DESIGN OF PERCEPTUAL VIDEO ENCRYPTION ALGORITHMS FOR CONTENT PROVIDERS

Abstract

There is high demand by content providers of multimedia services such as Pay-TV News, Pay-per-View (PPV) video and Video-on-Demand (VoD) to expand their customer based through advertisement. However, the threats of unauthorized access, reproduction and re-distribution of the high quality product by illegal users’ remains high thereby affecting the returns of the content providers/producers. In this regard, perceptual video cryptosystems have become a highly active research area; developing strategies to achieve the aspirations of content providers/producers while maintaining some level of security. Apart from the limitation in scope of existing perceptual video encryption algorithms, the embedded audio signals in the source video are normally not encrypted and transmission efficiency is not also addressed. Consequently, this thesis employs scrambling and encoding techniques, together with residue number system (RNS) to address these limitations. The first proposal is a perceptual video encryption algorithm that combines rotation and unit anti-diagonal matrices to effectively scramble source video into cipher video with varying degradations in visual quality and enhanced security. The second proposal uses specially formulated orthogonal matrices to encrypt the source video through block encoding. Thirdly, a scheme to secure embedded audio signals is proposed using K-shuffle technique. To crown it all, the thesis proposes an efficient scheme that integrates residue number system (RNS) to the proposed video encryption schemes to ensure efficient transmission of cipher video across networks.