Human-Activity Sensor Returns

ABSTRACT: To address the Air Forces need for software-synthesized multimodal sensor returns to explore data fusion for human-based activities, Physical Optics Corporation (POC) proposes to develop a new Human-Activity Sensor Returns (HASR) modeling software library and simulation framework. It integrates human, environment, and sensor models with physical effects modeling to generate scientifically rigorous synthetic sensor returns. Specifically, the innovation in synchronized simulation of multimodal sensors (e.g., radar, SAR, LIDAR, and hyperspectral) with a single input dataset representing the environment and human activities, will produce synthetic multimodal sensor return data that is consistent across modalities, and high fidelity in human-focused scenarios. With an architecture that scales easily with sensor and environmental model additions, HASR is the first solution for comprehensive and integrated high-fidelity modeling of human activities across multiple sensor modalities, which directly addresses the Air Force need for multimodal human-activity imagery baselines. In Phase I, POC will demonstrate the feasibility of HASR by surveying mathematical sensor and environmental models, developing an end-to-end systems architecture, and demonstrating a preliminary proof-of-concept system. In Phase II, POC plans to develop and demonstrate a functional prototype with a graphical user interface and multimodal capabilities.; BENEFIT: Current imagery baselines for radar, SAR, LIDAR, and hyperspectral sensors were developed independently, without a common 3D scene baseline, and without a focus on human-based activity. These limitations prevent their use for the direct comparison of sensor results in general (due to disparate 3D scenes), and prevent the application of the current imagery baselines to human-based activities. HASR applies a unified 3D scene to every sensor so sensor results can be compared directly, and adds a focus on human activities that will enable those sensor results to be applied to human activities such as walking, running, digging, and throwing. HASR is the first tool that can produce appropriate imagery baselines for evaluation of new sensors, sensor technologies, surveillance monitors, crowd monitors, and reconnaissance technologies in both military and commercial worlds. For example, HASR sensor data could be used to select an optimal sensor(s) and sensor processing technologies for monitoring of neutral civilians with embedded covert agents, for automated surveillance of crowds around military bases. In the commercial world, HASR could be applied to surveillance systems for homes and businesses, to improve automatic detection of potentially covert activities (e.g., mine burial and bomb throwing) while minimizing generation of false alarms from similar benign activities (e.g., gardening and waving).