About Us - Kestrel Corporation

CUSTOMER	PROJECT TITLE	DESCRIPTION
Air Force Research Lab - Directed Energy Directorate	Distorted Grating Wavefront Sensor (DG-WFS) for "Difficult" Conditions	Kestrel developed a new technology that offered the opportunity to use phase diversity wavefront sensing to compensate for high scintillation conditions and to operate using passive extended target references. Based on the use of a distorted optical grating technology, this project showed that it was feasible to provide, simultaneously, the images that are needed to execute phase diversity algorithms before and after the optical entrance pupil. The work demonstrated that these algorithms could be completed in real time frames that are compatible with the compensating atmospheric aberrations.
Air Force Research Lab - Directed Energy Directorate	Distorted Wavefront Grating Sensor - Enabling Robust Adaptive Optics Correction Systems	Kestrel demonstrated that a distorted grating based wavefront sensor (DG-WFS) is a viable technology for measuring wavefront distortions caused by multiple concurrent adverse conditions that could exist in an Air Borne Laser (ABL) environmental. Kestrel designed and transitioned the research of DG-WFS from a laboratory experiment to a well-understood and demonstrated sensor system. The resulting DG-WFS has the capability to enable and directly drive a complementary and corrective adaptive optics system operating in a highly aberrated wavefront propagation environment. The work was divided into two efforts; a basic set of predominantly experimental research, and a follow-on set of mostly theoretical modeling and simulation research based on the results of the initial analysis and experimentation.
Air Force Research Lab - Directed Energy Directorate	Ultra High Speed Aero Optics Compensation	Kestrel developed and demonstrated the ability of its Distorted Grating Wavefront Sensor (DG-WFS) to measure both low-speed and ultra-high speed aero-optics effects and to compute the need compensation for various applications.
Air Force Research Lab - Space Vehicles Directorate	High Band-width Atmospheric Turbulence Balloon Measurement System	This study provided a design for a high bandwidth balloon borne sensor suite to support detailed measurements of atmospheric turbulence. These turbulence measurements included the data required to formulate an understanding of the optical wave propagation environment in the vicinity of, and above, the troposphere.
Army Space and Missile Defense Command	Atmospheric Turbulence Measurement System	This project focused on the development of a totally new, non-intrusive optical method for making fundamental atmospheric turbulence measurements. A correlation between the movements of a pair of thin beams was used to define the inner and outer scale size independent of any assumed turbulence model and to calculate the index of refraction coefficient. During the effort, extensions were made to the underlying theory to adapt it to upper atmosphere ambient conditions and to determine the effect of a moving platform. Once completed the models were exercised to define the usable operating envelope. A set of laboratory experiments were accomplished that demonstrated the application of the technique to a moving platform and to demonstrate its use as a three-dimensional measurement tool. From these results two conceptual designs for a dual thin beam atmospheric turbulence measurement system were completed; one to operate from a balloon platform, the other from an aircraft.
Australian Proprietary Client	QuickMap-AirCam Imaging System (QUMACIS)	Kestrel developed its QuickMap-AirCam Imaging System (QUMACIS) as a unique high-tech solution to multi-dimensional image data collection. The system is a compact, relatively low-cost electronic imaging system suitable for a wide range of surveying, mapping, and imaging missions. Based on a combination of multispectral and true color imaging sensors, the system provides the capability to simultaneously collect data unmatched in the commercial industry. The modular hardware and software design allowed QUMACIS to be rapidly modified or expanded to suit specific mission requirements. The system's small size, weight and power requirements allows it to be easily implemented in lightweight aircraft and even in unmanned aerial vehicles (UAVs).
Ball Aerospace	Ultraspectral Imager (USI) Gas Concentration Detection	This project provided detection sensitivity data regarding the ability of Kestrel's Ultraspectral Imager (USI) in the identification and analysis of various gas concentrations against known backgrounds.
Boeing	High Altitude Balloon Experiment (HABE Program)	The government's High Altitude Balloon Experiment (HABE) is an autonomous laser pointing and tracking experiment that supported risk reduction for the Space Based Laser (SBL) program. Kestrel provided artificial intelligence software to meet the autonomous requirements of the balloon payload so the experiment could assess its own health and status, and also to manage the laser engagement with the target missile. The software delivered was an embedded system.
