This PDF file contains the front matter associated with SPIE Proceedings Volume 7457, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing.

In July 2006, the Department of Energy (DOE) National Nuclear Security Administration (NNSA) initiated a new project in their remote sensing test and evaluation program. Upwelling and downwelling radiance ground truth measurements have been advanced as a result of this project. Upwelling radiance measurements are used for the spectral characterization of calibration targets, and downwelling measurements are used to profile the temperature and moisture within the atmospheric column above these targets. These measurements will be used for the development and improvement of atmosphere compensations algorithms, as well as for the evaluation of the radiometric accuracy of other remote sensing systems. In order to meet stringent wavelength and radiometric calibration requirements, the selected technology is based on a Michelson interferometer spectrometer equipped with an internal calibration unit. The proposed configuration facilitates precise radiometric accuracy for target measurements, as well as concurrent temperature and moisture profiling of the atmosphere's Planetary Boundary Layer (PBL) above the target. In this paper we describe the instrument approach and its configuration. We also present results demonstrating the instrument performance. Atmospheric sounding results are compared to measurements made with other sounding systems at the ARM site in Oklahoma.

The need to provide high-quality large-area hyperspectral data for land use, geological, environmental, and mapping applications is critical to the US Government in the 21st Century. Technological advances with regard to larger focal plane arrays (FPA) along with maturing spectrometer designs have made it possible for the development of a next generation system beyond AVIRIS. This paper will introduce the MaRS system along with some data examples from Cuprite, NV and the National Arboretum.

This manuscript details the development and validation of a unique forward thinking instrument and methodology for monitoring terrestrial carbon dynamics through synthesis of existing hyperspectal sensing and Light Detection and Ranging (LIDAR) technologies. This technology demonstration is directly applicable to linking target mission concepts identified as scientific priorities in the National Research Council (NRC, 2007) Earth Science Decadal Survey; namely, DESDynI and HyspIRI. The primary components of the Hyperspec-LIDAR system are the ruggedized imaging spectrometer and a small footprint LIDAR system. The system is mounted on a heavy duty motorized pan-tilt unit programmed to support both push-broom style hyperspectral imaging and 3-D canopy LIDAR structural profiling. The integrated Hyperspec-LIDAR sensor system yields a hypserspectral data cube with up to 800 bands covering the spectral range of 400 to 1000 nm and a 3-D scanning LIDAR system accurately measuring the vertical distribution of intercepted surfaces within a range of 150 m with an accuracy of 15 mm. Preliminary field tests of the Hyperspec-LIDAR sensor system were conducted at a mature deciduous mixed forest tower site located at the Smithsonian Environmental Research Center in Edgewater, MD. The goal of this research is to produce integrated science and data products from ground observations that will support satellite-based hybrid spectral/structural profile linked through appropriate models to monitor Net Ecosystem Exchange and related parameters such as ecosystem Light Use Efficiency.

An airborne thermal hyperspectral imager is underdevelopment which utilizes the compact Dyson optical configuration and quantum well infrared photo detector (QWIP) focal plane array. The Dyson configuration uses a single monolithic prism-like grating design which allows for a high throughput instrument (F/1.6) with minimal ghosting, stray-light and large swath width. The configuration has the potential to be the optimal imaging spectroscopy solution unmanned aerial vehicles (UAV) due to its small form factor and relatively low power requirements. The planned instrument specifications are discussed as well as design trade-offs. Calibration testing results (noise equivalent temperature difference, spectral linearity and spectral bandwidth) and laboratory emissivity plots from samples are shown using an operational testbed unit which has similar specifications as the final airborne system. Field testing of the testbed unit was performed to acquire plots of emissivity for various known standard minerals (quartz). A comparison is made using data from the ASTER spectral library.

