• English
  • Deutsch
  • Log In
    Password Login
    Research Outputs
    Fundings & Projects
    Researchers
    Institutes
    Statistics
Repository logo
Fraunhofer-Gesellschaft
  1. Home
  2. Fraunhofer-Gesellschaft
  3. Konferenzschrift
  4. A 4k capable FPGA based high throughput binary arithmetic decoder for H.265/MPEG-HEVC
 
  • Details
  • Full
Options
2014
Conference Paper
Title

A 4k capable FPGA based high throughput binary arithmetic decoder for H.265/MPEG-HEVC

Abstract
High Efficiency Video Coding (HEVC) is the newest video coding standard approved by the ISO/IEC and ITU-T in January 2013. By providing a video coding efficiency gain of up to 50% compared to its predecessor H.264/MPEG-4 AVC, the complexity of the used algorithms has raised significantly. Targeting video formats with higher spatial and temporal resolutions - e.g. 4Kp60 in broadcast applications - make implementing encoders and decoders a challenging task. A well known bottleneck in the decoder architecture is the Context-Based Adaptive Binary Arithmetic Coding (CABAC), specified as entropy coding method in the H.265/MPEG-HEVC standard. Reaching high throughput for real time applications is a demanding task in terms of context modeling and serial bin-to-bin dependencies in the binary arithmetic coding engine. Especially the context selection and the update of the context model in binary decision mode requires a complex underlying control flow for the decoding process of each syntax element. This paper presents a pure hardware implementation of a Main Profile H.265/MPEG-HEVC compliant entropy decoder. The design has been optimized to achieve real time performance up to 4k video resolutions. By using state-of-the-art FPGA technology, data rates of up to 104 MBit/s can be achieved.
Author(s)
Hahlbeck, J.
Stabernack, B.
Mainwork
IEEE Fourth International Conference on Consumer Electronics - Berlin, ICCE-Berlin 2014  
Conference
International Conference on Consumer Electronics - Berlin (ICCE-Berlin) 2014  
DOI
10.1109/ICCE-Berlin.2014.7034335
Language
English
Fraunhofer-Institut für Nachrichtentechnik, Heinrich-Hertz-Institut HHI  
  • Cookie settings
  • Imprint
  • Privacy policy
  • Api
  • Contact
© 2024