Building an Omnidirectional 3-D Color Laser Ranging System Through a Novel Calibration Method

Abstract

Three-dimensional (3-D) color laser ranging technology plays a crucial role in many applications. This paper develops a new omnidirectional 3-D color laser ranging system. It consists of a two-dimensional (2-D) laser rangefinder (LRF), a color camera, and a rotating platform. Both the 2-D LRF and the camera rotate with the rotating platform to collect line point clouds and images synchronously. The line point clouds and the images are then fused into a 3-D color point cloud by a novel calibration method of a 2-D LRF and a camera based on an improved checkerboard pattern with rectangle holes. In the calibration, boundary constraint and mean approximation are deployed to accurately compute the centers of rectangle holes from the raw sensor data based on data correction. Then, the data association between the 2-D LRF and the camera is directly established to determine their geometric mapping relationship. These steps make the calibration process simple, accurate, and reliable. The experiments show that the proposed calibration method is accurate, robust to noise, and suitable for different geometric structures, and the developed 3-D color laser ranging system has good performance for both indoor and outdoor scenes.