Caterpillar Power	Kuwait Patriot Power	Kestrel performed on-site tests of newly delivered engine and power generation computer control interface modules (CIMs) installed on 150 KW generators in Kuwait in support of the Kuwait Patriot Power program. The continuous running tests over 5 days were designed to operate the generator through its different operating states, under multiple environmental conditions, and to check compatibility with loading and interfacing to the existing electrical power network.
Caterpillar Power	Kuwait Air Defense System - Power Generation Test and Qualification	Project to develop, test and evaluate (T&E), and accept 150kW and 50kW power generation units for the Kuwait Air Defense System.
Centers for Disease Control and Prevention (CDC) - National Institute for Occupational Safety and Health (NIOSH)	Computer Assisted Chest Radiograph Reader System CARRS II	The long term objective of this project was to provide a means for reducing inter- and intra-reader variability diagnosing interstitial lung diseases in chest radiographs through a computer-based system for analyzing digital images. Kestrel evolved its first generation solution into the Computer-Assisted Chest Radiograph Reader System (CARRS), which applied recognized principles in the psychophysics regarding human vision understanding and incorporated neural network-based image analysis and integrated with an appropriate graphical user interface. Advances in digital image processing and classification techniques made CARRS feasible for meeting screening, research and development, and clinical requirements.
Centers for Disease Control and Prevention (CDC) - National Institute for Occupational Safety and Health (NIOSH)	Computer Assisted Chest Radiograph (X-ray) Interpretation	The specific objective of this project was to apply Kestrel's existing, state-of-the-art Computer-Assisted Medical Digital Image Analysis System (CAMDIAS) to chest radiograph (X-ray) interpretation and to demonstrate that this physician-interactive system can reduce inter-and intra-reader variability. The ultimate goal of the project was to increase the usefulness of chest radiographs for 1) medical surveillance of workers exposed to agents that might cause lung disease, 2) epidemiological studies that require interpretation of large numbers radiographs and collection of data for quantitative analysis, and 3) documenting eligibility for medical compensation programs.
Computer Science Corp	Environmental Assessment - Roving Sands Military Exercise II	In support of Roving Sands 1997, the world's largest air & missile defense exercise, Kestrel collected pre-exercise/pre-engagement multispectral imagery of the exercise area to establish an environmental baseline. Subsequently, Kestrel collected post-exercise multispectral baseline imagery of the area, which was compared to the pre-exercise data to assess overall environmental impact.
Consumer Products - Skin Care Company	Skin Condition Image Analysis II	Kestrel developed and optimized its Artificial Intelligence software for automatically grading of skin condition.
Department of Energy	Micro, Low Cost, Airborne, Hyperspectral Imager	The Department of Energy needed a more affordable, airborne, micro-sized hyperspectral imager for use in its treaty verification and non-proliferation (weapons of mass destruction, WMD) remote sensing missions. This sensor needed to offer dramatic cost efficiency while retaining current state-of-the-art performance. Under this project a low cost, small size airborne hyper spectral imaging system was developed that met these needs. The same instrument will have application to a wide variety of agriculture, land uses and oceanographic surveys. The instrument was based on the technical advantages offered by Fourier transform imagers, transition of data management and processing technology from the Sandia National Laboratories' Interferometric Synthetic Aperture Radar System, and a unique utilization of existing commercial-off-the-shelf (COTS) hardware. A dual-spectrum sensor system that could operate over the 400 nm to 950 nm visible and near-infrared (VIS/NIR) and 1500 nm to 2700 nm shortwave infrared (SWIR) spectral bands with over 250 hyperspectral channels and 0.7 milliradians of spatial resolution was designed. During the first Phase of this project three objectives were accomplished that demonstrated the technical underpinnings for a low cost micro design. The first objective was to incorporate Department of Energy (DOE) operational requirements into existing commercial designs. The second objective was to demonstrate, by laboratory experiments, the underlying technology that transformed an existing Fourier system into a low-cost derivative. The final objective was to complete the conceptual design of a low cost hyperspectral imaging system.
