采用树结构矢量组合的小波图像分类矢量量化

本文提出了采用树结构矢量组合对小波图像进行分类矢量量化的新方法.该方法充分利用了子带系数的带间和带内的相关性,分类信息占用比特数少,并采用了基于人眼视觉特性的加权均方误差准则进行矢量量化,提高了量化增益.仿真结果表明,该方法实现简单,可达到很好的压缩效果.

Stereo Video Coding Based Disparity Compensated Prediction

A new method which performs classified vector quantization to wavelet image using tree structure vector combination is proposed. the method utilizes correlation between the subbands and within the subbands fully, the classified information need a little bites to express, and vector quantization adopt weighted mean square error rule that using mankind visual characteristic to improve quantization gain. Simulation states that this method can be realized easily and can compress the wavelet image efficiently.