文章基本信息

标题：Building Edge Extraction From LiDAR Based on Jump Detection in Non-parameter Regression Model
本地全文：下载
作者：Jun Li ; Jonathan Li ; Michael Chapman 等
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：2007
卷号：XXXVI-5/C55
出版社：Copernicus Publications
摘要：Light Detection a nd Ranging (LIDAR) technology has received great attention d ue to its ability to accurately measure the shape and height of objects suitable for a large range of applications such as generating Digital Elevation Models (DEM ' s) and modelling 3D city environment. The buildings are of particular interest due to their us efulness in 3D city modelling . Several techniques have been used to extract building with LIDAR systems. In this paper, the task of extracting significant built edge in raster LIDAR data is studied. The implemented method takes advantage of the detection of jump points in nonparametric models since points of building edges in LIDAT data usually describe sudden local changes, called jump points. Firstly, the LIDAR pixels are represented as observations from a regression model with identical distributed random errors with mean 0 and variance σ 2 in order to transform the extraction of building edge point problem into a nonparametric regression problem. Then the jump points corresponding to the building edge in the nonparametric regression model are detected by a locally weighted estimator. Fina lly, by a post ping procedure the thinning building edges are extracted. This algorithm has been examined by a set of simulated LIDAR data. The results show the efficiency of the proposed algorithm for extracting building edges in complex city environments
关键词：LIDAR data; Jump Point; Building extraction