Presentation + Paper
20 June 2021 HEVC improved performance with end-to-end super resolution
Mohamed Abdelazim, Ahmed Hamza, Abdelrahman Abdelazim, Djamel Ait-Boudaoud
Author Affiliations +
Abstract
High Efficiency Video Coding (HEVC), also known as H.265 and MPEG-H part2 is the state of art video encoder standard developed by the Joint Collaborative Team on Video Coding (JCT-VC). It achieves up to 50% better data compression at the same level of video quality than its predecessor (H.264-AVC). One of the main challenges of the HEVC coding is its overall performance as the improved quality and bit rate of the standard coding comes with significant performance overhead. This motivates significant research to reduce the overall complexity. In this paper, we explore the effect of applying down sampling and up sampling of coded video on the performance of the HEVC. The down sampling is applied to all video frames before the encoder process begins and the up sampling is applied to all the video frames after the decoding process is completed. We use an average filter in down sampling, and machine learning based network (SRCNN) for the up sampling. In contrast to other methods, the down sampling and up sampling are applied on frame level outside the encoding/decoding processes and not in block level. Our experiments show the down sampling and up sampling can improve the HEVC encoding/decoding performance by up to 50% in some sequences with limited impact on the output encoded bit rate and decoded video quality. The performance comparison is done using different quantization values.
Conference Presentation
© (2021) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Mohamed Abdelazim, Ahmed Hamza, Abdelrahman Abdelazim, and Djamel Ait-Boudaoud "HEVC improved performance with end-to-end super resolution", Proc. SPIE 11788, Digital Optical Technologies 2021, 117880M (20 June 2021); https://doi.org/10.1117/12.2589651
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Video

Computer programming

Video coding

Super resolution

Neural networks

Quantization

Video processing

RELATED CONTENT

Texture refinement framework for improved video coding
Proceedings of SPIE (January 18 2010)
An improved adaptive interpolation approach for H.264
Proceedings of SPIE (January 29 2007)
The H.264/MPEG4 advanced video coding
Proceedings of SPIE (August 05 2009)

Back to Top