25 October 2006 Modified Fisher's linear discriminant analysis for hyperspectral image dimension reduction and classification
Author Affiliations +
Abstract
In this paper, we present a modified Fisher's linear discriminant analysis (FLDA) to hyperspectral remote sensing image dimension reduction and classification. The basic idea of FLDA is to design an optimal transform which can maximize the ratio of between-class scatter matrix to within-class scatter matrix. The practical difficulty of applying the FLDA to hyperspectral images includes the unavailability of enough samples for all the classes. So the original FLDA is modified to avoid the requirement of class samples. In the following data classification using the FLDA-transformed low-dimensional data, a more powerful classifier generally is required. Fortunately, we find this is not difficult to achieve. A simple distance based classifier, such as Spectral Angle Mapper (SAM), can provide satisfactory classification performance. This approach is particularly useful to the data sets with small classes.
© (2006) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Qian Du, "Modified Fisher's linear discriminant analysis for hyperspectral image dimension reduction and classification", Proc. SPIE 6378, Chemical and Biological Sensors for Industrial and Environmental Monitoring II, 63781D (25 October 2006); doi: 10.1117/12.686399; https://doi.org/10.1117/12.686399
PROCEEDINGS
8 PAGES


SHARE
Back to Top