In 1980, development of the first imaging spectrometer for earth observation began at the Jet Propulsion Laboratory. At that time neither the detectors, optics, electronics, nor computers for rapid analysis were readily available. Subsequent developments led to the implementation of the airborne imaging spectrometer (AIS) and the airborne visible/infrared imaging spectrometer (AVIRIS), the present-day workhorse for hyperspectral imaging. Plans for a shuttle instrument and the Earth observing system (EOS) high resolution imaging spectrometer (HIRIS) did not materialize. The newest major airborne imaging system is HYDICE, sponsored by the Naval Research Laboratory, and currently undergoing flight tests. Data analysis techniques and software to deal with 200 channel images has recently become available and it has made it feasible for researchers and application specialists, not directly involved with sensor development, to make sensible use of hyperspectral image data.

Convex and affine geometry in n-dimensions provide powerful tools for the analysis, understanding, and visualization of hyperspectral data. The ubiquitous mixed pixel problem can be exploited as an advantage and is easily cast in an n-d convexity context. Convexity concepts can be used to identify the purest pixels in a given scene and to unravel spectral mixing, both fully and partially. Visualization techniques based on these concepts permit human interpretation of all spectral information of all image pixels simultaneously. Convex geometry forms a natural framework for the unique challenges associated with analysis of hyperspectral data.

Observations in remote sensing generally result in compositional measurements due to the finite sampling aperture of the sensor. The measurements are modeled as mixtures of distinct endmember signatures. Unmixing algorithms attempt to estimate the mixture components and their proportional contributions to the measurement, and by inference, to the composition of the measured scene. The choice of an unmixing algorithm depends on our knowledge about the mixture process, the number of endmembers present and their unmixed signatures, the variational structure of the signatures and proportions, and sources of nonsystematic errors and noise. A number of linear unmixing algortihms are surveyed and tested against synthetic hyperspectral reflectance data under a variety of conditions reflecting varying degrees of uncertainty in the endmember population and system noise structure.

Analysis methods for multispectral data have been under study for at least three decades. In spite of that fact, the state of the technology is still far from satisfactory for conventional multispectral data, and the advent of hyperspectral sensor systems raises the challenge substantially. Thus a focused effort was begun a few years ago to advance the technology of multispectral analysis to a more effective level, and especially to prepare suitable methods to yield full information extraction capabilities from the new hyperspectral data. This paper outlines some of what has been learned about this problem from the research effort.

A temporally and spatially nonscanning imaging spectrometer is described in terms of computed-tomography concepts, specifically the central-slice theorem. The critical system element is a sequence of three transmission sinusoidal-phase gratings rotated in 60 degree increments which achieve dispersion in multiple directions and into multiple orders. The dispersed images of the system's field-stop are interpreted as 2D projections of a 3D (x, y, (lambda) ) object cube. Due to finite focal-plane array size, this computed-tomography imaging spectrometer (CTIS) is an example of a limited-view-angle tomographic system. The imaging spectrometer's point-spread-function is measured experimentally as a function of wavelength and position in the field-of-view. Reconstruction of the object cube is then achieved via the maximum-likelihood expectation-maximization algorithm under the assumption of a Poisson likelihood law. Experimental results using a spatial/spectral 'University of Arizona' target indicate that the instrument performs well in the case of broadband and narrowband emitters. A relationship between an object's spatial size and spectral resolution characteristic of limited-view-angle systems is demonstrated.

A spectral imager constructs a 3D (two spatial and one spectral) image from a series of 2D images. This paper discusses a technique for spectral imaging that multiplexes 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 3-5 micrometers band, but can easily be applied to other spectral regions. A hardware description, the mathematical development and experimental results are presented.

Hyperspectral imaging can be a powerful tool for remote sensing of geologic, biologic, and ocean surfaces of atmospheres, and of rocket plumes. A hyperspectral imager provides a 3D (two spatial and one spectral) description from a sequence of 2D images. Common hyperspectral approaches using narrow band filters or imaging spectrometers are inefficient because photons outside the filter passband or the slit area are not detected. A new imaging technique called spectro-tomography collects all available photons and relies on computer tomography to reconstruct the 3D data cube of the image. A rotational spectro-tomographic (RST) imager is designed with a wide aperture, objective-grating camera, that is rotated in steps around its optical axis. The 2D projections of the object are analyzed using methods based on Fourier transforms. Both direct Fourier methods and filter-backprojection algorithms have been developed for 3D tomographic analysis. Numerical methods are employed to simulate and reconstruct a broad spectrum object with 64 spectral bands and 64 X 64 spatial resolution elements. For this example, the photon flux at the detector of the RST imager is 64 times that of a conventional spectral imager. The rotational spectro-tomographic imager has applications to detection of natural and artificial atmospheric emissions where large photon through-put is required.

