This PDF file contains the front matter associated with SPIE Proceedings Volume 6661, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing.

The Advanced Responsive Tactically-Effective Military Imaging Spectrometer (ARTEMIS) is under development for tactical military applications and is the primary payload for the TacSat-3 satellite. The optical design for the telescope, imaging spectrometer, and high resolution imager is described.

A persistent barrier to the wider use of the Computed Tomographic Imaging Spectrometer (CTIS) has been the extraordinary demands it places on computational resources. Raw images can be obtained at snapshot speeds, but reconstructed datacubes typically require minutes of reconstruction time each. We present a new approach to the CTIS reconstruction problem which makes use of the spatial shift-invariance in a CTIS system to greatly reduce the dimensionality of the matrix inversion process performed during reconstruction. Preliminary results indicate that a speedup by a factor of 4000 is possible.

A computed tomography imaging channeled spectropolarimeter (CTICS) is a combination of a computed tomography imaging spectrometer (CTIS) and a channeled spectropolarimeter (CHSP). The CTICS instrument can simultaneously obtain image spatial and spectral information as well as polarization Stokes vectors at each resolution element in a single focal plane array (FPA) integration time with no moving parts. An instrument has been designed and built for the visible wavelength region at the University of Arizona. Performance testing is underway. In this work, we present initial results from data acquired during testing of the CTICS instrument.

We introduce and demonstrate a simple spectrophotometer system insensitive to input polarization and with strong potential for compact and low-cost implementation. This technology has a wide variety of potential applications ranging from astronomy to medicine and even the cosmetics industry. To enable more powerful and portable microspectrometers we employ a novel design based on a tunable liquid crystal filter with polarization-independence, which is constructed of stacked liquid crystal polarization gratings (LCPGs). These switchable, anisotropic, thin diffraction gratings exhibit unique properties that include diffraction at visible and infrared wavelengths that can be coupled between only the zero- and first-orders (with nearly 100% and 0% experimentally verified efficiencies), depending on the applied voltage and wavelength of incident light. When combined with an elemental spatial filter, polarization-independent bandpass tuning can be achieved with minimum loss. Analogous to Lyot and Solc filters, several LCPGs are layered and introduced into a temporally resolved system using a single photodetector. The unique filter design enables improvement in terms of resolution and sensitivity by eliminating the polarization dependence present in all competing birefringence-based technologies. Also, the temporal detection system has a potential for improved miniaturization compared to any competing relevant approach and decreased cost by avoiding highly sensitive alignment, reflective diffraction components, Fabry- Perot cavities, and expensive detectors. In this work we describe the core principles of the tunable filter, present a representative spectrometer system design, report preliminary experimental data, and discuss the capabilities of the system in terms of spectral range, resolution, and sensitivity.

Imaging Fourier-transform spectrometers can quickly produce massive amounts of raw data, especially when paired with large focal plane arrays. As the spatial resolution is increased, overwhelming amounts of data must be managed properly. A suitable design of the data processing chain is thus required to minimize the dataload and deliver processed information in real-time. This paper reviews the work being done to tailor data processing pipelines for Fourier-transform spectrometers (FTS) coupled with externally triggered CCD cameras. Various sampling techniques as well as spectral calibration and line shape correction approaches will be reviewed. Since traditional sampling techniques are not well suited for an FTS operating with a CCD camera, a hybrid time-position sampling approach is presented to reduce the number of samples per pixel. Furthermore, the approach enables a sampling jitter correction algorithm that can account for velocity fluctuations and channel delays, such as the CCD integration time. A fast spectral calibration approach is also demonstrated, based on a rapid line shape integration scheme. The calibration algorithm brings all pixel spectra on the same spectral grid and allows the user to directly compare spectral features between pixels. Moreover, the correction method offers software field-widening capabilities by binning pixels after spectral calibration. A large single-pixel detector can thus be emulated from the CCD array, allowing the user to broaden the field of view and to increase the SNR.

In June 2005, a newly develop long wave Focal Pane Array (FPA), based on photo-voltaic technology was delivered to the Defense Research & Development of Canada (DRDC). This development was part of technological Demonstration program that was founded by the DRDC. This paper will describe the FPA configuration along with its performance assessment configured in the Air PIRATE FTIR spectrometer. Air PIRATE is an airborne version of the hyper spectral spectrometer used by the Canadian Defense for target identification, as well as chemical agent identification.

