Advances in the Analysis of Topographic Features on Discrete Images

  • Authors:
  • Pierre Soille

  • Affiliations:
  • -

  • Venue:
  • DGCI '02 Proceedings of the 10th International Conference on Discrete Geometry for Computer Imagery
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

By viewing the grey scale values of 2-dimensional (2-D) images as elevation values above the image definition domain, geomorphological terms such as crest lines, watersheds, catchment basins, valleys, and plateaus have long been used in digital image processing for referring to image features useful for image analysis tasks. Because mathematical morphology relies on a topographic representation of 2-D images allowing for grey scale images to be viewed as 3-D sets, it naturally offers a wide variety of transformations for extracting topographic features. This paper presents some advances related to the imposition of minima, the lower complete transformation, the hit-or-miss transform, and the extraction of crest lines by a skeletonisation procedure.