Volume image registration by template matching

A template-matching approach to registration of volumetric images is introd uced. The process automatically selects a dozen highly detailed and unique templates (cubic or spherical subvolumes) from one image and locates the te mplates in another image of the same scene. The centroids of the correspond ing templates are used as corresponding control points and the best four co rrespondences that minimize an error criterion are used to determine the tr anslational and rotational parameters needed to register the images. Use of different similarity measures in template matching and determination of co ordinates of corresponding control points with subpixel accuracy are also d iscussed. (C) 2001 Elsevier Science B.V. All rights reserved.