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：69-76
出版社：Copernicus Publications
摘要：The prime objective of this paper is to develop a new method for regularizing noisy building outlines extracted from airborne LiDAR data. For the last few decades, a lot of research efforts have been made towards the automation of building outline regularization, which include Douglas-Peucker's polyline simplication, least-square adjustment, model hypothesis-verification, and rule-based rectification. This paper presents a new method to rectify noisy building polylines by dynamically re-arranging quantized line slopes in a local shape configuration and globally selecting optimal outlines based on the Minimum Description Length principles. The optimality is achieved when a building polyline is maximally hypothesized as the repetition of identical line slope and inner angular transition are enhanced with minimal numbers of vertices. A comparative evaluation of the proposed regularization method is compared with existing methods using simulated building vectors with different random errors. The results show that the proposed method outperforms the selected existing algorithms at the most of noise levels
