Automated Multi-class Classification of Remotely Sensed Hyperspectral Imagery Via Gaussian Processes with a Non-stationa
- PDF / 3,470,920 Bytes
- 22 Pages / 439.37 x 666.142 pts Page_size
- 9 Downloads / 205 Views
Automated Multi-class Classification of Remotely Sensed Hyperspectral Imagery Via Gaussian Processes with a Non-stationary Covariance Function Anna Chlingaryan 1 · Arman Melkumyan 1 · Richard J. Murphy 1 · Sven Schneider 1
Received: 10 February 2015 / Accepted: 26 October 2015 © International Association for Mathematical Geosciences 2015
Abstract The ability to automatically classify hyperspectral imagery is of fundamental economic importance to the mining industry. A method of automated multi-class classification based on multi-task Gaussian processes (MTGPs) is proposed for classification of remotely sensed hyperspectral imagery. It is proved that because of the illumination invariance of the hyperspectral curves, the covariance function of the Gaussian process (GPs) has to be non-stationary. To enable multi-class classification of the hyperspectral imagery, a non-stationary multi-task observation angle-dependent covariance function is derived. In order to test MTGP, it was applied to data acquired in the laboratory and also in field. First, the MTGP was applied to hyperspectral imagery acquired under artificial light from samples of rock of known mineral composition. Data from a high-resolution field spectrometer are used to train the GPs. Second, the MTGP was applied to imagery of a vertical rock wall acquired under natural illumination. Spectra from hyperspectral imagery acquired in the laboratory are used to train the GPs. Results were compared with those obtained using the spectral angle mapper (SAM). In laboratory imagery, MTGP outperformed SAM across several metrics, including overall accuracy (MTGP: 0.96–0.98; SAM: 0.91–0.93) and the kappa coefficient of agreement (MTGP: 0.95–0.97; SAM: 0.88–0.91). MTGP applied to hyperspectral imagery of the rock wall gave broadly similar results to those from SAM; however, there were important differences. Some rock types were confused by SAM, but not by MTGP. Comparison of classified imagery with ground truth maps showed that MTGP outperformed SAM.
B 1
Anna Chlingaryan [email protected] Australian Centre for Field Robotics, University of Sydney, Rose Street Building (J04), Sydney, NSW 2006, Australia
123
Math Geosci
Keywords Hyperspectral imagery · Gaussian processes · Classification · Spectrum · Multi-class · Minerals · OAD covariance function
1 Introduction Hyperspectral data have been used to obtain quantitative information for many applications including studies of vegetation (Daughtry et al. 2004; Elvidge 1988; Kokaly et al. 2003), soil mapping (Ben-Dor et al. 2002; Chabrillat et al. 2002; Murphy and Wadge 1994) and mapping the distribution of rocks and minerals (Bierwirth et al. 2002; Kruse et al. 1993; Murphy 1995). The hyper-dimensional nature of hyperspectral data has required new approaches to be developed for its analysis in order to take advantage of this property (Goetz 2009; Vane and Goetz 1988). Over the past few decades, many approaches to classifying hyperspectral data have been developed (reviewed by Cloutis 1996; Plaza et al. 20
Data Loading...
