A Weighted Extrinsic Calibration of a Camera and a LRF

Robots are often equipped with cameras and Laser Rangefinders (LRFs), and it is important to obtain their relative pose since they are separated. In this paper, an algorithm for calibrating the extrinsic parameters between a camera and a 2D LRF is proposed, and a system consisting of 2D LRF and a rotating unit is designed to obtain a 3D scanning map of the surroundings. Since the LRF beam is invisible, the proposed calibration method uses a common calibration object and establish geometric constraints through observations from both sensors. The calibration setup geometry is described by the normal vector of the plane. Besides, we add weights to the formula to represent the uncertainty of the image and LRF data. The simulation and real experiments show that our method is more robust in noise.