高效率视频编码快速模式判决算法

A Fast Mode Decision Algorithm for High Efficiency Video Coding

为了降低高效率视频编码(HEVC)的编码复杂度,提出一种新的快速模式判决算法。考虑到视频帧的纹理特性和编码中所采用的量化参数影响最优编码单元(CU)模式的选择,首先提取前一帧的平均分割层数和其最大编码单元(LCU)的最小分割层数来预测当前帧对应位置处LCU的最小分割层数,以跳过部分大块CU的模式判决;然后计算当前CU的运动矢量差值和预测残差系数来决定CU的最大分割层数,以避免小块CU的模式判决。实验结果表明,与原始的HEVC编码方法相比,高效率视频编码快速模式判决算法可平均降低51%的编码复杂度,而编码比特率平均只上升了0.69%。

A fast mode decision algorithm is proposed to reduce the complexity in high efficiency video coding （HEVC） encoder. The potential effects of frame texture and its related quantization parameter on the final decision of optimal coding unit （CU） modes is considered, and the average depth of the previous frame and the minimum depth of its related large coding unit （LCU） are extracted to estimate the minimum depth of collocated LCU in the current frame to skip partial larger CUs in mode decision process. Then the maximum depth for current CU is determined by calculating the parameters of motion vector difference and predicting residues so that the unnecessary mode decision of smaller CUs can be avoided. Experimental results and comparisons with the conventional HEVC show that the proposed algorithm for HEVC achieves an average bit-rate increment of 0.69% whereas the coding complexity has a significant reduction of about 51%.