Fraunhofer Line Discrimination (FLD) is a passive optical spectroscopy technique with potential for battlefield remote sensing of aerosol targets, as well as other military and academic applications. The Spatial Heterodyne Interferometer for Emergent Line Discrimination Spectroscopy (SHIELDS) will provide real-time remote sensing using FLD. The unit will be contained in a man-portable box to provide heads-up detection of dangerous chemicals in target clouds. The spectrometer employed will be the monolithic Spatial Heterodyne Spectrometer (SHS). One SHIELDS unit will feature a monolithic SHS to look at the 589-nm Solar Fraunhofer doublet. A second monolith will be built, using novel designs, to look at several different Fraunhofer lines of interest, all in the visible (H-b, Mg, H-a). The finished monoliths will be tested on laboratory targets, and the final complete SHIELDS unit will be further tested in the field.

We present the recent development of a compact Michelson-like interferometer for an imaging Fourier Transform Spectrometer (IFTS). The interferometer has a mass of less than 600 g and dimensions of about 60 mm x 90 mm x 100 mm. It is designed to be stiff to reduce its sensitivity to vibrations. Its maximum optical path difference is 1 cm. Despite its small size it can support an etendue of 9.2x10^-7 m² sr. This interferometer is well suited to serve as the modulator for a small IFTS when mass and volume are restricted such as onboard planetary probes, UAV, etc. This interferometer can be adapted to a wide variety of infrared imaging detectors. It is a building block upon which can be designed a large range of custom infrared imaging spectrometers.

DRDC Valcartier is continuing to developed infrared spectral imagery systems for a variety of military applications. Recently a hybrid airborne spectral imager / broadband imager system has been developed for ground target interrogation (AIRIS). This system employs a Fourier Transform Interferometer system coupled to two 8x8 element detector arrays to create spectral imagery in the region from 2.0 to 12 microns (830 to 5000 cm^-1) at a spectral resolution of up to 1 cm^-1. In addition, coupled to this sensor are three broadband imagers operating in the visible, mid-wave and long-wave infrared regions. AIRIS uses an on-board tracking capability to: dwell on a target, select multiple targets sequentially, or build a mosaic description of the environment around a specified target point. Currently AIRIS is being modified to include real-time spectral imagery calibration and application processing. In this paper the flexibility of the AIRIS system will be described, its concept of operation discussed and examples of measurements will be shown.

Large crystals of Hg₂Cl₂ and Hg₂Br₂ (48 mm in diameter, up to 600g in weight) have been grown by self-seeded contactless Physical Vapor Transport (PVT) technique in closed ampoules in vertical configuration. Seed selection was accomplished in the small diameter tubing at the ampoule tip. Material purification was done by resublimation processing. The crystals show high transparency, with good transmittance and good crystallographic quality. An acousto-optic modulator built from the Hg₂Cl₂ crystal showed good performance consistent with predicted device parameters.

BAE SYSTEMS has developed a high-resolution 2D imaging laser radar (LADAR) system that has proven its ability to detect and identify hard targets in occluded environments, through battlefield obscurants, and through naturally occurring image-degrading atmospheres. Limitations of passive infrared imaging for target identification using medium wavelength infrared (MWIR) and long wavelength infrared (LWIR) atmospheric windows are well known. Of particular concern is that as wavelength is increased the aperture must be increased to maintain resolution, hence, driving apertures to be very larger for long-range identification; impractical because of size, weight, and optics cost. Conversely, at smaller apertures and with large f-numbers images may become photon starved with long integration times. Here, images are most susceptible to distortion from atmospheric turbulence, platform vibration, or both. Additionally, long-range identification using passive thermal imaging is clutter limited arising from objects in close proximity to the target object.

The performance of hyperspectral target detection is known to vary with the light level. To characterize this variation, the effect of synthetically varying the amount of light is studied, starting with real images. Camera calibration coefficients are divided out to obtain the number of detector photoelectrons corresponding to each radiance sample. It is then straightforward to scale down the photoelectron numbers to represent a camera with lower light throughput and introduce additional noise by drawing corresponding Poisson-distributed photoelectron values. To first order, the resulting data may alternatively be considered to represent the increase in noise expected from a reduction in scene illumination. Images are generated with reduction of the light level down to a fraction of a percent of the original image. Sample spectra are shown after correction to photoelectron count, and it is argued that this representation of the data is useful because it permits direct estimation of signal to noise ratio. Spectral anomaly detection is applied to original and modified images, and the variation of false alarm rate with reduction factor is characterized. The results indicate that photon noise is the dominating cause of false alarms. There is no clear sign of false alarm contributions from the actual background clutter down to the data limit of false alarm probability, about 10^-6 per pixel.

