Real-time Ortho-Mosaicing for Awareness and Navigation (ROMAN)

Unmanned aircraft systems (UAS) engaged in maritime operations are equipped with electro-optical (EO) and infrared (IR) sensors for mission profiles, including: intelligence, surveillance, and reconnaissance; anti-submarine warfare; and countermine operations. The Navy wants to use these platforms to build and maintain a sea surface map that consists of latest-time-available imagery over a broad area. It must detect objects, such as periscopes and mines, and make this information available to sensor operators and mission Commanders in real time. Charles River Analytics proposes to design and prototype Real-Time Ortho-Mosaicing for Awareness and Navigation (ROMAN) as a system for ortho-mosaicing and object detection using EO/IR imagery from UAS to improve situational awareness. ROMAN will ingest full-motion video from the sensor aboard UAS and geo-register sequences of imagery, stitching frames of video together to form a continuously updated ortho-mosaic. ROMAN will detect, classify, and track maritime objects visible to the sensor. It will include planning tools to avoid areas obscured due to weather, and it will use automated image analysis techniques to detect and reject unsatisfactory images. ROMAN will include a sensor package that enables the ortho-mosaic to be generated in a local coordinate system relative to the mothership in GPS-denied environments.