Abstract
Subpixel-edge detection is the first stage in processing many high-level vision algorithms. However, the study of the statistical properties of such data has remained incomplete in most research. We present a method for estimating the second-order moments of the subpixel-edge position, computed by a deterministic algorithm based on three quadratic interpolations. The algorithm is tested on different types of noise (Gaussian, impulse, colored) and compared with two methods in the literature. The results show that this approach remains accurate even at high noise levels.
© 2004 Optical Society of America
Full Article | PDF ArticleMore Like This
Frédéric Bouchara and Sofiane Ramdani
J. Opt. Soc. Am. A 26(4) 820-832 (2009)
Vincent Samson, Frédéric Champagnat, and Jean-François Giovannelli
Appl. Opt. 43(2) 257-263 (2004)
Jorge Ares and Justo Arines
Appl. Opt. 43(31) 5796-5805 (2004)