Detection of Seam Carving and Contrast Enhancement Operation Chain

Abstract

With the wide use of digital image processing software, it is easier to edit images. Digital image forensic technology is becoming vitally necessary to identify the content authenticity. Although numbers of detection algorithms have been proposed in the past years, there is not a perfect solution for detecting multi-operation problems. However, multi-operation in image editing process is more practical, so detection for multi-operation has more significance. In this paper, we focus on forensics of operations involving contrast enhancement and seam carving. In order to solve the two-manipulation forensic problem, we propose to calculate Markov one-step transition probability matrix on the second order difference array, and use the high dimensional feature of DCT coefficients for further classification. Both contrast enhancement followed by seam carving and seam carving followed by contrast enhancement can be detected correctly through the classification. A series of experiments are conducted to prove the accuracy of the proposed algorithm.