A training algorithm for optimal margin classifiers
COLT '92 Proceedings of the fifth annual workshop on Computational learning theory
The nature of statistical learning theory
The nature of statistical learning theory
Introductory Digital Image Processing: A Remote Sensing Perspective
Introductory Digital Image Processing: A Remote Sensing Perspective
A Tutorial on Support Vector Machines for Pattern Recognition
Data Mining and Knowledge Discovery
International Journal of Remote Sensing
Hi-index | 0.00 |
In this study we show that multiangle remote sensing is useful for increasing the accuracy of vegetation community type mapping in desert regions. Using images from the National Aeronautics and Space Administration (NASA) Multiangle Imaging Spectroradiometer (MISR), we compared roles played by Bidirectional Reflectance Distribution Function (BRDF) model parameters with those played by topographic parameters in improving vegetation community type classifications for the Jornada Experimental Range and the Sevilleta National Wildlife Refuge in New Mexico, USA. The BRDF models used were the Rahman-Pinty-Verstraete (RPV) model and the RossThin-LiSparseReciprocal (RTnLS) model. MISR nadir multispectral reflectance was considered as baseline because nadir observation is the most basic remote sensing observation. The BRDF model parameters and the topographic parameters were considered as additional data. The BRDF model parameters were obtained by inversion of the RPV model and the RTnLS model against the MISR multiangle reflectance data. The results of 32 classification experiments show that the BRDF model parameters are useful for vegetation mapping; they can be used to raise classification accuracies by providing information that is not available in the spectral-nadir domain, or from ancillary topographic parameters. This study suggests that the Moderate Resolution Imaging Spectroradiometer (MODIS) and MISR BRDF model parameter data products have great potential to be used as additional information for vegetation mapping.