Progressive band selection (PBS) reduces spectral redundancy without significant loss of information, thereby reducing hyperspectral image data volume and processing time. Used onboard a spacecraft, it can also reduce image downlink time. PBS prioritizes an image's spectral bands according to priority scores that measure their significance to a specific application. Then it uses one of three methods to select an appropriate number of the most useful bands. Key challenges for PBS include selecting an appropriate criterion to generate band priority scores, and determining how many bands should be retained in the reduced image. The image's Virtual Dimensionality (VD), once computed, is a reasonable estimate of the latter. We describe the major design details of PBS and test PBS in a land classification experiment.

The management and storage of spectroradiometer data are important issues, especially in regards of long-term use, data quality and shareability. The SPECCHIO spectral database system developed at the Remote Sensing Laboratories (RSL) provides a solution for the organized storage of spectral data and associated metadata and for the spectral processing based on interactive, customizable and generic processing chains. Optimized data structures and graphical user interfaces combined with intelligent file parsing routines enable the efficient entry of spectral data and metadata. The system can be operated in a heterogeneous computing environment, offering multiuser access to a centralized database and enabling easy data sharing within and even across research groups.

An empirical (target-) BRDF normalization method has been implemented for Imaging Spectrometry data processing, following the approach of Kennedy, published in 1997. It is a simple, empirical method with the purpose of a rapid technique, based on a least-squares quadratic curve fitting process. The algorithm is calculating correction factors in either multiplicative or additive manner for each of the identified land cover classes, per spectral band and view angle unit. Image pre-classification is essential for successful anisotropy normalization. This anisotropy normalization method is a candidate to be used as baseline correction for future data products of APEX, a new airborne Imaging Spectrometer suitable for simulation and inter-calibration of data from various other sensors. A classification algorithm, being able to provide anisotropy class indexing that is optimized for the purpose of BRDF normalization has to be used. In this study, the performance of the standard Spectral Angle Mapper (SAM) approach with RSL's spectral database SPECCHIO attached is investigated. Due to its robustness regarding directional effects, SAM classification is estimated to be the most efficient. Results of both the classification and the normalization process are validated using two airborne image datasets from the HyMAP sensor, taken in 2004 over the "Vordemwald" test site in northern Switzerland.

This paper presents a comparison of different methods for structural modeling of hyperspectral imagery for target detection. We study structured models, based on linear subspaces and convex polyhedral cones, and their application for target detection. Different training methods are studied: Singular Value Decomposition (SVD) is used for subspace modeling, and Maximum Distance (MaxD) and Positive Matrix Factorization (PMF) for convex polyhedral modeling. We study different detectors based on orthogonal and oblique projections for subspace and convex polyhedral cones and evaluate their performance. Experimental results using HYDICE imagery are presented.

This project is an application of hyperspectral classification and unmixing in support of an ongoing archaeological study. The study region is the Oaxaca Valley located in the state of Oaxaca, Mexico on the southern coast. This was the birthplace of the Zapotec civilization which grew into a complex state level society. Hyperion imagery is being collected over a 30,000 km2 area. Classification maps of regions of interest are generated using K-means clustering and a novel algorithm called Gradient Flow. Gradient Flow departs from conventional stochastic or deterministic approaches, using graph theory to cluster spectral data. Spectral unmixing is conducted using the RIT developed algorithm Max-D to automatically find end members. Stepwise unmixing is performed to better model the data using the end members found be Max-D. Data are efficiently shared between imaging scientists and archaeologists using Google Earth to stream images over the internet rather than downloading them. The overall goal of the project is to provide archaeologists with useful information maps without having to interpret the raw data.

Jet engine exhaust radiates strongly in the midwave infrared due to line emission from combustion byproducts such as CO₂, CO, and H₂O. Imaging Fourier-transform spectrometers (IFTS) have the potential to measure spatial variations in plume temperature and density. However, the turbulent flow yields rapid, stochastic fluctuations in radiance during interferometric measurements which corrupt corresponding spectra. A novel, statistics-based method of interpreting a time-sequence of interferograms collected from a stochastic blackbody source is presented which enables good estimation of the underlying temperature distribution. It is shown that the median (and various other quantiles) interferograms afford unbiased spectral estimates of temperature upon Fourier transformation, in contrast to temperature estimates based on spectra obtained from mean interferograms. This method is then applied to IFTS data (200×64 pixels at 1cm^-1 resolution) of a turbulent exhaust plume from a small turbojet engine. Spatial maps of brightness temperature and estimates of turbulence-induced temperature distribution are presented.