Jamieson Science and Engineering	Advanced Clutter Suppression Techniques for Hyperspectral Imagery	Designed a test and evaluation (T&E) plan to measure the performance of Kestrel's Fourier Transform Hyperspectral Imager (FTHSI) to determine which generic background clutter mitigation algorithms were suitable for use with FTHSI data. Hyperspectral scene data was collected by the Kestrel FTHSI and evaluated with different scene clutter mitigation algorithms to determine optimal results.
Joslin Diabetes Center (Harvard Medical School affiliated)	Automated Micro-Aneurism Segmentation for Diabetic Retinopathy	Continuing its original work for the National Eye Institute, under this project Kestrel validated its computer software algorithm for automatically segmenting (detecting lesion and defining the margin) of microaneurysms (MAs) as caused by diabetic retinopathy. Kestrel developed and implemented software to screen digital retinal images and met the present clinical diagnosis standard of a minimum of 70% sensitivity and 70% specificity (as demonstrated in a blind test with images provided by and to Joslin). The standards used by Joslin include the Early Treatment Diabetic Retinopathy Study (ETDRS) seven standard 35-mm stereoscopic color fundus photographs and its three non-simultaneous 45-degree field of view stereoscopic digital color images (the standard modality for its Joslin Vision Network).
Lighthawk	Light Hawk Airborne Multi-Spectral Camera System	The scope of this effort included the design, fabrication, test, installation, and support for the flight test of an airborne multispectral camera system in a Cessna 206 aircraft. Kestrel furnished all parts, materials, and labor to build and incorporate the system into aircraft. Kestrel acted as Program Manager and was responsible for all aircraft modifications and FAA approvals and certifications (FAA Form 337s).
Lions Eye Bank	Donor Cornea Tissue Evaluation System	Kestrel designed a Donor Cornea Tissue Evaluation System in collaboration with the University of New Mexico Health Sciences Center.
Lockheed Martin	Coal-bed Gas Methane Application - Commercial Remote Sensing Satellite Program	Multispectral imagery is limited in its ability to define vegetation communities and stressed vegetation. However, hyperspectral imagery (e.g., from Kestrel's Fourier transform hyperspectral imaging system) has the potential to delineate plant communities and stressed vegetation to a level sufficient to identify new areas where methane gas seepage may be occurring. This hyperspectral imaging study included, but was not limited to, vegetation stress, alterations of soils, and geological fracture analysis. Four small study areas were examined to develop the application. This analysis included the surface and sub-surface geology and the production practices found in coal-bed gas-methane fields. The purpose of this study was to characterize the hyperspectral signatures for coal fields with high potential for methane gas production. The application was tested by extrapolating to the remainder of the coal outcrop areas of the San Juan Basin.
Lockheed Martin	Southern Pine Beetle Application - Stennis Operation	Kestrel's Fourier transform hyperspectral imagery was acquired over the Stennis Space Center to test for potential southern pine beetle infestation sites.
Lockheed Martin	Environmental Baseline Application - US Army	Kestrel deployed its AirCam multispectral imager over a site near Wagon Mound, NM to collect data over a site environmentally similar to a US military exercise location. The imagery was used to develop a corresponding environmental baseline application of vegetation, habitat, soils, etc., that could be used by the military to later determine the extent of environmental damage that had occurred during large military exercises. The imagery was also compared to the utility of NASA satellite imagery for the same purpose.
Lovelace Respiratory Research Institute	Multispectral Microscope Imaging of Bacillus Bacterial Spores	Performed an experiment to determine whether live versus dead bacillus bacterial spores presented spatially-resolved spectral signatures. Captured microscope images of bacilli in various states (defined by Lovelace Respiratory Research Institute), background noise (dust particles, etc.), and other specimens of interest.
