Symmetry Detection Using Gradient Information

Abstract

C. Sun, “Symmetry Detection Using Gradient Information”, Pattern Recognition Letters, vol.16, no.9, pp.987-996, 1995.

Symmetry detection is important in the area of computer vision. A simple and fast symmetry detection algorithm has been developed in this paper. The algorithm employs only the original image and the gradient information. The direction of the symmetry axis is obtained from the gradient orientation histogram; and the position of this symmetry axis is decided either by the center of gravity or by the profile of the image projection along the direction of the symmetry axis. This method works directly on the grey scale image and does not require any prior segmentation of the input image. Both simulated and real images have been tested and the results are very convincing.