A novel error robust video coding concept using motion vectors and parity bits

The paper investigates the concept of developing an error robust encoding mechanism with the minimum level of redundancy. The traditional prediction based video coding architectures have high susceptibility for channel errors. Most of these techniques fail to deliver acceptable error recovery at the receiving end at high error rate transmission networks. However, the parity based Wyner-Ziv paradigm exhibits acceptable level of error recovery in highly error prone environments, at the expense of higher processing power. The proposed method attempts to integrate all of the advantages of these two techniques in order to mitigate the disadvantages associated with each and every technique while providing high error robustness to the transmitted bit stream.