Chinese Journal of Computers   Full Text
Title	Research on Trustworthiness Evaluation Model for Digital Image in Blind Environment
Authors	HU Dong-Hui1),3) WANG Li-Na1),2) JIANG Xia-Qiu1)
Address	1)(School of Computer Science, Wuhan University, Wuhan 430079) 2)(Key Laboratory of Aerospace Information Security and Trust Computing of Ministry of Education, Wuhan 430072) 3)(Institute of Network and Information Security, School of Computer Science and Information, Hefei University of Technology, Hefei 230009)
Year	2009
Issue	No.4(675—687)
Abstract &   Background	Abstract With the emerging of various digital image process software and informatation hiding software, digital image are facing credit crisis. This paper discusses the security problem of digital image from the point of trustworthiness, and proposes the concept of digital image trustworthiness evaluation. Firstly, the digital image trustworthiness judging model and the digital image trustworthiness measuring model are proposed. The former is researched on different phases of digit image life cycle. The latter is based on hidden Markov model (HMM), and the digital image trustworthiness synthesis measuring model is proposed based on the method of fuzzy analytic hierarchy process. Digital image trustworthiness evaluation system in blind environment is further proposed. Finally, experimental results show that the digital image trustworthiness synthesis measuring model is effective. Keywords digital image; evaluation model; trustworthiness judging model; trustworthiness history measuring model; trustworthiness synthesis measuring model Background Many technologies have been developed aim to assure the credibility of digital image such as digital watermark and digital image Hash. But as described in many papers, those technologies may reduce image quality, increase produce cost, and the fact that the overwhelming majority of images on the internet do not contain digital watermark or have image Hash, so it is important to find other possible technology to measure the origin, integrity and security of digital images. Digital image forensic is a new developed technology which may resolve this problem, especially blind image forensic technology. But due to the present image forensic technology can only tell whether or not an image has been tampered, it can not measure the origin, integrity and security of digital images synthetically，neither can tell the degree of the trustworthiness of an suspect image. In this paper the authors think the trustworthiness evaluation of digital image is a synthesis measurement, for example the synthesis measurement can composes of image source reliability, tamper integrity and steganography security. They also think digital image manipulation history is also an important trustworthy nature of digital image, take blind digital image forensic as a kind of digital image trustworthiness evaluation method and raise it to the model level, called trustworthiness judging model. After the authors present the concept of trustworthiness evaluation, they propose two models, one is trustworthiness judging model, the other one is trustworthiness measuring model. The authors pay more attention to trustworthiness measuring model, based on HMM model propose a trustworthiness history measuring model, and based on method of fuzzy analytic hierarchy process propose a trustworthiness synthesis measuring model. The whole digital image trustworthiness evaluation system is proposed later. Experimental results prove the effective of digital image trustworthiness synthesis measuring model. This work is supported by the Major Research Project of National Natural Science Foundation of China (grant No.90718006), National Basic Research Program(973 Program) of China (grant No.2007CB310801), and the Ph.D. Programs Foundation of Ministry of Education of China (grant No.20070486107).