25 October 2012 Hyperspectral image reconstruction of an array of extended targets using chromotomosynthesis
Randall L. Bostick, Glen P. Perram
Author Affiliations +
Abstract
The ability of a chromotomosynthetic imaging (CTI) system to accurately represent spatial and spectral characteristics of a spectrally diverse set of extended sources was evaluated. The goal is to evaluate CTI hyperspectral imagery for the remote classification of widely separated combustion events against a dark background. Accuracy is determined by comparing results to those of a liquid crystal tunable filter (LCTF). The reconstructed CTI two-dimensional (2-D) image was equivalent to a wideband, undispersed image for images reconstructed using 1 deg projection separation. The Colsher discrepancy metric degrades from 0.44 for the 1 deg sampling to 0.72 for a 10 deg sampling, with 5 deg sampling still equivalent to the image obtained by integrating the spectral frames of the LCTF. Hyperspectral data cubes are reconstructed with image quality dependent primarily on source intensity, not wavelength. Through nearly the entire range, narrower spectral lineshapes are attained from CTI data with little degradation in results from the 1 deg, 5 deg, and 10 deg sampled projections. The measured spectral locations were in agreement between the CTI and LCTF for most elements in the mosaic. Spectral and spatial artifacts are observed in the CTI hyperspectral imagery particularly for dim objects adjacent to bright and spectrally similar sources. As a scene becomes more continuous with spatially and spectrally similar sources, more complex processing will be necessary to avoid degraded performance.
© 2012 Society of Photo-Optical Instrumentation Engineers (SPIE) 0091-3286/2012/$25.00 © 2012 SPIE
Randall L. Bostick and Glen P. Perram "Hyperspectral image reconstruction of an array of extended targets using chromotomosynthesis," Optical Engineering 51(10), 103205 (25 October 2012). https://doi.org/10.1117/1.OE.51.10.103205
Published: 25 October 2012
Lens.org Logo
CITATIONS
Cited by 1 scholarly publication and 1 patent.
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Optical filters

Image filtering

Hyperspectral imaging

Prisms

Image restoration

Spectral resolution

3D image reconstruction

Back to Top