Enhanced coder with error resilience

A scalable video coder produces a bit stream, decodable at different bit rates. It can substantially improve the quality of video transmitted over error prone channels such as Internet or wireless. The scalable video coding utilizes motion prediction to achieve higher compression efficiency. Since motion compensation is used to reduce temporal redundancy, the quality of the decoded video may decay due to the propagation of errors in the temporal domain. Hence, the double-vector motion compensation (DMC) is used in the scalable video coder, which is implemented in this work. Search strategies based on gradient-descent methods are used. In this study, error concealment techniques are used, which estimate the lost motion vectors by exploiting correlation within the video sequences.