Hyperspectral image analysis using Hilbert-Huang transform
Journal
32nd Asian Conference on Remote Sensing 2011, ACRS 2011
Journal Volume
1
Pages
285-290
Date Issued
2011
Author(s)
Huang, X.-M.
Abstract
Hyperspectral images, which contain rich and fine spectral information, can improve land use/cover classification accuracy, while traditional statistics-based classifiers cannot be directly used on such images with limited training samples. The commonly used method to solve this problem is dimensionality reduction, and this can be done by feature extraction for hyperspectral images. There are two types of feature extraction methods. The first type is based on the statistical property of data, such as principal component transform (PCT), discriminant analysis feature extraction (DAFE) and decision boundary feature extraction (DBFE). The other type of feature extraction methods is based on time-frequency analysis. For example, it has been proven that wavelet-based feature extraction provide an appropriate and effective tool for spectral feature extraction. However, these methods have some disadvantages; for instance, it still needs adequate training samples, or it has to select the wavelet basis function in advance. Hilbert-Huang transform (HHT), consisting of empirical mode decomposition (EMD) and Hilbert spectral analysis (HSA), is a relatively new adaptive time-frequency analysis tool, and has been used extensively in nonlinear and nonstationary data analysis. In this study, the HHT is implemented on the hyperspectral data for physically spectral analysis. The spectral features are then extracted based on the results of physically spectral analysis, so that we can get a small number of salient features, reduce the dimensionality of hyperspectral images and keep the accuracy of classification results. In our experiment, a real hyperspectral data set is used to test the effectiveness of HHT which was applied on feature extraction and classification. Finally, the results are also compared with other feature extraction methods.
Subjects
Classification; Feature extraction; Hilbert-Huang transform; Hyperspectral image
SDGs
Other Subjects
Adaptive time-frequency analysis; Classification accuracy; Classification results; Decision boundary; Dimensionality reduction; Discriminant analysis feature extractions; Effective tool; Empirical Mode Decomposition; Feature extraction and classification; Feature extraction methods; Hilbert Huang transforms; Hilbert spectral analysis; Hyper-spectral images; Hyperspectral Data; Hyperspectral image analysis; Land use/cover; Nonstationary data; Principal Components; Salient features; Spectral feature; Spectral feature extraction; Spectral information; Statistical properties; Time frequency analysis; Training sample; Wavelet basis functions; Wavelet-based Feature; Classification (of information); Discriminant analysis; Independent component analysis; Principal component analysis; Remote sensing; Sampling; Signal processing; Spectrum analysis; Statistical tests; Feature extraction
Publisher
journal article
Type
journal article