Layered Scalable and Low Complexity Video Encoding : New Approaches and Theoretic Analysis

Liu, Yuxin. Ph.D., Purdue University, August, 2004. Layered Scalable and Low Complexity Video Encoding: New Approaches and Theoretic Analysis. Major Professors: Edward J. Delp and Paul Salama. Transmission of digital video signals over current data networks demands efficient, reliable, and adaptable video coding techniques due to the heterogeneous nature of current wired and wireless networks. In this dissertation, we focus on the scalable video coding structure, in particular leaky prediction layered video coding (LPLC), and low complexity video encoding. Scalable video coding facilitates channel adaptive and error resilient performances. Low complexity video encoding shifts the computational complexity from the encoder to the decoder, which addresses applications with scarce resource at the encoder. We highlight a deficiency inherent in LPLC, namely that the enhancement layer cannot always “enhance” the rate distortion performance. We develop a general framework that applies to both LPLC and a multiple description coding scheme using motion compensation, and use this framework to confirm the existence of the deficiency. We propose an enhanced LPLC based on maximum-likelihood estimation to address the deficiency in LPLC. We further develop theoretic analysis of LPLC with respect to the leaky factor. We obtain two sets of rate distortion functions in closed form for LPLC, through the xxiv use of rate distortion theory and the use of a quantization noise model. Theoretical results of both closed form expressions are evaluated, which conform with the operational results. We describe a low complexity video encoding technique, which is developed for applications where resources are scarce at the video encoder whereas resources at the decoder are relatively abundant. We develop a low complexity video encoding approach that uses new B-frame direct coding modes. Experimental results have shown that our approach has a competitive rate distortion performance compared to the conventional high complexity video encoding approach. We also discuss the reliable transmission of digital video over an error-prone environment. We present a thorough evaluation of a joint source-channel video coding technique over wireless networks. We obtain reliable video transmission using adaptive packetization and a two-layer rate-distortion optimization scheme.