• DocumentCode
    3046828
  • Title

    A poselet-based approach for fall detection

  • Author

    Lv, Qingcong

  • Volume
    2
  • fYear
    2011
  • fDate
    9-11 Dec. 2011
  • Firstpage
    209
  • Lastpage
    212
  • Abstract
    Thousands of elderly people are victim of a fall incident every year. So it is important to have a fall detection system. This paper gives a poselet-based approach to detect a fall in video. Combined with aspect ratio, the centroid of the torso computation and interaction objects reorganization, the method may detect falls with partial occlusion and pose variations.
  • Keywords
    hidden feature removal; object recognition; pose estimation; video signal processing; aspect ratio; fall detection system; interaction object reorganization; occlusion; pose variation; poselet based approach; torso computation fall centroid; Accelerometers; Feature extraction; Humans; Senior citizens; Sensors; Torso; Training; Fall Detection; Lying Pose Recognition; Poselets;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    IT in Medicine and Education (ITME), 2011 International Symposium on
  • Conference_Location
    Cuangzhou
  • Print_ISBN
    978-1-61284-701-6
  • Type

    conf

  • DOI
    10.1109/ITiME.2011.6132023
  • Filename
    6132023