Research On Image Compression Technology Based On Enbedded Coding

With the development of Multimedia technique, image compression plays a more important role in the storage and transfer of mass image data. Embedded image coding is a new coding method and can provide the possibility to progressively reconstruct reproductions of the image better and better as the coding image bit stream is received more and more. The output bit stream of the embedded code can be truncated and the images for practical applications can be reconstructed to achieve varying degrees of scalability and different target bit rates or distortion. Embedded image coding is especially fit for the changeful and fallible environment, such as internet/wireless transmission, image browse, and etc.The basic theory of wavelet transform, information and coding are introduced in this paper. Several image compression techniques and standards are compared. Then, the theory and methods of embedded image coding are studied. After introducing the Embedded Zerotree Wavelet (EZW) algorithm in detail, the Set Partitioning in Hierarchical Trees (SPIHT) algorithm is presented. SPIHT algorithm, which is based on the zerotree structure, has been performed more effectively in the embedded image coding. It is the most popular method at present. But in the process of SPIHT coding, three lists must be used. They consume a lot of memory to save the significant information of the wavelet coefficients and wavelet coefficients sets. The SPIHT algorithm is difficult to be applied in hardware because of its large space of storage. The problem of SPIHT without lists is studied to save the memory volume. And based on analysis of the Listless Zerotree Coding (LZC) algorithm and the No List SPIHT (NLS) algorithm, a new scheme of NLS has been put forward in this paper. The scheme use a function to judge weather the wavelet coefficients sets are significant, instead of the vector which save the maximum value of the wavelet coefficients sets. The improved scheme of NLS consumes less memory than the classic NLS algorithm.In the research process, the new scheme of NLS has been programmed by the C language. The program has been compiled by Visual C++6.0 compiler in Windows platform. The results show that in the same bit rate, the quality of image restructured by the improved NLS algorithm is close to the image restructured by SPIHT, and save memory markedly. The reconstructed image quality of the improved NLS algorithm is better than the LZC.