No abstract

In the technique of copy move forgery, a certain portion of the digital image is copied and pasted into the same image, in order to hide this target portion. The tampered image looks very much original, since the target region despite being forged, has inherited the basic characteristics of the same image itself. Thus the detection for such kind of forgery is pretty difficult. This paper proposes a method to detect such forged regions using Hybrid Wavelet Transform. The combination of transforms used for the same is of Kekre Wavelet Transform and Hadamard Transform. The Hybrid Transform obtained is then applied to even order overlapping blocks, and the features extracted from each block are then compared with features of all other blocks. For the same image, the overlapping block sizes are varied and a subjective selection is done from the results obtained for each block size.