| As one of the the best still image compression standards, JPEG2000has many newfeatures superior to JPEG and has a great variety of applications. JPEG2000also hassome shortages, such as it is difficult to achiewe precise rate control and the streamerror-resilience is weakened，so the research and implementation of rate controlalgorithm and fault-tolerant decoding method based on JPEG2000has importantscientific and practical significance.This paper first describes the characteristics, algorithm coding framework and corealgorithms of the JPEG2000standard, focuses on the options of the quantization stepsize parameter, then describes the Tier-2algorithm: rate distortion optimizationinterception algorithm, stream organization modules in detail, focused on research andanalysis of the encoding method of the header information, by analying the range ofeach of the header information, this paper presents a packet header informationencoding method combined with the rate control. Use the the least fixed number of bitsto represent the four kinds of header informations when rate distorted optimizationtruncation is processed, the code length of these header information can be reservedaccurately, precise rate control can be reached. The experimental result shows that theperformance of the algorithm is good as the kakadu software and achieving a relativelyprecise control of the rate at the same time.Finally, the paper describes the JPEG2000own fault-tolerant technology, thenpresents an effective method to improve the decoding robustness on the base of thestructure characteristics of the JPEG2000codestream:Transfer the important globalinformation multiply which is on the head of the codestream to improve the reliability;Detect the positions of the packet identifiers, and use the positions to let the error ofcodestream be limited in the current packet. The experimental results shows that theproposed algorithm can effectively improve the decoding fault-tolerance of theJPEG2000. |
PDF Full Text Request