An error robust distortion model for depth map coding in error prone network

Error robustness becomes an important issue when the compressed depth map is transmitted over error prone network. There have been some algorithms to improve the error robustness of color video in the past years. However, their extension to depth map are not reasonable due to the difference between depth map and color video. In this paper, a novel error robust distortion model is proposed to enhance the error robustness of depth map, in which the end-to-end distortion of virtual view is estimated. More specifically, the proposed distortion model recursively computes the expected decoded depth for each pixel by considering the channel condition and error concealment method, and then the expected decoded depth is used to estimate the end-to-end distortion of virtual view. Experimental results show that the proposed distortion model consistently outperforms the conventional distortion model and outperforms the random intra updating algorithm in most cases.