Image Copy-Move Forgery Detection Based on SIFT-BRISK

Abstract

Copy-move forgery is the most common type of image forgery. SIFT is widely used in copy-move forgery detection due to its excellent scale invariance and rotation invariance. However, the detection efficiency of the traditional SIFT-based method has not performed very well because its high-dimensional feature descriptors leads to a long time of feature extracting and matching. In this paper we propose an efficient method for copy-move forgery detection. First, for the forged image, we determine the SIFT keypoints with scale and position information. Then, we use BRISK algorithm to generate a binary feature descriptor for each keypoint. Finally, we can use Hamming distance to quickly match similar keypoints. The experimental results show that the proposed method obtains a significant improvement in the speed of forgery detection under the premise of better robustness.