Abstract: This paper proposes a new CAVLC decoding algorithm which based on the analysis of the structure of CAVLC code table, aiming to solve the problem of large computing and high complexity during lookup the code table. The basic idea of this algorithm is to take the numbers of zero in CAVLC code prefix as the primary index and to take the CAVLC code suffix as the secondary index, so to get the decoded output quickly by the primary index and the secondary index. The results show that the algorithm has improved significantly in the decoding time and storage space, comparing with the original algorithm.

Key words: CAVLC codebook, index techniques, H.264, entropy coding

摘要： 针对CAVLC解码算法中码表查找算法存在运算量大和复杂度高的问题，在分析研究CAVLC码表结构特征的基础上提出一种新CAVLC解码优化算法。算法基本思路是对CAVLC码字前缀0的个数进行一级索引，对码字后缀进行二级索引，由一二级索引查询快速得到解码输出。测试结果表明，相比原算法，该优化解码算法在解码时间、存储空间方面都有显著的提高。

关键词: CAVLC码表, 索引技术, H.264标准, 熵编码