IMSS utilizes a very simple optical design that enables a robust and low cost hyperspectral imaging instrument. This technology was developed under a phase II SBIR with the Air Force Philips Lab. (Dr. Paul LeVan), for the midwave IR to perform clutter rejection and target identification based upon IR spectral signatures. The prototype instrument has been field tested on numerous occasions and successfully measured background, aircraft, and misile plume spectra. PAT currently has several contracts to commercialize this technology both for the DoD and the commercial market. Under contract to the BMDO, (Paul McCarley), with matching funds from Amber Engineering, we are developing an F/2.3 system that will be sold by Amber Engineering as an accessory to the Radiance 1 camera. PAT is also under contract with ONR (Mr. Jim Buss), to develop a longwave IR version of IMSS as well as an MWIR version tuned to operate as a 'little sister sensor for target identification' for the Navy's IRST's. The purpose of this paper is to briefly describe the hyperspectral image data that was collected in the field at Long Jump '94, and Santa Ynez Peak using IMSS prototype hyperspectral imager. Examples of spectral images, as well as spectra of different aircraft at various ranges, power settings, and aspect angles, an Atlas liquid hydrocarbon burning missile, and a solid beester. All data presented in this paper are a result of a single spectral scan. The limitation in digital storage of the prototype system do not allow multiple scans in order to improve signal to noise. In spite of this limitation, the performance of the prototype system has proven to be excellent.

This paper describes a system for performing high resolution spectroscopy with 2D spatial imaging. We motivate the discussion by describing the application of our system to the astronomical study of dust in the interstellar medium. Our methodology can be implemented on a wide variety of optical systems from a wide variety of platforms. Several such configurations are discussed.

Spectrotomography is a new branch of optical tomography, which allows to study the internal spectral and spatial structure of the polychromatic objects. The paper describes methods and algorithms applied to spectrotomography of 2D and 3D polychromatic objects, which were proposed and tested by the authors.

The Moderate Resultion Imaging Spectroradiometer (MODIS) is a key instrument scheduled for flight on the AM (1030 hours equator crossing time) and the PM (1330 equator crossing time) platforms of the Earth Observing System (EOS). It has a considerable multispectral capability for observations of land, ocean, and atmospheric features. The 36 bands sampling the visible, near IR, and thermal IR portions of the spectrum along with 250 to 1000 meter resolution will offer very powerful observing advantages over heritage instruments such as the Advanced Very High Resolution Radiometer on the NOAA environmental satellites. The engineering model of the MODIS is under construction and so far the performance being observed is generally meeting specifications. Preliminary analyses of components of the engineering model indicate that it will meet most of the spectral specifications with some variances relative to the specifications being required and approved. Overall MODIS fabrication is on schedule for its first flight in 1998.

An imaging spectrometer concept called digital array scanned interferometry (DASI) is being explored in our laboratories for terrestrial remote sensing applications. The essence of DASI operation is that interferograms are resolved spatially over one coordinate at the detector plane, and spatial information is obtained over the orthogonal coordinate. In this paper we focus on recent developments for approaching the fundamental capabilities of the DASI's performance, specifically the signal-to-noise ratio. We also describe selected land observations acquired from an airborne DASI operating in the 4550-9090 cm^-1 (1.1-2.2 micrometers ) spectral region with a spectral resolution of 266 cm^-1.

A large variety of optical concepts for imaging spectrometers with high geometrical and spatial resolution have been studied at Dasa/Ottobrunn in various projects, which were funded by national agencies (DLR) and the European Space Agency (ESA). The imaging spectrometers emphasized herein measure spatial images of the upwelling spectral radiance from 400 to 2400 nm at 5 to 15 nm spectral intervals. All concepts are prism/grating designs based on pushbroom imaging, and are designed to fulfill stringent requirements on spatial and spectral registration accuracy. Such imaging spectrometers comprise several critical and challenging subunits such as frontend calibration, pointing, baffling, telescope and spectrometer optics, focal plane assembly, etc. Of these subunits, the paper emphasizes the driving requirements and constraints of the optics. In particular, methods to control and optimize the most critical parameters like polarization, spatial and spectral purity/accuracy, transmission, and image quality are presented. The achieved performances and design inherent properties of all concepts are given.

