The Realization And Study Of A Scalable Data Encoder Based On SPIHT

With the rapid development of Multimedia Technology and lnternet,the computer is more and more used in Proeessing videos and other multimedia information. However, the current internet is a best-effort work where loss,variable delays and bandwidth fluctuation occur during data delivery in a streaming session.So,the traditional non-scalable streaming cannot adapt to variation of available network bandwidth.Moreover,the diversities of terminals also challenge the popularization and development of streaming media.So, a new coder is asked to provide network adaptive streaming service.By using the DWT and the SPIHT,a scalable data decoder is designed and realization in this paper.The paper first introduces the video encoding systems.The encoding system is divided into three parts which include the time domain model, the spatial model and the entropy coding model to analyse. The wavelet theory's principles and advantages are intrduced in this paper. The paper also introduces and compares the EZW, SPIHT two encoding algorithms which are used after the wavelet.By analysing these two algorithms, the reason they compress the data and make the data scalable is clear. The paper shows the overall structure of the encoder. The process and the key point of each module is also intrduced. The lifting wavelet structure and how to construct a 9 / 7 lifting wavelet is also discussed. SPIHT coding process and many key implementation code is also given in this paper.The organization oh the encoder data is introduced.The test result of each part of the encoder is given at the end of the paper.By analysing the test result,the encoder indeed compressed the data and make it scabable with the DWT and the SPIHT.