• DocumentCode
    3682611
  • Title

    Inter-prediction optimisations for fast HEVC encoding of ultra high definition content

  • Author

    Ivan Zupancic;Saverio G. Blasi;Ebroul Izquierdo

  • Author_Institution
    School of Electronic Engineering and Computer Science, Queen Mary University of London, E1 4NS London, United Kingdom
  • fYear
    2015
  • Firstpage
    85
  • Lastpage
    88
  • Abstract
    Thanks to technological advances in devices for video signal capturing and displaying, the usage of large resolution video formats is becoming more and more popular in consumer and professional applications. The new Ultra High Definition (UHD) format aims to provide enhanced viewing experience with its spatial resolution of at least 3840 × 2160 luma samples per frame and temporal resolution of up to 120 frames per second. Such a large amount of raw data requires extremely high levels of compression efficiency. The latest state-of-the-art High Efficiency Video Coding (HEVC) standard was developed also to respond to these needs. HEVC outperforms previous video coding standards in terms of coding efficiency, especially while encoding high resolution video content. However, this comes at very high computational costs, making current HEVC encoder implementations of very limited usage. In this paper, a fast inter-prediction scheme is proposed targeting fast mode decision based on reverse Coding Unit (CU) visiting order, and early termination of both the integer- and fractional-pel motion estimation processes. Experimental results show that the proposed algorithm achieves on average 72.9% time savings, for limited coding losses.
  • Keywords
    "Encoding","Video coding","Image coding","High definition video","Testing","Optimization","Complexity theory"
  • Publisher
    ieee
  • Conference_Titel
    Systems, Signals and Image Processing (IWSSIP), 2015 International Conference on
  • ISSN
    2157-8672
  • Electronic_ISBN
    2157-8702
  • Type

    conf

  • DOI
    10.1109/IWSSIP.2015.7314183
  • Filename
    7314183