The requirements concerning the radiometric accuracy of optical remote sensing systems for earth and environmental observations especially to high resolution imaging spectro- radiometers are increasing more and more. Accurate and conscientious on-ground and in-flight calibration of the sensors is one of the baselines to meet this requirement. From this point of view the polarization sensitivity of the sensors plays an important role because it is present more or less every time. Polarization sensitivity and its changes affect directly the radiometric accuracy of the estimated radiances of the polarized radiation coming from the scenes under investigation. In this paper an equipment for in-flight monitoring the polarization sensitivity of the sensor as part of the calibration procedure is presented. It can be used for measuring the plarization state of the incoming radiation too.

The European Ariborne Remote Sensing Capabilities (EARSEC) is a program of the Commission of the European Union in coordination with the European Space Agency. Its goal is to establish an independent European state-of-the-art capability in remote sensing for a wide range of applications. The core instrument of the 'Optical' component of EARSEC is an Imaging Spectrometer (the Digital Airborne Imaging Spectrometer 7915 or DIAS 7915) built by Geophysical & Environmental Research Corporation (GER) and operated by DLR, Oberpfaffenhofen, in collaboration with JRC. The 79 channel high resolution Imaging Spectrometer (IS) covers the 0.4 to 12.3 micrometers wavelength range with a spectral resolution varying from 16 to 2000 nm. Operated from a Dornier 228 aircraft, the spatial resolution can vary between 3 and 15 m. The instrument is calibrated and improved at DLR and should be operational in 1995 for campaigns over Europe. The 'Optical' component of EARSEC also includes ground facilities, mainly an electro-optical (EO) processor developed for JRC by Earth Observation Sciences Limited in the United Kingdom for DAIS 7915 data processing. This processor will generate four levels of products, from simple ingestion, to calibration into physical units of radiance, to geolocation and geocoding. Collaborations are foreseen with European groups operating other advanced optical sensors such as the Italian LARA project, operating the MIVIS (IS comparable to the DAIS), and DLR, operating the ROSIS (developed for marine applications, covering the 430-850 nm region). The EO processor could be adapted in the future to handle other IS data for a more universal use.

The Visible Infrared Mapping Spectrometer--Visible Channel (VIMS-V) has been designed to produce high resolution multispectral images, in the optical waveband, of different planetary bodies. VIMS-V, presently under test, has been developed by Officine Galileo on behalf of the Agenzia Spaziale Italiana (Italian Space Agency) and will cover the spectral range from 300nm to 1050nm. This range will allow the detailed investigation of the mineralogy of Saturn satellites surfaces, searching for those components capable of affecting their evolution; studies of Saturn and Titan cloud structure and haze layers by identifying chemical components; searches for lighting and analysis their spectra. Light weight, thermal stability, and capability to operate with different mission scenarios have been the imposed design criteria of the instrument. Two further versions of VIMS-V are presently under study: one for a cometary mission and the other for a lunar detailed exploration mission.

Persistent spectral hole burning (PSHB) makes it possible to store images of the sun spectrally and spatially in a single exposure step at very high resolution. The current system consists of a chlorin-doped polymer film (polyvinylbutyral), cooled to 2 K. It has a spectral resolution of 300 MHz (0.0004 nm) and may be used in the range of about 628 to 638 nm. Theoretically the spatial resolution is confined to molecular dimensions. In solar observations, however, it is determined by the optical setup and atmospheric conditions. The exposure is done by imaging the sun onto the sample (exposure energy: 6 mJ/cm² GHz). Afterwards the stored information is read out by scanning a tunable dye-laser across the spectal range of interest. The laser light is used to image the sample at each frequency point onto a cooled 12 bit CCD- camera. For acquisition, archiving, processing, and visualization of the huge amount of data (up to 10 GByte per experiment), a parallel processor system is used.

