Multimodal Sensor Fusion and Exploitation for Proliferation Detection

Detection and identification of new and emerging illicit activities and threats in remote or restricted locations is a national priority for the United States, its people, and various government offices and law enforcement agencies. To address this need, this project is developing a data fusion framework to rapidly ingest, process, and analyze satellite imagery from multiple providers. Within the next five years, more than 100 images per day will be available over most sites on Earth, requiring new algorithms to automatically process and extract useful information. This project is developing a data fusion framework that is agnostic to data provider and modality, and which can combine data sources into a common geo-referenced projection. Once data are placed into the same coordinate system, traditional and newly-developed algorithms can be, providing greater analyst and decision-maker confidence. During Phases I and II, this program advanced in three distinct ways. First, a registration framework was created that projects imagery into a common coordinate system. Second, key software was expanded to generate 3D models using imagery from any provider. Lastly, Synthetic Aperture Radar processing abilities were expanded to allow for faster and superior products and intelligence. The Phase IIA project plans to expand the fusion framework by further developing algorithms to exploit Synthetic Aperture Radar in combination with 3D models. The culmination of this effort will be a demonstration at a relevant site and a business transition plan. The benefits of this fusion framework will be realized in improved analyst effectiveness and efficiency across many industries. Increased automation will serve to expand the envelope of existing systems and workflows. Improved efficiency will result in reduced analysis timelines, resulting in cost savings for existing and future programs and applications.