Improving Deep Hyperspectral Image Classification Performance with Spectral Unmixing
This addresses overfitting issues in hyperspectral image classification for remote sensing applications, but it is incremental as it builds on existing spectral unmixing and autoencoder techniques.
The paper tackles overfitting in hyperspectral image classification by converting images to an abundance domain using a dataset-specific autoencoder, enabling the use of simple classifiers and enlarged training data, which improved classification performance as verified by ablation and comparative experiments.
Recent advances in neural networks have made great progress in the hyperspectral image (HSI) classification. However, the overfitting effect, which is mainly caused by complicated model structure and small training set, remains a major concern. Reducing the complexity of the neural networks could prevent overfitting to some extent, but also declines the networks' ability to express more abstract features. Enlarging the training set is also difficult, for the high expense of acquisition and manual labeling. In this paper, we propose an abundance-based multi-HSI classification method. Firstly, we convert every HSI from the spectral domain to the abundance domain by a dataset-specific autoencoder. Secondly, the abundance representations from multiple HSIs are collected to form an enlarged dataset. Lastly, we train an abundance-based classifier and employ the classifier to predict over all the involved HSI datasets. Different from the spectra that are usually highly mixed, the abundance features are more representative in reduced dimension with less noise. This benefits the proposed method to employ simple classifiers and enlarged training data, and to expect less overfitting issues. The effectiveness of the proposed method is verified by the ablation study and the comparative experiments.