Advancements in FPA density and associated wide-field-of-view infrared sensors (>=4000x4000 detectors) have
outpaced the current-art HWIL technology. Whether testing in optical projection or digital signal injection modes,
current-art technologies for infrared scene projection, digital injection interfaces, and scene generation systems simply
lack the required resolution and bandwidth. For example, the L3 Cincinnati Electronics ultra-high resolution MWIR
Camera deployed in some UAV reconnaissance systems features 16MP resolution at 60Hz, while the current upper limit
of IR emitter arrays is ~1MP, and single-channel dual-link DVI throughput of COTs graphics cards is limited to
2560x1580 pixels at 60Hz. Moreover, there are significant challenges in real-time, closed-loop, physics-based IR scene
generation for large format FPAs, including the size and spatial detail required for very large area terrains, and multi -
channel low-latency synchronization to achieve the required bandwidth. In this paper, the author’s team presents some
of their ongoing research and technical approaches toward HWIL testing of large-format FPAs with wide-FOV optics.
One approach presented is a hybrid projection/injection design, where digital signal injection is used to augment the
resolution of current-art IRSPs, utilizing a multi-channel, high-fidelity physics-based IR scene simulator in conjunction
with a novel image composition hardware unit, to allow projection in the foveal region of the sensor, while non-foveal
regions of the sensor array are simultaneously stimulated via direct injection into the post-detector electronics.
Vehicles concealed in highly-cluttered, vegetated scene environments pose significant challenges for passive sensor systems and algorithms. System analysts working hypersectral exploitation research require and at-aperature simulation capability that allows them to reliably investigate beyond ther highly-limited scenarios that expensive field data sets provide. To be useful to the analyst, such a simulation should address the following requirements: (1) the ability to easily generate scene representations for abritrary Earth regions of tactical interests; (2) the ability to represent scene components, like terrain, trees and bushes, to an extremely high spatial resolution for calculation of accurate multiple spectral reflections, occlusions and shadowing; (3) the ability to stimulate the 3D scene with realistic natural irradiances for arbitrary model atmospheres; (4) the ability to appropriately integrate improving, rigorous thermal, spectral signature and atmospheric propogation models; (5) the ability to effectively render at-apurature hyperspectral data sets in a reasonable run-time. herein the authors describe their continuing work toward a comprehensive ray-tracer-based simulation archetecture and prototype capability that addresses these requirements, with emphasis on new techniques for high fidelity thermal modeling, and recent improvements in atmospherically scattered irradiance modeling, manmade light source modeling, and GIS-based database generation, including automated material classification of terrain and scene elements.
Proc. SPIE. 5431, Targets and Backgrounds X: Characterization and Representation
KEYWORDS: Reflection, Databases, Computer programming, 3D modeling, Geographic information systems, Bidirectional reflectance transmission function, Solids, Chemical elements, Atmospheric modeling, RGB color model
Vehicles concealed in highly cluttered, vegetated scene environments pose significant challenges for passive sensor systems and algorithms. System analysts working hyperspectral exploitation research require an at-aperture simulation capability that allows them to reliably investigate beyond the highly-limited scenarios that expensive field data sets afford.
To be useful to the analyst, such a simulation should address the following requirements: (1) the ability to easily generate scene representations for arbitrary Earth regions of tactical interest; (2) the ability to represent scene components, like terrain, trees and bushes, to an extremely high spatial resolution for calculation of accurate multiple spectral reflections, occlusions and shadowing; (3) the ability to stimulate the 3D scene with realistic natural spectral irradiances for arbitrary 3D model atmospheres; (4) the ability to appropriately integrate constantly improving, rigorous thermal, spectral signature and atmospheric propagation models; (5) the ability to efficiently render at-aperture hyperspectral data sets in a reasonable run-time.
Herein the authors describe their work toward a comprehensive ray-tracer-based simulation architecture and prototype capability that addresses these requirements. They describe their development of a GIS-based toolset for database generation, tools for 3D vegetated terrain-model development, and a prototype raytracer-based spectral scene generator.
A compelling figure-of-merit for synthetic sensor data generation is target acquisition performance relative to field data of similar scenarios. That is, if one synthesized a sensor data set from the ground truth of an actual field measurement, one would expect a realistic simulation to give rise to probabilities of target detection and identification, and false alarm rates comparable to that of the field data. This correlation would, of course, be expected to extend to both human and machine-based performance. Key to this correlation is realistic background synthesis, providing appropriate spatial and spectral competition with the target signatures for both man and algorithm. Hyperspectral target signature synthesis is fairly mature, while background modeling for target- competitive clutter leaves much to be desired. The authors herein detail a ray-tracing approach for rigorous hyperspectral background signature synthesis that focuses on trees and forest canopies, and synthesis techniques for producing spatial-spectral statistics consistent with field data. In addition, the authors present some of the hyperspectral synthesis components, including the signature model, which can be used in a multi-spectral mode for real- time EO/IR image synthesis.