Biological Motion Processing for Low-Light-Level Multi-Spectral Sensors

Modern warfare is increasingly conducted during conditions of low/no visibility, often using small autonomous weapons platforms requiring miniature sensor systems for navigation and guidance. Flying insects provide excellent models from which guidance,navigation, and moving target detection/tracking principles can be derived and applied for military usage.Many animals have visual sensing and processing that operate under very weak ambient illumination. Nocturnal hawkmoths are capable of precise visually guided hovering and rapid flight under starlight conditions.Strategies for developing bio-vision for military use under low light conditions include temporal and spatial integration. The motion detectors of nocturnal species employ radically different temporal filters than diurnal insects, and elementary motiondetection (EMD) involves nonlinear processing, therefore, neural integration along with nonlinear EMD operations may yield a more nearly optimal approach to motion processing under low light levels.In Phase I, the academic partner will study and model motion processing in nocturnal hawkmoths, and the commercial partner will develop an approach to implement this enabling technology in analog VLSI, and (optionally) plan its integration with other keytechnologies (e.g., tunable MEMS etalons) into a multi-spectral sensor system. We will leverage our ongoing R&D on insect-based EMD and tunable MEMS etalons. Military/commercial applications currently exist for use in medium caliber autonomous munitions; and, in miniature UAV GN&C systems for near-ground flight. Similarly, compact multi-spectral sensors enable low-light moving target detection from miniatureautonomous platforms.