Curvilinear Feature Detection from Curvature Estimation
One method of detecting features on digitized image lines is to first estimate the local curvature along the lines, then analyze the curvature plot for peaks corresponding to corners and curves. The DOS sup (+) method, a particular case of the difference of slopes (DOS) approach, has been shown to be effective for estimating curvature in terms of signal detectability. In this paper, relationships are established between the DOS sup (+) parameters and feature parameters to distinguish corner and curve features and accurately determine the locations of corners, transitions between straight lines and curves, and curve centers.