Abstract

Most computer vision systems use one-dimensional operators to calculate the curvature of boundaries or image contours. However, the more sophisticated biological systems use two-dimensional operators. Here we analyze how curvature can be computed by using two-dimensional operators and show the relation of this approach to the one-dimensional methods currently in use.

© 1988 Optical Society of America

Mechanisms of contour curvature discrimination

Hugh R. Wilson and Whitman A. Richards
J. Opt. Soc. Am. A 6(1) 106-115 (1989)

Encoding contour shape by curvature extrema

Whitman Richards, Benjamin Dawson, and Douglas Whittington
J. Opt. Soc. Am. A 3(9) 1483-1491 (1986)

Curvature and separation discrimination at texture boundaries

Hugh R. Wilson and Whitman A. Richards
J. Opt. Soc. Am. A 9(10) 1653-1662 (1992)

Edge-curvature discriminability argues against explicit curvature detectors

Erik de Haan
J. Opt. Soc. Am. A 12(2) 202-213 (1995)

Pattern and motion vision without Laplacian zero crossings

John G. Daugman
J. Opt. Soc. Am. A 5(7) 1142-1148 (1988)