Fourier transform spectroscopy (FTS) employs a single aperture and a moving delay line to estimate the spectral radiance as a function of wavelength of the source but integrated over an angular field-of-view (FOV). Double-Fourier spatial-spectral interferometry mixes two, or more, apertures against each other while changing the aperture spacing (baseline) and moving the delay line to yield spatial-spectral information within the FOV (typically ~λ/D) subtended by a single detector pixel. Introduction of a 2D format detector allows the double-Fourier approach to operate in such a manner that each detector pixel forms its own interferometer. This is readily understood by looking at an off-axis pixel and moving the delay line such that the peak of the interferometer fringe falls on this off-axis pixel; i.e. a different delay line position for each pixel. The approach operates by collecting ‘baseline cubes’ each of which is at a particular baseline and each frame of the 3D cube forms a 2D detector image but with a different peak fringe location. The set of baselines are judiciously chosen to form a synthetic aperture. The aggregate set of cubes is subsequently reduced via algorithmic processing to a single 3D data cube. Each frame of the resulting 3D data cube represents an image at resolution ~λ/2Bmax << ~λ/D where Bmax is the length of the widest interferometric baseline, and each frame is at the different wavelength λ.
NASA has sponsored a project to advance the wide-field double-Fourier approach and Goddard Space Flight Center has developed the Wide-Field Imaging Interferometry Testbed (WIIT). Herein are discussed the mathematical, algorithmic and computational aspects of the data processing for WIIT to construct a hyperspectral cube from the set of baseline cubes. Results will be shown using the WIIT lab data.
© 2013 Optical Society of AmericaPDF Article