Michigan Technological University	Upper Atmosphere Turbulence System	Kestrel designed, constructed, and tested a balloon-borne anemometer system to measure atmospheric turbulence based on fine temperature data. The system was designed to eliminate artifacts due to making strong assumptions about the turbulence statistics, and due to the placement of the instrument package. A second instrument package to relay the raw data to a ground-based recording system was constructed and incorporated into the balloon-borne system. The balloon data package included a "standard" US military/US Air Force atmospheric turbulence instrument package suspended the- appropriate distance below the envelope in order to simultaneously collect "traditional" data for comparison purposes. Two balloon systems were produced and flown.
Missile Defense Agency	Space Object Detection, Identification and Tracking	Kestrel investigated the advantages of a distorted grating wavefront curvature sensor (DG-WFS) in combination with phase diversity and curvature sensing as a means for aligning large, segmented, space observing telescopes. Kestrel's approach, measured the intensity in two planes through which the wavefront propagates, and the difference gave a measure of the axial intensity gradient. This approach differed from conventional phase diversity analyses, which uses information primarily around the image plane and requires an iterative solution. By using the pupil plane Kestrel was able to work with a linear relationship, the Intensity Transport Equation. This technique is the basis for wavefront curvature sensing and needs only two image planes to calculate the wavefront and required no additional sensors. In addition, each data point in the wavefront is independent, so the calculation can be performed in parallel, making very high speed throughput feasible. A novel use of boundary conditions also meant that there was no requirement for separate sensors to estimate wavefront tilt at the pupil edge. These differences distinguished Kestrel's approach from earlier attempts at real-time, high-speed, and highly accurate pointing and tracking to support Space Object Detection, Identification and Tracking.
Missile Defense Agency	High Temperature Spectral Resolution Laser Beam Diagnostics	Kestrel, along with our collaborator Boeing SVS developed an infrared hyperspectral imagine technology to remotely sense the spatially distributed spectral signature fingerprints of a high energy laser against a distant target. This technology' offered .high temporal and spatial resolution for the purpose of measuring laser beam characteristics during airborne High Energy Laser (HEL) engagements, while simultaneously collecting target response (damage assessment) data. Kestrel defined the requirements for the hyperspectral sensor and the use of an MWIR hyperspectral imager to address the key technical issues. This technology was coupled with an existing sensor pointing and tracking system from Boeing. Kestrel showed that its sensor system can be used to observe the HEL beam size, energy distribution, location on the target, and target spectral response, with centimeter spatial distribution at rates in excess of 30 full samples per second. The techniques used took advantage of Kestrel's developed IR hyperspectral imaging to simultaneous observe reflected HEL energy with the IR radiation and absorption associated with the interaction of the beam on, and with the target. The sensor offered a spectral, spatial, and signature map created at milli-second rates with immediate presentation of the engagement results in real time.
Missile Defense Agency	High Efficiency Spectral Imager	Kestrel demonstrated a new design that increases the throughput of a Fourier transform hyperspectral interferometric imager by a factor of two while retaining the spectral and spatial resolution of the original instrument. The new technique overcomes an existing limitation in virtually all interferometric instruments by recovering the light lost through the beam splitter, thereby providing increased signal to noise, improved optical efficiency, and greatly extended dynamic range. Such improvements open up multiple new applications for optical spectral imaging, both in military and commercial market segments.
Missile Defense Agency	Chromatic Correction for Multi- and Hyperspectral Images	This project modeled a distorted grating into an existing analytical description for use in Multispectral (MSI) and Hyperspectral (HSI) system. The development of this modeling tool serves as the method for conducting parametric analysis of the improvements that can be expected from adapting a distorted grating chromatic corrector to real world instruments.
