基于记忆和预测机制的自适应矢量量化及其在图像压缩编码中的应用

本文提出了一种用于图像压缩的低比特率自适应矢量量化技术，该方法有以下三个特色：１）引入了有效的记忆预测机制，使量化器和编码器有很好的自适应性。２）采用二次寻址方法对矢量的地址进行编码，大大提高了编码效率；３）算法运算复杂度低，便于ＶＬＳＩ实现。

Adaptive Vector Quantization Based on Memory and Prediction and Its Application in Image Coding

An adaptive low bitrate vector quantization scheme for image compression is developed.The features of this scheme are: (1)Good adaptation is achieved due to a novel and effective prediction mechanism. (2)The efficiency of encoding the vector index is significantly improved by using dual addressing technique. (3)Simple implementation suitable for VLSI architecture.