Video copy-move forgery detection and localization based on Tamura texture features

Copy-move forgery is one of the most common types of video forgeries. This paper proposed a new algorithm to detect and locate the copy-move forgery based on Tamura texture features. In the algorithm, the eigenvector matrix of the video is made up by extracting the Tamura texture features of every video frames. Then the eigenvector matrix of video is sorted by dictionary ordering and computing the differences between the feature vector and the adjacent feature vectors. And when the differences between two feature vectors less than the threshold, compare their distance of the serial number with the threshold, record the pairs of the serial numbers which is greater than the distance threshold. Finally, locate the copy-move sequences. The experimental results show that the precision of this method is higher than the previous relevant study.