Low Cost 3D Radar Imaging and 3D Association Method from Low Count Linear Arrays for All Weather Autonomous Vehicle Navigation

摘要

A low cost, all weather, high definition imaging system for an autonomous vehicle is described. The imaging system generates true target object data suitable for imaging, scene understanding, and all weather navigation of the autonomous vehicle. Data from multiple arrays is fed to a processor that performs data association to form true target detections and target positions. A Boolean associator uses an association method for determining true target detections and target positions to reduce many of the ghosts or incorrect detections that can produce image artifacts. The imaging system provides near optimal imaging in any dense scene for autonomous vehicle navigation, including during visually obscured weather conditions such as fog. The system and method can be applied to variety of imaging technologies, including an RF system, a Lidar system, a sonar system, an ultrasound system, and/or an optical system.