Imaging spectrometers make possible remote detection of individual spectral features that can be attributed to specific physical characteristics of geologic targets. Spectral variability measured in the field or laboratory can be directly related to mineralogy, however, for airborne systems, variability is compounded by within-pixel mixing. The research described here evaluated the combined use of an absorption-feature-based analysis approach with linear spectral unmixing for analysis of Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data. The feature-based approach allowed direct identification of individual materials in mixed pixels based on comparison of AVIRIS absorption band feature characteristics with facts and rules compiled using a spectral library. Probability images were created using the AVIRIS data through simultaneous assessment of multiple absorption features for multiple materials. The areas with the highest probabilities for each potential endmember were used to generate image-based average endmember spectra. Average spectra were also extracted using n- dimensional geometric techniques from areas with the 'purest' pixels and the feature-based approach was used to identify the endmembers. Linear spectral unmixing of the AVIRIS data for each material of interest provided estimates of mineral abundances and their spatial distributions. Contour maps of individual absorption feature characteristics such as absorption band depth were overlain on the abundance images to compare spectral variability to estimated mineral abundances. These images showed a strong spatial correlation between the deepest absorption features for specific minerals and the highest mineral abundances for those mineral from the unmixing results. The results suggest a methodology for analysis of imaging spectrometer data, where rather that applying feature-based methods to the entire imaging spectrometer data set, these methods are used instead only to identify materials extracted using the unmixing concepts.

The normalized difference vegetation index (NDVI), which is equal to (NIR- RED)/(NIR+RED), has been widely used for remote sensing of vegetation for many years. One weakness of this index is that the reflectance of RED channel has no sensitivity to changes in lead area index changes when the leaf area index is equal to 1 or greater due to strong chlorophyll absorption near 0.67 micron. In this paper, another index, namely the normalized difference water index (NDWI), is proposed for remote sensing of vegetation liquid water from space. NDWI is equal to [R(0.86 micrometers ) - R(1.24 micrometers )]/[R(0.86 micrometers ) + R(1.24 micrometers )], where R represents the apparent reflectance. At 0.86 micrometers and 1.24 micrometers , vegetation canopies have similar scattering properties, but slightly different liquid water absorption. The scattering by vegetation canopies enhances the weak liquid water absorption at 1.24 micrometers . As a result, NDWI is sensitive to changes in liquid water content of vegetation canopies. Spectral imaging data acquired with Airborne Visible Infrared Imaging Spectrometer (AVIRIS) over Jasper Ridge, California and Holland, Maine are used to demostrate the usefulness of NDWI. Comparisons between NDWI and NDVI images are also given. Because aerosol scattering effects in the 0.86-1.24 micrometers region are weak, NDWI is less sensitive to atmospheric effects that NDVI.

Interpretation of remotely sensed data is difficult in coastal regions compared to the open ocean, where optical signals are highly coupled to phytoplankton/chlorophyll. In estuarine and coastal areas, terrigenous colored dissolved organic matter (CDOM) does not covary with chlorophyll, and if the water column is optically shallow, bottom reflectance confounds interpretation of remote-sensing reflectance (R_TS) signals. In order to more accurately model R_TS in nearshore environements, bottom reflectance must be adequately characterized. During research cruises to the Florida Keys region in July, 1994 and March, 1995, reflectance spectra were obtained of various bottom types. R_TS measurements, obtained during these cruises, and R_TS measurements made with the Airborne Visible and Infrared Imaging Spectrometer (AVIRIS) were compared against R_TS derived from a hyperspectral model that decoupled water column and bottom contributions to the R_TS signal.

A new type of imaging telescope-spectrometer for surviving the sky aboard a satellite is described. A static Michelson interferometer in front of an objective with 2D-arrays in its focal plane is capable of providing interferograms both for point and extended sources. As an example, the telescope-spectrometer based on the 15-cm telescope of the IKON project and a plane-parallel Ge plate as a beamsplitter may have approximately equals 30 cm(superscript -1 spectral resolution in the range 3 - 20 micrometers . For higher resolution, such an objective interferometer has advantage over a dispersion spectrometer in the signal-to-noise ratio and is free from the disadvantage of an objective prism not providing spectra of extended sources.

This paper presents an overview of the HYDICE hyperspectral system, sponsored by the Naval Research Laboratory, built by Hughes Danbury Optical Systems and flown by the Environmental Research Institute of Michigan. HYDICE stands for the Hyperspectral Digital Imagery Collection Experiment. The sensor component of this experiment was procured as a dual use initiative. This unique sensor has pushbroom imaging optics with a prism spectrometer and InSb focal plane array detector. It represents a significant advance in signal- to-noise ratio, spatial and spectral resolution, and radiometric accuracy. This paper describes the system that has been built and flight tested. Differences between the original design and the actual hardware are indicated. The integration of the sensor with the aircraft is explained, including an overview of other on-board capabilities. The way in which the sensor system can be operated is illustrated. Flight test data are currently being analyzed, but selected laboratory performance tests are shown. The overall system flight performance is assessed qualitatively, with reference to laboratory data.

