Research On Up-sampling Filtering In Spatial Scalable Video Coding

With the fast development of network and multimedia techniques, multimedia applications based on networks have been widely applied to daily life. There are mainly three characteristics in the current multimedia application environment: heterogeneous networks, diversity of receiving terminals and complexity of multimedia applications. These features lead to difficulties as well as challenges to multimedia applications. Scalable video coding, which contains multiple sub-bitstreams for different temporal, spatial and quality scalabilities, differs greatly from traditional single-layer coding, and is adaptive to above environments.Spatial scalable video coding can provide various resolution sequences for a single encoded bit-stream. The various resolutions can be achieved by up-sampling a lower image resolution sequence. Firstly, an efficient inter-layer up-sampling filtering algorithm in spatial scalable video coding was proposed. Based on different component-sensitive of the human vision system, scheme assigns up-sampling filters depending on the sensitivity priority of the image component in order to increase the coding efficiency was proposed. Secondly, a classified multi-filter algorithm was proposed, which divides an image into edges and non-edges. Conventional up-sampling filter is applied to non-edges, while Wiener filter is applied to edges, which could preserve more edge infomation in the image sequence.The experiment results show that scheme I could deeply reduce the computational complexity, without noticeable bit-rate and PSNR performance loss. Scheme II could improve the bit-rate and PSNR performance, at the cost of introducing a few computation efforts.