-->

Scale-space: A framework for handling image structures at multiple scales

Tony Lindeberg

Computational Vision and Active Perception Laboratory (CVAP)
Department of Numerical Analysis and Computing Science
KTH, (Royal Institute of Technology)
S-100 44 Stockholm, Sweden

http://www.csc.kth.se/~tony
Email: <tony@csc.kth.se>

In: Proc. CERN School of Computing,
Egmond aan Zee, The Netherlands,
September 8--21, 1996.

Abstract:

This article gives a tutorial overview of essential components of scale-space theory -- a framework for multi-scale signal representation, which has been developed by the computer vision community to analyse and interpret real-world images by automatic methods.

 
• The need for multi-scale representation of image data
• Scale-space representation: Definition and basic ideas
• Axiomatic scale-space formulations
  • Original formulation.
  • Causality.
  • Non-creation of local extrema and semi-group structure.
  • Non-enhancement of local extrema and infinitesimal generator.
  • Scale invariance.
  • Relations to biological vision.
• Multi-scale feature detection
  • Feature detectors expressed in terms of local directional derivatives
    • Edge detection.
    • Ridge detection.
    • Corner detection and blob detection.
• Automatic scale selection
  • Edge detection.
  • Ridge detection.
  • Corner detection.
  • Blob detection.
• Summary and outlook
• References
• About this document ...