Research of Blind Forensics Algorithm on Digital Image Tampering

Abstract

With rapid development of the internet and the multimedia technology, the digital images tampering with blind forensics technology become a new research direction in the study of information security. The technology is based on the digital image, and it is only to realize image integrity and authenticity of the certification, so it is widely used in the civil fields such as news reports, judicial proof and military fields such as military intelligence analysis. So the research of it has great significance and broad application prospect.  In the paper, the exploration and research on the paste tampering and splicing tampering with blind forensics algorithm are made. The blind forensics algorithm based on radial Krawtchouk copy-and-paste invariant moment is proposed. As the current copy-and-paste blind forensics algorithm has low localization accuracy, and poor robustness of the post-processing in solving the problem. Based on sliding window block matching method and the radial Krawtchouk invariant moment, it proposes a copy-and-paste tampering with blind forensics algorithm. The experimental results show that the algorithm can effectively locate the tampered area, and it has very strong robustness of the rotating operation, JPEG compression, Gaussian noise, etc.