Distributed video through telecommunication networks using fractal image compression techniques

Abstract

The research presented in this thesis investigates the use of fractal compression techniques for a real time video distribution system. The motivation for this work was that the method has some useful properties which satisfy many requirements for video compression. In addition, as a novel technique, the fractal compression method has a great potential. In this thesis, we initially develop an understanding of the state of the art in image and video compression and describe the mathematical concepts and basic terminology of the fractal compression algorithm. Several schemes which aim to the improve of the algorithm, for still images are then examined. Amongst these, two novel contributions are described. The first is the partitioning of the image into sections which resulted insignificant reduction of the compression time. In the second, the use of the median metric as alternative to the RMS was considered but was not finally adopted, since the RMS proved to be a more efficient measure. The extension of the fractal compression algorithm from still images to image sequences is then examined and three different schemes to reduce the temporal redundancy of the video compression algorithm are described. The reduction in the execution time of the compression algorithm that can be obtained by the techniques described is significant although real time execution has not yet been achieved. Finally, the basic concepts of distributed programming and networks, as basic elements of a video distribution system, are presented and the hardware and software components of a fractal video distribution system are described. The implementation of the fractal compression algorithm on a TMS320C40 is also considered for speed benefits and it is found that a relatively large number of processors are needed for real time execution.