帧内预测模式的快速决策算法研究

Research on Fast Decision Algorithm of Intra Prediction Mode

HEVC是由视频编码联合协作小组(JCT-VC)开发出的一种先进的视频编解码标准。在帧内编码的亮度预测中,HEVC引入了35种预测模式,使得编码性能显著提高,但是计算复杂度急剧增加。为了有效地降低HEVC的计算复杂度,研究人员已经研究了许多快速模式决策算法。许多传统的快速模式决策算法一方面在计算梯度或者用特定的算子提取信息时,会不可避免地引入复杂度较高的运算;另一方面,梯度纹理信息也仅仅反映当前块的大致纹理信息,用其指导的预测模式与最终实际最优预测模式会有一定的偏差。本论文在帧内模式决策的过程中充分利用角度模式之间的相关性,提出了一种新的快速帧内预测模式决策的算法。

HEVC is an advanced video coding and decoding standard developed by the Joint Collaborative Group on Video Coding (JCT-VC). In the brightness prediction of intra coding, HEVC introduces 35 prediction modes, which significantly improves the coding performance, but the computational complexity increases sharply. In order to effectively reduce the computational complexity of HEVC, researchers have studied many fast mode decision algorithms. Many traditional fast mode decision algorithms, on the one hand, will inevitably introduce more complicated operations when calculating gradients or extracting information with specific operators;on the other hand, gradient texture information only reflects the rough texture information of the current block. The prediction mode guided by it will have a certain deviation from the final actual optimal prediction mode. This paper makes full use of the correlation between angle modes in the process of intra mode decision-making, and proposes a new fast intra prediction mode decision-making algorithm.