Hybrid error concealment method for H.264 video transmission over wireless networks

Changhoon Yim, Wonjung Kim, Hyesook Lim

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

Error concealment (EC) has been adopted to reconstruct lost information for video transmission over wireless networks. For intra-frame EC in H.264 video, only spatial correlations between neighboring macroblocks have been applied for spatial interpolation of the lost macroblock. In this paper, we first propose a simple temporal EC,which uses the temporal correlations between neighboring frames and improves the EC performance for intra-frame in H.264. Then we propose a new hybrid EC method for intra-frame in H.264, which combines the spatial EC and the temporal EC adaptively. Simulation results show that the proposed temporal EC shows substantial PSNR improvement compared to the previous spatial EC in H.264 intra-frame when the amount of motion is small. The proposed hybrid EC achieves further PSNR improvement compared to the temporal ECwhen the amount of motion is rather large.

Original languageEnglish
Title of host publicationIWCMC 2007
Subtitle of host publicationProceedings of the 2007 International Wireless Communications and Mobile Computing Conference
Pages606-611
Number of pages6
DOIs
StatePublished - 2007
EventIWCMC 2007: 2007 International Wireless Communications and Mobile Computing Conference - Honolulu, HI, United States
Duration: 12 Aug 200716 Aug 2007

Publication series

NameIWCMC 2007: Proceedings of the 2007 International Wireless Communications and Mobile Computing Conference

Conference

ConferenceIWCMC 2007: 2007 International Wireless Communications and Mobile Computing Conference
Country/TerritoryUnited States
CityHonolulu, HI
Period12/08/0716/08/07

Keywords

  • Error concealment
  • Video transmission
  • Wireless networks

Fingerprint

Dive into the research topics of 'Hybrid error concealment method for H.264 video transmission over wireless networks'. Together they form a unique fingerprint.

Cite this