Journal of Remote Sensing / 2021 / Article / Fig 18

Research Article

Shearlet-Based Structure-Aware Filtering for Hyperspectral and LiDAR Data Classification

Figure 18

Trento dataset: (a) ground-truth map, the rest are classification maps obtained by (b) Raw-H, (c) Raw, (d) 3D-CNN, (e) miniGCN, (f) SAE-LR, (g) NMFL, (h) EPCA, (i) GGF, (j) EMAP, (k) OTVCA, and (l) ShearSAF when the number of training samples is five per class (the percentage in the brackets is the corresponding accuracy).