Tree species classification from hyperspectral data using graph-regularized neural networks
This addresses the problem of accurate tree species mapping for forestry and ecology, using a semi-supervised approach that is incremental over existing methods.
The paper tackled tree species classification from hyperspectral data by proposing a graph-regularized neural network (GRNN) algorithm, which achieved high accuracy (approx. 92%) on a new French Guiana dataset with less than 1% labeled pixels.
We propose a novel graph-regularized neural network (GRNN) algorithm for tree species classification. The proposed algorithm encompasses superpixel-based segmentation for graph construction, a pixel-wise neural network classifier, and the label propagation technique to generate an accurate and realistic (emulating tree crowns) classification map on a sparsely annotated data set. GRNN outperforms several state-of-the-art techniques not only for the standard Indian Pines HSI but also achieves a high classification accuracy (approx. 92%) on a new HSI data set collected over the heterogeneous forests of French Guiana (FG) when less than 1% of the pixels are labeled. We further show that GRNN is competitive with the state-of-the-art semi-supervised methods and exhibits a small deviation in accuracy for different numbers of training samples and over repeated trials with randomly sampled labeled pixels for training.