Study Of Low Bit Rate Video Coding And Optimization Of Compressing Algorithm On TM1300 DSP

With the development of digital commnication and multimedia technology, it is becoming more important to study the video codec technology than ever. As well we known, a significant appliaction for digital comunication is "What You See Is What You Get", which need to transform picture and video sequences; similarly, the outlook of multimedia technology reckons on the future network. So, low-bitrate video coding technology is the key in this field, under the state of the extremely limited bandwidth at present.The low frame rate of the monitoring video makes it possible to use more complex algorithm to gain better compression ratio. That is to sty, monitoring video signals have more compression potential than the ordinary ones. Great emphasis in the dissertation is put on the following three parts: first, after studying of MVFAST, PMVFAST and AMVFAST, an efficient and applied algorithm is presented named novel predictive motion vector field adaptive search technique, in which the characteristic of monitoring video signals and VLIW structure of special dsp are taken intoaccount; second, a multimedia processing platform based on TM1300 is built, on which low-bitrate video coding algorithm is implemented and optimized; third, CABAC alogrithm is improved slightly.This dissertation is organized as follow: in chapter 1, video codec technology is summarized; the main part of theoretic innovation in chapter 2, including NPMVFAST and CABAC alogrithm; in chapter 3,TMN8 is introduced and tested on dsp platform; in chapter 4, the platform based on TM1300, on which low-bitrate video coding alogrithm is implemented and optimized, is set up; the research is summarized in chapter 5, and some advices to give directions to the following work are proposed.