Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2004 · 167 citations · 0 references
Natural SciencesAstronomyPhysicsOpticsRemote SensingSpectroscopyAdaptive OpticsAtmospheric OpticsHyperspectral ImagingMultispectral ImagingTelescope OpticsAdaptive Optics InstrumentationConvolutionExpectation-maximization AlgorithmHyperspectral AnalysisHyperspectral Data ProcessingHyperspectral Image ProcessingHyperspectral Sensors
We present results of field evaluations of computed tomography imaging spectrometer (CTIS) systems, using an adaptive optics (AO) compensated telescope, for applications in astronomical multispectral imaging and imaging spectroscopy in both visible (0.4 - 0.7 μm) and near infrared (1.2 - 2.1 μm) bands. We develop and analyze a convolution model for this hyperspectral imaging method which includes a convolution representation of the effects of the AO system. We use expectation maximization, with additional prior constraints derived from the physics of the instrumental, statistical and atmospheric effects inherent in such observations, to extract details of features in the hyperspectral images. Data for this study were obtained using the Advanced Electro-Optics System (AEOS) -- a facility of the Air Force Maui Space Surviellance System.