• DocumentCode
    2650983
  • Title

    Distortion correction on gun bore panoramic image

  • Author

    Lei, Jie ; Fu, Jianping ; Guo, Qi

  • Author_Institution
    First Dept., Ordnance Eng. Coll., Shijiazhuang, China
  • fYear
    2011
  • fDate
    17-19 June 2011
  • Firstpage
    956
  • Lastpage
    959
  • Abstract
    The observing system made up of panoramic annular lens (PAL) can observe objects around the lens, which fits for the quality inspection of gun bore. But there is serious distortion in the tangential and radial directions on the annular image formed through twice reflection. The paraboloid perspective projection model is proposed to correct the image distortion according to the imaging feature of PAL. Firstly, the PAL correction model which includes distortion parameters is constructed. The annular image is transformed back into a cylinder in three-dimensional space under this model. Then spreads out the curved image plane into a flat one. Meanwhile, the distortion parameters are obtained by adopting modified genetic algorithm according to the labelled image. Simulation and panoramic gun bore image experiments demonstrate that this algorithm could greatly correct the tangential and radial distortions and enhance the classified precision of gun bore flaw.
  • Keywords
    genetic algorithms; image processing; object detection; weapons; PAL; annular image; distortion correction; genetic algorithm; gun bore flaw; gun bore panoramic image; image distortion; image plane; observing system; panoramic annular lens; paraboloid perspective projection model; quality inspection; radial directions; radial distortions; tangential directions; tangential distortions; Adaptive optics; Calibration; Genetic algorithms; Imaging; Lenses; Nonlinear distortion; Optical distortion; PAL; distortion parameter; perspective projection; radial distortion;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Quality, Reliability, Risk, Maintenance, and Safety Engineering (ICQR2MSE), 2011 International Conference on
  • Conference_Location
    Xi´an
  • Print_ISBN
    978-1-4577-1229-6
  • Type

    conf

  • DOI
    10.1109/ICQR2MSE.2011.5976762
  • Filename
    5976762