Coding unit complexity-based predictions of coding unit depth and prediction unit mode for efficient HEVC-to-SHVC transcoding with quality scalability

Chia Hung Yeh, Wen Yu Tseng, Li Wei Kang*, Cheng Wei Lee, Kahlil Muchtar, Mei Juan Chen

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)

Abstract

To support good video quality of experiences in heterogeneous environments, transcoding an existed HEVC (high efficiency video coding) video bitstream to a SHVC (scalability extension of HEVC) bitstream with quality scalability is highly required. A straightforward way is to first fully decode the input HEVC bitstream and then fully re-encode it with the SHVC encoder, which requires a tremendous computational complexity. To solve the problem, in this paper, a coding unit complexity (CUC)-based prediction method for predictions of CU (coding unit) depth and PU (prediction unit) mode for efficient HEVC-to-SHVC transcoding with quality scalability is proposed to significantly reduce the transcoding complexity. The proposed method contains two prediction techniques, including (i) early termination and (ii) adaptive confidence interval, and predicts the CU depth and PU mode relying on the decoded information from the input HEVC bitstream. Experimental results have shown that the proposed method significantly outperforms the traditional HEVC-to-SHVC method by 74.14% on average in reductions of encoding time for SHVC enhancement layer.

Original languageEnglish
Pages (from-to)342-351
Number of pages10
JournalJournal of Visual Communication and Image Representation
Volume55
DOIs
Publication statusPublished - 2018 Aug

Keywords

  • Coding unit complexity
  • Early termination
  • HEVC (high efficiency video coding)
  • Scalable video coding
  • SHVC (scalability extension of HEVC)
  • Video transcoding

ASJC Scopus subject areas

  • Signal Processing
  • Media Technology
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Coding unit complexity-based predictions of coding unit depth and prediction unit mode for efficient HEVC-to-SHVC transcoding with quality scalability'. Together they form a unique fingerprint.

Cite this