Lossless Reference Frame Compression Combined with Read and Write Behaviors of HEVC and VVC Codecs

Tingting Li, Leilei Huang, Yibo Fan
{"title":"Lossless Reference Frame Compression Combined with Read and Write Behaviors of HEVC and VVC Codecs","authors":"Tingting Li, Leilei Huang, Yibo Fan","doi":"10.1109/ICCS56666.2022.9936212","DOIUrl":null,"url":null,"abstract":"With the emergence of immersive videos like 8K@l20fps, memory bandwidth induced by reference frames is gradually becoming the bottleneck of codecs. This paper analyzes related data caching strategies and then proposes a lossless reference frame compression algorithm and its hardware implementation, which combines read and write behaviors of High Efficiency Video Coding (HEVC) and Versatile Video Coding (VVC) codecs. The basic block of this algorithm has adaptive sizes, which would be predicted with the median edge detector (MED) and encoded with optimized semi-fixed-length (SFL) coding. For HEVC and VVC sequences, results demonstrate an average reduction of 68.22% and 74.49% in memory bandwidth. The corresponding compressor and decompressor take 36. 9K and 50. 6K gates when implemented with TSMC 65nm technology, which could achieve high throughputs of11.47 and 6.27 Gpixels/s respectively, equivalent to 8K@267fps and 8K@l88fps.","PeriodicalId":293477,"journal":{"name":"2022 IEEE 4th International Conference on Circuits and Systems (ICCS)","volume":"420 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE 4th International Conference on Circuits and Systems (ICCS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCS56666.2022.9936212","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1

Abstract

With the emergence of immersive videos like 8K@l20fps, memory bandwidth induced by reference frames is gradually becoming the bottleneck of codecs. This paper analyzes related data caching strategies and then proposes a lossless reference frame compression algorithm and its hardware implementation, which combines read and write behaviors of High Efficiency Video Coding (HEVC) and Versatile Video Coding (VVC) codecs. The basic block of this algorithm has adaptive sizes, which would be predicted with the median edge detector (MED) and encoded with optimized semi-fixed-length (SFL) coding. For HEVC and VVC sequences, results demonstrate an average reduction of 68.22% and 74.49% in memory bandwidth. The corresponding compressor and decompressor take 36. 9K and 50. 6K gates when implemented with TSMC 65nm technology, which could achieve high throughputs of11.47 and 6.27 Gpixels/s respectively, equivalent to 8K@267fps and 8K@l88fps.
结合HEVC和VVC编解码器读写行为的无损参考帧压缩
随着8K@l20fps等沉浸式视频的出现,由参考帧引起的内存带宽逐渐成为编解码器的瓶颈。在分析相关数据缓存策略的基础上,结合高效视频编码(HEVC)和通用视频编码(VVC)编解码器的读写行为,提出了一种无损参考帧压缩算法及其硬件实现。该算法的基本块具有自适应大小,使用中值边缘检测器(MED)进行预测,并使用优化的半定长编码(SFL)进行编码。对于HEVC和VVC序列,内存带宽平均降低了68.22%和74.49%。对应的压缩、减压机取36。9K和50。采用台积电65nm技术实现6K栅极,分别可实现11.47和6.27 Gpixels/s的高吞吐量,相当于8K@267fps和8K@l88fps。
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
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信