Feature Mining for Hyperspectral Image Classification

TLDR

Hyperspectral sensors capture high‑resolution reflectance across the solar spectrum, producing high‑dimensional data where only a subset of features is informative, making feature mining—generation, selection, and extraction—a critical task that has driven extensive research into supervised, unsupervised, parametric, non‑parametric, linear, and nonlinear techniques. This paper reviews conventional and advanced feature reduction methods, detailing commonly used techniques for hyperspectral data analysis. The authors present a general framework that unifies several linear and nonlinear feature extraction methods and illustrate selected feature selection and extraction approaches through experiments on two widely available hyperspectral datasets. The experiments show that the selected feature selection and extraction methods improve classification performance on the two datasets.

Abstract

Hyperspectral sensors record the reflectance from the Earth's surface over the full range of solar wavelengths with high spectral resolution. The resulting high-dimensional data contain rich information for a wide range of applications. However, for a specific application, not all the measurements are important and useful. The original feature space may not be the most effective space for representing the data. Feature mining, which includes feature generation, feature selection (FS), and feature extraction (FE), is a critical task for hyperspectral data classification. Significant research effort has focused on this issue since hyperspectral data became available in the late 1980s. The feature mining techniques which have been developed include supervised and unsupervised, parametric and nonparametric, linear and nonlinear methods, which all seek to identify the informative subspace. This paper provides an overview of both conventional and advanced feature reduction methods, with details on a few techniques that are commonly used for analysis of hyperspectral data. A general form that represents several linear and nonlinear FE methods is also presented. Experiments using two widely available hyperspectral data sets are included to illustrate selected FS and FE methods.