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：2008
卷号：XXXVII Part B3a
页码：179-184
出版社：Copernicus Publications
摘要：This paper describes an efficient method for detecting and modeling roof superstructures (detailed volumes present on roofs such as chimneys, dormer windows...) using only a Digital Elevation Map (DEM) and an initial building model without superstructures. This problem is made challenging since both detection and reconstruction should be performed simultaneously. Our proposed method is modular. The first stage (initialization stage) provides a fast and coarse detection of possible superstructures. The second stage (refine- ment stage) refines the parametric models of the coarse superstructures using either an iterative improvement scheme or a stochastic diffusion where both heuristics attempt to maximize the benefit of adding a superstructure model to the whole building model. The final stage selects the most consistent set of superstructures. Experiments as well as method comparisons show the efficiency of the proposed method
关键词：Building modeling; Digital Elevation Map; model-based reconstruction; parameter fitting; building superstructure
