A Registration Method of Multi-View Point Clouds Using Turntable and Monocular Structured Light

This paper describes an effective coarse registration method for multi-view point clouds which benefits from the high accuracy of the structured light 3D measurement system. In this method, we place a calibration ball in the center of the turntable, and then use the structured light system to obtain the point cloud of one side of the calibration ball. The center coordinates of the calibration ball are then calculated by spherical fitting to establish the relationship between the camera and turntable coordinate systems. In addition, the coordinate transformation among multi-view point clouds is calculated according to the preset rotation angle. Based on the data obtained in the previous two steps, multi-view point clouds are unified into the same coordinate system to achieve coarse registration as a reasonable initial value for the iterative closest point (ICP) algorithm. Finally, through the ICP algorithm for precise registration, a complete point cloud of the object is obtained. Experiments on 3D reconstruction for multiple objects confirm the efficiency of the proposed registration method. Code has been made available at: https://github.com/ZengZhiK/Structured-Light-MultiView-Reconstruction.