Curves Detection Using Delaunay Triangulation Based on Perceptual Grouping Factors

Abstract

In this paper, we propose a method to detect perceptual continuous curves in a binary image including discontinuous curves and noises based on perceptual grouping factors. Firstly, making a Delaunay diagram from a binary image, adjacency relation is obtained. Secondly, the method selects a Delaunay edge as the starting line and connects a certain number of points. When the points have maximum evaluation value based on perceptual grouping factors, they construct an appropriate partial curve. Evaluation value of an appropriate partial curve is added to Delaunay edges selected by it. Finally, Delaunay edges that have high evaluation value are extracted. Experimental results show that proposed method has yielded results that are similar to visual curves and better than conventional methods.