Corner detection using difference chain code as curvature

Detection of discontinuity plays an important role in image registration, comparison, segmentation, time sequence analysis and object recognition. In this paper we address some aspects of analyzing the content of an image. A point is classified as a corner where there are abrupt changes in the curvature of the curve. This paper presents a new approach for corner detection using first order difference of chain codes called shape numbers. Since the proposed method is based on integer operations it is simple to implement and efficient. Preliminary results are presented and evaluation with respect to standard corner detectors is done as a benchmark.