RESEARCH INTERESTS

• modeling of physical phenomenology for synthetic image generation
• development of novel ground truth collection devices and techniques
• field and laboratory longwave infrared data collection
• temperature-emissivity separation techniques
• temperature extraction algorithms for multi- and hyperspectral image data
• algorithm development for the detection, identification and quantification of gaseous effluents
• radiation propagation models
• atmospheric correction techniques
• multi-sensor data merging
• spectral/feature band optimization methods
• three-dimensional geometry extraction from multi-view imagery and integration with image simulation and geographic information systems

CURRENT-FUNDED RESEARCH CONTRACTS/GRANTS/GIFTS

Department of Energy

RIT will provide directed development support and interface definition to allow the DIRSIG environment to serve as an image-rendering engine for the three-dimensional physics-based modeling being carried out by the DOE. As DIRSIG is already a first-principles physics-based environment for optical interactions at the target surface, propagation through the atmosphere, interactions with the mechanics of the sensing platform, and sensor response and artifact modeling for most available types of imaging systems in use today, the further opening of this environment to allow input from the DOE modeling community with respect to phenomenological scenarios involving nuclear nonproliferation would allow a full, end-to-end system for which the imaging product produced can serve as data for a variety of purposes such as algorithm development and testing, system design trade-off studies, and scenario-driven modality selection. Given current DIRSIG capabilities to model passive systems such as spectral and hyperspectral devices, thermal infrared sensors as well as active systems such as LIDAR and RADAR, and the physical process modeling available from current DOE researchers, this integration would enable very comprehensive system trade studies and future sensor system simulations directed at the complex nonproliferation problem. As part of this effort, RIT will:
1. Enhance the current capabilities of DIRSIG for the prediction of target surface temperatures by incorporating the MuSES infrared signature prediction model. MuSES will allow DIRSIG to model active targets in a complex meteorological environment with far greater fidelity than is currently possible. The interface of this model will be accomplished in such a way that other current or future temperature prediction codes may be easily used in its place.
2. Improve the current automated vehicular traffic capabilities in DIRSIG by allowing for anomalous movements to occur such as unusual stops or off-road movements. RIT sees this as advantageous for a dynamic persistent surveillance scenario.
3. Develop a simplified flight-planning interface to allow the modeler to quickly plan a collection scenario using a UAV type collection platform. DIRSIG can currently follow a flight plan but this requires a great deal of user interaction and modification of configuration files. RIT will evaluate existing flight planning methods in use by the DOE and the community at large and apply a best effort at integrating these existing tools. If this is not possible, a custom approach will be developed that will emulate the best features found in the existing approaches.
4. Create automatic scene building for DIRSIG based upon contextual inputs including geographic information system descriptive layers and/or multi-view airborne imagery collections. The intent of this task is to greatly accelerate and automate the most time consuming task in the DIRSIG workflow, the scene creation process.

RIT Project Number:	30958
Funding:	$700,000
Period of Performance:	September 19, 2008 to September 18, 2012

Department of Energy

A physics-based target space approach to small target temperature determination is being developed that will allow targets that are the same size as a thermal infrared sensor ground sampling distance to be interrogated for subsequent analysis.

Current approaches to this problem employ traditional frequency domain restoration techniques that are based upon an estimate of the point spread function of the sensing system. This approach, however, is subject to error for a number of reasons. The point spread function of the entire collection system includes spatial blurring due to atmospheric scattering and sensor motion. These parameters are typically not included in the solution and as such the process produce errant answers. Additionally, these traditional techniques work for targets that occupy several pixels in a scene, not for single or subpixel sized targets.

The proposed approach will use the DIRSIG synthetic image modeling code to produce hundreds or thousands of possible candidate images of the target/background under examination. Different positioning of the target relative to the sensor sampling array will be generated as well as varying background and target temperature combination. This create synthetic target space will be developed so as to include the actual scenario encountered. Image matching techniques will be used to compare the actual thermal image to each synthetic image in the target space to find the best match and therefore the most likely set of physical parameters.

A complete validation and verification of the developed technique will be carried out using modeled data, data collected from a fixed imaging platform with full control over all target space parameters, and in a real-world airborne image collection experiment. Data will be collected so that an assessment of accuracy of this proposed technique may be assessed.

RIT Project Number:	30830
Funding:	$656,751
Period of Performance:	September 19, 2008 to September 18, 2012

Central Inteligence Agency IC Post Doctorate Fellowship Program (FY08)

This fellowship will focus on the development of novel experimental techniques and complementary modeling tools to accurately predict the polarized bi-directional reflectance distribution function (BRDF) of a variety of materials of interest to the remote sensing community. In addition, the research will develop a collection methodology for measuring polarized BRDF remotely for denied targets. The experimental techniques will take advantage of an image based approach to BRDF measurement with emphasis on deriving the polarized Mueller scattering matrix form of the BRDF.

