Robust Region Duplication Detection on Log-Polar Domain Using Band Limitation

Region duplication is one of the image-tampering techniques in which a part of image is copied and pasted into another region of the same image. In this paper, a robust duplication detection algorithm is proposed against severe degradations such as illumination changes, blurring, large scaling, noise contamination and JPEG compression. We introduce an adaptive phase correlation scheme in the log-polar domain, which is effective to locate the most salient local feature of an image patch on frequency band. By using the information collected from the band limitation, duplicated regions can be correctly located. Our contribution is to present a robust image duplication detection algorithm which can handle large scaling manipulation while preserving detection performance under other manipulations or degradations. We perform degradation and comparison test on various tampered images, and experimental results show that the proposed algorithm achieves satisfactory performance.