Fusion of Airborne Discrete-Return LiDAR and Hyperspectral Data for Land Cover Classification

摘要: Accurate land cover classification information is a critical variable for many applications. This study presents method to classify using the fusion data of airborne discrete return LiDAR (Light Detection and Ranging) CASI (Compact Airborne Spectrographic Imager) hyperspectral data. Four LiDAR-derived images (DTM, DSM, nDSM, intensity) (48 bands) with 1 m spatial resolution were spatially resampled 2, 4, 8, 10, 20 30 resolutions nearest neighbor resampling method. These thereafter fused layer stacking principal components analysis (PCA) methods. Land was classified by commonly used supervised classifications in remote sensing images, i.e., support vector machine (SVM) maximum likelihood (MLC) classifiers. Each classifier applied four types datasets (at seven different resolutions): (1) data; (2) PCA (3) alone; (4) alone. In this study, category into classes, buildings, road, water bodies, forests, grassland, cropland barren land. A total 56 results produced, accuracies assessed compared. The show that produced from two higher than single at all resolutions. Moreover, we find overall both SVM MLC highest accuracy obtained (OA = 97.8%, kappa 0.964) on resolution. Compared best alone, improved 9.1% 19.6%, respectively. Our findings also demonstrated generally performed better when classifying multisource however, none classifiers consistently