abstract |
The invention discloses a hyperspectral image classification method based on deep learning, which includes a pre-training stage and a fine-tuning stage. On the basis of improving the deep convolutional network, the parameters of the deep network and the shallow network are adjusted, and the parameters of the deep network and the shallow network can be adjusted as far as possible. It avoids overfitting in the case of possibly few labeled samples, improves the classification effect of deep networks in the case of small samples, and has certain advantages when training the network with the same amount of data. |