Scalable Fidelity - Performance IR Scene Generation Technologies for the Sensor Analyst

JRM Technologies proposes to develop a comprehensive, PC-based, high-fidelity physics-based IR scene generation system as a reliable alternative to field data for the evaluation of emerging IR sensor designs. In order to provide the sensor analyst a complete end-to-end product that will afford streamlined, but highly-realistic IR scenes of targets in difficult terrain settings, JRM will make substantial enhancements to its hyper-spectral and near-IR scene generation tools currently being developed under related AFRL and US Army CECOM Phase II SBIRs. The enhanced design will include key technology components that address the challenges in data adequacy, modeling and performance that are needed to support sensor system parameter trade studies. They include (1) a Material Classifier--a robust, high-resolution signature-predictive material classification algorithm that uses high-resolution remote-sensed imagery (RGB photo, multi- or hyper-spectral data) to "reverse synthesize" the surface materials, or pixel-mixture of materials that give rise to that pass-band signature in a Bayesian maximum-likelihood estimator; (2) 3D High Resolution Database Building Tool--a GUI-based commercial 3D database generation tool (enhanced TerraTools) for construction of highly realistic geo-specific 3D terrain backgrounds from GIS data, including trees and vegetation modeled as individual-branched/leaved plants; (3) Scalable Fidelity-Performance Thermal Models-advanced 3D thermal solvers for complex targets and backgrounds, spanning the full range of signature performance from highest-fidelity non-real-time, to high-fidelity near-real-time to medium-fidelity real-time; (4) BRDF-based spectral directional radiance synthesis; (5) validated, optimized Modtran 4.0-based spectral irradiance and atmospheric propagation modeling in 3D-modeled weather; (6) scalable performance-resolution scene rendering options, including optimized non-real-time high-resolution ray-tracing, and real-time medium resolution PC-based graphics rendering; and (7) an advanced IR sensor design tool. The technologies to be developed as part of this effort have very broad Phase III commercial application, particularly in the fields of remote sensing for mineral exploration, and for highly-realistic spectral scene generation for industry. A low-cost version of this simulator would provide industry a crucial technology component for remote-sensing algorithm improvement test-beds, and a catalyst for substantially improved GIS products for geology, land use monitoring, agricultural, and mineral exploration. The defense industry could leverage this technology into improved fidelity sensor simulators for a broad range of applications, including training, mission planning and rehearsal.