Fusion of imaging spectrometer LIDAR data using support vector machines for land cover classification

Benjamin Koetz, Felix Morsdorf, T Curt, S Van Der Linden, L Borgniet, Daniel Odermatt, S Alleaume, C Lampin, C Jappiot, Britta Allgöwer
2007
A combination of the two remote sensing systems, imaging spectrometry (IS) and Light Detection And Ranging (LiDAR), is well suited to map fuel types, especially within the complex wildland urban interface. LiDAR observations sample the spatial information dimension describing geometric surface properties. Imaging spectrometry on the other hand samples the spectral dimension, which is sensitive for discrimination of species and surface types. As a non-parametric classifier Support Vector
more » ... port Vector Machines (SVM) are particularly well adapted to classify data of high dimensionality and from multiplesources as proposed in this work. The presented approach achieves an improved land cover mapping based on a single SVM classifier combining the spectral and spatial information dimensions provided by imaging spectrometry and LiDAR. Commission VII, WG 1 KEY WORDS: Support Vector Machines, land cover classification, hyperspectral, LiDAR, multi-sensor fusion ABSTRACT: A combination of the two remote sensing systems, imaging spectrometry (IS) and Light Detection And Ranging (LiDAR), is well suited to map fuel types, especially within the complex wildland urban interface. LiDAR observations sample the spatial information dimension describing geometric surface properties. Imaging spectrometry on the other hand samples the spectral dimension, which is sensitive for discrimination of species and surface types. As a non-parametric classifier Support Vector Machines (SVM) are particularly well adapted to classify data of high dimensionality and from multiple-sources as proposed in this work. The presented approach achieves an improved land cover mapping based on a single SVM classifier combining the spectral and spatial information dimensions provided by imaging spectrometry and LiDAR.
doi:10.5167/uzh-77980 fatcat:y3vvmemdvzgflcqxhgnnrhursm