Research In The Technology Of Digital Image Tamper Traces Detection

The occurrence of digital image technology has brought a lot of benefits to human life.Digital images are so vivid that they are widely used almost everywhere.The technology of digital image processing initially born in the 1950s,it was mainly for improving the quality of image,and becomes a special science system later.A lot of excellent image processing software comes out due to the development of digital image processing science.There is no doubt that these software objects are not created for image forgery,but they provide convenience in a certain degree.In recent years,digital image fraud events emerge in an endless stream,which causes a bad influence.The image credibility is greatly suspected by the public.So the research of digital image forgery detection is of great importance.In this paper,some image forgery detection technologies are studied and improved.The job can be mainly divided into three parts:Firstly,we investigate the deficiency of the copy-move tamper detection algorithm which is based on SIFT.The copy-move detection algorithm based on SIFT is well performed in most cases,but it may be invalid when the tamper areas are smoothly enough.We upgrade this situation by image outline matching detection algorithm,and the detection result is advanced under certain conditions.Secondly,we propose a new tamper detection method for clipped double JPEG compression image based on the statistical characteristics of DCT coefficients and block effect caused by double JPEG compression.The offset of the two DCT grids can be obtained by computing the loss of information.Then we will trim the image according to the offset,and it will be available to estimate the primary compression quality factor.The tampered area can be located according to the difference of DCT coefficients.Experiments results show that the proposed method is robust and effective in detecting the clipped double JPEG compression image.Furthermore,the method is also applicable to the unclipped double JPEG compression image.Thirdly,we study the tamper detection technology of multiple JPEG compression image.When creating a digital forgery to a JPEG image,it is often involved uncertain times of compression,which may lead to the limitation of some current method for JPEG tamper detection.Multiple JPEG compression detection algorithm based on the consistency of quantization error is proposed here to deal with the situation that tampered images are compressed with quality factors ranging from low to high.The algorithm makes use of the consistency of the DCT coefficients quantization error between the unrecovered image and the recovered image when compressed with special quantization matrix to detect the compression traces.This technique is used to detect tampered image,and obtain a good result.