Paper
25 October 2006 Variable-size variable-band selection for spectral feature characterization in hyperspectral data
Author Affiliations +
Abstract
This paper presents a band selection technique for spectral feature characterization in hyperspectral data, referred to as band selection for hyperspectral signature feature characterization (BSHSFC). Since a hyperspectral signature is characterized by its spectral profile, the number of bands to be selected is totally determined by spectral features that uniquely characterize the signature. As a result, two hyperspectral signatures may require different sets of bands for spectral characterization. The proposed BSHSFC is a variable-size variable-band selection (VSVBS) where the number of selected bands varies with a hyperspectral signature to be processed. In order for BSHSFC to select an appropriate set of bands for a hyperspectral signature, a new band prioritization criterion, referred to as orthogonal subspace projectorbased band prioritization criterion (OSP-BPC) is derived for this purpose. It assigns a different priority score to each spectral band of a hyperspectral signature such that various features can be captured by the BSHSFC from the original set of bands. Accordingly, the BSHSFC can be interpreted as a spectral feature extraction technique for signature characterization. Finally, experiments using two sets of data are conducted to demonstrate that the BSHSFC-based BS can improve the performance of hyperspectral signature characterization.
© (2006) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Su Wang and Chein-I Chang "Variable-size variable-band selection for spectral feature characterization in hyperspectral data", Proc. SPIE 6378, Chemical and Biological Sensors for Industrial and Environmental Monitoring II, 63781B (25 October 2006); https://doi.org/10.1117/12.684911
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Platinum

Projection systems

Feature extraction

Hyperspectral target detection

Remote sensing

Target detection

Image processing

Back to Top