Precise Targeting of Tomahawk Cruise Missile Using Low-Cost, Low-Quality Miniature Unmanned Aerial Vehicle (UAV) Sensor Data

We propose methodologies for using Miniature unmanned Air Vehicle (MAV) sensor data to enhance Tomahawk Land Attack Missile (TLAM) targeting. Our approach uses 3D contextual information of the area available from existing DOD databases to determine MAV sensor pose. Optimization is performed to refine the camera orientation to improve pose accuracy. For relatively flat targets, the 3D world coordinates of a target selected in the 2D imagery are determined using the projection properties of the sensor. For targets with significant elevation, multi view processing is performed to determine accurate 3D positioning. We demonstrate additional capabilities for accurate targeting without pre-existing 3D models when GPS and IMU data are available. Error bounds relating the required system resolution and control point accuracy to targeting position estimation accuracy will be quantified. During Phase I, we will build on our past work on image to 3D registration to provide reliable targeting from videos acquired by MAVs. Phase II efforts will focus on developing a prototype of the proposed system.