Regular Article
Edge Detector Evaluation Using Empirical ROC Curves

https://doi.org/10.1006/cviu.2001.0931Get rights and content

Abstract

We demonstrate a method for evaluating edge detector performance based on receiver operating characteristic (ROC) curves. Edge detector output is matched against ground truth to count true positive and false positive edge pixels. A detector's parameter settings are trained to give a best ROC curve on one image and then tested on separate images. We compute aggregate ROC curves based on 1 set of 50 object images and another set of 10 aerial images. We analyze the performance of 11 different edge detectors reported in the literature.

References (42)

  • K.W. Bowyer et al.

    Empirical Evaluation Techniques in Computer Vision

    (1998)
  • D.J. Bryant et al.

    Evaluation of edge operators using relative and absolute grading

    IEEE Conf. on Pattern Recognition and Image Processing

    (1979)
  • J. Canny

    A computational approach to edge detection

    IEEE Trans. Pattern Anal. Machine Intell.

    (1986)
  • K. Cho et al.

    Quantitative evaluation of performance through bootstrapping: Edge detection

    Int. Symp. on Computer Vision

    (1995)
  • H. Christensen et al.

    guest editors, Special issue on performance evaluation

    Machine Vision Appl.

    (1997)
  • S. Dougherty and K. W. Bowyer, Objective evaluation of edge detectors using a formally defined framework, in Empirical...
  • E.S. Deutsch et al.

    A quantitative study of the orientation bias of some edge detector schemes

    IEEE Trans. Comput.

    (1978)
  • P.W. Eichel et al.

    Quantitative analysis of a moment-based edge operator

    IEEE Trans. Systems Man Cybernet.

    (1990)
  • W.E.L. Grimson et al.

    Comments on “Digital step edges from zero crossings of second directional derivatives”

    IEEE Trans. Pattern Anal. Machine Intell.

    (1985)
  • R.M. Haralick

    Digital step edges from zero crossings of second directional derivatives

    IEEE Trans. Pattern Anal. Machine Intell.

    (1984)
  • Cited by (203)

    View all citing articles on Scopus
    View full text