Error-Resilient Video Communication Using H.264/AVC Coding Standard

Multimedia communication over wired and wireless packet based network is exploding these years. There has been a growing need for multimedia service but unfortunately, network channels are characterized by varying bandwidth and bursty errors. H.264/AVC offers significantly improved compression efficiency and flexibility compared to previous standards. The mechanism for H.264/AVC video compression utilizes a hybrid of temporal and spatial prediction, transform coding, and variable length coding. All these methods provide high compression gain but resulted encoded stream becomes more sensitive to such error-prone transmission channels. Error resilience techniques help to cope with transmission errors and to keep acceptable video quality at decoder.Techniques for error resilience transmission of H.264/AVC video streaming over error-prone network are investigated in this thesis. There are some interactive and non-interactive methods that can help to prevent error propagation. Interactive methods result in increased end-to-end delay which is unsuitable for conversational applications e.g. video conference. Error concealment techniques are based on estimation and interpolation procedure that do not require additional information. It makes them an attractive choice for resilient video communication where bit rate and delay are very critical issues. The thesis discusses the choice of useful error resilient techniques that lead to a better concealment.The thesis presents methods to improve error resilience from the level achievable by earlier methods. These methods are designed for Flexible Macroblock Ordering (FMO), which is among the most striking new tools for error resilience within the H.264/AVC specification. A new approach in FMO is discussed that disperse possible errors throughout the video frame and help concealment for its better performance. The focus is on coping with errors caused by video data losses during the communication and enhancing the error resilient tools to get better objective and subjective quality of received video.