One of the biggest issues with the Linear Mixing Model (LMM) is that it is implicitly assumed that each of the individual material components throughout the scene may be described using a single dimension (e.g. an endmember vector). In reality, individual pixels corresponding to the same general material class can exhibit a large degree of variation within a given scene. This is especially true in broad background classes such as forests, where the single dimension assumption clearly fails. In practice, the only way to account for the multidimensionality of the class is to choose multiple (very similar) endmembers, each of which represents some part of the class. To address these issues, we introduce the endmember subgroup model, which generalizes the notion of an 'endmember vector' to an 'endmember subspace'. In this model, spectra in a given hyperspectral scene are decomposed as a sum of constituent materials; however, each material is represented by some multidimensional subspace (instead of a single vector). The dimensionality of the subspace will depend on the within-class variation seen in the image. The endmember subgroups can be determined automatically from the data, or can use physics-based modeling techniques to include 'signature subspaces', which are included in the endmember subgroups. In this paper, we give an overview of the subgroup model; discuss methods for determining the endmember subgroups for a given image, and present results showing how the subgroup model improves upon traditional single endmember linear mixing. We also include results that use the 'signature subspace' approach to identifying mixed-pixel targets in HYDICE imagery.

Endmember extraction has received considerable interest in recent years. Many algorithms have been developed for this purpose and most of them are designed based on convexity geometry such as vertex or endpoint projection and maximization of simplex volume. This paper develops statistics-based approaches to endmember extraction in the sense that different orders of statistics are used as criteria to extract endmembers. The idea behind the proposed statistics-based endmember extraction algorithms (EEAs) is to assume that a set of endmmembers constitute the most un-correlated sample pool among all the same number of signatures with correlation measured by statistics which include variance specified by 2^nd order statistics, least squares error (LSE) also specified by 2^nd order statistics (variance), 3^rd order statistics (skewness), 4^th order statistics (kurtosis), k^th moment, entropy specified by infinite order of statistics and statistical independency measured by mutual information. Of particular interest are Independent Component Analysis-based EEAs which use statistics of various orders such as variance, skewness, kurtosis the kth moment and infinite orders including entropy and divergence. In order to substantiate proposed statistics-based EEAs, experiments using synthetic and real images are conducted in comparison with several popular and well-known EEAs such as Pixel Purity Index (PPI), N-finder algorithm (N-FINDR).

Remote sensing often utilizes models to predict the ability of an optical system to collect data optimally prior to costly sensor testing and manufacturing. Significant effort is required to create an accurate model, and therefore most designs focus on either radiometric or spatial precision rather than a combination of the two. We present a case study in which a model has been created to satisfy both radiometric and spatial fidelity requirements. Terrain, vegetation, targets and other components of the model were designed with high precision. Hyperspectral imagery was generated using the Digital Imaging and Remote Sensing Image Generation Model (DIRSIG) based on numerous spectral and spatial ground-truth measurements. These included spectral reflectance of targets and the environment, atmospheric variables, as well as geometry and distribution of objects within the scene. Imagery was collected by airborne systems for accuracy assessment. The generated data has been validated by qualitative evaluation of the spectral characteristics and comparisons of results from PC transform and the RX anomaly detection algorithm. Validation results indicate that the model achieved a desired level of accuracy.

Many algorithms exist to invert airborne imagery from units of either radiance or sensor specific digital counts to units of reflectance. These compensation algorithms remove unwanted atmospheric variability allowing objects on the ground to be analyzed. Low error levels in homogenous atmospheric conditions have been demonstrated. In many cases however, clouds are present in the atmosphere which introduce error into the inversion at unacceptable levels. For example, the relationship that is defined between sensor reaching radiance and ground reflectance in a cloud free scene will not be the same as in a scene with clouds. A novel method has been developed which utilizes ground based measurements to modify the empirical line method (ELM) approach on a per-pixel basis. A physics based model of the atmosphere is used to generate a spatial correction for the ELM. Creation of this model is accomplished by analyzing whole-sky imagery to produce a cloud mask which drives input parameters to the radiative transfer (RT) code MODTRAN. The RT code is run for several different azimuth and zenith orientations to create a three-dimensional representation of the hemisphere. The model is then used to achieve a per-pixel correction by adjusting the ELM slope spatially. This method is applied to real data acquired over the atmospheric radiation measurement (ARM) site in Lamount, OK. Performance of the method is evaluated with the Hyperspectral Digital Imagery Collection Experiment (HYDICE) instrument as well as a simulated multi-spectral system.

