Chinese Journal of Computers   Full Text
Title	An Image Authentication Scheme Separating Tamper Detection from Tamper Location
Authors	WANG Guo-Dong LIU Fen-Lin WANG Ping GENG Nan-Nan
Address	(Information Engineering Institute, PLA Information Engineering University, Zhengzhou 450002)
Year	2007
Issue	No.10(1880—1888)
Abstract &   Background	Abstract An image authentication scheme is proposed to locate tampered area on the watermarked image and improve the security of existing watermarking algorithms. In this scheme, the original image is permuted in pixel-level and the Hypo-LSB and LSB of each pixel are set to zero, the location watermark is embedded into the Hypo-LSB of each pixel, then, the image is inversely permuted and divided into blocks, afterwards, the detection watermark of each block is embedded into its LSB. Theoretical analysis and simulation results show that the proposed algorithm can not only resists all kinds of attacks including Oracle and VQ attack, but also locate tampered area in pixel-level accurately as well as ensure the security. keywords fragile watermarking; tamper detection; tamper location; Oracle attack; vector quantization attack background The image authentication technology based on fragile watermarking has emerged as a research spotlight. Generally speaking, for fragile watermarking, the important factors including perceptual, localization capability and security. In recent years, a number of fragile watermarking algorithms have been proposed to defeat the VQ attack. Nonetheless, most of these methods, either sacrifice tamper localization accuracy or lead to security problem. In the former researches, the authors followed and analyzed a great majority of the relevant information. Meanwhile, the authors proposed a novel fragile watermarking scheme based on singular value decomposition(SVD) and 2D chaotic mapping in 2006, which aimed at improving the tamper localization accuracy and its security. And this paper was published in the 2nd Information Security Practice and Experience Conference(ISPEC 2006). This paper is a further extension of the above-mentioned work. After making a lot of relevant researches about image authentication, the authors present a novel method separating tamper detection from tamper localization. Extensive experiments show this algorithm can accurately locate tampered region in pixel-level as well as ensure the security. This research is supported by the National High Technology Research and Development Program (863 Program) of China under grant No.2006AA10Z409 and the Science-Technology Project of Henan Province of China under grant No.0623021500.