Light Pen Calibration for a Monocular-Vision-Based Coordinate Measuring System

A monocular-vision-based coordinate measuring system obtains the 3-D coordinate via calculating the position of a light pen in camera coordinate system which is disturbed seriously by the locational precision of target points. In this paper, a concise onsite light pen calibration method is proposed. It includes three steps. First, the intrinsic and extrinsic camera parameters are estimated from several views of a known calibration pattern and then the target points are placed on the plane at the last view. Second, the captured image with target points is projected onto a canonical fronto-parallel plane using the extrinsic parameters above. And feature detection is executed on the canonical fronto-parallel plane. As the canonical images have been both undistorted and unprojected, feature of target points can be detected easily and localized precisely. Finally, points on the canonical plane are projected onto 3D plane. The proposed method is conducted in our monocular-vision-based coordinate measuring system. The results confirm the accuracy and effectiveness of the proposed method.