Hand-Eye Calibration
Overview
Hand-eye calibration estimates how a camera or other sensor is positioned relative to a robot, so visual measurements can guide accurate motion. Methods include solving the AX=YB geometry with Lie algebra optimization, using fiducials or external metrology, and marker-free approaches based on RGB-D point clouds or differentiable rendering. Researchers evaluate these methods in settings such as robotic manipulation, surgical instruments, ultrasound, and optical scanning, measuring transform accuracy and downstream task success. A central challenge is achieving reliable calibration with few observations and without specialized markers or trackers, while also adapting to changing scenes, viewpoints, and uncertainty.
Latest papers
13 papers1 paper in the last four weeks, down 67% on the four weeks before. 0.0% of all new papers.