Hyperspectral bands reduction based on rough sets and fuzzy C-means clustering

Abstract

A method of hyperspectral bands reduction based on rough sets and Fuzzy C-Means clustering is proposed, which consists of two steps: first, Fuzzy C-Means (FCM) clustering algorithm is used to classify the original bands into equivalent band groups, which employs the concept of attribute dependency in Rough Sets (RS) to define the distance between a group and the cluster center, viz. the correlatives of adjacent bands; then the data is reduced by selecting only the one with maximum grade of fuzzy membership from each of the groups. So the great number of bands is decreased while preserving most of the wanted information. Simulation results prove the effectiveness of this approach.