Automated detection of chemical threats is essential for an early warning of a potential attack. Harnessing plants as bio-sensors allows for distributed sensing without a power supply. Monitoring the bio-sensors requires a specifically tailored hyperspectral system. Tobacco plants have been genetically engineered to de-green when a material of interest (e.g. zinc, TNT) is introduced to their immediate vicinity. The reflectance spectra of the bio-sensors must be accurately characterized during the de-greening process for them to play a role in an effective warning system. Hyperspectral data have been collected under laboratory conditions to determine the key regions in the reflectance spectra associated with the degreening phenomenon. Bio-sensor plants and control (nongenetically engineered) plants were exposed to TNT over the course of two days and their spectra were measured every six hours. Rochester Institute of Technologys Digital Imaging and Remote Sensing Image Generation Model (DIRSIG) was used to simulate detection of de-greened plants in the field. The simulated scene contains a brick school building, sidewalks, trees and the bio-sensors placed at the entrances to the buildings. Trade studies of the bio-sensor monitoring system were also conducted using DIRSIG simulations. System performance was studied as a function of field of view, pixel size, illumination conditions, radiometric noise, spectral waveband dependence and spectral resolution. Preliminary results show that the most significant change in reflectance during the degreening period occurs in the near infrared region.

Unmixing of the seabed for benthic habitat mapping in shallow coastal waters is a difficult problem due to the confounding effects of space variant bathymetry and water optical properties, which result in signatures for the same habitat classes to look different at the water surface across the image. This paper discusses different approaches to modify the linear unmixing approach to account for variable water optical properties and bathymetry and their implementation in the Hyperspectral Coastal Image Analysis Toolbox (HyCIAT). This toolbox allows the processing of hyperspectral imagery of shallow coastal areas to estimate water column optical properties, bathymetry, and perform unmixing for bottom composition. HyCIAT has been developed as part of the UPRM Hyperspectral Solutionware project to develop software tools for hyperspectral image processing. The tool has been developed under the MATLAB^TM environment and it includes a series of algorithms developed by UPRM researches under a graphical user interface that facilitates its use by the remote sensing community. The paper describes algorithms implemented in the toolbox, gives an overview of the graphical user interface, and presents results of its applications to AVIRIS and AISA hyperspectral imagery collected over Kaneohe Bay in Hawaii and over Southwestern Puerto Rico, respectively.

Significant advances have been made in developing normal-incidence sensitive quantum-dot infrared photodetectors (QDIPs) for midwave- and longwave-infrared imaging systems. QDIPs with nanoscale asymmetrical structures of the quantum dots can exhibit spectral responses tunable through the bias voltages applied. This makes it possible to build spectral imaging system in IR range based on single QDIP, without any spectral dispersive device upfront. Further more, unlike conventional systems whose spectral bands are fixed for various tasks which leads to data redundancy, the QDIP based system can be operated as being adaptive to scenes if different sets of operating bias voltages are selected for different tasks. To achieve such adaptivity, optimization algorithms must be developed to find the scene-based operation bias voltages set which maximizes the spectral context inside the output data while reducing the data redundancy. In this paper, we devise a series of optimization methods based on a recently developed geometrical spectral imaging model (Wang et al., 2007):1 In the beginning, an scene-independent set of bias voltages is selected to maximize the average signal-to-noise ratio (SNR) of the sensor. Then, some bias voltages are added or removed based on the captured data. This dynamic optimization process is performed throughout the imaging process so that the balance between data information and data volume is always achieved. Due to the universality of the algorithm, this optimization process can be applied to any spectral sensor whose spectral response functions are known.

