A study of scalability in video compression: Rate -distortion analysis and parallel implementation

Theoretical rate-distortion performance bounds are derived and evaluated for both layered and continuously rate scalable video compression algorithms which use a single motion-compensated prediction (MCP) loop. These bounds are derived using rate-distortion theory, and are based on an optimum mean-square error (MSE) quantizer. Consequently, the theory serves as a bound to all possible implementations of MCP scalable video coders which use MSE as a distortion measure. Parametric versions of the rate-distortion functions are derived which are based solely on the input power spectral density and accuracy of the MCP loop. The theory is applicable to scalable video coders which allow prediction drift, such as the data-partitioning and SNR-scalability schemes described in the MPEG-2 standard, as well as those with zero prediction drift such as fine granularity scalability MPEG-4. For video coders which allow prediction drift, MCP performed optimally in the encoder is shown to be a sufficient condition for stability of the decoder. Simulation of the optimal methods correspond well with the published results of actual system implementations. The theory is significant because it separates the effects of scalability from individual scalable video coder implementation artifacts, and can serve as a guide for potential increases in scalable video coder performance. The problem inherent with any digital image or digital video system is the large amount of bandwidth required for transmission or storage. This has driven the research area of image compression to develop more complex algorithms that compress images to lower data rates with better fidelity. One approach that can be used to increase the execution speed of these complex algorithms is through the use of parallel processing. The problem addressed here is the parallel implementation of the JPEG still image compression standard on the MasPar MP-1, a massively parallel SIMD computer. Developed here are two novel byte alignment algorithms which are used to efficiently input and output compressed data from the parallel system. Results are presented which show real-time performance is possible. Also discussed are several applications, such as motion JPEG, that can be used in multimedia systems