Search

A novel approach to image alignment or registration for three-dimensional reconstruction of medical images is introduced. The scheme is based on recent advances in computer vision. A point-to-point disparity analysis is used in the alignment algorithm. It is assumed that a sequence of images to be aligned is related by a linear affine transformation. Based on this model, a disparity function is defined by considering the image shape structure or the image intensity changes. The function parameters are estimated simultaneously from the computation of shape or intensity disparity. The decomposition of the disparity function parameters reveals the object rotation, deformation, scaling, and translation individually in a 3D space. These parameters provide enough information for image alignment as well as image matching. Experiments on 3D reconstruction of autoradiographic images demonstrate the effectiveness of the disparity analysis method. >

The generalized Hough transform (GHough) is a useful technique for detecting and locating 2D shapes. However, GHough requires a 4D accumulator array to detect objects of unknown scale and orientation. In this paper, we propose an extension of GHough, the virtual line segment-based Hough transform (VHough) that requires much less storage than GHough to accurately determine the scale and orientation of an object instance. VHough takes O(N/sup 2/) time, where N is the number of edge pixels in an image, but requires only 2D accumulator array for the detection of arbitrarily rotated and scaled objects. We present an experimental result to show that VHough is well-suited to recognition tasks when no a priori knowledge about parameters is available.