Retrieval of missing frames in a given video sequence

Reconstruction of missing frames in a video sequence is of considerable interest in recent times. In this paper we present our work on reconstruction of missing information based on local and global coherence of given video sequence such that, the missing information is visually coherent with available information. The reconstruction is based on both, the spatial and temporal domain study of the missing information. The intensive computation, due to large dimensionality of the video data, is greatly reduced by applying some preprocessing and separating the foreground and background information, and then applying the algorithm on the foreground information.