FEATURE-EXTRACTION USING THE CONSTRAINED GRADIENT

Low-level operators are needed in most computer vision applications in order to get relevant image primitives. In this paper, we present a l ine and a corner detector. Both operators use specific constraints on the gradient of the image intensity. The operators are applied to sate llite and aerial images. The line detector is very useful for extracti ng roads even on the noisy SAR images, while the corner detector enabl es to detect salient points such as corners of buildings in aerial ima ges. The information brought by these detectors completes the edge-bas ed description of an image. (C) 1998 Elsevier Science B.V. All rights reserved.