Blind Detection of Copy-Move Forgery in Digital Images Based on Dyadic Wavelet Transform
This paper proposed a detection algorithm for copy-move in same image based on dyadic wavelet transform. First of all, four sub images could be got through the decomposition of detecting image by dyadic wavelet transform. Secondly, high-frequency and low-frequency sub image were decomposed into blocks without any overlap and two sub image’s dyadic wavelet coefficients were regarded as the eigenvalue of the image block. At Last, both the high similarity among the low-frequency sub image blocks and the low similarity among the high-frequency sub image blocks were selected as a distorted image block. A kind of image edge processing methods was used to improve the tampering region at the same time. Through the experiments, it shows that the algorithm got the higher detection rate and lower error rates.
S.Z. Cai, Q.F. Zhang, X.P. Xu, D.H. Hu and Y.M. Qu
R. F. Zhang et al., "Blind Detection of Copy-Move Forgery in Digital Images Based on Dyadic Wavelet Transform", Advanced Materials Research, Vols. 989-994, pp. 4127-4131, 2014