Piecewise DC prediction in HEVC

  • Kim, Kibaek
  • Jeon, Gwanggil
  • Jeong, Jechang
Citations

WEB OF SCIENCE

5
Citations

SCOPUS

6

초록

In this paper, we propose an improved DC prediction method for high-efficiency video coding (HEVC) intra coding. The technique involves the application of pixel-wise predictors rather than block-based predictor in the DC mode. In HEVC, the pixels of neighboring reconstructed blocks are used to support multiple directional spatial prediction modes and reduce spatial redundancy. For lossless coding, pixel-by-pixel differential pulse code modulation (DPCM) is applied to directional predictions. Consequently, residuals are reduced and coding efficiency is improved. Since DC prediction still employs block-based coding, pixel-wise DC prediction is needed for better coding efficiency. When compared to HEVC lossless intra coding, the proposed algorithm reduces the bit rate by 5.95%. Furthermore, with pixel-by-pixel DPCM, the average bit rate reduction is 10.64% when compared to HEVC lossless intra coding. (C) 2014 Elsevier B.V. All rights reserved.

키워드

HEVCIntra predictionLossless codingDifferential pulse code modulationPixel-wise predictorRATE-DISTORTION OPTIMIZATIONSTANDARDCOMPRESSION
제목
Piecewise DC prediction in HEVC
저자
Kim, KibaekJeon, GwanggilJeong, Jechang
DOI
10.1016/j.image.2014.07.002
발행일
2014-10
유형
Article
저널명
Signal Processing: Image Communication
29
9
페이지
945 ~ 950