• DocumentCode
    255539
  • Title

    Human detection based on explicit skin colour space thresholding and minor motion detection

  • Author

    Joshi, D. ; Purwar, A. ; Dasgupta, M.S.

  • Author_Institution
    Dept. of Electr. & Electron. Eng., Birla Inst. of Technol. & Sci., Pilani, India
  • fYear
    2014
  • fDate
    11-13 Dec. 2014
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    In this paper, a novel approach for human presence detection from digital image is presented. This approach is based on combined clue from skin colour and minor motion detection. For skin colour detection, explicit thresholding of CbCr colour channels is used and for minor motion detection morphological differences of multiple consecutive images has been used. MATLAB® is used for processing. Relatively small part of exposed human body in the frame of an image is sufficient for detection. The approach is computationally efficient, potentially real time and robust. The combined technique gives improved performance compared to each individual technique. Location of the object (human) within the frame is also simultaneously identified.
  • Keywords
    image classification; image colour analysis; image segmentation; motion estimation; object detection; skin; Matlab; colour channels; digital image; explicit skin colour space thresholding; explicit thresholding; human body; human detection; human presence detection; minor-motion detection; multiple consecutive images; object location identification; performance improvement; skin colour detection; Algorithm design and analysis; Classification algorithms; Image color analysis; Lighting; MATLAB; Motion detection; Skin; MATLAB®; colour channels; human presence detection; image segmentation; motion detection; skin colour identification; temporal differencing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    India Conference (INDICON), 2014 Annual IEEE
  • Conference_Location
    Pune
  • Print_ISBN
    978-1-4799-5362-2
  • Type

    conf

  • DOI
    10.1109/INDICON.2014.7030518
  • Filename
    7030518