An accommodating edge follower

  • Authors:
  • Karl K. Pingle;Jay M. Tenenbaum

  • Affiliations:
  • Computer Science Department, Stanford University, Stanford, California;Computer Science Department, Stanford University, Stanford, California

  • Venue:
  • IJCAI'71 Proceedings of the 2nd international joint conference on Artificial intelligence
  • Year:
  • 1971

Quantified Score

Hi-index 0.00

Visualization

Abstract

We describe an operational program that locates objects in a television image and traces their edges. The program accommodates the television camera, maximizing dynamic range during acquisition and sensitivity during tracing, to obtain the most appropriate image for each phase. If the trace routine loses an edge, various heuristics diagnose the difficulty and tune both the camera and software to recover contrast at the point of difficulty. Experimental evidence of the effectiveness of accommodation is provided.