最大概率匹配的矢量量化编码算法

文章提出了一种最大概率匹配的矢量量化编码算法，它为码书中的每一码字增加一个计数器，统计在编码图象时每个码字的出现的频数，并进行排序；在量化矢量时，根据当前码字出现频数大小依次选择侯选码字，即频数大的码字优先选为候选码字。该算法可以和已有的预测法结合，形成预测加最大概率匹配的联合矢量量化编码算法。实验表明，联合算法的效率较高，在最初几次的搜索中就能以较高的命中率命中最佳匹配码字。

Maximum Probability Matching Algorithm for Vector Quantization

In this paper, a novel maximum probability matching algorithm for vector quantization is presented. In the process of encoding. one counter is added for every codeword in the codebook, it counters the times that the codeword emerged in the encoding image. The most often emerged codeword by now is taken as the first candidate codeword. The second often-emerged codeword is taken as the second candidate codeword. In this way, authors select the rest candidate codewoul if it is necessary. The simulation shows that this algorithm is efficient.