Angularly multiplexed spectral imager

Abstract

A spectral imager constructs a three dimensional (two spatial and one spectral) image from a series of two dimensional images. This paper discusses a technique for spectral imaging that multiplexes the the spatial and spectral information on the focal plane, then demultiplexes the resulting imagery to obtain the spectral image. The resulting spectral image consists of 184 x 184 spatial pixels and 40 spectral bands. The current implementation operates over the 3-5pmband, but can easily be applied to other spectral regions. A hardware description, the mathematical development and experimental results are presented. BACKGROUNDMost spectral imagers slice the three dimensional image into a sequence of two dimensional images. Two common approaches are: 1) image in the two spatial dimensions while using a sequence of spectral filters to slice in the spectral domain, or 2) image in one spatial and one spectral dimension while scanning a slit over the remaining spatial direction. Each of these approaches are inherently inefficient. In the first case only one color is imaged at a time. If N,, spectral bands are to be imaged, then the efficiency is (1/NA). In the second case only one slit of the scene is imaged at a time. If NxN spatial resolution elements are to be imaged by scanning in the x direction, then the efficiency will be (1/Nr).Color visible CCDs achieve spectral imaging by using dichroic filters such that each pixel detects either red, green or blue light. The efficiency of these devices is at best 1/3, and they also suffer from a degradation in spatial resolution by 1/v. This approach is limited to imaging no more than a few colors.Though most spectral imaging techniques suffer from a degradation in efficiency, resolution or both; there are two noteworthy exceptions. One is to use a cascade of spectral beam splitters with a separate imaging focal plane for each color. The efficiency of this approach is determined by that of the beam splitters and in theory can be 100%; however, practical constraints limit application of this approach to resolution of no more than a few spectral bands. The other exception is Fourier transform spectroscopy. A Fourier transform spectrometer uses an interferometer to multiplex the spectral information. Both lateral shear and longitudinal displacement interferometers are used. Lateral shear interferometers use common path interferometers that multiplex the spatial and spectral information in space; they are relatively insensitive to vibration and often find application in field measurements. Longitudinal displacement interferometers multiplex the spectral information in time; they are susceptible to vibrations and usually find application in laboratories. A Fourier transform imaging spectrometer based on a lateral shear interferometer would have the same efficiency limit as a scanned slit spectrometer, while that based on a longitudinal displacement interferometer could have very high efficiency but would be useful only in low vibration environments.Each of these imaging techniques ...

Review of snapshot spectral imaging technologies

“…Its main disadvantage is that the detector is not used efficiently in comparison to alternative methods. 49 …”

Pseudo imaging

“…Focal plane arrays (FPAs) coated with a filter matrix (such as a Bayer pattern) and fiber-optic conversion of a two-dimensional image to a slit constitute examples of multiplexing at the image plane, which may lead to aliasing problems or sensor complexity problems, respectively. Finally, spectral sensing techniques based on tomographic projection include the CTIS system 7 developed by the University of Arizona based on a diffraction grating, and the CTHIS sensor [8][9] developed by the Air Force Research Laboratory (AFRL/SNHI) based on a spinning direct vision prism. In this last case of tomographic projection using a spinning prism, when the sensor operates as a hyperspectral imager the large temporal period required for scene acquisition make it impractical for the study of rapidly evolving events that require temporal resolution many times better than the sensor can measure.…”

A Diffractive Optic Image Spectrometer (DOIS)

“…This represents low frequency spatial/spectral information that isn't imaged by DOIS or recoverable with processing. Plotted in Figure 8.21 as a cross-section of each spectral frequency, the frequently discussed missing cone which plagues such imaging systems is seen [Barrett 1981, Mooney 1995.…”

“…The task is to invert this H matrix, however it can have various problems which cause it to be singular and difficult to invert [Mooney 1995 This is the method of choice for solving most linear least squares problems [Press 1988]. …”