Fusion of 4D Point Clouds From a 2D Profilometer and a 3D Lidar on an Excavator

Abstract

The research aimed to integrate a solid-state 2D-pulsed time-of-flight (TOF) flash Lidar profilometer and a mechanically scanning 3D Lidar in order to monitor the 4D (3D + signal intensity) environment at a construction site from an excavator. Utilizing the excavator's GNSS and IMU sensor data enables to fuse the profilometer and 3D Lidar point clouds into the same coordinate system. Both measurement devices have quite a narrow field of view, but the overall viewing angle is expanded due to the movement of the excavator boom and rotation of the excavator cab. The major advantage of this combination method is that the profilometer provides centimetre-accurate details of the worksite within a range of approximately 1-10 m, while the commercial 3D Lidar scanner provides a larger scale (roughly 1-64 m) description of the worksite with an accuracy of ±6 cm. Additionally, the profilometer's intensity information can be used to identify people on the basis of varying reflective vest patterns. The experimental results indicate that our combination system can accurately build a quality 4D point cloud map better than either individual device on its own. The combination of Lidars could prove to be an interesting tool, with the potential to improve construction site management processes.