High spectral resolution Fourier transform imaging spectroscopy has been demonstrated at the Lockheed Martin Advanced Technology Center. A testbed was built using a Michelson interferometer with a two-stage end-mirror control system. Homodyne laser metrology was used to sense relative tip, tilt and piston in the interferometer, and a 3-degree of freedom fast steering mirror in conjunction with a linear actuator stage provided sub-nanometer actuation control over 20 millimeters of piston range. The range of piston over which signal was present allowed for spectral resolution at the nanometer level in the visible / near infrared (VNIR) band for every pixel in the reconstructed image.

The background and beginning of this project have been described in a previous SPIE paper1. Since then, the Na monolithic Spatial Heterodyne Spectrometer (SHS) units were constructed, and the housing for the full SHIELDS unit designed and built. The dual-wavelength SHIELDS was designed, and its construction begun, while the LC reflectors used in the selection between wavelengths for the dual-wavelength monolith were tested for efficacy in an instrument-like configuration. Optical modeling and procurement of optical components was completed, making the Na unit nearly ready for lab tests with a low-pressure sodium source, and then appropriate Na-wavelength fluorophores. Atmospheric modeling showed the importance of both dealing with the Ring effect -- as it is at least equal to the fluorescence effect to be measured -- and selecting the best wavelength to observe to mitigate the effects of vegetative fluorescence and water vapor absorption. The full SHIELDS unit has been assembled and initially tested in May 2009, and the dual-wavelength monolith will be completed in July 2009.

An imager based upon an etched liquid-crystal Fabry-Perot (LCFP) dispersive element is able to simultaneously sample four distinct resolution elements in the region 800 nm - 1100 nm, and tune in milliseconds to any one of 30,661 possible four-color scene images with a spectral resolution of 0.67 nm. Independently tunable quadrants of a single LCFP etalon are created by etching a transparent conducting layer on the etalon substrate, and one image from each quadrant is formed on a focal-plane array detector. Designed to weigh less than 20 lbs. in production, the portable, solid-state camera system is designed to provide fast RGB images of transient spectral phenomena, but many applications are possible. The fourth image in each four-element image is intended to be a background channel for contrast enhancement in bright background environments. The LCFP hyperspectral imager provides high-spectral resolution, allowing detection of short-lifetime atomic spectral line emissions characteristic of excited or ablating constituents against a bright, broadband, greybody background. High luminosity via the characteristic Fabry-Perot étendue advantage and an f/0.9 optical system accommodate the tactical need for a lightweight device with a small footprint. The LCFP dispersive element is tuned with battery-pack power, 0- 10V DC and mA current. The LCFP hyperspectral technology is easily adapted to Doppler imaging by enhancing the etalon gap and sampling over a narrower instrument passband. Operation in the MWIR and LWIR is also possible. The camera design creates multi-spectral images with a small but simultaneously sampled data-cube of narrow bandwidth.

The hypersensor camera operates with a unique multispectral imaging modality developed recently at Surface Optics Corporation. The Hypersensor camera is small, low cost, rugged, and solid state, using micro-optics and an array of spectral filters, which captures a complete multispectral cube of spatial and spectral data with every focal plane exposure. The prototype VNIR Hypersensor camera captures full cubes of 588x438 (spatial pixels) x 16 (spectral bands) at frame rates up to 60 Hz. This paper discusses the optical design of the Hypersensor camera, the measured performance, and the design and operation of a custom video-rate hyperspectral processor developed for this system.

Paper discusses an accuracy, reliability and reproducibility of "in vivo" laser fluorescent diagnostics (LFD) in medicine. Modern equipment and calibration standards for LFD are discussed as well. It is shown, that, in spite of the fact that formal random errors of "in vivo" instantaneous fluorescent measurements have been previously evaluated on a level of 30-40%, the medical accuracy and reliability of LFD could reach a quite high and informative level. Most of the formal "random" disperses in results of the single "snapshot" measurements are associated with a changeable and alive character of the object of diagnostics.

