期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2006
卷号：XXXVI Part 3
出版社：Copernicus Publications
摘要：A method for accurate 3D reconstruction of road features from multiple calibrated aerial images of urban areas is proposed in this paper. We here focus on road-marks and in particular on zebra-crossings and discontinuous road-marks separating circulation lanes. The approaches used here are generic and based on a-priori external knowledge and thus constrain the extraction of image features. As we will explain, two strategies are adopted depending on the object size. For zebra-crossings, we first build 3D segments representing stripes' borders by 2D segments matching. For discontinuous lanes, we build a graph describing the network in each image and then match nodes in order to obtain 3D position of stripes' centers. This provides in both cases an initial solution in 3D space. Using geometric and radiometric modeling to obtain a set of plausible models, e then look for an optimal solution. The last step yields us to choose the best one in adequacy with images data. A correlation based energy and template matching strategy achieve this in a hierarchical frame. The algorithm is finally evaluated with ground control points surveyed with a millimetric precision
