Recently, the growth in mobile communication and Internet usage has lead to an increase in demand for robust transmission of compressed video over error prone channels. The quality of decoded video in an erroneous environment depends on efficient detection, localization and concealment of errors. In fact, the role of error detection and localization is very important because undetected erroneous macroblocks may not be concealed and correct macroblocks may be termed as erroneous. Given these challenges, MPEG-4 has emerged as a powerful video standard for transmission over a lossy transmission channel due to its error resilience tools and compression efficiency. We propose improved methods for error detection and localization using syntactic, semantic and statistical consistency methods in MPEG-4. Also, the improvements are suggested for spatial error concealment techniques in intra-frames.
[1]
Aggelos K. Katsaggelos,et al.
Signal Recovery Techniques for Image and Video Compression and Transmission
,
1998,
Springer US.
[2]
Yao Wang,et al.
Maximally smooth image recovery in transform coding
,
1993,
IEEE Trans. Commun..
[3]
Fernando Pereira,et al.
Error resilience and concealment performance for MPEG-4 frame-based video coding
,
1999,
Signal Process. Image Commun..
[4]
Ness B. Shroff,et al.
Error Concealment in Encoded Video Streams
,
1998
.
[5]
Pamela Cosman,et al.
Comparison of error concealment strategies for MPEG video
,
1999,
WCNC. 1999 IEEE Wireless Communications and Networking Conference (Cat. No.99TH8466).