The measurement of the spectral distibution of radiation from individual points on the earth's surface provides important information of the material composition of those points. To collect data over large areas, it is necessary to use either airborne or spaceborne platforms. In this paper, different approaches to collecting such spectral information from a lightweight airborne platform will be reviewed. The use of spectral tuning technologies such as acousto-optical filters, liquid crystal filters, and linearly variable filters, as well as slit spectrometer systems, will be rebiewed. Features and benefits of each tuning technology and their limitations are presented. Various system configurations are also reviewed as well as constraints placed on system level performance parameters. In particular, trade studies of system level parameters, such as field of view and optical throughput, which are directly constrained by tunable filter technologies, are presented.

The next generation Wedge Imaging Spectrometer (WIS) instruments currently in integration at Hughes SBRD incorporate advanced features to increase operation flexibility for remotely sensed hyperspectral imagery collection and use. These features include: a) multiple linear wedge filters to tailor the spectral bands to the scene phenomenology; b) simple, replaceable fore-optics to allow different spatial resolutions and coverages; c) data acquisition system (DAS) that collects the full data stream simultaneously from both WIS instruments (VNIR and SWIR/MWIR), stores the data in a RAID storage, and provides for down-loading of the data to MO disks; the WIS DAS also allows selection of the spectral band sets to be stored; d) high-performance VNIR camera subsystem based upon a 512 X 512 CCD area array and associated electronics.

TRW has been involved in hyperspectral imaging since late 1989. The first instruments were constructed from commercially available components and were restricted in wavelength response to the visible and near IR (i.e., about 0.48 micrometers to 0.88 micrometers). They were used to take data from airborne platforms to support phenomenology studies. An instrument was then constructed to make measurements in the SWIR (i.e., out to 2.5 micrometers). It used mostly commercial components and contained some custom developments such as the foreoptics. These early instruments all recorded data using videotape recorders. A real time processor has been constructed which performs real time spectral template matching on six spectral templates. This significantly reduces operator load for systems where spectrally known targets are being sought. We are currently developing three new systems using custom components. The first is a high performance, aircraft based instrument called TRWIS III; the second, called HSI, will be the first hyperspectral imager in space, and is being developed for the NASA Small Satellite Technology Initiative; and the third is an ocean color instrument, known as the Low Resolution Camera, using the hyperspectral approach. Each of these instruments will be briefly described.

Data is presented from several types of infrared spectral imaging systems: narrowband, linear variable filter, transmission grating, and hyperspectral. A comparison review of existing infrared hyperspectral systems is provided.

The Department of Defense, through the US Air Force's Wright Laboratory, Armament Directorate is sponsoring the development of two types of IR imaging spectroradiometers (project name: IRIS) to measure the spatial/spectral characteristics of various military targets. Design and analysis of several technical approaches were conducted during an initial phase of the program. The technical approaches investigated included: a dispersive imaging spectrometer design utilizing a fiber-optic reformatter (contractor: ERIM); an imaging acousto-optic tunable filter (AOTF) design (contractor: Westinghouse); a spatial/spectral Fourier transform infrared (FTIR) spectrometer (contractor: Bomem Inc./Canada); a spatially modulated imaging fourier transform spectrometer (contractor: Daedalus Enterprises); an imaging Fabry-Perot design (contractor: Physical Sciences Inc.). Two of these designs were selected for brass board prototype fabrication. An FTIR prototype being built by Bomem Inc., offers an instrument with high sensitivity and high spectral resolution with modest spatial performance. An imaging Fabry-Perot prototype being built by Physical Sciences Inc., offers high spatial resolution with moderate sensitivity and spectral resolution.

A design approach for developing a staring-dispersive infrared imaging spectroradiometer utilizing a fiber-optic image formatter to simultaneously capture, at rates greater then 100 Hz, 256 high-resolution spectral images of a 2D scene is described. Detailed performance analyses confirm that the spectral radiance within the image scene can be measured to high absolute accuracy. The proposed low-risk prototype, based on commercial off-the-shelf components to minimize the development risk, can be cost-effectively upgraded to a final field instrument through replaceable optical, calibration, and detector modules. Producibility of the design, as described by mechanical layouts and manufacturing drawings of the optical, fiber-optic, and dispersive elements, has been confirmed by independent component manufacturers. The rugged design is sufficiently compact for use in high-performance military aircraft, and can be applied to numerous civilian applications such as environmental cleanup, mineral surveys, vegetation monitoring, and combustion analysis.

