DocumentCode :
3017108
Title :
Enhanced Backward Error Concealment for H.264/AVC Videos on Error-Prone Networks
Author :
Ping-Chin Wang ; Chow-Sing Lin
Author_Institution :
Dept. of Comput. Sci. & Inf. Eng., Nat. Univ. of Tainan, Tainan, Taiwan
fYear :
2013
fDate :
2-5 July 2013
Firstpage :
62
Lastpage :
66
Abstract :
Transmitting compressed video data is quite common over wireless and wired networks. In error-prone networks, however, packet loss can lead to incorrect decoding. With the new generation standard, H.264/AVC, such issues are prevalent and drastically degrade the quality of decoded video. To solve these problems, error concealment is applied in the decoder. However, traditional error concealment does not give satisfactory results in all cases, and can result in whole frame loss. This issue is addressed by backward error concealment that conceals corrupted frames by using succeeding frames that have been correctly received. This backward error concealment efficiently conceals most of the corrupted pixels, however, it excludes the unreferenced pixels. In this paper, we propose an enhanced backward error concealment method. Based on the continuity of moving objects, the proposed method provides estimated motion vectors to conceal unreferenced pixels. Experimental results show that the proposed method achieves better performance in terms of distortions.
Keywords :
data compression; decoding; motion estimation; vectors; video coding; video communication; H.264/AVC videos; compressed video data; corrupted frames; corrupted pixels; decoded video quality; enhanced backward error concealment method; error-prone networks; estimated motion vectors; incorrect decoding; moving objects; new generation standard; packet loss; traditional error concealment does; unreferenced pixels; whole frame loss; wired networks; wireless networks; Decoding; Extrapolation; Propagation losses; Resilience; Vectors; Video coding; Videos; Backward decoding; Error concealment; Frame loss; H.264/AVC; Video communication;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Biometrics and Security Technologies (ISBAST), 2013 International Symposium on
Conference_Location :
Chengdu
Print_ISBN :
978-0-7695-5010-7
Type :
conf
DOI :
10.1109/ISBAST.2013.12
Filename :
6597667
Link To Document :
بازگشت