Image Authentication Using Distributed Source Coding a Dissertation Submitted to the Department of Electrical Engineering and the Committee on Graduate Studies of Stanford University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

Image authentication is important in content delivery via untrusted intermediaries, such as peer-to-peer (P2P) file sharing. Many differently encoded versions of the original image might exist. In addition, intermediaries might tamper with the contents. Distinguishing legitimate diversity from malicious manipulations is the challenge addressed in this dissertation. We propose an approach using distributed source coding for the image authentication problem. The key idea is to provide a Slepian-Wolf encoded quantized image projection as authentication data. This version can be correctly decoded with the help of an authentic image as side information. Distributed source coding provides the desired robustness against legitimate variations while detecting illegitimate modification. The decoder incorporating expectation maximization (EM) algorithms can authenticate images which have undergone contrast, brightness, and affine warping adjustments. Our novel authentication system also offers tampering localization by using inference over a factor graph that represents tampering models. Video quality monitoring is closely related to the image authentication problem. We contribute an approach using distributed source coding. The video receiver sends the Slepian-Wolf coded video projections to the quality monitoring server which has access to the original video. Distributed source coding provides rate-efficient encoding of the projection by exploiting the correlation between the projections of the original and received videos. We show that the projections can be encoded at a low rate of just a few kilobits per second. Compared to the ITU-T J.240 Recommendation for remote PSNR monitoring, our scheme achieves a bit-rate which is lower by at least one order of magnitude.