• DocumentCode
    2674588
  • Title

    A robust object-specified active contour model for tracking smoothly deformable line-features and its application to outdoor moving image processing

  • Author

    Nakazawa, Yuji ; Komatsu, Takashi ; Saito, Takahiro

  • Author_Institution
    Dept. of Electr. Eng., Kanagawa Univ., Yokohama, Japan
  • Volume
    1
  • fYear
    1996
  • fDate
    16-19 Sep 1996
  • Firstpage
    689
  • Abstract
    By restricting object tracking, we form a robust object-specified active contour model which adapts itself to the properties of the specified object, that is to say, the objective smoothly deformable line-feature. The proposed model allows us to track the objective deformable line-feature throughout the entire observed noisy moving image sequence automatically. Furthermore we apply it to the surveillance image processing task that suspended power-transmission wires swinging due to strong winds are automatically tracked in an observed outdoor noisy moving image sequence. The experimental results demonstrate that the proposed method is considerably robust to noise
  • Keywords
    edge detection; image sequences; motion estimation; noise; object recognition; power engineering computing; power overhead lines; power transmission lines; surveillance; tracking; automatic tracking; experimental results; noisy moving image sequence; object recognition; outdoor moving image; robust object-specified active contour model; smoothly deformable line-features tracking; strong winds; surveillance image processing; suspended power-transmission wires; Active contours; Active noise reduction; Deformable models; Degradation; Image processing; Image recognition; Image sequences; Noise robustness; Wires; Working environment noise;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Image Processing, 1996. Proceedings., International Conference on
  • Conference_Location
    Lausanne
  • Print_ISBN
    0-7803-3259-8
  • Type

    conf

  • DOI
    10.1109/ICIP.1996.560972
  • Filename
    560972