An advanced infrared imaging spectroradiometer for hot targets (100 to 1500 C) has been designed to give real-time spectrally resolved images in the 2-5 micrometers band and to measure irradiances with high accuracy. A tripod-mounted off-axis afocal telescope with a 2.5 degree diagonal field of view directs the input radiation through a Tl₃AsSe₃ acousto-optic tunable filter (AOTF). The AOTF deflects away from the main beam a narrow spectral bandwidth beam with its central wavelength determined by the acoustic frequency of the AOTF. The AOTF is electronically controlled and can change the wavelength of the deflected beam within approximately 25 microsecond(s) to any other wavelength. The deflected beam is focused onto an 128 X 128 InSb focal plane array which has a frame rate electronically adjustable from 1 to 217 frame/s. With this versatility: 1) key wavelength discriminators of potential targets can be rapidly accessed, 2) the signal-to-noise ration can be improved by increasing the integration time for point targets, and 3) detector saturation can be avoided by reducing the integration time and AOTF diffraction efficiency for very hot targets. Calculations indicate that irradiance measurement errors should usually be less than 1% and often less than 0.1%.

Many imaging applications require quantitative determination of a scene's spectral radiance. This paper describes a new system capable of real-time spectroradiometric imagery. Operating at a full-spectrum update rate of 30Hz, this imager is capable of collecting a 30 point spectrum from each of three imaging heads: the first operates from 400 nx m to 950 nm, with a 2% bandwidth; the second operates from 1.5 micrometers to 5.5 micrometers with a 1.5% bandwidth; the third operates from 5 micrometers to 12 micrometers , also at a 1.5% bandwidth. Standard image format is 256 X 256, with 512 X 512 possible in the VIS/NIR head. Spectra of up to 256 points are available at proportionately lower frame rates. In order to make such a tremendous amount of data more manageable, internal processing electronics perform four important operations on the spectral imagery data in real-time. First, all data in the spatial/spectral cube of data is spectro-radiometrically calibrated as it is collected. Second, to allow the imager to simulate sensors with arbitrary spectral response, any set of three spectral response functions may be loaded into the imager including delta functions to allow single wavelength viewing; the instrument then evaluates the integral of the product of the scene spectral radiances and the response function. Third, more powerful exploitation of the gathered spectral radiances can be effected by application of various spectral-matched filtering algorithms to identify pixels whose relative spectral radiance distribution matches a sought- after spectral radiance distribution, allowing materials-based identification and discrimination. Fourth, the instrument allows determination of spectral reflectance, surface temperature, and spectral emissivity, also in real-time. The spectral imaging technique used in the instrument allows tailoring of the frame rate and/or the spectral bandwidth to suit the scene radiance levels, i.e., frame rate can be reduced, or bandwidth increased to improve SNR when viewing low radiance scenes.

The Advanced Airborne Hyperspectral Imaging System (AAHIS) is a compact, lightweight visible and near IR pushbroom hyperspectral imaging spectrometer flown on a Piper Aztec aircraft. AAHIS is optimized for use in shallow water, littoral, and vegetation remote sensing. Data are collected at up to 55 frames/second and may be displayed and analyzed inflight or recorded for post-flight processing. Swath width is 200 meters at a flight altitude of 1 km. Each image pixel contains hyperspectral data simultaneously recorded in up to 288 contiguous spectral channels covering the 432 to 832 nm spectral region. Pixel binning typically yields pixels 1.0 meter square with a spectral channel width of 5.5 nm. Design and performance of the AAHIS is presented, including processed imagery demonstrating feature detection and materials discrimination on land and underwater at depths up to 27 meters.

Lawrence Livermore National Laboratory is currently operating a hyperspectral imager, the Livermore Imaging Fourier Transform Infrared Spectrometer. This instrument is capable of operating throughout the infrared spectrum from 3 to 12.5 micrometers with controllable spectral resolution. In this presentation we report on its operating characteristics, current capabilities, data throughput, and calibration issues.

Design considerations and experimental measurements from an imaging Fourier transform spectrometer are presented. The system is based on the Bomem MB-series of Fourier transform interferometer and is capable of more than 8 frames/second at 4 cm(superscript -1 apodized spectral resolution. The interferometer features dual output beams, allowing for example, the coverage of two different spectral ranges using a short-wave array and a long- wave array. The present system uses a set of two 8 X 8 InSb detector arrays to cover the 2 to 5.3 micrometers spectral range on two coaligned fields of view of 4 mrad X 4 mrad and 1 mrad X 1 mrad. Predicted noise equivalent spectral radiance as well as instrument lineshape are compared to measurements on the actual system. Particular emphasis is devoted to the behavior of the instrument lineshape with respect to off-axis position in the focal plane.