There are two broad classes of hyperspectral detection algorithms.1, 2 Algorithms in the first class use the spectral covariance matrix of the background clutter; in contrast, algorithms in the second class characterize the background using a subspace model. In this paper we show that, due to the nature of hyperspectral imaging data, the two families of algorithms are intimately related. The link between the two representations of the background clutter is the low-rank of the covariance matrix of natural hyperspectral backgrounds and its relation to the spectral linear mixture model. This link is developed using the method of dominant mode rejection. Finally, the effects of regularization

Military targets such as aircrafts and flares do not exhibit unique infrared signatures; their emissions are dominated by combustion products (mostly water vapor, carbon dioxide and hydrogen chloride) and hot metal greybody emissions. An algorithm has thus been developed to categorize target signatures based on their emission source components. The signatures are then partitioned, based on their emission components, into groups of similar emission characteristics. Using previous trial data, seven unique flare categories were defined. A second algorithm was finally developed to exploit this signature description and interrogate individual field measurements for target detection and categorization.

Many hyperspectral measures such as Spectral Angle Mapper (SAM), Euclidean Distance (ED), Spectral Information Divergence (SID) calculate values that can be used to measure the closeness between two hyperspectral signatures in terms of spectral similarity. When a hyperspectral measure is used for target discrimination or classification, a commonly used approach assigns a data sample to a spectral signature with the maximum spectral similarity or a class whose mean is most similar to its spectral signature. As a result, such a hyperspectral measure is called a hard decision hyperspectral measure and its performance is evaluated by its confusion matrix. This paper develops a new class of hyperspectral measures, called soft-decision hyperspectral measures which use the similarity value between a data sample and a target signature or a class as an indicator of the likelihood of the data sample assigned to this particular signature or class instead of signature or class map resulting from hard decisions. In order for a soft-decision to perform target discrimination or classification, the soft-decision hyperspectral measure-generated likelihood values are normalized to probabilities so that a threshold can be used to make hard decisions via a recently developed 3D ROC analysis. Experimental study demonstrates that these two approaches yield different results.

The detection of sub-pixel materials in a hyperspectral scene is often accomplished using spectral matched filters or subspace projection. These methods rely on estimates of background second order statistics or subspaces in a scene that are usually based on either on global statistics of the entire scene or on adaptive local statistics. Global statistics have the disadvantage of including materials of interest in the background estimate and this implies the method assumes these materials occupy an insignificant portion of the scene. Adaptive methods that use a small number of samples surrounding the pixel of interest to estimate a background covariance eliminate much of this disadvantage, but this comes at the cost of significantly increasing computation time and potentially unstable estimates for some backgrounds. A number of spectral matched filter methods have been developed with increasing sophistication, but experience indicates that the method used to compute the background statistics may have a greater impact on overall detector performance. This research investigates the use of a neural network approach to estimate the background statistics needed for certain spectral matched filters requiring global statistics. The context of the effort is terrain, urban, and shallow-water mapping using hyperspectral imagery, where the materials of interest inherently occupy a significant portion of a scene or where certain background classes have problematic second-order statistics. Results of experiments within this context are shown.

A smart spectral imaging detection method based on the integration of electrically tunable liquid-crystal(LC) Fabry-Perot(FP) microstructure array is proposed. It has very broad application in many fields with advantages of low cost, highly compact integration and working without mechanical part. The device can get hundreds of spectral bands simultaneously in one frame of picture in theory. This paper proposes the structure of smart spectral imaging array device, and analyses some key issues of liquid crystal Fabry-Perot structures for imaging application and calibration. Prototypes of 4 × 4 LC-FP array with the cavity thickness ranging from 4 to 20 μm for the working wavelength in the range of 800~900nm, are made by lithography and wet etching. Test is carried out and analysed.