Abstract: The development of LiDAR ranging technology provides a new technical means for the acquisition of 3D spatial data, which enables people to realize the transformation from traditional manual single-point data acquisition to continuous automatic data acquisition, greatly improving the speed and accuracy of measurement. However, the current 3D LiDAR data acquisition equipment has the disadvantages of expensive price, poor mobility, large data volume and long time for 3D reconstruction. In this paper, a two-dimensional laser scanner is proposed to obtain the three-dimensional point cloud through the spatial coordinate transformation alg orithm. The rotation and translation matrix obtained from the point cloud registration is optimized through the quaternion algorithm, and the optimal rotation and translation matrix is obtained through multiple iterations, so as to improve the speed and accuracy of point cloud registration. The experimental application analysis results show that the real-time 3D reconstruction method of two-dimensional laser radar proposed in this paper has the characteristics of low cost, high accuracy and fast 3D reconstruction, and meets the mapping requirements of on-site low-cost, real-time and accurate 3D reconstruction.

Key words: 2D laser radar, 3D point cloud, point cloud registration, quaternion, 3D reconstruction