Development of an in-situ real-time laser targeting algorithm for a 3D LiDAR-integrated galvanometric laser scanner system

Abstract

A Galvanometric laser scanner utilizes mirrors attached to dual galvanometers to direct laser beams. GLS systems are widely employed across various industrial sectors due to their compact size, high resolution, and high positional repeatability. Extensive research has been conducted to calibrate these systems to steer lasers to desired positions accurately. However, much of this research has traditionally focused on fixed working distances, which does not align well with the portable capabilities that GLS systems offer in non-destructive evaluation fields. This study explores an algorithm that enables real-time laser targeting using a 3D LiDAR camera integrated into a GLS system. The research involved developing and calculating a projection matrix that projects 3D camera coordinates into 2D bit coordinates input to each galvanometer, based on linear assumptions. To address the non-linear errors that emerge from the linear assumption process, compensation was applied through 3rd-order regression and Artificial neural networks. As a result, the algorithm facilitated highly accurate laser targeting within a range of 0.5 m to 2.5 m based on coordinates acquired by the LiDAR camera.