• DocumentCode
    3063246
  • Title

    Spatial segmentation based on modified morphological tools

  • Author

    Tancharoen, D. ; Jitapunkul, S. ; Chompun, S.

  • Author_Institution
    Dept. of Electr. Eng., Chulalongkorn Univ., Bangkok, Thailand
  • fYear
    2001
  • fDate
    36982
  • Firstpage
    478
  • Lastpage
    482
  • Abstract
    The MPEG-4 visual standard supports the trend in video object segmentation for the spatial segmentation process in only one way. Therefore, this paper presents the spatial segmentation process according to MPEG-4. The spatial segmentation includes the image simplification, gradient operation and watershed algorithm. This paper presents the morphological tools to use in the segmentation process. Firstly, the morphological filters are used to simplify the image by the open-close operation by partial reconstruction while preserving the boundary information. Secondly, the multiscale gradient operator is used to reduce the oversegmentation and also preserve the homogeneous region boundary. This simplification and labeling of the region method give the homogeneous region and the small number of regions. Finally, the uncertain pixels are reduced for use with the watershed algorithm
  • Keywords
    code standards; data compression; image segmentation; telecommunication standards; video coding; MPEG-4 standard; gradient operation; image simplification; modified morphological tools; morphological filters; multiscale gradient operator; spatial segmentation; uncertain pixels; video object segmentation; watershed algorithm; Digital signal processing; Image reconstruction; Image segmentation; Information filtering; Information filters; Labeling; Laboratories; MPEG 4 Standard; Object segmentation; Signal processing algorithms;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Technology: Coding and Computing, 2001. Proceedings. International Conference on
  • Conference_Location
    Las Vegas, NV
  • Print_ISBN
    0-7695-1062-0
  • Type

    conf

  • DOI
    10.1109/ITCC.2001.918842
  • Filename
    918842