A multi-stage vector quantization algorithm based on wavelet transform for image coding is proposed in this paper. According to the theory of multiresolution for signal decomposition in the wavelet representation, and the characteristic of multistage vector quantization. 结合小波变换中多分辨率分析特性以及多级矢量量化复杂度低、量化效果较好的特点提出了一种基于小波变换的多级矢量量化图像编码方法。
Firstly three directional high frequent multi-scale coefficients vector is formed by wavelet decomposition, then self-organizing neural network is used to compress coding. 首先通过小波分解得到三个方向的高频多尺度系数矢量，分别利用自组织特征映射神经网络对三个方向的多尺度系数矢量进行加权矢量量化压缩编码。
Intra frame, on the earth, is still picture. There are many compression methods of still picture, for example, DPCM, Vector Quantization, Block Transform Coding, Fractal Coding, ANN Coding and Wavelet Transform Coding etc. Ⅰ帧属于静止图像，通常静止图像的压缩编码方法有预测编码、矢量量化、块变换编码、分形编码、人工神经网络编码和墓于小波变换的编码方法等。
This paper introduces several new data compression techniques in multimedia : vector quantization coding, generalized transformation coding based neural networks, structure coding, wavelet transformation coding and mould based coding. 简要介绍了若干知名度较高的多媒体数据压缩新技术：矢量量化编码、神经网络广义变换编码、结构编码、小波变换编码和基于模型的编码。
This chapter first introduces the principium of trellis coded vector quantization, then focuses on analyzing the algorithm of wavelet image classified weighted TCVQ and its realization process, also gives TCVQ application to wavelet image coding and its simulation results. 首先分析了网格编码矢量量化的原理，在第三章矢量量化的基础上详细介绍了小波图像分类加权TCVQ算法的原理和实现过程，并给出了TCVQ在小波图像量化中的应用实例和仿真结果。
