• DocumentCode
    2627968
  • Title

    A hybrid Particle filter-CAMSHIFT model based solution for aerial maritime survivor search

  • Author

    Kin Hong Wong ; Yibo Gong ; Hung Kwan Fung

  • Author_Institution
    Dept. of Comput. Sci. & Eng., Chinese Univ. of Hong Kong, Hong Kong, China
  • fYear
    2013
  • fDate
    9-11 May 2013
  • Firstpage
    178
  • Lastpage
    182
  • Abstract
    The speed for finding and locating the survivors from shipwrecks can never be considered to be too fast. Existing methods are either too slow or inefficient. In this paper, we proposed an effective solution by employing multi-camera cooperation mechanism and hybrid Particle filter-CAMSHIFT framework to realize the detection for survivors wearing Iifejackets from the airplanes. The multi-camera system can significantly enlarge the surveillance region while preserving the imaging quality with the zoomed cameras; and the hybrid framework can locate the survivors in the sea accurately and reliably based on the color clues. Moreover, this multi-camera cooperation mechanism has made our proposed system highly extendable, thus by including more cameras in the system, the imaging quality and monitoring power can be further enhanced.
  • Keywords
    cameras; computer vision; emergency services; image colour analysis; marine accidents; nonparametric statistics; object detection; particle filtering (numerical methods); surveillance; Iifejackets; aerial maritime survivor search solution; airplanes; color clues; continuous adaptive MEANSHIFT; hybrid particle filter-CAMSHIFT model; imaging quality; monitoring power; multicamera cooperation mechanism; shipwrecks; surveillance region; survivor detection; zoomed camera; Cameras; Image edge detection; Optimized production technology; Reliability;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Technological Advances in Electrical, Electronics and Computer Engineering (TAEECE), 2013 International Conference on
  • Conference_Location
    Konya
  • Print_ISBN
    978-1-4673-5612-1
  • Type

    conf

  • DOI
    10.1109/TAEECE.2013.6557218
  • Filename
    6557218