Research On The Scalable Video Coding Of The H.264/AVC Standard

Scalable Video Coding has been developed a new kind of video coding recently. ITU-T and ISO/IEC have established extending video coding standard—SVC based on H.264 since January,2005. The goal of SVC is to provide scalability with good compression efficiency by allowing combinations of scalable layers, including the spatial layer, temporal layer and Signal-to-Noise Ratio layer. This enables a simple and flexible solution for transmission over heterogeneous networks, additionally providing adaptability for bandwidth variations. Therefore, it is of great significance to research on encoding algorithm of H.264 Scalable Video Coding.The article firstly summarized limitations in terms of supporting scalability of video coding systems by introducing the development process of video compression standards. Then it introduced coding structure of H.264 and put emphasis on analyzing the key technologies of H.264, including intra-prediction, inter-prediction, entropy coding etc. Secondly, it researched the overall implementation framework based on the extending standard scalable video coding of H.264, making particular discussion on key technologies to achieve temporal scalability, spatial scalability and Signal-to-Noise Ratio scalability, with temporal , spatial and extended spatial, and coarse, fine and medium granular scalable, and hybrid scalable coding principle included. In the end, based on JSVM9.19.7 JVT has released and some standard video sequence , it researched on the coding performance of SVC by simulating , having focus on analyzing its key parameters, including GOP, InterLayerPred, MGSControl, MGSVector, SearchRange and Renumber. And making deep research on medium granular scalability, comparing the primary algorithm to adaptive enhancement layer medium granular scalability, it is the result that verified that the proposed algorithm through simulation can improve the coding efficiency while achieved fine SVC.