• DocumentCode
    2837680
  • Title

    A New Pseudo-Zernike Boundary Moment for Marine Targets Feature Extraction

  • Author

    Shi-long Wang ; Yu-ru Xu ; Yong-jie Pang

  • Author_Institution
    Nat. Key Lab. of Sci. & Technol. on Autonomous Underwater Vehicle, Harbin Eng. Univ., Harbin, China
  • fYear
    2011
  • fDate
    17-18 July 2011
  • Firstpage
    1
  • Lastpage
    5
  • Abstract
    In the marine environment, the underwater images would have low SNR and the detail is fuzzy due to scattering and absorption of a variety of suspended matter in sea water and the water itself. For the three types of characteristic marine targets, the novel moments called Pseudo-Zernike boundary moments are proposed. By expanding the scope of the calculation to any rectangle and the use of fast algorithm and the normalization method, the invariants of discrete moments are constructed. Thus, the moments can be used as the descriptors of the samples. Experimental results of compared with the traditional HU moments and Pseudo-Zernike regional moments show that the new moment invariants not only can reduce the calculation in data processing to a large extent, but also improve the robustness and timeliness for engineering applications. When applying to the practical engineering, that is particularly approval for AUV to complete a certain mission.
  • Keywords
    feature extraction; marine engineering; remotely operated vehicles; underwater vehicles; AUV; Pseudo-Zernike boundary moment; discrete moment invariants; marine targets feature extraction; normalization method; underwater images; Algorithm design and analysis; Feature extraction; Image recognition; Image segmentation; Pattern recognition; Polynomials; Target recognition;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Circuits, Communications and System (PACCS), 2011 Third Pacific-Asia Conference on
  • Conference_Location
    Wuhan
  • Print_ISBN
    978-1-4577-0855-8
  • Type

    conf

  • DOI
    10.1109/PACCS.2011.5990238
  • Filename
    5990238