Shape Segmentation Using Arc/Chord Properties

Abstract

This report investigates a class of shape segmentation methods in which, for each arc of the shape's contour, we consider the region bounded by the arc and its chord; compute a simple geometrical property of this region; and choose arcs for which this property's value is a local extremum. The characteristics of this approach are analyzed for several such properties, and examples of the segmentations obtained in this way are given. It is concluded that such methods will sometimes yield results that are not perceptually plausible.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1979
Accession Number
ADA086103

Entities

People

  • W. S. Rutkowski

Organizations

  • University of Maryland

Tags

Communities of Interest

  • C4I

DTIC Thesaurus Topics

  • Applied Computer Science
  • Artificial Intelligence
  • Boundaries
  • Coefficients
  • Computer Science
  • Computer Vision
  • Computers
  • Curvature
  • Displacement
  • Equations
  • Geometry
  • Maryland
  • Night Vision
  • Shape
  • Square Roots
  • Universities

Readers

  • Computer Vision.
  • Graph Algorithms and Convex Optimization.