Research, Based On Jpeg2000 Joint Source Channel Coding

With the rapid development of computer, communication and multimedia technology, image transmission has become an important part in multimedia communication. Traditional communication system is based on Shannon theory whose source and channel coding are considered individually. This design approach can not achieve the efficient and reliable goal in image communication. Therefore joint source and channel coding (JSCC) scheme has become one of the hottest research points. There are many kinds of joint source and channel coding technique. A leading approach nowadays is to keep the source coder and the channel coder separate, but optimize their parameters jointly. This scheme called parameter optimized joint source-channel coding (POJSCC) is studied in this paper.The key problems of POJSCC are: (1) Source coding scheme; (2) Rate-distortion function and estimation of its parameters; (3) Channel coding and unequal error protection (UEP) mechanism; (4) Bit allocation between source and channel coder. This paper aims to discuss and research these four issues above. The main structure is listed as follow:1. Introduced the image compression standard JPEG2000 and analyzed its error resilient mechanism and bit stream features.2. Discussed the UEP scheme and some classic rate allocation algorithms. On the basis, Heuristic Graph Search (HGS) algorithm is introduced. With the analytical rate-distortion model Y. Charfi, and R. Hamzaoui proposed in 2003, HSG is used in UEP rate optimized allocation when channel coding. Simulation result proved its shorter search time and the flexibility of packet cell.3. Researched the excellent channel coding approach-Turbo codes. A multi-parameter optimized UEP Turbo scheme is proposed with the rate allocation, interleaving length, number of iterate decoding are jointly optimized4. The JSCC scheme discussed above is implemented in Matlab7.0 by employing a step-like bit allocation algorithm which finds the best bit rates point between source and channel. Intensive simulation is conducted to compare with the existed approach. The results validate that the proposed image transmission system has a better improvement in recovery image quality.