• DocumentCode
    649727
  • Title

    A homogenizing filter for depth map compressive sensing using edge-awarded method

  • Author

    Taemin Cho ; Youngwoo Lee ; Jitae Shin

  • Author_Institution
    Coll. of Inf. & Commun. Eng., Sungkyunkwan Univ., Suwon, South Korea
  • fYear
    2013
  • fDate
    14-16 Oct. 2013
  • Firstpage
    591
  • Lastpage
    595
  • Abstract
    The edge-awarded method for depth map compressive sensing (CS) was proposed. The edge-awarded depth map CS showed better visual quality than H.264/AVC which are widely used. This method can reconstruct efficiently the boundary of objects. However, this makes staircase artifact around edges unfortunately. These artifact makes pixel value around edge to distorted value. Edge is most important information in depth map. So, this artifact cause really critical problem especially when image rendering process. In this paper, to eliminate staircase artifact generated by edge-awarded depth map CS, we proposed a homogenizing filter to relive the staircase effects. The experimental results show the proposed filter provides better visual image quality than compressive sensing using edge-adaptive method and H.264.
  • Keywords
    compressed sensing; filtering theory; image coding; rendering (computer graphics); H.264/AVC; boundary reconstruction; depth map compressive sensing; edge-adaptive method; edge-awarded depth map CS; homogenizing filter; image rendering; pixel value; staircase artifact; visual image quality; Compressed sensing; Educational institutions; Filtering algorithms; Filtering theory; Image coding; Image edge detection; Image reconstruction; Compressive sensing(CS); Depth Map; Homogenous filter; edge-adaptive method;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    ICT Convergence (ICTC), 2013 International Conference on
  • Conference_Location
    Jeju
  • Type

    conf

  • DOI
    10.1109/ICTC.2013.6675427
  • Filename
    6675427