H.264中CAVLC算法改进

The improvement of CAVLC algorithm in H.264

简要分析了H.264中的上下文自适应变长编码(CAVLC),并研究其采用的Golomb编码,提出了一种基于最优Golomb编码参数选择的改进算法.利用前一个需编码的数值的Golomb编码最优参数,对当前编码参数进行预测,使当前编码参数更逼进它的最优编码参数.试验结果表明:该算法较原有算法能更好的压缩残差数据,在量化参数QP较小时,能带来一定的压缩增益.

After a simple analysis of context-based adaptive variable length coding （CAVLC） in H. 264, and a deeper study of Golomb coding in CAVLC, we proposed an advanced algorithm based on an optimal rule of choosing parameters in this paper. The optimal coding parameter of previous value is utilized to predict the coding parameter of current value thus will make it approaching better of the optimal coding parameter of current value. Results of experiment show that the advanced algorithm compresses the residual data better than original algorithm, especially in low quantization parameter （QP）, it could bring compression gain.