Conference Paper

A spatial-domain error concealment method with edge recovery and selective directional interpolation

Integrated Media Syst. Center, Univ. of Southern California, Los Angeles, CA, USA
DOI: 10.1109/ICME.2003.1221574 Conference: Multimedia and Expo, 2003. ICME '03. Proceedings. 2003 International Conference on, Volume: 2
Source: IEEE Xplore


A low complexity spatial-domain error concealment method is proposed to reconstruct still images and intra-coded (I) frames in video when they are transmitted through unreliable channels in this work. The proposed concealment algorithm works with the following steps. First, missing edges in a lost macroblock (MB) are detected and recovered using gradient data. Then, the lost MB is implicitly divided into several segments along the recovered edges. Finally, each pixel in a segment is directionally interpolated from boundary pixels adjacent to the segment. Experimental results show that the proposed algorithm can recover high as well as low frequency information in lost MBs and provide better visual quality in comparison with the conventional spatial domain interpolation.

  • Source
    • "Different from the approaches in [17], [18], and [20]–[22] "
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we propose an edge-directed error concealment (EDEC) algorithm, to recover lost slices in video sequences encoded by flexible macroblock ordering. First, the strong edges in a corrupted frame are estimated based on the edges in the neighboring frames and the received area of the current frame. Next, the lost regions along these estimated edges are recovered using both spatial and temporal neighboring pixels. Finally, the remaining parts of the lost regions are estimated. Simulation results show that compared to the existing boundary matching algorithm [1] and the exemplar-based inpainting approach [2] , the proposed EDEC algorithm can reconstruct the corrupted frame with both a better visual quality and a higher decoder peak signal-to-noise ratio.
    Full-text · Article · Apr 2010 · IEEE Transactions on Circuits and Systems for Video Technology
  • Source
    • "While this is possibly sufficient for higher resolution, it brings limited improvement for low resolutions where often more edges meet in one missing area. In [6] a method considering all detected edges is proposed. This method matches the edges entering and leaving the missing area, partitions the area accordingly and smoothes preserving the edges. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The losses of packets cannot he avoided if real-time video is transported over error prone environments. To conceal missing parts of video pictures, the spatial and temporal correlation feature of natural video sequences is used. However, in some cases - for instance in case of a scene change - there is no temporal correlation available and thus spatial error concealment has to be used. This article proposes flexible spatial error concealment based on directional interpolation method that performs well also if only two neighboring boundaries are used as common for H.264 spatially predicted frames. The proposed method was implemented and tested in a H.264 codec together with other error concealment methods to evaluate their performance
    Preview · Conference Paper · Jul 2005
  • [Show abstract] [Hide abstract]
    ABSTRACT: Thelosses ofpackets cannot beavoided ifreal-time vuideo istransported overerror prone environments. Toconceal missing parts ofvideo pictures, thespatial andtempo- ralcorrelation feature ofnatural video sequences isused. However, insomecases -forinstance incaseofascene change -there isnotemporal correlation available and thusspatial error concealment hastobeused. Thisarti- cleproposes aflexzble spatial error concealment based on directional interpolation method that performs well also if only twoneighboring boundaries areusedascommonfor H.264 spatially predicted frames. Theproposed method wasimplemented andtested inaH.264 codectogether withother errorconcealment methods toevaluate their performance.
    No preview · Article ·
Show more