An efficiency enhanced cluster expanding block algorithm for copy-move forgery detection

Abstract

The proposed scheme detects the copy-move forgery detection regions through the invariant features extracted from each block. First, an image is divided into overlapping blocks, and seven invariant moments of the maximum circle area in each block are calculated as moment features. Two clustering features, denoted by mean and variance of these seven moment features, are acquired for block comparison to reduce computation time. Therefore, the proposed scheme takes limited computation time because the seven moment features in each block are only compared to other blocks under the intersection of closed mean and variance features. The copy-move forgery regions can be found by matching the detected blocks with relative distance calculation. Experimental results show that the adopted moment features are efficient for detecting rotational or flipped duplicated regions.