Problem-based band selection for hyperspectral images

Abstract

This paper addresses the band selection of a hyperspectral image. Considering a binary classification, we devise a method to choose the more discriminating bands for the separation of the two classes involved, by using a simple algorithm: single-layer neural network. After that, the most discriminative bands are selected, and the resulting reduced data set is used in a more powerful classifier, namely, stacked denoising autoencoder. Besides its simplicity, the advantage of this method is that the selection of features is made by an algorithm similar to the classifier to be used, and not focused only on the separability measures of the data set. Results indicate the decrease of overfitting for the reduced data set, when compared to the full data architecture.

Supervised band selection in hyperspectral images using single-layer neural networks

“…When it comes to filter-based approaches, there are some different criteria for the band selection, such as, distances measures (Keshava 2004), class separability measures (Cui et al 2011), information, dependence (Camps-Valls, Mooij, and Scholkopf 2010), correlation, searching strategies (Jahanshahi 2016;Su, Yong, and Du 2016) and classification measures (Habermann, Fremont, and Shiguemori 2017). For example, in (Damodaran, Courty, and Lefevre 2017), the authors propose a class separability-based approach.…”

Unsupervised Cluster-Wise Hyperspectral Band Selection for Classification

“…In terms of the available data set to train the algorithm, a band selection framework can be considered supervised [16][17][18][19], semi-supervised [20][21][22][23], or unsupervised [24][25][26]. The latter ends up being the most feasible in real applications due to the difficulty in acquiring labeled samples [27].…”