An Authentication Method for Copy Areas of Images

This paper presents an effective automatic detection technique for copy areas of digital images. This technique adopts primary components analysis for small image blocks with fixed size, and generates the degraded dimensions representation. This representation is robust for small difference caused by noises or lossy compression. This technique detects the copy areas according to the ascending order of all image blocks. The effectiveness and robusticity for resisting authentic forgery, and the sensitivity for lossy compression or additive noises of this technique are analyzed in this paper. Some simulation results are also provided to show the advantages of this technique.