Missile Defense Agency	Ultraspectral Imager Program	The Missile Defense Agency (MDA) (formerly the Ballistic Missile Defense Organization (BMDO)) required an imaging spectral radiometer with wavenumber spectral resolution and milliradian spatial resolution that operates over the 8nm to 12nm long-wave infrared, (LWIR), 3nm to 5nm mid-wave infrared (MWIR), and 0.5nm to 1.05nm visible and near-infrared (VIS/NIR) bands. This research was based on a spatially modulated Fourier transform spectral imager to take advantage of the inherent benefits in these devices in the MWIR and L WIR. The research discovered optical techniques that could be merged with a Sagnac interferometer to create conceptual designs for a L WIR imaging spectrometer that has a 0.4 cm-1 spectral resolution, a 2 mrad IFOV, and a 0.25 rad FOV using a commercially available HgCdTe detector. These same techniques produced an MWIR imaging spectrometer with 1.5 cm-1 spectral resolution, 2 mrad IFOV, and a 0.25 FOV, again using a commercial HgCdTe detector. Application of the techniques to the VIS/NIR spectrum produced spectral resolutions on the order of 10 cm-1. Key to the exceptional LWIR and MWIR performance was the insertion of a paired set of gratings into the long leg of the interferometer. By varying the grating parameters the interferometer could be tailored to a specific band and spectral resolution significantly increased to the best spectral resolution. A side benefit to one of the optical folding configurations that was used to expand the spectral resolution were methods for electronically varying the spectral resolution, which allowed a spectral zoom mode to be developed. This work concluded with the creation of conceptual designs for two different airborne-based sensor systems that had wavenumber ultraspectral imaging capability in the MWIR and LWIR spectrums. A sensor of this caliber was shown to have wide Government and Commercial interest with the unique ability to image a feature's molecular absorption signature.
National Aeronautics and Space Administration (NASA)	High Temperature Ultraspectral Imaging - Propulsion System "Health" Monitoring	This project developed a new technology that could generate spatially-distributed, wave number-resolution spectral signatures or "fingerprints" to determine when rocket ignitions deviate from nominal. The real-time purpose is to note these abnormalities in time to abort a launch or prevent initiation of auxiliary or secondary functions. The concept uses the rocket's exhaust as a source of performance and health data that had not previously been fully exploited. Kestrel demonstrated that a new technology called Ultraspectral Imaging (USI) could be used as a imager to observe the mid-wave infrared (MWIR) to long-wave infrared (LWIR) spectral bandwidths and could be used to measure the varying flow and spatial location for a liquid rocket engine and that these data could be processed to create a non-intrusive rocket health and performance monitoring system.
National Aeronautics and Space Administration (NASA)	High Temperature Ultraspectral Imaging -Propulsion System "Health" Monitoring II	During this project, Kestrel performed a test of its Ultraspectral Imager (USI) in the mid-wave infrared (MWIR) portion of the spectrum to optimize requirements. Kestrel tested the USI on a single downstream station to demonstrate system performance. Kestrel subsequently identified performance opportunities for requirements input and testing situations for a commercial system that could be designed to collect USI data from multiple down range stations to maximize available signatures and determine on-going/in-flight rocket performance.
National Geospatial-Intelligence Agency	Dual Use of Airborne Hyperspectral Imagery	The objective of this program was to generate a highly attributed geospatial terrain database using hyperspectral imagery (HSI) as collected by Kestrel's Fourier Transform Hyperspectral Imager (FTHSI). To the greatest extent possible, the data collected included features that met or exceeded existing national map product accuracy and content standards. These data conformed to data collection and format standards as agreed upon by both the Terrain Modeling Project Office (TMPO) and Kestrel. Digital terrain feature data was provided to TMPO for product comparison and evaluation to determine how well HSI-derived terrain feature data supports DoD modeling and simulation requirements.
National Institutes of Health/National Eye Institute	Multispectral Fundus Imaging System	Advances in underlying sensor technology has resulted in the corresponding evolution of sophisticated hardware for high resolution multispectral imaging applicable to a broad class of applications. Multispectral imaging and associated cutting edge multispectral image and data fusion processing software have led to the realization of techniques that add significantly to the ability of identify and characterize the nature of features of everything from locations on the earth as seen from space to the identification of very small microscopic objects based upon their multispectral and spatial signatures. A multispectral image fusion system was designed to provide additional diagnostic information to a clinical or research ophthalmologist. Using state-or-the-are multispectral image capture systems, advanced multispectral image fusion techniques inspired by the human vision system, and neural network feature classification algorithms, a Multispectral Image Fusion System for detection and identification of ocular pathological features was developed.
