An Efficiency Enhanced Cluster Expanding Block Algorithm for Copy-Move Forgery Detection

This paper presents an efficient scheme for detecting copy-move forgery tampering attacks. The copy-move forgery attack is defined as a region of an image is replaced by a copy of other region in the same image. This detection is useful for malicious modifying an image. The proposed scheme improves previous cluster expanding block scheme to clustering by mean and variance for reducing the computation time. Experimental results show that the proposed scheme requires fewer computation time. Although the overhead of preprocessing is an extra load that takes more time than previous cluster expanding block scheme, but the total computation time is still improved at least 10% comparing with previous study. Moreover, the using of block variance reduces the false positive rate.