Fast global reflectional symmetry detection for robotic grasping and visual tracking

Abstract

Many objects found in domestic environments are symmetrical. In this paper, a novel method to detect axis of reflectional symmetry in digital images is presented. The proposed approach utilizes image gradients and edge pixel matching to find lines of symmetry. Preliminary results, comparing the symmetry detector with the Generalized Symmetry Transform [Reisfeld et al., 1995] is provided. 1