This paper presents an evaluation of edge detector performance. We use the
task of structure from motion (SFM) as a "black box" through which to evalu
ate the performance of edge detection algorithms. Edge detector goodness is
measured by how accurately the SFM could recover the known structure and m
otion from the edge detection of the image sequences. We use a variety of r
eal image sequences with ground truth to evaluate eight different edge dete
ctors from the literature. Our results suggest that ratings of edge detecto
r performance based on pixel-level metrics and on the SFM are well correlat
ed and that detectors such as the Canny detector and Heitger detector offer
the best performance.