DOI: 10.1016/j.jag.2016.06.007
Scopus记录号: 2-s2.0-84997795021
论文题名: A building extraction approach for Airborne Laser Scanner data utilizing the Object Based Image Analysis paradigm
作者: Tomljenovic I ; , Tiede D ; , Blaschke T
刊名: International Journal of Applied Earth Observation and Geoinformation
ISSN: 15698432
出版年: 2016
卷: 52 起始页码: 137
结束页码: 148
语种: 英语
英文关键词: Building delineation
; Building identification
; DTM generation
; LiDAR
; Object-Based Image Analysis
; Point cloud analysis
Scopus关键词: accuracy assessment
; algorithm
; digital elevation model
; digital terrain model
; identification method
; image analysis
; lidar
; performance assessment
; remote sensing
; satellite imagery
; Germany
英文摘要: In the past two decades Object-Based Image Analysis (OBIA) established itself as an efficient approach for the classification and extraction of information from remote sensing imagery and, increasingly, from non-image based sources such as Airborne Laser Scanner (ALS) point clouds. ALS data is represented in the form of a point cloud with recorded multiple returns and intensities. In our work, we combined OBIA with ALS point cloud data in order to identify and extract buildings as 2D polygons representing roof outlines in a top down mapping approach. We performed rasterization of the ALS data into a height raster for the purpose of the generation of a Digital Surface Model (DSM) and a derived Digital Elevation Model (DEM). Further objects were generated in conjunction with point statistics from the linked point cloud. With the use of class modelling methods, we generated the final target class of objects representing buildings. The approach was developed for a test area in Biberach an der Riß (Germany). In order to point out the possibilities of the adaptation-free transferability to another data set, the algorithm has been applied “as is” to the ISPRS Benchmarking data set of Toronto (Canada). The obtained results show high accuracies for the initial study area (thematic accuracies of around 98%, geometric accuracy of above 80%). The very high performance within the ISPRS Benchmark without any modification of the algorithm and without any adaptation of parameters is particularly noteworthy. © 2016 The Author(s) Citation statistics:
资源类型: 期刊论文
标识符: http://119.78.100.158/handle/2HF3EXSE/80021
Appears in Collections: 气候变化事实与影响
There are no files associated with this item.
作者单位: Department of Geoinformatics – Z_GIS, University of Salzburg, Schillerstrasse 30, Salzburg, Austria
Recommended Citation:
Tomljenovic I,, Tiede D,, Blaschke T. A building extraction approach for Airborne Laser Scanner data utilizing the Object Based Image Analysis paradigm[J]. International Journal of Applied Earth Observation and Geoinformation,2016-01-01,52