Modeling is an important part of any instrument development. For the shearing imaging spectrometer in use at the Malabar Test Facility, PL/LIMN a comprehensive model was developed. Preliminary testing and verification were done. The model was then used to perform a parametric study of CCD cameras for this instrument. The results indeed agree with intuition and serve as another verification of this model code.

Hyperspectral sensing systems are being developed for applications spanning astronomy, space object identification, remote sensing, and surveillance. We've developed an interactive, spreadsheet-based computer model which can be used to predict the performance for a type of hyperspectral sensor referred to as spatially modulated imaging Fourier transform spectrometers. The Hyperspectral Imager Model Program (HIMP) includes parameters which allow the specification of numerous target, atmospheric, instrumental, geometrical, and detector characteristics, as well as a variety of graphical outputs. HIMP may easily be modified or altered for a wide range of applications and scenarios.

We describe a band selection process based on wavelet analysis of hyperspectral data which naturally decomposes the data into sub-bands. Wavelet analysis allows the control of the position, resolution, and envelope of the specific spectral sub-bands which will be selected. The sub-band sets are selected to maximize the Kullback-Liebler distance between specific classes of materials for a specific dimensionality contraint or discrimination performance goal. A sequential construction of the sub-band sets is used as an approximation to the global maximization operation over all possible sub-band sets. A max/min strategy is also introduced to provide a robust framework for sub-band selection when faced with multiple materials. We show band selection and material classification results of this technique applied to Fourier transform spectrometer data.

In this article, recent measurements made with LIFTIRS, the Livermore Imaging Fourier Transform Infrared Spectrometer, are presented. The experience gained with this instrument has produced a variety of insights into the tradeoffs between signal to noise ratio (SNR), spectral resolution, and temporal resolution for time multiplexed Fourier transform imaging spectrometers. This experience has also clarified the practical advantages and disadvantages of Fourier transform hyperspectral imaging spectrometers regarding adaptation to varying measurement requirements on SNR versus spectral resolution, spatial resolution, and temporal resolution.

We use chlorin-doped polymer films at low temperatures as the primary imaging detector. Based on the principles of persistent spectral hole burning, this system is capable of storing spatial and spectral information simultaneously in one exposure with extremely high resolution. The sun as an extended light source has been imaged onto the film. The information recorded amounts to tens of GBytes. This data volume is read out by scanning the frequency of a tunable dye laser and reading the images with a digital CCD camera. For acquisition, archival, processing, and visualization, we use MUSIC (MUlti processor System with Intelligent Communication), a single instruction multiple data parallel processor system equipped with the necessary I/O facilities. The huge amount of data requires the developemnt of sophisticated algorithms to efficiently calibrate the data and to extract useful and new information for solar physics.

The space-borne imaging spectrometers to be flown in the near future will have the spatial resolution of a few hundred meters to one kilometer. The usefulness of their data for land- oriented applications is limited due to the problem of 'mixed pixels'. A technique is proposed for 'unmixing' the data of an imaging spectrometer by combined processing of its data with the data of a high resolution multispectral camera. It allows the retrieval of the spatial distribution of classes with the resolution of the multispectral camera and their spectral signatures, with the detail of spectral measurements of the imaging spectrometer. The technique has been tested, using the data of the airborne imaging spectrometer GER-II, obtained over an agricultural area. Various resolutions of the imaging spectrometer were simulated from tens of meters to approximately 1 km. The accuaracy of the retrieved spectra proved to be a few to ten percent in most cases, even when the mean area of 'homogeneous units' was significantly smaller than the pixel area of the imaging spectrometer. The proposed approach makes it possible to use a combination of a high resolution imaging spectrometer and to combine a high resolution information of the camera with the detailed spectral information of the imaging spectrometer during data processing. The advantages of this approach are: simpler and cheaper instruments can be used, including the instruments of the currently planned missions; data fluxes are significantly reduced; the swath width can be increased.

An evaluation of Russian satellite analog sensor systems providing panchromatic and multispectral imagery is given. Results will focus on the information content and the geometric quality of the Russian KVR-1000, TK-350, and KFA-1000 systems. Discussion and conclusions concerning potential design characteristics and the utility of the sensors to support existing and future mapping applications are provided.

