), judgement distorted image block of pixels: the periodic feature distributing according to correlative character between sub-block coefficient and histogram bar (i), distinguish the otherness (comparison diagram 3 and Fig. 4) that second-compressed and first compression coefficient histogram distribute, while join probability Principle of Statistics, the probability of the sub-block that calculating may be distorted;
Step6), judge tampered region: the judgement whether result that the threshold values Threshold obtaining by statistics rule processes Step5 distorts, threshold values Threshold obtains in experiment according to statistics in advance, the scope of threshold values Threshold is 0.51-0.60, during practical application, according to the situation of a class photo, specifically determine definite value; Judge that the method whether block of pixels distorts is: lower than above-mentioned threshold values Threshold, think to distort piece, higher than threshold values Threshold, think the non-piece of distorting;
Step7), eliminate interfering noise: with wave filter, abate the noise dry;
Step8), positioning tampering region: according to above-mentioned steps label tampered region, as shown in Figure 5, Figure 6, the tampered region that in Fig. 6, red area is mark.
Method of the present invention can judge whether it distorts through region to given digital picture, and the means of distorting comprise and copy, paste, splice, scratch figure, synthetic etc., also can orient the region of distorting, thereby effectively confirm the true or false of view data if having.Compare with more existing detection methods, the present invention has effectively overcome the defect that traditional algorithm calculated amount is large, computing velocity is slow, means of identification is limited, provides a kind of dual compression jpeg picture based on statistics judgement to distort blind checking method.
Described above is only preferred embodiment of the present invention, and above-mentioned specific embodiment is not limitation of the present invention.In technological thought category of the present invention, can there is various distortion and modification, the retouching that all those of ordinary skill in the art make according to above description, revise or be equal to replacement, all belong to the scope that the present invention protects.
Claims (4)
1. the dual compression jpeg picture based on statistics judgement is distorted a blind checking method, it is characterized in that: it comprises the following steps:
Step1), obtain picture primary data information (pdi) I (M, N), I (M, N) is carried out to spatial alternation, be converted to brightness space Y(M, N), wherein, Y represents the Y component space in YCrCb space, MN is respectively row pixel count and the row pixel count of picture;
Step2), the cutting apart of jpeg image: by Y(M, N) be divided into nonoverlapping sub-block blk(i), i=1,2,3
Step3), calculate DCT coefficient: each sub-block blk (i) is carried out to DCT(discrete cosine transform), obtain the DCT coefficient of each sub-block;
Step4), extraction and the histogrammic foundation of vector: the DCT coefficient of each sub-block blk (i) is arranged to (getting a front n coefficient) by Z-type, obtain new sub-block sequence vecter (i), i=1,2,3 Extract the coefficient of each sub-block vecter (i) same sequence number, set up DCT coefficient histogram bar (i) model, i=1 wherein, 2,3
Step5), block of pixels is distorted in judgement: the periodic feature distributing according to correlative character between sub-block coefficient and histogram bar (i), the otherness that differentiation second-compressed and first compression coefficient histogram distribute; Join probability Principle of Statistics, calculates the probability that each sub-block is distorted simultaneously;
Step6), judge tampered region: the judgement whether result that the threshold values obtaining by statistics rule is processed step 5 distorts;
Step7), eliminate interfering noise: with wave filter, abate the noise dry;
Step8), positioning tampering region: according to upper step label tampered region.
2. a kind of dual compression jpeg picture based on statistics judgement according to claim 1 is distorted blind checking method, it is characterized in that: in described Step2: big or small Wei8*8， unit sub-block blk(i) is pixel.
3. a kind of dual compression jpeg picture based on statistics judgement according to claim 1 is distorted blind checking method, it is characterized in that: in described Step4: n is 21.
4. a kind of dual compression jpeg picture based on statistics judgement according to claim 1 is distorted blind checking method, it is characterized in that: in described Step6, threshold values Threshold obtains in experiment according to statistics in advance, and the scope of threshold values Threshold is 0.51-0.60; Judge that the method whether block of pixels distorts is: lower than above-mentioned threshold values Threshold, think to distort piece, higher than threshold values Threshold, think the non-piece of distorting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210243242.4A CN103544692A (en) | 2012-07-13 | 2012-07-13 | Blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on basis of statistical judgment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210243242.4A CN103544692A (en) | 2012-07-13 | 2012-07-13 | Blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on basis of statistical judgment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103544692A true CN103544692A (en) | 2014-01-29 |
Family
ID=49968112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210243242.4A Pending CN103544692A (en) | 2012-07-13 | 2012-07-13 | Blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on basis of statistical judgment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103544692A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106303524A (en) * | 2016-08-10 | 2017-01-04 | 上海交通大学 | Video dual compression detection method based on prediction residual abnormal patterns |
| CN108269221A (en) * | 2018-01-23 | 2018-07-10 | 中山大学 | A kind of JPEG weight contract drawing is as tampering location method |