Two major issues encountered in unsupervised hyperspectral image classification are (1) how to determine the number of spectral classes in the image and (2) how to find training samples that well represent each of spectral classes without prior knowledge. A recently developed concept, Virtual dimensionality (VD) is used to estimate the number of spectral classes of interest in the image data. This paper proposes an effective algorithm to generate an appropriate training set via a recently developed Prioritized Independent Component Analysis (PICA). Two sets of hyperspectral data, Airborne Visible Infrared Imaging Spectrometer (AVIRIS) Cuprite data and HYperspectral Digital Image Collection Experiment (HYDICE) data are used for experiments and performance analysis for the proposed method.

Upon collecting hyperspectral image (HSI) data, several steps are usually required prior to comparing sensor data to a target reflectance spectrum of interest. A common practice is the application of an atmospheric compensation routine, which converts a spectral cube from the radiance domain to the reflectance domain. Such routines may prove to be problematic when predicting reflectance spectra for subpixel targets under varying illumination conditions. An alternative to atmospheric compensation is to employ physics-based forward models to predict the ways that a target spectrum might appear at the sensor. Instead of generating a single target radiance vector, a target vector space is created, which theoretically spans all possible target manifestations in the sensor radiance cube. Typically, a background vector space is also generated using in-scene radiance vectors that are significantly unlike the target space. Target detection then occurs in the radiance domain by comparing a scene pixel's radiance spectrum to the target and background vector spaces. One disadvantage of using such physics-based model approaches is that the complexity of the radiance model drives the span of the target vector space. It may be possible to optimize the volume of target spaces on a local basis by incorporating spatial information, as provided by a geo-registered, co-temporal topographical Lidar data set. Such data may serve to eliminate geometric ambiguity at any given location in a scene; thus, the local target vector space may be constrained relative to the global target space. In doing so, target detection performance may be improved. A description of the various image processing techniques used to constrain target vector spaces is presented, including estimation of shadows, ground plane orientation, skydome visibility, and pixel purity. Finally, target detection performance resulting from the constrained vector space approach is discussed.

Airborne passive hyperspectral infrared spectral measurements of chemical vapors in the atmosphere have been completed over a wide variety of locations throughout the United States. These measurements are part of the US EPA emergency response chemical disaster mitigation capability. Analysis and regional comparison of these atmospheric measurements reveals a glycol constituent, which has been noted during flooding conditions along the Southern Gulf Coast Region and the Midwestern United States. This discussion will describe several differences in the natural atmospheric background for vapor species identified in various regions of the country. There are two possible sources for this constituent in these regions one is a natural source the other is an anthropogenic source. The paper will highlight the usefulness of passive infrared spectral measurements to determine key atmospheric indicators correlated with locations of major flooding along with the identification of naturally occurring species.

Standoff detection, identification and quantification of chemical agents are fundamental needs in several fields of applications. Additional required sensor characteristics include high sensitivity, low false alarms and high-speed (ideally real-time) operation, all in a compact and robust package. The thermal infrared portion of the electromagnetic spectrum has been utilized to implement such chemical sensors, either with spectrometers (with none or moderate imaging capability) or with imagers (with moderate spectral capability). Only with the recent emergence of high-speed, large format infrared imaging arrays, has it been possible to design chemical sensors offering uncompromising performance in the spectral, spatial, as well as the temporal domain. Telops has developed an innovative instrument that can not only provide an early warning for chemical agents and toxic chemicals, but also one that provides a "Chemical Map" of the field of view. To provide to best field imaging spectroscopy instrument, Telops has developed the FIRST, Field-portable Imaging Radiometric Spectrometer Technology, instrument. This instrument is based on a modular design that includes: a high performance infrared FPA and data acquisition electronics, onboard data processing electronics, a high performance Fourier transform modulator, dual integrated radiometric calibration targets and a visible boresighted camera. These modules, assembled together in an environmentally robust structure, used in combination with Telops' proven radiometric and spectral calibration algorithms make this instrument a world-class passive standoff detection system for chemical imaging. This paper presents chemical detection and identification results obtained with the FIRST sensor.

Measuring the Modulation Transfer Function (MTF) of a hyperspectral imaging spectrometer focused at infinity requires a longer optical path than is available in a typical laboratory. We describe a technique that uses images of rooflines on buildings of opportunity in a knife-edge-based MTF measurement. This technique only measures the MTF along one dimension. However, the hyper-spectral imaging systems characterized in this paper are particularly suited to a knife-edge technique, as imaging only takes place in one dimension of the array and spectral separation takes place along the other. The sharp edges needed in these measurements were provided by dark rooftops backlit by a uniformly cloudy sky. We have applied this technique to hyperspectral imagers that operate in the visible-near infrared (VNIR) and short-wave infraRed (SWIR) spectral bands. The data presented in this paper focuses on the characterization of the SWIR imaging spectrometer developed by Resonon Inc.