Improvement on error concealment for H.264 spatial scalable video coding

One of the scalability features in H.264/SVC is spatial scalability, and it is found to be useful in distribution of video content to a variety of consumers. Transmission of an H.264/SVC stream over a packet network may introduce losses, at worst case would involve whole frame losses. It is necessary to conceal the errors in the decoded bitstream. We propose an improvement in the error concealment scheme in H.264/SVC by improved motion estimation and post processing of the concealed frames in the pixel domain. The proposed method successfully conceals the lost frame and was found to be effective for low bitrate video transmission over conventional methods.