28 lines
5.7 KiB
TeX
Executable File
28 lines
5.7 KiB
TeX
Executable File
|
|
\selectlanguage{english}
|
|
|
|
\chapter*{\HIGHLIGHT{Abstract}}
|
|
\addcontentsline{toc}{chapter}{Abstract}
|
|
|
|
% The increasing diversity of services, the demands for high quality multimedia and the growing popularity of Ultra-HD formats are pushing technologies forward and opening new research directions.
|
|
% The recently developments in video coding techniques resulted in the novel standard, referred to as High Efficiency Video Coding (HEVC).
|
|
% The high level of compression efficiency achieved by \ac{hevc} coding techniques decreases the error resilience performance under error prone conditions.
|
|
%
|
|
% In the scope of this research work the error resilience of the \ac{hevc} standard was firstly analysed. The results reveal the lack of robustness of the motion vector (MV) coding, leading to quality degradation due to error propagation. It is also shown that the temporal dependency of motion information is comparatively higher than that in the H.264/AVC standard.
|
|
% Based on this evidence, this paper proposes a method to make intelligent use of temporal motion vector (MV) candidates during the motion estimation process, in order to decrease the temporal dependency, and improve the error resiliency without penalising the rate-distortion performance.
|
|
% The simulation results show that the proposed method improves the error resilience under tested conditions, compared to the reference method that always enables temporal \ac{mv} candidates.
|
|
%
|
|
% Moreover, a method based on the prediction dependency of \acs{mv}is proposed, to select the most relevant ones for redundant coding with reduced overhead.
|
|
% The \ac{mv} dependencies are analysed in the encoder to prioritise the motion information , based on the number of subsequent dependent coding units.
|
|
% Then a subset of prioritised \acs{mv}selected with the previous is transmitted as redundancy, to reduce the use and propagation of mismatched \ac{mv} predictions in case of transmission errors or data loss.
|
|
% The simulation results show that the proposed \acs{mv} selection method can effectively identify the most relevant motion field, achieving improved error robustness with a reduced redundancy overhead.
|
|
%
|
|
% Further research will be pursed to improve the error robustness of the \ac{hevc} standard and develop novel concealment methods using combined info between the encoder and decoder side.
|
|
|
|
The increasing diversity of services and the demand for ever higher quality multimedia at wider resolutions, which leads to the growing popularity of Ultra-HD (UHD) format, are pushing technologies forward and opening new research directions. The recent developments in video coding techniques have resulted in a novel standard called \ac{hevc} that is the next breakthrough in multimedia compression after the globally adopted MPEG-4/\ac{h264} standard. However, the significantly improved levels of compression efficiency achieved by the \ac{hevc} standard lead to an inherent reduction in the error resilience performance under error prone conditions. It should be aimed at ensuring an adequate level of protection against network and transmission induced impairments (e.g., packet losses), without sacrificing the compression efficiency thanks to the novel block level operations done by \ac{hevc}.
|
|
|
|
My research work takes the \ac{hevc} standard to its core and builds new techniques on top of it to enhance its error resilience performance, and develop new processing techniques to improve decoder side error concealment performance. The first scope of this research work is to make an extensive analysis of various error resilience techniques developed for the \ac{hevc} standard, but also for the previous video coding standards (most notably \ac{h264}); identify their shortcomings and address those shortcomings. Up to date work has revealed that the \ac{mv} coding technique lacks robustness, which leads to quality degradation due to error propagation in lossy transmission environments. It has also been seen that the temporal dependency of motion information is comparatively higher than that in the \ac{h264} standard. Based on this evidence, a method was developed to make intelligent use of temporal \ac{mv} candidates during the motion estimation process, in order to decrease the temporal dependency, and improve the error resiliency without penalising the rate-distortion performance. The simulation results have shown that the proposed method improves the error resilience under tested conditions, compared to the reference method that always enables temporal \acs{mv} candidates. Another proposed method has exploited the prediction dependency of \glspl{mv}in order to select the most relevant ones for redundant coding with reduced overhead. Simulation results have shown that the proposed \gls{mv} selection method can effectively identify the most relevant motion fields, achieving improved error robustness with a significantly reduced redundancy overhead.
|
|
|
|
The second scope of this research work is to develop new processing techniques for the concealment of the coding elements lost during the transmission and delivery system, using combined information between the encoder and decoder sides. Error concealment techniques mitigate the detrimental effects of network impairments on the perceptual multimedia quality and the performance these techniques determine the quality of a system directly. The third and the ultimate aim of this research work is to strike a balance between the designed error resilience and error concealment techniques so as to meet a set of QoS/QoE requirements associated with current and future multimedia-based services (e.g., real-time multi-party video conferencing, Ultra-HD video on demand, etc.), such as bandwidth and delay. It is envisaged to produce a set of configuration profiles for chosen services based around the developed tools for \ac{hevc}.
|
|
|