Naval Infrared Scene Simulation STTR

ThermoAnalytics proposes to develop and demonstrate the generation of synthetic infrared imagery for objects in a realistic maritime background to improve operator training, train AI/ML algorithms and test infrared cameras. Many existing simulation programs focus on the visible spectrum where phenomenology such as natural lighting, reflections, diffractive media and RGB coloring are primary; temporal history is unimportant, and only present geometric/optical/lighting conditions must be considered. Thermal wavebands present a different challenge. LWIR and MWIR cameras detect radiation, both emitted and reflected, that depends on the temperatures of scene elements. The calculation of transient scene temperatures requires knowledge of prior temperatures, so the thermal history of objects must be combined with thermal material properties, optical surface properties and heat rates from various modes of heat transfer (including engine and exhaust heating). We will investigate a computationally tractable approach to physics-based temperature and radiation calculation for the generation of infrared synthetic maritime data. This solution will utilize modern processing techniques and proven heat transfer calculation methods to compute the transient temperatures of objects at sea and generate synthetic imagery of maritime scenarios. The proposed work will leverage our existing transient thermal/infrared prediction software, MuSES, which has been validated and employed by many DoD agencies over the decades. MuSES calculates maritime object temperatures with weather data and mission profiles. Spectral surface properties are combined with surface temperatures and surrounding environmental components (e.g., sky, sun) to compute accurate radiances from the perspective of a user-defined virtual sensor. These sensor-detected radiances correctly incorporate surface emissions, reflections, atmospheric extinction and path radiance using MODTRAN. The MuSES physics-based maritime background model includes ambient sea surface temperature, gravity and capillary waves, foam, ship wakes, sky radiance gradients and resultant reflections.During the Phase I effort, we will explore several approaches and document the strengths and weaknesses of each to determine the most viable approach. We will investigate expediting portions of our existing (and validated) MuSES infrared simulation suite while adding any missing ship/sea physics critical for naval applications of interest. We will also explore coupling our surface temperature and radiosity results with a computationally-efficient 3rd party ocean simulator (e.g., NVIDIA’s WaveWorks) that can be modified to achieve sufficient infrared accuracy, using MuSES as an evaluation standard. We believe that our existing physics-based infrared simulation capabilities, combined with state-of-the-art computation techniques, will deliver the correct balance of accuracy and simulation speed for the US Navy.