Skip to main content
Fig. 6 | Plant Methods

Fig. 6

From: Atmospheric correction of vegetation reflectance with simulation-trained deep learning for ground-based hyperspectral remote sensing

Fig. 6

A subset of the normalized simulated at-sensor spectral radiance (left) from the training (top), testing (middle), and validation (bottom) sets, and their associated \(\Delta\)reflectance (deviation of spectral reflectance from the median, right). After splitting both reflectance and solar and atmosphere simulations into 3 independent sets (50% training, 30% testing, and 20% validation) each, we generate 360,000 training instances, 129,600 testing instances, and 57,600 validation instances. Each of the 3 sets are separated into 10 equal subsets, where each subset is used to train, validate, and test a neural network model independently for a total of 10 models

Back to article page

Plant Methods

ISSN: 1746-4811