Spatial error concealment with sequence-aligned texture modeling and adaptive directional recovery

  • Authors:

  • Wuttipong Kumwilaisak;C. -C. Jay Kuo

  • Affiliations:

  • Communications and Multimedia Laboratory, Department of Electronics and Telecommunication, King Mongkut's University of Technology Thonburi, Bangkok 10140, Thailand;Signal Processing and Image Processing Institute, Department of Electrical and Computer Engineering, University of Southern California, LA, CA 90007, USA

  • Venue:
  • Journal of Visual Communication and Image Representation
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

A spatial error concealment technique based on the sequence-aligned texture modeling and the adaptive directional recovery is proposed in this work. The sequence alignment technique captures the local variation and the global trend of image textures with surrounding uncorrupted pixels, and provides the best texture model under a given cost function. With the derived texture model, geometric interpolation is used to recover lost pixels adaptively based on pixel locations. There are four candidate modes of pixel sequences to recover lost pixels, and one of them is selected for the concealment purpose. The selection criterion is based on the texture pattern modes of surrounding uncorrupted blocks. The pixel sequences used for error concealment can be obtained from the computation of the decoder or the side information from the encoder. Extensive experimental results are given to demonstrate that our error concealment technique outperforms several benchmark methods in both objective and subjective tests.