The collection of spatial-spectral data of space objects is unique from most applications of spectral imaging technology. Space object identification requires collection of multiple frames of relatively low intensity, fast moving targets. The objective is to collect a single aggregate reflectance spectrum over the satellite and correlate this to orientation and solar phase angle. Higher statistical signal-to-noise can be achieved by averaging over the spatial dimension of a data frame. Blur was introduced by the tracking conditions in only the spatial dimension due to the design of the instrument. This blur allows some minimization of camera noise that cannot be removed through conventional techniques.

Diffractive optical elements (DOEs) have been proposed for many applications. One of the principle limitations of these lenses is abundant chromatic aberration that prohibits broadband use without design compensation. I present a novel configuration that exploits this typically unwanted effect to create an image spectrometer for visible or IR applications. The DOE provides both imaging and dispersion. A monochromatic CCD is scanned along the optical axis through a region described by the image locations of the desired wavelengths. Each detector position is an image location of one wavelength. However, the images captured by the CCD are a superposition of this infocus image and the images of the other wavelengths at various stated of defocus. Post-detection processing with computer tomography techniques is used to remove the unwanted blurred components, leaving an image for each spectral componenet. This paper presents the design, algorithms, experimental demonstration, and computer simulations of this image spectrometer's performance.

Kestrel Corporation and the Florida Institute of Technology have designed, and are now manufacturing, a Fourier transform visible hyperspectral imager system for use in a single engine light aircraft. The system is composed of a Sagnac-based interferometer optical subsystem, a data management system, and an aircraft attitude and current position sybsystem. The system is designed to have better than 5 nm spectral resolution at 450 nm, operates over the 440 nm to 1150 nm spectral band and has a 2D spatial resolution of 0.8 mrad. An internal calibration source is recorded with every frame of data to retain radiometric accuracy. The entire system fits into a Cessna 206 and uses a conventional downward looking view port located in the baggage compartment. During operation, data are collected at a rate of 15 Mbytes per second and stored direct to a disk array. Data storage has been sized to accommodate 56 minutes of observations. Designed for environmental mapping, this Fourier transform imager has uses in emergency response and military operations.

Space-based methods have become one of the major components in the whole set of the Earth exploration methods and means being developed for a long period of history. They played break-point roles in such fileds as hydrometeorology, geodesy and mapping, in study of the solar-terrestrial links, etc. Unfortunately, for another numerous fields of application, space means remained just an addition to the traditional data sources. Here we will talk about space- based observation systems for imaging of Earth by means of electronic scanning. Such systems have developed with continuous increase of the resolution of the Earth observation instruments, of the number of spectral channels, and the swath width. That has resulted in the enlarging of the amount of data obtained by the onboard instrumentation together with the data transmission capacity of the communications channels. As a consequence, data reception from the main space systems which observe the Earth from space can only be carried out by special receiving stations equipped with sophisticated and sometimes unique devices. Parallel to, and independent of, the development of space-based systems, an evolution of computers has been taking place on Earth. Achievements in microelectronics and personal computer development allow us now to process considerable amounts of data at very high rates which can be compared with data rates of the Earth imagery sources. The centralized reception of spaceborne data in systems of Earth exploration from space and the decentralization of data processing by a multitude of users has resulted in a certain contradiction which cannot be compensated by the ground communications means.

Our spectrograph for photometric imaging with numeric reconstruction uses a novel technique to produce 2D spectral images using a 1D imaging spectrometer. By varying the dispersion direction with respect to the center of the field-of-view, we produce a series of images with the intensity of each pixel equal to the integrated intensity of the field-of-view along the dispersion direction. Reconstruction of the 2D spectral image from this series of profiles can be accomplished with a variety of numerical methods. We present results obtained using the maximum entropy algorithm on lab tests, which include expected instrument noise, and compare this method to analytical backprojection methods. Finally, in each of these tests, we present improvements in speed over previous results through code optimization, which improves the viability of using the maximum entropy algorithm for spectroscopic imaging.

We describe the design of a silicon immersion grating spectrograph for the remote detection of chemicals in the atmosphere. The instrument is designed to operate in the two atmospheric windows from 2.3 to 2.5 and 2.8 to 4.2 microns at a resolution of 0.1 cm^-1. This is achieved by cross dispersing a high order silicon immersion echelle (13.5 grooves/mm) and a first order concave grating operating in a reflective configuration to generate a 2D spectrum in the image plane with diffraction limited performance.