Hybrid spatio-temporal error concealment technique for image/video transmission

Compressed bit streams are, in general, very sensitive to erroneous channel. Compressed image or video transport over error prone wireless channels may results in loss or erroneous decoding and due to that bit stream may cause severe degradation on picture quality. Various error control coding techniques like, error resilience, forward error control coding and error concealment are used to control the error. This paper proposes hybrid error concealment technique. In spatial domain three methods weighted average interpolation (WAI), Directional interpolation (DI) and transformed based concealment (TBC) are implemented and compared for various noises. For video, one more temporal concealment is implemented based on motion vector. To reduce the computational complexity and better quality, we proposed a new hybrid spatio-temporal video error concealment technique with mode selection based on boundary matching decision. It performs well compare to spatial and temporal methods separately and provides satisfactory results over a large set of conditional channels, where significant information loss is present.