Research Of Image Compression Based On Wavelet Transform And Vector Quantization

The most important content in this paper is research in methods of digital image compression coding. Both still image compression coding based on wavelet transform and video image compression coding based on telemeeting protocol H.261 are deeply researched and discussed. In chapter 1, the task background of this paper is introduced first and the importance of image compression at present is pointed. Here, image compression is also described, that is including the foundation theory, methods and appraising of image compression. At the end of this chapter, the task that has been accomplished in this paper is also introduced. In chapter 2, the theory of wavelet analysis is mainly introduced. The advancement of wavelet transform is pointed here and the maths foundation of wavelet transform is also introduced by discussing Fourier transform and short time Fourier transform. Then , the theory of multi-resolution analysis and the fast wavelet transform-Mallat algorithm are introduced. Based on the theory of wavelet analysis, how to realize image analysis and synthesis by wavelet transform is discussed including two dimension multi-resolution analysis and Mallat algorithm. At last, the image analysis and synthesis by wavelet transform are realized by programming and the characteristics of wavelet coefficients are discussed in this chapter. Based on chapter 2, in chapter 3, how to use vector quantization in wavelet transform is researched. The theory and methods of vector quantization are introduced, and the key algorithms of vector quantization compression are mainly discussed, including Classified Vector Quantization, Structure Matrix, Codebook Design and so on. Then, a structure matrix vector quantization scheme based on wavelet transform and the advanced one are proposed. The two schemes are both realized by programming and the results of the experiments are analyzed and compared. In the last chapter, the methods of telemeeting video image coding are discussed. The telemeeting and the international standards are also introduced here, especially the H.26 I protocol. The algorithms of intraframe and interframe are proposed. How to control the bit flow and memory is discussed too. In the end, a video image coding scheme is proposed. This scheme is also realized by programming and the results are analyzed. At last, the development of telemeeting is put forward.