基于感知的帧内快速模式决策算法研究  被引量:1

Research on Intra Fast Mode Decision Algorithm Based on Perception

在线阅读下载全文

作  者:谢豪 姜周 张隽祺 吴坤任 周旋 Xie Hao;Jiang Zhou;Zhang Junqi;Wu Kunren;Zhou Xuan(State Key Laboratory of Nuclear Reactor System Design Technology Nuclear Power Institute of China,Chengdu Sichuan 610213)

机构地区:[1]中国核动力研究设计院核反应堆系统设计技术国家级重点实验室,四川成都610213

出  处:《中国仪器仪表》2022年第2期67-72,共6页China Instrumentation

摘  要:AVS2是由中国数字音视频编解码技术标准工作组独立提出的一种视频编码标准。帧内编码是AVS2的关键技术之一。在帧内编码中,AVS2采取了33种帧内预测模式,使得编码性能显著提高,但是计算复杂度急剧增加。为了有效地降低帧内编码的计算复杂度,研究人员已经研究了许多帧内模式快速决策算法。然而,传统的帧内模式快速决策算法主要的关注点在于保持视频的客观性能几乎不变的条件下,减少编码的时间,而没有对视频的主观性能进行考虑。这就意味着可以对基于感知的帧内模式快速决策算法进行进一步的研究。本文提出了一种基于感知的帧内快速模式决策算法。实验结果表明,在RD17.0的全帧内测试配置下,该方法平均可减少16.93%的时间,基于SSIM的BD-Rate损失为1.07%。AVS2 is a video coding standard independently proposed by the China Digital Audio and Video Codec Technical Standard Working Group. Intra coding is one of the key technologies of AVS2.In intra coding, AVS2 adopts 33 intra prediction modes, which significantly improves the coding performance, but the computational complexity increases sharply. In order to effectively reduce the computational complexity of intra-frame coding, researchers have studied many intra-mode fast decisionmaking algorithms. However, the main focus of the traditional intra-mode fast decision algorithm is to reduce the encoding time while keeping the objective performance of the video almost unchanged,without considering the subjective performance of the video. This means that further research can be carried out on the perception-based fast decision-making algorithm of intra mode. This paper proposes a perception-based intra fast mode decision algorithm. Experimental results show that under the full frame test configuration of RD17.0, this method can reduce the time by 16.93% on average, and the BD-Rate loss based on SSIM is 1.07%.

关 键 词:AVS2 感知编码 CU尺寸决策 帧内编码 

分 类 号:TN919.81[电子电信—通信与信息系统]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象