提出了一种矢量量化编码的快速匹配算法。利用码书的拓扑结构和柯西 施瓦兹不等式的性质推导出了消除不必要匹配操作的条件,并根据这个条件提出了快速搜索算法。实验结果表明,该算法不仅明显降低了矢量编码过程的复杂程度,同时也保证了与全搜索编码相同的图像编码质量。与其它搜索方法的操作次数和存储空间进行了比较。 A fast matching algorithm of vector quantization coding is proposed. Based on the topological structure of codebook and Cauchy Schwarz inequality, the conditions for eliminating unnecessary matching operations are deduced. Based on this condition, a fast search algorithm is proposed. Experimental results show that this algorithm not only reduces the complexity of vector coding process obviously, but also guarantees the same quality of image coding as full search coding. Compared with other search methods, the number of operations and storage space were compared.