Enhanced Global and Local Curvature Properties for Corner Detection

Abstract

Abstract Corner detection is basically a methods used to extract certain kind of features in images which could produce some information including the location or position of the corner points. Thus, in this paper an enhancement shape corner detection method is proposed to detect true corners of shape images. The overall performance of the proposed enhanced shape corner detector and six other existing shape detectors and descriptors including Harris, SUSAN, Harris-Laplace, CSS, SIFT and global and local curvature properties is presented. The experimental results of corner detection methods are tested using the benchmark binary image MPEG-7 Core Experiment Shape-1 Part B dataset. To measure the performance of corner detection evaluation, an appropriate number of true corners were determined.