A Fast HEVC Encoding Method Using Depth Information of Collocated CUs and RD Cost Characteristics of PU Modes

Kuang Han Tai, Min Yuan Hsieh, Mei Juan Chen, Chia Yen Chen, Chia Hung Yeh*

*此作品的通信作者

研究成果: 雜誌貢獻期刊論文同行評審

25 引文 斯高帕斯(Scopus)

摘要

The latest video coding standard, high efficiency video coding (HEVC), aims to achieve better coding efficiency than the H.264/AVC standard. To improve the coding performance of HEVC, tools and structures are included that also increase the computational complexity of encoding, especially for the mode decisions in the coding unit (CU), which is a structural element in the HEVC. For each CU size, the HEVC encoder performs mode decision and motion estimation using all prediction unit (PU) types and incurs massive computational burdens. In this paper, we investigate the depth correlation between the current and the collocated CU to avoid irrelevant CU procedure and truncate some PU predictions as a result. We also investigate the relationship of the rate-distortion (RD) costs after the Merge/SKIP prediction. By analyzing different CU sizes and quantization parameters, we build a mathematical model to represent the relationship for the RD costs after the Merge/SKIP prediction with an adaptive termination of the CU procedure. In addition, the relationship of the RD costs after the Merge/SKIP prediction and 2N 2N mode is also explored. Search range reduction is used to further speed up our algorithm. Experimental results show that our algorithm can reduce the coding time by up to 80%, and provide average coding time savings of 46%, 45%, and 52%, for low delay B, low delay P, and random access configurations, respectively. Moreover, the proposed algorithm is able to maintain coding performance similar to HM 11.0. The proposed scheme outperforms previous work in terms of both the coding speed and the RD performance.

原文英語
文章編號7990599
頁(從 - 到)680-692
頁數13
期刊IEEE Transactions on Broadcasting
63
發行號4
DOIs
出版狀態已發佈 - 2017 十二月

ASJC Scopus subject areas

  • 媒體技術
  • 電氣與電子工程

指紋

深入研究「A Fast HEVC Encoding Method Using Depth Information of Collocated CUs and RD Cost Characteristics of PU Modes」主題。共同形成了獨特的指紋。

引用此