National Institutes of Health/National Eye Institute	Stereoscopic (3D) Grading System for Age-Related Macular Degeneration (AMD) II	Kestrel developed a computer-based grading system for both 2 dimension (2D) and stereo (resulting in 3 dimensional or 3D) visualization and analysis of digital fundus images with a user-selectable level of automation geared toward clinical application and for use in epidemiological studies of age-related macular degeneration (AMD). The specific aims of this project were to develop and validate a commercial-quality stereo-based AMD grading system (AMD-GS) and to perform a longitudinal study that demonstrated clinical and research capabilities not available through conventional light-box-based methods. The ability to display digital images in a familiar 3D format provided great enhancement to traditional methods without sacrificing the benefits of manual grading. Kestrel enhanced its first generation Stereo AMD-GS through an evolutionary path that provided improvements in stages. Initial enhancement included automatic registration of time series images to facilitate longitudinal study and user-directed semi-automatic segmentation of lesions. The final stage was able to automatically segment the entire image at once without user direction. The various levels of system automation was designed to allow a user a choice as to the most appropriate tool for a given patient. The final system was demonstrated and validated through a retrospective longitudinal study of 40 subjects.
National Institutes of Health/National Eye Institute	Optical Imaging Device of Retinal Function (OID-RF)	Kestrel developed and demonstrated an optical imaging device of retinal function (OID-RF) to measure the change in the reflectance of the retinal that occurs as a result of the changes in oxyhemoglobin saturation in response to the metabolic demands of active neurons. The OID-RF exploits recent technological advances in high sensitivity imaging detectors and multispectral imaging. The OID-RF produces objective, quantitative, and localized information in the form of a functional image of neuronal activity across the retina. The project assembled an OID-RF visualization system and tested it by detecting and measuring pathologies of the optic nerve head (ONH) and nerve fiber layer (NFL) in the research and clinical study of glaucoma. The specific aims demonstrated the value of functional imaging for detecting stimulus dependent retinal activity using spectral imaging in optical wavelengths and produced 2D OID-RF maps for normal subjects and glaucomatous eyes, and demonstrated a path for a commercial clinical device.
National Institutes of Health/National Eye Institute	Donor Cornea Screening System (DCSS)	Prior to surgery a donated cornea must be screened for a number of pathologies and its history examined for contra-indicated events, such as refractive surgery, which makes a donated cornea undesirable for transplant. Current techniques, such as topological measurement, produce significant errors, resulting in the rejection of viable corneas, or worse the unknown use of modified corneas. With over a million corneal refractive procedures (e.g., laser-assisted in-situ keratomileusis (LASIK)) performed yearly, the potential of these modified corneas entering into the tissue bank system has increased significantly.
National Institutes of Health/National Eye Institute	Spectral Fundus Imager (SR-SFI)	Kestrel designed a retinal imaging system that combined multispectral fundus imaging capabilities with adaptive optics and image deconvolution capabilities to provide microscopic, in-vivo images of the retina to resolutions not previously obtainable. Previous systems, such as Williams University of Rochester, had developed adaptive optics systems for laboratory use and had achieved 2-4um resolution images of the retina, but not for in-vivo clinical use. Kestrel's solution used a combination of enabling technologies, such as wavefront sensing, adaptive optics, and image deconvolution, to solve the technical issues that had limited the use of adaptive optics to research instruments. In addition, Kestrel increased the field of view (FOV) from 1-degree in standard laboratory systems to 20-degrees for use in the clinic. Thus, Kestrel's solution can collect the same area of image in a single flash exposure compared to the 50 or more frames required by competing systems.
National Institutes of Health/National Eye Institute	Optical Imaging Device of Retinal Function (OID-RF) II	Detection of nerve fiber loss in diseases like glaucoma presents a significant challenge to ophthalmologists. Visual field tests, the standard for detecting nerve fiber loss, are low in sensitivity requiring greater than 40% nerve fiber loss before detection. Kestrel demonstrated the ability of its Optical Imaging Device of Retinal Function (OID-RF) to objectively record retinal function with nerve fiber losses significantly less than 40%. As OID-RF is a non-invasive medical device it is ideal for clinical applications.
