Open Collections

Scalable coding of H.264 video

University of British Columbia

Real-time transmission of digital video over media, such as the Internet and wireless networks has recently been receiving much attention. A big challenge of video transmission over such networks is the variation of available bandwidth over time. Traditional video coding standards whose main objective is to optimize the quality of transmitted video at a given bitrate, do not offer effective solutions to the bandwidth variation problem. To deal with this problem, different scalable video coding techniques have been developed. The latest video coding standard, H.264, provides superior compression efficiency over all previous standards. This standard, however, does not include tools for coding the video in a scalable fashion. In this thesis, we introduce methods that allow encoding and transmitting of H.264 video in a scalable fashion. The method we propose is an adaptation of the existing MPEG-4 Fine Granular Scalability structure (FGS) to the H.264 standard. Our proposed algorithm minimizes the added number of the bits needed in adapting the advanced features of H.264 to the FGS system. Our proposed system has the advantages of being highly error resilient and having low computational complexity. Due to its structure, the FGS standard has low coding efficiency when compared to single layer coding. To overcome this problem, we also introduce a hybrid method that combines our proposed H.264 based FGS approach with the stream-switching approach employed in the H.264 standard. By combining different techniques, our proposed system offers a complete solution for all kinds of applications. The proposed system outperforms existing systems by offering optimum bandwidth utilization and improved video quality for the end user.

Thesis/Dissertation

application/pdf

2009-11-24

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/15728

Electrical and Computer Engineering

University of British Columbia

UBCV

DSpace