RIT Project Number:	30866
Funding:	$239,978
Period of Performance:	September 4, 2008 to October 20, 2010

Savannah River National Laboratory / Department of Energy

RIT is collaborating with the Savannah River National Laboratory (SRNL) to extend the capabilities of the ALGE hydrodynamic code to include simulations of surface ice formation and melting in cooling lakes that receive heated effluent from nuclear reactors operating in cold climates and collect the data needed to validate this extended version of the ALGE code at suitable locations in the northern US or Canada. SRNL uses the ALGE code to perform technical analyses of heat-generating industrial facilities for DOE and other government agencies. At present, ALGE simulations of cooling lakes and other bodies of water being used to dissipate waste heat from nuclear reactors are restricted to ice-free conditions. This restriction prevents imagery analysis for more than one-half the year at some northern sites. The extension to the ALGE code proposed here will allow year-round applications of the code which will result in more rapid completion of technical analyses.

RIT Project Number:	30789
Funding:	$849,971
Period of Performance:	January 22, 2008 to January 21, 2011

PAST-FUNDED RESEARCH CONTRACTS/GRANTS/GIFTS

Central Inteligence Agency IC Post Doctorate Fellowship Program (FY06)

This fellowship will perform a comparative analysis of ship-based thermal infrared spectroradiometer measurements against MODTRAN simulations incorporating humidity dependent aerosol nucleation effects.

It is commonly assumed that scattering effects are negligible in the thermal infrared region. This, however, is based on standard atmospheric and aerosol models that have particle size distributions that are heavily weighted and have maximum concentrations in the sub-micron range. While particle concentrations in the super-micron range are orders of magnitude smaller than its submicron counterparts, certain conditions of high humidity may significantly increase their concentrations to levels that may impart scattering effects in the thermal infrared regions.

This hypothesis will be tested using the MODTRAN radiative transfer model (Berk 1989) coupled with the NOVAM - Navy Oceanic Vertical Aerosol Model (Gatham 1993). Given the nature of the data set from Explorer of the Seas ARM facility, this is a logical approach since oceanic aerosols will be inherent to the environment of the surface based measurements. The NOVAM model will be used to obtain first order estimates of potential aerosol effects and indicate possible aerosol nucleation mechanisms describing the concentrations of larger hygroscopic aerosols and their particle size distributions (Leeuw 1992). Currently, the NOVAM model supports three mode radii (peak of the particle size distribution) of 0.03, 0.24, and 2.0 microns. Modifications to the model to include larger mode radii will be investigated to support the thermal scattering hypothesis.

RIT Project Number:	30632
Funding:	$239,926
Period of Performance:	August 17, 2006 to February 16, 2009

DP Instruments / Environmental Protection Agency

RIT will perform a system integration of the DP Instruments MARLIN high-speed FTIR spectrometer with a FLIR Systems GasFindIR camera to create a bore-sighted or common-optic system capable of imaging effluent gas plumes and making spectral signature measurements at a central location in the field of view. RIT will further perform a proof-of-concept field study with this system against controlled gas releases. Once the data is collected, RIT will continue to analyze this data by producing an analysis software capability to perform gas identification and volume/rate-of-release estimations.

RIT Project Number:	30603/30652
Funding:	$47,400
Period of Performance:	May 1, 2006 to November 30, 2006

Savannah River National Laboratory / Department of Energy

RIT will provide support to gain insight into the phenomenology that influences the radiance field leaving the interior of a mechanical-draft cooling tower (MDCT). The DIRSIG modeling capability will be enhanced such that the models produced reflect, as accurately as possible, the actual data gathered with real airborne infrared imaging systems. These modeling efforts will focus on the phenomenology associated with "cavern-like" targets with numerous material types internally contained. This effort will be cyclical in nature with modeling approaches continually modified based on newly discovered phenomenology observed in real image data. The desired outcome of the modeling will be accurate internal-element emissivities and temperatures for the components that comprise the cooling tower for use with an external process model developed by Savannah River National Laboratory.

RIT Project Number:	30571
Funding:	$457,820
Period of Performance:	February 2, 2006 to June 30, 2009

MITRE Corporation / Central Intelligence Agency

RIT will provide support to the government as a critical participant in the National Signatures Program (NSP) working group monthly meetings. The NSP is responsible for the establishment of spectral measurements standards for solid, liquid, and gaseous materials. Standards in the form of measurement methodologies, formats, metadata, etc. fall within the purview of this group. RIT has been asked to be the representative for the government organization that funds and oversees this program to provide guidance, experience, and critical review to/of this programs efforts and to report back to the supporting organization with impressions and advice.

