Context Similarity-Enabled CU Partitioning Algorithm in VVC

2023 5th International Conference on Communications, Information System and Computer Engineering (CISCE) Pub Date : 2023-04-14 DOI:10.1109/CISCE58541.2023.10142615

Jiayuan Jin, Xiantao Jiang, Tian Song, F. Richard Yu, WeiLi, Jin Liu

{"title":"Context Similarity-Enabled CU Partitioning Algorithm in VVC","authors":"Jiayuan Jin, Xiantao Jiang, Tian Song, F. Richard Yu, WeiLi, Jin Liu","doi":"10.1109/CISCE58541.2023.10142615","DOIUrl":null,"url":null,"abstract":"A new Quadtree with nested multi-type tree(QTMT) partitioning method is adopted in the latest video coding standard Versatile Video Coding(VVC) to achieve effective encoding. Compared with Quadtree(QT) in HEVC, QTMT can provide enhanced encoding performance. However, this improved coding performance is obtained at the cost of additional computational load from recursive and nested searches for the best CU structure. A context similarity-enabled CU partitioning algorithm is proposed to achieve a balance between coding efficiency and compression quality in VVC. According to the difference between frames, the method of frame difference algorithm is used to calculate the possibility that the current block can follow the same subblock division mode of the previous frame, so as to skip the unnecessary division selection process. The experimental results show that the proposed method can save the encoding time up to 18.22% with about 0.91% BD-rate degradation on average compared with VVC software reference VTM-9.3","PeriodicalId":145263,"journal":{"name":"2023 5th International Conference on Communications, Information System and Computer Engineering (CISCE)","volume":"418 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 5th International Conference on Communications, Information System and Computer Engineering (CISCE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CISCE58541.2023.10142615","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

A new Quadtree with nested multi-type tree(QTMT) partitioning method is adopted in the latest video coding standard Versatile Video Coding(VVC) to achieve effective encoding. Compared with Quadtree(QT) in HEVC, QTMT can provide enhanced encoding performance. However, this improved coding performance is obtained at the cost of additional computational load from recursive and nested searches for the best CU structure. A context similarity-enabled CU partitioning algorithm is proposed to achieve a balance between coding efficiency and compression quality in VVC. According to the difference between frames, the method of frame difference algorithm is used to calculate the possibility that the current block can follow the same subblock division mode of the previous frame, so as to skip the unnecessary division selection process. The experimental results show that the proposed method can save the encoding time up to 18.22% with about 0.91% BD-rate degradation on average compared with VVC software reference VTM-9.3

查看原文本刊更多论文

VVC中启用相似度的CU分区算法

在最新的视频编码标准VVC (Versatile video coding)中，采用了一种新的四叉树嵌套多类型树(QTMT)划分方法来实现有效的编码。与HEVC中的四叉树(QT)相比，QTMT可以提供更好的编码性能。然而，这种改进的编码性能是以通过递归和嵌套搜索获得最佳CU结构的额外计算负载为代价的。为了在VVC中实现编码效率和压缩质量的平衡，提出了一种基于上下文相似度的CU分区算法。根据帧间的差异，采用帧差算法的方法计算当前块与前一帧遵循相同子块划分模式的可能性，从而跳过不必要的划分选择过程。实验结果表明，与VVC软件参考VTM-9.3相比，该方法可节省高达18.22%的编码时间，平均bd率下降约0.91%

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

2023 5th International Conference on Communications, Information System and Computer Engineering (CISCE)

自引率

0.00%

发文量