Font Size: a A A

Research Of Adaptive Error Concealment In Video Communication Based On H.264

Posted on:2011-06-13Degree:MasterType:Thesis
Country:ChinaCandidate:H B XiaFull Text:PDF
GTID:2178360308460228Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Multimedia communications especially video communications will be of importance support services for the 3G and NGN, the quality of transmitted video is a key to business development. Because of channel defects, the error or packet loss will be unavoidable, and the same time, but the video data via efficient compression is very sensitive to the error. In order to meet the business quality of service of video stream, we must take some error control scheme on encoder and decoder. This is called error concealment that it make use of relativity on spatial or temporal of video sequences, which compensate and restore the damaged data depended on the adjacent data. The paper is focused on error post-processing of the decoder-side on the ground of H.264 standard.Firstly, this paper analyses and studies kinds of algorithm on spatial or temporal in common use, then one better algorithm is proposed for improving the accuracy of components of high frequency in error concealment on spatial-domain, which is called Concealment algorithm on sub-pixel. Then we run the simulation on the platform of JM 11.0. Secondly, another algorithm is proposed for improving the effect of concealment for the object boundary or a large number of pieces of image on temporal-domain, which is called Enhanced many weights boundary matching errors. Then we run the simulation on the platform of JM 11.0. Thirdly, the adaptive scheme is the combination of two new algorithms, adaptive scheme selection of temporal and spatial, which is decided on macro block damaged belong to new object of image at the same time. Then on condition that the different rate of packet loss in IP-based network and actual video scenarios, we run the simulation on the common test circumstances based on the H.264. At last, the results show that the proposed adaptive algorithm is superior to the original algorithm.
Keywords/Search Tags:Relativity, Error Concealment, H.264, Concealment algorithm on sub-pixel, Enhanced many weights boundary matching errors, More adaptive
PDF Full Text Request
Related items