一种改进的相关图像矢量量化编码算法

A Modified Correlation Vector Quantization Algorithm for Image Coding

本文提出了一种改进的图像矢量量化编码算法．该算法通过采用增加预测块，以及根据相邻块状态进行控制信息的条件熵编码等改进方法，进一步提高了算法的编码效率．为了便于算法实现，本文对图像的边缘引进了一致的处理方法．测试结果表明，相对于无记忆矢量量化编码算法，比特率约下降４０％，相对于文献［５］算法，比特率约下降１５％．

In this paper,a new image coding algorithm based on the correlation vector quantization scheme is presented. Improvements are developed in the proposed algorithm to reduce bit rate,such as increasing anticipation blocks, entropy coding of side information according to the condition of the adjacent blocks. The test results show that the reduction in the bit rate of the proposed algorithms is about 4O% against the memoryless vector quantizer,and about 15% against the algorithm proposed in[5].