摘要
基于边缘方向分类的ＣＶＱ能够减少在较低位码率进行图象信号矢量量化时的边缘失真和计算的复杂程度。根据静止图象小波变换后小波系数的特征提出了一种在ＷＴ域进行分类的分类算法。此方法可以容易地扩展到ＷＴ变换编码。实验结果表明，与同类方法相比，在相同或稍高信噪比的情况下，此方法的位码率大大减小。
s A classified VQ(CVQ) tehchnique,based on the edge oriented classifier,can reduce the edge degradation as well as the encoding complexity.In this paper,according to the WT coefficients feature of a still image transformed,we propose a classification algorithm in the wavelet transform(WT) domain for the CVQ.This approach can be easily extended to the wavelet transform coding technique.Coded images based on this classification are shown to enjoy considerable bpp reduction over three existing methods in the same or a little higher PSNR conditions.
出处
《中国图象图形学报（A辑）》
CSCD
1997年第8期585-588,593,共5页
Journal of Image and Graphics
关键词
分类
矢量量化
小波变换
图象编码
Classification,Vector quantization,Wavelet transform
