An Efficient Detection of Copy-Move Forgery Using Phase Correlation

L. Chitirap Paavai^1,*, V. Vadivu¹, L. Krishnan²

¹Assistant Professor, Department of ECE, Knowledge Institute of Technology, Salem, Tamil Nadu, India

²Assistant Professor, Department of Civil, Knowledge Institute of Technology, Salem, Tamil Nadu, India

Emails: lcece@kiot.ac.in; vvaece@kiot.ac.in; lkcivil@kiot.ac.in

Abstract

Creating images is one of the main focuses of digital image processing. There are multiple techniques to spot image fraud. This work proposes a new approach to detect attacks that mimic Copy-Move forgeries. The proposed method applies DWT on the input image to create a reduced dimensional representation of the image. After that, the compressed image is divided into overlapping blocks. After these blocks are sorted, phase correlation is utilized as a similarity criterion to find duplicate blocks. Due to DWT usage, the lowest-level picture representation is first employed for detection. This work also covers the examination of numerous limits that are imposed to the input image, and the results are used in the analysis that follows.

Keywords: Copy-Move forgery; Digital forensics; DWT; Phase correlation