基于小波变换的空间约束矢量量化方法

Space-Conatrained VectorQuantization Using Wavelet Transform

图像的小波变换能同时提供空间─—频率局部化信息，而且小波变换域内矢量量化数据压缩已得到广泛应用．经过小波变换后，各子带小波分量存在相关性和空间约束，同时考虑到人类视觉对水平和垂直方向高频分量比对角方向更加敏感，本文提出了基于空间约束的矢量量化方法，该算法能同时提高编码效率和改善重构图像质量．

The wavelet transform of images can supply information about localization of both space andfrequency.Data compression of vector quantization involved in the wavelet transform domain has been widelyused.After the multresolution pyramidal decomposition of wavelet transform, there aze correlation and spatialconstraints among the wavelet coefficients of subbands. Based upon the more sensivity of human vision to hori-zontal and vertical high-frequency components than diagonal ones, in this paper space-constrained vector quanti-zation using wavelet transform is proposed, which supplies improvements on both coding efficiency and quality ofdecoded images.