改进的分级图像矢量量化

Improved Hierarchical Vector Quantizer of Images

提出了一种基于Ｐｅａｎｏ扫描的分级矢量量化算法，通过有效地利用数据间的相关性，改进了ＶＱ的性能．该算法先通过能比光栅扫描更好地保存二维数据间的相关性的Ｐｅａｎｏ扫描对原始图像进行预处理，再根据信号特性进行分级ＶＱ；还提出了基于Ｐｅａｎｏ扫描的平滑算法。

To improve the performance of a compression algorithm based on VQ, a hierarchical VQ based on the characteristics of image data is presented. A HR (high rate) vector quantizer and a LR (low rate) vector for variable rate compression are designed with various vector dimensions and thresholds for segmentation, using quad tree and Bin tree respectively in segmentation. Classified data are encoded using TC/VQ and spatial VQ according to their characteristics. A smoothing algorithm reduces the blocking effect. Compared with standard algorithms, better objective and subjective performances are shown in the algorithms.