National Institutes of Health/National Eye Institute	Automated Micro-Aneurism	Kestrel developed, tested and tuned a computer automated micro-aneurism segmentation algorithm with our collaborator the Joslin Diabetes Center (Harvard Medical School affiliated). See Joslin Diabetes Center section, Project 0126 for additional information.
National Institutes of Health/National Eye Institute	Optimized Retinal Imager (ORI)	Kestrel developed a new, improved and affordable high resolution and high contrast digital camera optimized for eye doctors (both optometrists and ophthalmologists).
National Institutes of Health/National Eye Institute	Donor Cornea Characterization System (DCCS) II	Kestrel developed additional automated software capabilities to screen for a number of cornea disease pathologies.
National Institutes of Health/National Eye Institute	Stereoscopic (3D) Grading System for Age-related Macular Degeneration (AMD)	Kestrel developed a Stereoscopic Grading System for age-related macular degeneration (AMD) using an affordable stereoscopic visualization device and modern computer software techniques. The objective of this project was to develop a computer-aided system for grading retinal images which display volumetric lesions, and to increase efficiency and quantification of the diagnosis while reducing reader variability. Revolutionary advances in technology made it possible to move from the exiting "light-table and stereoscopic viewing" to the computer screen and still meet the University of Wisconsin grading standards.
National Institutes of Health/National Eye Institute	Digital Fundus Image Diagnostic System II	This project developed the Digital Fundus Image Diagnostic System (DFIDS) capability. DFIDS is a computer-aided software assistant to aid clinicians in the grading of fundus/retinal images for the diagnosis of eye disease. DFIDS applies three key technologies: 1) high performance computing, 2) state-of-the-art interfaces, and 3) artificial neural networks for automated image processing. DFIDS integrates these computer-based technologies with the unmatched attributes for pattern recognition of biological vision system.
National Medical Test Bed (NMTB) - Department of the Army - Loma Linda University Medical Center	Hyperspectral Imaging Disease Modality	Diabetic retinopathy and age-related macular degeneration are the two leading causes of blindness in the US. Hyperspectral imaging with digital image processing has the potential to provide a reasonable cost, minimally non-invasive, highly informative data source to permit more effective patient care.
National Medical Test Bed (NMTB) - Department of the Army - Loma Linda University Medical Center	Hyperspectral Imaging Disease Modality - Improved	Today's medical marketplace, particularly as it affects the development, manufacturing, and sale of medical technologies, is strongly influenced not only by the actual medical needs of patients but also by the recent trends in healthcare economics. The explosive rate of technological advances in imaging science, along with the changing healthcare environment, clearly indicate that the success of new imaging technologies will depend on their ability to contain the cost of the diagnostic process while broadening the scope of the diagnostic information provided. One of the hypotheses of significant interest of this particular project was to consider the value of definitive automated diagnosis compared to the costs of delayed treatments and whether more modern processes (e.g., preventive medicine) could significantly reduce costs.
Navy Sea Systems Command	Wavefront Sensing Applications in Maritime Environments	Kestrel demonstrated that its Distorted Grating Wavefront Sensor (DG-WFS) was able to successfully reconstruct wavefronts in severely scintillated condition resemblance of maritime environments in which existing Shack-Hartmann (SH) sensors typically fail to give a good reconstruction. The results of this study also considered the amplitude effects of scattering aerosol particles on the relative performance of a SH and DGWFS system. Two wavefront sensors were convened into a single instrument with synchronized detectors allowing frame-to-frame and statistical comparison of the performance. A fog chamber was constructed that used water-based aerosols to provide the simulation of the maritime environment. The visibility through the fog was varied from 1.2km (good visibility) to 200m (poor).

New Mexico Dept of Forestry	Fire Management Imaging System II	This effort determined the feasibility of developing a relatively inexpensive airborne imaging system for fire management and mapping efforts that could be easily used by both fire response teams and forestry personnel for post-fire damage assessment...Kestrel collected visible and various infrared imagery over active forest fire areas.. Sufficient data was collected to accurately locate a fire's perimeter and to identify hot spots inside the fire area needing suppression..

