Research on 3D point-cloud registration technology based on Kinect V2 sensor

A 3D point-cloud registration algorithm based on the Kinect V2 sensor is proposed for the point-cloud registration in 3D reconstruction. By the Microsoft Kinect V2 sensor to acquire the multi-view depth map of the object, the target area is extracted and transformed into 3D point cloud. In order to improve the robustness of ICP algorithm, an initial registration stage is added, before the accurate registration is performed by the improved ICP algorithm. Based on the initial registration, an iterative algorithm (ICP) is used for fine registration. The experimental results show that the registration method can not only guarantee the registration quality of 3D point cloud, but also reduce the computational complexity. It has higher maneuverability and robustness.