Hybrid frame-recursive block-based distortion estimation model for wireless video transmission

In wireless environments, video quality can be severely degraded due to channel errors. Improving error robustness towards the impact of packet loss in error-prone network is considered as a critical concern in wireless video networking research. Data partitioning (DP) is an efficient error-resilient tool in video codec that is capable of reducing the effect of transmission errors by reorganizing the coded video bitstream into different partitions with different levels of importance. Significant video performance improvement can be achieved if DP is jointly optimized with unequal error protection (UEP). This paper proposes a fast and accurate frame-recursive block-based distortion estimation model for the DP tool in H.264.AVC. The accuracy of our model comes from appropriately approximating the error-concealment cross-correlation term (which is neglected in earlier work in order to reduce computation burden) as a function of the first moment of decoded pixels.Without increasing computation complexity, our proposed distortion model can be applied to both fixed and variable block size intra-prediction and motion compensation. Extensive simulation results are presented to show the accuracy of our estimation algorithm.