Title :
Robust video decoding by exploiting residual source redundancy
Author :
Lulu Sun;Shaoshuai Gao
Author_Institution :
School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing, China
Abstract :
To achieve high compression efficiency, traditional video coding standards like H.264/AVC usually use entropy coding and prediction coding. As a result, a single channel error may lead to artifacts in decoding the following bitstream and even affect the following frames, which will deteriorate the video quality greatly. A robust video decoding method called the iterative verification decoding (IVD) scheme for H.264/AVC coded bitstream is proposed in this paper. By only exploiting residual redundancy of source coding, this scheme can recover high-quality video when some erasure errors occur. Moreover, a novel method based on bitstream linear mixing (BLM) is proposed, which can greatly enhance the error detection capability of the video decoder. Performance evaluation results for the straight decoding (SD), the proposed IVD and BLMIVD schemes with different erasure error rates are presented. It has been shown that our proposed schemes can perform much better than the SD scheme. And for the BLMIVD scheme, it can even achieve up to 18 dB PSNR gain when compared with the SD scheme.
Keywords :
"Decoding","Iterative decoding","Syntactics","Redundancy","Encoding","Buffer storage","Complexity theory"
Conference_Titel :
Wireless Communications & Signal Processing (WCSP), 2015 International Conference on
DOI :
10.1109/WCSP.2015.7341082