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 B3b
页码：319-324
出版社：Copernicus Publications
摘要：3D city model reconstruction is one of the main application of airborne laser scanning system. Roof detection is a primary step which needs to find each roof from irregular point clouds with noise points involved. Another problem existed in roof detection is that the detected sequence of roofs will affect the size and shape of detected roofs even though a lot fancy algorithm are applied. In this paper, we first analyzed the reason of roof size changing caused by detected sequence, then an weighted points normal voting based algorithm is proposed to detect roofs with multiple threshold applied. After than a boundary points competition approach is investigated to optimize the gained roof which can re-adjust the size and shape of detected roof and reduce the influence of roof detection sequence. Experiments is given to prove the algorithm proposed
