METHOD FOR POSITIONING IMAGE TAMPERING REGION BASED ON DCT COEFFICIENT

摘要

Disclosed is a method for positioning an image tampering region based on a DCT coefficient. The method comprises the following steps: establishing a DCT coefficient probability distribution model of an image to be detected; estimating an unknown parameter of the DCT coefficient probability distribution model; calculating to obtain a probability of tampering with a DCT coefficient of a certain DCT block in the image to be detected; using prior knowledge of a tampering region to obtain the probabilities of tampering with different DCT blocks in the image to be detected, and thus obtaining a final positioning result of the tampering region in the image to be detected. Since it is not necessary to embed watermark information in an image in advance, only the image per se is required to be analyzed without using a large number of training libraries to conduct machine learning, and the present invention can be effectively applied in content forensics environments of large-scale data communication, multimedia transmission and the like.