• DocumentCode
    2817471
  • Title

    A fast object segmentation approach based on integral image in ALV system

  • Author

    Qian, Shenyi ; Chen, Xiaolei ; Xia, Yongquan

  • Author_Institution
    Dept. of Comput. Sci. & Eng., Zhengzhou Univ. of Light Ind., Zhengzhou, China
  • Volume
    1
  • fYear
    2010
  • fDate
    17-18 April 2010
  • Firstpage
    357
  • Lastpage
    360
  • Abstract
    In ALV system, the object segmentation approach is one of the most important research fields for object detection. In this paper, a simple and fast object segmentation approach is proposed for ALV system. Firstly, a binarizaton method based on integral image is applied to binarize the input images; Secondly, in order to save the computation time of gray mean of each pixels in squared window, a method of reduce the redundant computation is applied in our algorithm; After that, the “holes” in object are filled by inner-filling algorithm; Lastly, the filled images are segmented by using easy edge detection approach. Several experiment images of ALV system is used to verify the proposed algorithm, the result indicate that the approach is valid and feasible.
  • Keywords
    artificial intelligence; edge detection; image segmentation; mobile robots; object detection; robot vision; vehicles; ALV system; autonomous land vehicle; binarizaton method; edge detection; fast object segmentation approach; gray mean; inner filling algorithm; integral image; object detection; redundant computation; squared window; Filling; Image edge detection; Image motion analysis; Image segmentation; Lighting; Object detection; Object segmentation; Optical sensors; Pixel; Roads; binarization; iIntegral image; inner filling; object detection; object segmentation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    E-Health Networking, Digital Ecosystems and Technologies (EDT), 2010 International Conference on
  • Conference_Location
    Shenzhen
  • Print_ISBN
    978-1-4244-5514-0
  • Type

    conf

  • DOI
    10.1109/EDT.2010.5496563
  • Filename
    5496563