Image registration using a point-line duality based line matching method

Abstract

In the paper, a Point-Line Duality (PLD), i.e. a line in the image (x–y) space corresponds to a point in the dual (θ–ρ) space, based line matching method is proposed for image registration. First, edge points are detected in a template image and a target image. The edge points are linked and segmented into chains. The chains are fitted to lines, and the lines are mapped to dual points in the dual space. To improve stability and efficiency, a point merging algorithm is proposed to deal with the fragmentary line segments that should belong to a single line. As a result, a line matching problem is converted to a point pattern matching problem. Finally, a point pattern matching algorithm is proposed to determine registration parameters and to determine matched line pairs. Experimental results demonstrate that the proposed method is effective for images under rigid body transformation, occlusion, and illumination change.