Fig 1.
Map of study area.
Table 1.
List of 6 tree species samples of the study area.
Fig 2.
Hybrid-CS HSI tree classification model.
Fig 3.
Sandglass block with bottleneck structure.
Table 2.
The layer types, output map dimensions, and number of parameters of the Hybrid-CS model.
Table 3.
The AA of kernel size in the experiment.
Fig 4.
Training time.
Fig 5.
Accuracy with the proportion of training samples.
Fig 6.
Trees species classification map.
(a) 3D-CNN. (b) DBMA. (c) DBDA. (d) DBSSN. (e) Hybrid-CS. (f) Ground truth.
Table 4.
Classification results of different models (%).
Table 5.
Classification accuracy of tree species with different models (%).
Fig 7.
Training time and OA of the Hybrid-CS model with sandglass block and none.