DELINEATION OF VEGETATION AND BUILDING POLYGONS FROM FULL-WAVEFORM AIRBORNE LIDAR DATA USING OPALS SOFTWARE

被引:0
作者
Hollaus, M. [1 ]
Wagner, W. [1 ]
Molnar, G. [1 ]
Mandlburger, G. [1 ]
Nothegger, C. [1 ]
Otepka, J. [1 ]
机构
[1] Vienna Univ Technol, Inst Photogrammetry & Remote Sensing, A-1040 Vienna, Austria
来源
GEOSPATIAL DATA AND GEOVISUALIZATION: ENVIRONMENT, SECURITY, AND SOCIETY | 2010年 / 38卷
关键词
Three-dimensional; LIDAR; Retrieval; Roughness; Land Cover; Full-Waveform;
D O I
暂无
中图分类号
P9 [自然地理学];
学科分类号
0705 ; 070501 ;
摘要
Full-waveform LiDAR is an active remote sensing technique that provides the scattering properties of the targets i.e. amplitude and echo width (EW) in addition to 3D point clouds. The amplitude provides information on the target's reflectance and the EW is a measure for the range variation of scatterers within the laser footprint contributing to a single echo and is, therefore, an indicator for surface roughness. For the delineation of high vegetation (i.e. trees and bushes) and buildings the normalized digital surface model provides the main input. Based on a height threshold, areas covered with elevated objects can be classified. For the differentiation between buildings and high vegetation the surface roughness is used, where roofs are assumed to be smooth and vegetation rough. The surface roughness can be described with the standard deviation (SD) of detrended LiDAR points. However, for low LiDAR point densities and for very dense deciduous forests or building parts overgrown with tree crowns the separability is limited if only the SD is used. Within the presented approach additionally the EW, featuring small values for smooth roof areas, and the echo ratio (ER), a measure for the penetrability of a surface, are used. For the final delineation of high vegetation and building areas morphologic operations are applied. The classified areas are vectorized and a minimum mapping unit is applied. The presented workflow is implemented into the scientific software package OPALS (Orientation and Processing of Airborne Laser Scanning data). The results show that the EW, the SD and the ER are useful quantities to delineate high vegetation and building polygons based on full-waveform LiDAR point clouds with high accuracy. The high degree of automation and the scripting capability OPALS promise a high potential for operational large area LiDAR data processing. Examples from different test sites in Austria are shown and discussed.
引用
收藏
页数:7
相关论文
共 16 条
[1]   A linear-time component-labeling algorithm using contour tracing technique [J].
Chang, F ;
Chen, CJ ;
Lu, CJ .
COMPUTER VISION AND IMAGE UNDERSTANDING, 2004, 93 (02) :206-220
[2]   A Comprehensive Automated 3D Approach for Building Extraction, Reconstruction, and Regularization from Airborne Laser Scanning Point Clouds [J].
Dorninger, Peter ;
Pfeifer, Norbert .
SENSORS, 2008, 8 (11) :7323-7343
[3]  
Gonzalez R. C., 1992, DIGITAL IMAGE PROCES, V2nd
[4]  
Haala N., 1999, INT ARCH PHOTOGRAMME, V32
[5]  
Hfle B., 2009, P GEOINF FOR SALZB, P66
[6]  
Hollaus M, 2010, INT ARCH PHOTOGRAMM, V38, P287
[7]  
Hollaus M., 2010, INT ARCH PHOTOGRAMME, V39, P6
[8]  
Mandlburger G., 2009, INT ARCH PHOTOGRAMM, P55
[9]   Cities from space: potential applications of remote sensing in urban environmental research and policy [J].
Miller, RB ;
Small, C .
ENVIRONMENTAL SCIENCE & POLICY, 2003, 6 (02) :129-137
[10]  
*OPALS, 2010, OPALS OR PROC AIRB L