North Atlantic Treaty Organization (NATO)	Ultra-high Resolution Adaptive Optics-based System for Medical Applications	This Program developed an ultra-high-resolution digital retinal imaging system to aid the early detection of many retinal diseases and specifically for early diabetic/diabetes diagnostics. Such unique diagnostic information would allow physicians to start treatment during the earliest stages of a disease, which not only improves the result of treatment, but also leads to substantial economy of financial resources. The apparatus and technology also has a dual use; in addition to its use in fundus imaging disease detection, the very accurate visual refraction error measurement has utility in eye surgery applications. The device may also find an application in routine optometrist practices as well, where many patients first go when they have a vision problem. The technology can be modified for use in scanning laser ophthalmoscopes (SLO) and confocal microscopes.

Pfizer Inc.	Hyperspectral Fundus Imager for Small Animals	Kestrel designed, fabricated and delivered a hyperspectral fundus imager (HSFI) that allowed it to operate with a small animal pupil. Kestrel collaborated with Pfizer to determine the appropriate pupil size for Pfizer's application and characterized system by conducting live subject testing.
Pfizer Inc.	Computer -Aided Evaluation of Choroidal Neovascularization (CNV)	Kestrel developed a computer-aided choroidal neovascularation (CNV) system for automatically evaluating multi-mode retinal images. Kestrel assessed the utility of fill-rate characteristics for automatically classifying the choroidal neovascularization (CNV) lesion and determining their margins. This objective has important significance to achieving the ultimate goal of computer-based objective "grading" of clinical trial images for drug concentrations, effectiveness, and drug toxicity resemblance /testing.
Rensselaer Polytechnic Institute (RPI)	Scheimpflug Camera Evaluation	Kestrel Corporation designed and determined the specifications for a digital camera which would capture and store 40 frames per second at 1000 pixels by 1000 pixels for a digital Scheimpflug Camera. Kestrels designed was based upon testing the sensitivity of measurements made in-situ of a Scheimpflug image collection system and provided data.
Tyler Camera Systems	Tyler SkyGyro Helicopter Camera System	This project prepared the electronics, (hardware and software systems), on-site test and measurement activities, as well as follow-up on analysis and design activities for the Tyler SkyGyro Camera System for use various FAA certified helicopters for both search and rescue, police surveillance, and military applications
Univ. of Iowa	Dual Imager Optical Assembly	Kestrel designed and fabricated a Dual Imager Optical Assembly (DIOA) to directly attach to a standard FF4 Fundus Imager. This assembly allowed a simultaneous attachment and operation of a Standard 35-mm film back mechanism and an Escalon CFA Digital Imager for direct image comparison.
Vanguard Research, Inc.	Atmospheric Turbulence for Directed Energy Applications	This project addressed the problem of utilizing spatial- hyperspectral imaging capabilities of space-borne sensors to detect and characterize regions of atmospheric turbulence and cirrus cloud clutter, which can impact the employment and/or performance of space-based laser and airborne high energy laser weapons systems. This work focused on the worldwide detection, characterization and mapping of atmospheric turbulence and cirrus clouds by the inclusion of spectral and hyperspectral imagers operated in the UV to MWIR spectral range in a virtual triangulation geometry. Properly deployed spectral and hyperspectral sensors allow "altitude sounding" of atmospheric clutter from turbulence and cirrus clouds. The triangulation geometry provides precise altitude determination by cross correlation of the backscatter signals. The combination of altitude and Fourier-space background spectral discrimination provide an altitude resolved measurement of atmospheric clutter from clear air turbulence and from cirrus clouds, both of which can affect performance of a SBL (Space-Based Laser) and an Airborne Ballistic Laser (ABL) system.
World Bank	Developing Nation Remote Sensing Program	Kestrel supported the World Banks' efforts in environmental assessment, natural resource census and international development project planning for the countries of Haiti and Belize. Kestrel's hyperspectral imaging and analysis capabilities proved invaluable in helping to quickly and cost-effectively completing the necessary inventories and surveys required for these countries.

.: Our Motto is: