Fast Detection of Copy-Move Forgery Image using Two Step Search Algorithm

Abstract

In this paper, we proposed a new fast detection method of copy-move forgery image using two step search algorithm in the spatial domain. We proposed a new two step search algorithm for copy-moved forgery image detection. The performance of the proposed method is experimented on several forged images. Our two step search algorithm reduced 96.82% computational complexity more than conventional algorithms for copy-move forgery image detection. The block distortion measure (BDM) of the two step search algorithm is sufficient by 2 pixels checking points instead of 64 pixels checking points for exhaustive search in the not copy-moved forgery image regions. Most blocks of the images constitute not copy-moved forgery images. Therefore, we can reduce computational complexity more than conventional methods. We didn’t use any exhaustive search method and frequency domain (ex. DCT, Wavelet Transform) to reduce computational complexity in this paper.