• DocumentCode
    3770246
  • Title

    Fast rate distortion optimized quantization for HEVC

  • Author

    Yongfei Zhang;Rui Tian;Jiangang Liu;Ning Wang

  • Author_Institution
    Beijing Key Laboratory of Digital Media, School of Computer Science and Engineering, Beihang University, Beijing 100191, China
  • fYear
    2015
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    As the new generation video coding standard, the High Efficiency Video Coding (HEVC) achieves significantly better coding efficiency than existing video coding standards, which is however at the cost of a much higher computational complexity. Rate-Distortion Optimized Quantization (RDOQ) is one of the powerful coding tools employed in HEVC to pursue high coding efficiency. Nevertheless, RDOQ requires an exhaustive search over multiple candidates to determine the optimal quantized level through Rate-Distortion Optimization (RDO), which leads to considerable complexity in practice. This paper addresses this issue from a new point of view and presents a fast RDOQ scheme for transformed DCT coefficients quantization in HEVC. More specifically, we first proposed a RDOQ call removal algorithm which can effectively remove the redundant RDOQ calls and thus speed up the RDOQ process. Secondly, an all-zero block-based RDOQ skip algorithm is developed to skip unnecessary RDOQ. Compared to HEVC reference software, the proposed algorithms can save on average 34% of the RDOQ time with only slight performance degradation. Besides, our proposed algorithms can be incorporated with existing fast RDOQ algorithms to further improve the encoding efficiency.
  • Keywords
    "Quantization (signal)","Encoding","Discrete cosine transforms","Algorithm design and analysis","Computational complexity","Distortion"
  • Publisher
    ieee
  • Conference_Titel
    Visual Communications and Image Processing (VCIP), 2015
  • Type

    conf

  • DOI
    10.1109/VCIP.2015.7457854
  • Filename
    7457854