This paper presents a method for registering multi-sensor images such as visible and infrared images. Registration is achieved by iteratively minimizing an objective function through updating transformation parameters of deformation. A new entropy-based objective function is proposed on the basis of a 3-D joint histogram incorporating intensity information and edge orientation information. To complement the entropy-based objective function, a weighting function is also introduced through the use of a coincidence measure of edge orientations. Experimental results show that the proposed method provides more robust registration results than the existing approaches. (C) 2008 Elsevier Ltd. All rights reserved.