RIT Project Number:	30517/30654
Funding:	$56,289
Period of Performance:	February 1, 2005 to July 31, 2007

National Reconnaissance Office (NRO) - LASS 2004/05 (CLIN 002)

RIT will provide support through the Laboratory for Advanced Spectral Sensing for the development of an enhanced full spectrum material reflectance database to support more realistic simulations in the DIRSIG model. Current database holdings will be evaluated and test plan developed and executed to provide additional spectra to fill in voids in the current database.

RIT Project Number:	30483
Funding:	$40,000
Period of Performance:	November 1, 2004 to May 31, 2006

Mission Research Corporation / Air Force Research Laboratory (WPAFB)

RIT will be providing modeling support in the development of calibration test targets for longwave infrared (IR) test chamber characterization. RIT will utilize its DIRSIG model to represent the interior environment of the IR test chamber located at the WPAFB facility and the potential calibration targets that will be used, performing trade-off and optimization studies as top the efficacy of the designs to choose the proper target to manufacture. RIT will also place certain targets of interest into a cluttered IR background to assist in the design of future chambers with enhanced background capability.

RIT Project Number:	30463
Funding:	$150,009
Period of Performance:	June 15, 2004 to November 11, 2005

ITT Industries / Space Systems Division (ICAP)

RIT will provide support for the development of new and/or evaluation of existing temperature extraction methodologies for infrared imagery. Techniques using a single band, multiple band, or multispectral/hyperspectral image data will be within the scope of this task. RIT will provide modeling support to produce synthetic imagery for evaluation of these algorithms using DIRSIG as requested. This support will may be in the form of assistance in the development of specific sensor models, specific imaging scenarios, and/or to provide a robust data set as a test-bed for new and existing algorithms. RIT will attend community meetings to report on the algorithm and modeling tasks as well as to provide critical review and/or insight into other proposed methodologies.

RIT Project Number:	30459
Funding:	$188,873
Period of Performance:	May 1, 2004 to September 30, 2006

New York State Office of Cyber Security and Critical Infrastructure Coordination

RIT will be providing a software tool that will allow the NYS OCSCIC to selectively degrade their online statewide library of aerial photographs to protect critical infrastructure information. As part of the Homeland Security initiative afoot in New York State, the OCSCIC desires a tool that would let them selectively degrade the resolution of their archives of air photos that are available to the public so that potential organizations that would aspire to do harm to these facilities do not have high quality data available to them, while maintaining the quality of the data in other regions for legitimate users. RIT will develop this software tool to key off of OSCCIC supplied shape files and apply the degradation to the library of imagery.

RIT Project Number:	33315
Funding:	$21,039
Period of Performance:	April 4, 2004 to September 4, 2004

BAE Systems

RIT will be utilizing funds provided to the current capital campaign by BAE Systems to support a graduate student in Imaging Science. The student will work on a project of interest to both RIT and BAE Systems for their thesis research project. The student will spend a Summer quarter working at BAE Systems with their scientist and engineers to gain real-world experience during their academic work.

RIT Project Number:	20157
Funding:	$34,306
Period of Performance:	December 1, 2003 to November 30, 2004

Pictometry International Corporation / CEIS

RIT is determining the feasibility of automated tie point identification on numerous oblique air photos collected with the Pictometry system. Once a feasible approach is identified, RIT will be developing prototype software to assist Pictometry in this now largely manual process. This research was funded in part by CEIS, a NYSTAR-designated Center for Advanced Technology.

RIT Project Number:	36291/C3320/33320
Funding:	$26,300
Period of Performance:	September 5, 2003 to September 4, 2004

LaSen, Incorporated / Army Research Laboratory

RIT performed two (2) tasks for LaSen, Incorporated in support of their aircraft-based LIDAR system. RIT investigated the feasibility of modeling a LaSen midwave infrared LIDAR system using DIRSIG with particular attention to spectral resolution, pulse generation, ground resolution, and atmospheric interaction. In addition, RIT made several high-spectral resolution reflectance measurements in the 3 to 5 micron region to aid in current exploitation task being carried out by the sponsor.

RIT Project Number:	30903
Funding:	$7,000
Period of Performance:	May 1, 2003 to September 30, 2003

General Dynamics / National Air Intelligence Center

RIT is supporting the NAIC Spectral Exploitation Center (NSEC) Spectral Library Support initiative by developing laboratory and field spectral library protocols to assure high quality spectral measurements are provide to the library expansion task. In addition, RIT personnel and students are measuring the spectral signatures of materials supplied by the library expansion team as well as those encountered during field exercises using laboratory and field spectrometers. Spectral reflectance measurements are being made from 0.35 through 20 microns for all materials.

RIT Project Number:	30340
Funding:	$470,540
Period of Performance:	October 16, 2002 to September 20, 2004