• DocumentCode
    3488095
  • Title

    SUSAN 3D operator, principal saliency degrees and directions extraction and a brief study on the robustness to noise

  • Author

    Walter, Nicolas ; Aubreton, Olivier ; Fougerolle, Yohan D. ; Laligant, Olivier

  • Author_Institution
    Lab. Le2i, Le Creusot, France
  • fYear
    2009
  • fDate
    7-10 Nov. 2009
  • Firstpage
    3529
  • Lastpage
    3532
  • Abstract
    In order to complete the work on the 3D SUSAN operator, an extraction of principal saliency degrees and direction is proposed. This will lead the proposed operator to a complete classification of vertices in five different classes: cavity points, valley points, flat points, prong points and salient points. To assess the operator, we propose two different studies. The robustness to the resolution and the regularity of meshes is presented for three different operators. To complete the evaluation, the robustness to noise is analysed, for the Stokely and the SUSAN operator. Our conclusion presents the benefits, drawbacks of the technique and future works.
  • Keywords
    feature extraction; image classification; mesh generation; noise; SUSAN 3D operator; Stokely operator; cavity points; directions extraction; flat points; mesh regularity; noise; principal saliency degree extraction; prong points; salient points; smallest univalue segment assimilating nucleus; valley points; vertices classification; Computer industry; Image segmentation; Noise measurement; Noise reduction; Noise robustness; Pixel; Production; Storage area networks; SUSAN; curvature; point characterization; robustness to noise; saliency; salient point detection;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Image Processing (ICIP), 2009 16th IEEE International Conference on
  • Conference_Location
    Cairo
  • ISSN
    1522-4880
  • Print_ISBN
    978-1-4244-5653-6
  • Electronic_ISBN
    1522-4880
  • Type

    conf

  • DOI
    10.1109/ICIP.2009.5414078
  • Filename
    5414078