• DocumentCode
    1337402
  • Title

    Structural analysis of DEM´s by intersection of surface normals in a three-dimensional accumulator space

  • Author

    Baudemont, Frédéric ; Parrot, Jean-François

  • Author_Institution
    LGGST, UPMC, Paris, France
  • Volume
    38
  • Issue
    3
  • fYear
    2000
  • fDate
    5/1/2000 12:00:00 AM
  • Firstpage
    1191
  • Lastpage
    1198
  • Abstract
    This paper proposes a new methodology in DEM analysis in order to extract relevant information. The method proposed generates a set of 3D accumulators by using the DEM data. Each accumulator contains the number of intersections of the normals derived from the DEM surface. An alternative procedure has been developed in order to minimize the huge memory space required by a complete set of voxels. The principle consists of scanning the discrete space by a vertical plane. This plane, called a Voxel Wall, is composed of all the local accumulators. Two matrices are generated. In the first, named Values of the Accumulator Matrix (VAM), the positions x,y and the corresponding maximum value μ of intersections encountered in each column of the Voxel Wall are registered. The depth ζ of each μ is coded on the second matrix, called Depth of the Accumulator Maxima (DAM). The results provided by regular shapes such as hemispheres, cones, and pyramids, as well as those produced by deformations and/or linear translations, are analyzed. An exhaustive interpretation of these results requires the creation of a catalogue of all the criteria for describing a given shape, whether it be isolated or not. The procedure has been applied to some “real” DEM containing volcanic edifices, crest lines, and bulging shapes
  • Keywords
    geodesy; topography (Earth); 3D accumulator; DAM; DEM; DEM analysis; Depth of the Accumulator Maxima; VAM; Values of the Accumulator Matrix; Voxel Wall; digital elevation model; geodesy; geodetic network; intersection; land surface topography; local accumulators; matrices; matrix; methodology; structural analysis; surface normal; three-dimensional accumulator space; Data mining; Image sampling; Information analysis; Interpolation; Morphology; Pattern recognition; Pixel; Polynomials; Shape; Statistics;
  • fLanguage
    English
  • Journal_Title
    Geoscience and Remote Sensing, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0196-2892
  • Type

    jour

  • DOI
    10.1109/36.843011
  • Filename
    843011