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：2005
卷号：XXXVI-3/W19
页码：227-232
出版社：Copernicus Publications
摘要：The paper focuses on comparing accuracies obtained with photogrammetry and laser scanning in building extraction and partly com- pares results obtained with various methods. The objective of the EuroSDR Building Extraction comparison was to evaluate the quality, accuracy, feasibility and economical aspects of semi-automatic building extraction based on photogrammetric techniques with the emphasis on commercial and/or operative systems, semi-automatic and automatic building extraction techniques based on high density laser scanner data and semi-automatic and automatic building extraction techniques based on integration of laser scan- ner data and aerial images (hybrid technique). The project consists of three test sites by Finnish Geodetic Institute (FGI), namely Senaatti, Hermanni and Espoonlahti and one test site by Institut Geographique National (IGN), namely Amiens. For each test site fol- lowing data was provided to the partners: aerial images, camera calibration and image orientation information, ground control point coordinates and jpg images of point locations (not for Amiens), laser scanner data and cadastral map vectors of selected buildings. Participants were requested to create the vectors of 3D city models. 3D-models were obtained from 11 participants. Paper confirms with experiments that laser scanning is more suitable in deriving building heights, extracting planar roof faces and ridges of the roof whereas the photogrammetry and aerial images are more suitable in building outline and length determination. CyberCity, Stuttgart and TerraScan (performed by ICC) solutions provided the highest accuracy. There seemed to be a higher variation in the quality of other models depending on test site or remotely sensed information
关键词：Laser scanning; photogrammetry; building extraction; accuracy; interquartile range; building outline; point cloud