| CN108537762A (en) * | 2017-12-29 | 2018-09-14 | 西安电子科技大学 | Secondary jpeg compressed image evidence collecting method based on the multiple dimensioned network of depth |
| WO2019024160A1 (en) * | 2017-08-04 | 2019-02-07 | 平安科技（深圳）有限公司 | Image tampering detection method and system, electronic device, and readable storage medium |
| WO2019061661A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技（深圳）有限公司 | Image tamper detecting method, electronic device and readable storage medium |
| CN116740015A (en) * | 2023-06-12 | 2023-09-12 | 北京长木谷医疗科技股份有限公司 | Medical image intelligent detection method and device based on deep learning and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706944A (en) * | 2009-11-03 | 2010-05-12 | 上海大学 | Quantization table evaluation based method for detecting JPEG image tampering |
| CN101989355A (en) * | 2009-07-31 | 2011-03-23 | 北京邮电大学 | Detection method for secondary compression of JPEG image |
| CN102208096A (en) * | 2011-05-26 | 2011-10-05 | 西安理工大学 | Image tamper detection and tamper localization method based on discrete wavelet transformation |
| CN102521606A (en) * | 2011-11-29 | 2012-06-27 | 中南大学 | Method for classifying pixel blocks of JPEG images and image falsification detecting and falsified area positioning methods based on same |
-
2012
- 2012-07-13 CN CN201210243242.4A patent/CN103544692A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101989355A (en) * | 2009-07-31 | 2011-03-23 | 北京邮电大学 | Detection method for secondary compression of JPEG image |
| CN101706944A (en) * | 2009-11-03 | 2010-05-12 | 上海大学 | Quantization table evaluation based method for detecting JPEG image tampering |
| CN102208096A (en) * | 2011-05-26 | 2011-10-05 | 西安理工大学 | Image tamper detection and tamper localization method based on discrete wavelet transformation |
| CN102521606A (en) * | 2011-11-29 | 2012-06-27 | 中南大学 | Method for classifying pixel blocks of JPEG images and image falsification detecting and falsified area positioning methods based on same |
Non-Patent Citations (2)
| Title |
|---|
| 刘助龙等: "《基于JPEG压缩的数字化地质资料篡改探测方法及应用》", 《中国有色金属学报》 * |
| 扈文斌等: "《基于量化表不一致性的JPEG图像篡改盲检测》", 《中国图象图形学报》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106303524A (en) * | 2016-08-10 | 2017-01-04 | 上海交通大学 | Video dual compression detection method based on prediction residual abnormal patterns |
| CN106303524B (en) * | 2016-08-10 | 2020-01-07 | 上海交通大学 | Video Double Compression Detection Method Based on Predictive Residual Variation Mode |
| WO2019024160A1 (en) * | 2017-08-04 | 2019-02-07 | 平安科技（深圳）有限公司 | Image tampering detection method and system, electronic device, and readable storage medium |
| WO2019061661A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技（深圳）有限公司 | Image tamper detecting method, electronic device and readable storage medium |
| CN108537762A (en) * | 2017-12-29 | 2018-09-14 | 西安电子科技大学 | Secondary jpeg compressed image evidence collecting method based on the multiple dimensioned network of depth |
| CN108269221A (en) * | 2018-01-23 | 2018-07-10 | 中山大学 | A kind of JPEG weight contract drawing is as tampering location method |
| CN108269221B (en) * | 2018-01-23 | 2021-08-06 | 中山大学 | A JPEG Recompression Image Tampering Location Method |
| CN116740015A (en) * | 2023-06-12 | 2023-09-12 | 北京长木谷医疗科技股份有限公司 | Medical image intelligent detection method and device based on deep learning and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gou et al. | Noise features for image tampering detection and steganalysis | |
| Mushtaq et al. | Digital image forgeries and passive image authentication techniques: a survey | |
| CN107067389B (en) | A kind of blind evidence collecting method of distorted image | |
| CN103561271B (en) | The moving target of static camera shooting is removed video spatial domain altering detecting method | |
| CN103544692A (en) | Blind detection method for tamper with double-compressed JPEG (joint photographic experts group) images on basis of statistical judgment | |
| Hussain et al. | Image forgery detection using multi-resolution Weber local descriptors | |
| CN104504669B (en) | A kind of medium filtering detection method based on local binary patterns | |
| CN103345758A (en) | Joint photographic experts group (JPEG) image region copying and tampering blind detection method based on discrete cosine transformation (DCT) statistical features | |
| CN102693522A (en) | Method for detecting region duplication and forgery of color image | |
| CN102567987A (en) | Method for detecting manual fuzzy operation trace in image synthesis tampering | |
| CN101916442A (en) | A robust localization method for tampered images using GLCM features | |
| CN104408728A (en) | Method for detecting forged images based on noise estimation | |
| CN110782442B (en) | Image artificial fuzzy detection method based on multi-domain coupling | |
| Muhammad et al. | A non-intrusive method for copy-move forgery detection | |
| CN109178831B (en) | A kind of belt conveyor longitudinal tear detection method | |
| CN104243970A (en) | 3D drawn image objective quality evaluation method based on stereoscopic vision attention mechanism and structural similarity | |
| CN110516527B (en) | Visual SLAM loop detection improvement method based on instance segmentation | |
| CN105374010A (en) | A panoramic image generation method | |
| Li et al. | Distinguishing computer graphics from photographic images using a multiresolution approach based on local binary patterns | |
| CN107507140A (en) | The outdoor scene vehicle shadow disturbance restraining method of highway of feature based fusion | |
| CN107563970A (en) | Method for detecting video image mosaic | |
| CN106021610B (en) | A kind of method for extracting video fingerprints based on marking area | |
| CN111429437B (en) | Object Detection-Oriented Image No-Reference Sharpness Quality Detection Method | |
| CN102609891B (en) | A Passive Blind Forensics Method of Digital Image Based on Texture Feature | |
| CN110796073B (en) | A method and device for detecting a specific target area in a textureless scene video |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140129 |