Kalman filter based error resilience for H.264 motion vector recovery

Ki Hong Ko, Seong Whan Kim

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

1 Scopus citations

Abstract

We propose an error concealment technique to recover lost motion vectors at the H.264 decoder side. To recover the lost motion vectors, there are two simple techniques: (1) no prediction, where the lost motion vectors are set to zeros, and (2) the prediction using the average or median of spatially adjacent blocks' motion vectors [1]. In this paper, we propose a Kalman filter based scheme for motion vector recovery, and experimented with two test image sequences: Mobile&Calendar and Susie. The experimental results show that our Kalman filter based motion vector recovery scheme improves at average 2 dB PSNR over conventional H.264 decoding with no error recovery. We also improve our scheme using Hilbert curve scan order for Kalman input, and we get 0.512 - 1.652 dB PSNR improvements with better subjective quality over line-by-line scan order.

Original languageEnglish
Title of host publicationAdvances in Mulitmedia Information Processing - PCM 2005 - 6th Pacific Rim Conference on Multimedia, Proceedings
Pages664-674
Number of pages11
DOIs
StatePublished - 2005
Event6th Pacific Rim Conference on Multimedia - Advances in Mulitmedia Information Processing - PCM 2005 - Jeju Island, Korea, Republic of
Duration: 13 Nov 200516 Nov 2005

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3767 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference6th Pacific Rim Conference on Multimedia - Advances in Mulitmedia Information Processing - PCM 2005
Country/TerritoryKorea, Republic of
CityJeju Island
Period13/11/0516/11/05

Fingerprint

Dive into the research topics of 'Kalman filter based error resilience for H.264 motion vector recovery'. Together they form a unique fingerprint.

Cite this