Automated Uav-based Video Exploitation for Mapping and Surveillance

Airborne intelligence, surveillance and reconnaissance (ISR) can provide essential information for both military and civilian applications. Such capabilities are critical for situational awareness, mapping and monitoring, and mission planning. UAVs gather large amounts of video data, but it is extremely labour-intensive for operators to analyse the hours and hours of data collected. At MDA, we have developed a suite of video exploitation tools to process UAV video data, including tools for mosaicking, change detection and 3D reconstruction. The main objective of this work is to improve the robustness of these tools, to integrate these tools within a standard GIS framework, to automate the tools, and to present these tools in user-friendly ways to operational users. Our mosaicking tool produces a 2D map from the sequence of video frames contained in a video clip. For scene monitoring applications, our change detection tool identifies differences between two videos taken from two passes of the same terrain at different times. Our 3D reconstruction tools create calibrated geo-referenced photo-realistic 3D models, which are useful for situational awareness. The semi-automated approach allows the user to create surfaces interactively by marking correspondences between frames, while the automated approach generates 3D models automatically. MDA provides UAV services to the Canadian and Australian forces, offering important ISR information to front-line soldiers. Our suite of semi and fully automated video exploitation tools can aid the operators to analyse the huge amount of UAV video data to support battle-space mapping and monitoring applications. * Corresponding author