Transform error concealment on video coding

Abstract

Recovering corrupted or lost regions of images and videos to maintain acceptable perceptual quality is one of the most challenging tasks in multimedia transmission. The goal of error concealment is to exploit correlation between frames to recover missing data. Error concealment schemes typically rely on producing a replacement for erroneous data that is similar to the original. This is possible since information such as video exhibits large amounts of self similarity. This project deals with implementing an error concealment method in video sequences in the frequency domain. To perform error concealment in videos, videos first need to be converted to frames and then transformed into frequency domain images using one of the may transform available. Lost or erroneous data in videos can be obtained by using successive frames. Lost information in predictive(P frames) and bi-directional predictive frames (B frames) can be reconstructed using intracoded frames (I frames). Lost information in intra-coded frames can be reconstructed using previous intra-coded frames. Error concealment is used to hide the effects of errors detected within digital video information.