干旱区生态环境知识资源中心

Image registration is a prerequisite for multisource data fusion. In this paper the problem of registering aerial images with lidar point clouds in desert areas is addressed. Compared with urban areas, registration in desert regions is difficult due to the lack of man-made features which are typically used in traditional methods. However, sand ridges can be used as registration primitives. Firstly, sand-ridge information is extracted from both the aerial image and the lidar point cloud. Secondly, by extending the iterative closest point (ICP) approach, a perspective-ICP algorithm is proposed that achieves data registration through matching sand ridges. To automatically deal with outliers, an adaptive weighting strategy is adopted. Experiments and assessment using data from Dunhuang, Gobi Desert, China, demonstrate that the method can achieve efficient and reliable registration for desert areas.