Rédiger un PREreview

This paper presents the Micro Video Guidance Sensor (UVGS-2), an advanced evolution of the Smartphone Video Guidance Sensor (SVGS), a vision-based 6-DOF pose estimation system for proximity operations in spacecraft, UAVs, and mobile robot platforms. The system is based on the photogrammetric estimation of a beacon’s relative position and attitude in a camera coordinate system, by processing images of illuminated 4-points in the beacon with known non-coplanar geometry. Estimation of the beacon’s 6-DOF position and attitude (XYZ, RPY) in a coordinate system attached to the camera is achieved with higher accuracy compared to standard localization methods based on mapping or inertial navigation. Image acquisition, feature extraction, and state estimation are executed using hardware resources (camera and CPU) of the host robot, resulting in a low-mass, low-power, computationally efficient sensing architecture suitable for embedded and resource-constrained systems. As case study, SVGS and UVGS-2 have been deployed as real-time guidance, navigation, and motion control sensor through integration with NASA’s Astrobee free-flying robot aboard the International Space Station (ISS), supporting autonomous proximity operations and formation flight. Sensor fusion based on the use of existing localization pipelines improves robustness against line-of-sight interruptions and illumination disturbances, and improves accuracy compared to the native Astrobee localization system (AstroLoc). Other case studies have demonstrated high-precision positioning performance in autonomous UAV landing experiments, with reliable target tracking over extended ranges and low angular estimation error, outperforming existing infrared-based landing approaches.