论文部分内容阅读
为了降低高效视频编码(HEVC)的复杂度,提出了一种基于纹理方向和空域相关性的帧内快速编码算法。一方面,利用相邻编码单元(CU,coding unit)的最佳分割尺寸确定相关性的权重,并根据已编码CU的率失真代价值定义了CU分割尺寸的相关性因子,通过比较该因子提前终止CU分割;另一方面,利用Sobel算子求取预测单元(PU,prediction unit)子块的纹理方向,通过判定其纹理方向显著性确定相应的候选模式集,然后根据PU大小对所得的预测模式修正处理,最后遍历这些候选模式选取最优模式。实验结果表明:本文算法相对于原始HEVC编码方法,在全I帧模式下编码时间平均减少36.84%,BDBR(Bjntegaard delta bit rate)上升约0.81%,BDPSNR(Bjntegaard delta peak signal-to-noise rate)降低了0.047dB。
In order to reduce the complexity of high-efficiency video coding (HEVC), an intra-frame fast coding algorithm based on texture direction and spatial correlation is proposed. In one aspect, the weight of the correlation is determined by using the optimal partition size of a CU (coding unit), and the correlation factor of the CU partition size is defined according to the value of the rate distortion of the CU being encoded. By comparing the factor advance On the other hand, the Sobel operator is used to get the texture direction of PU sub-block, and the corresponding candidate pattern set is determined by judging the significance of texture direction, and then the prediction is made according to the PU size Mode correction processing, the last traversal of these candidate modes to select the optimal mode. The experimental results show that compared with the original HEVC coding method, the average coding time in full I-frame mode is reduced by 36.84%, BDBR (Bjntegaard delta bit rate) is increased by about 0.81%, BDPSNR (Bjntegaard delta peak signal-to -noise rate) is reduced by 0.047dB.