An efficient method for reflective surface scanning using Kinect v2

Abstract

ToF cameras, also identified as range imaging cameras, are a recent generation of sensors. As they are desirable for measuring distances to objects at a high frame rate, such sensors are increasingly used for 3D acquisitions, and more generally for applications in robotics or reverse engineering. In this study, the authors presented an efficient method for reflective surface scanning using Kinect v2. The approach is based on the principles of the ICP (Iterative Closest Point) algorithm. During the Kinect v2 scanning, the obtained point cloud is incomplete and includes flying pixels. We solve the flying pixel problem by capturing point clouds as multiple viewpoints. Each point cloud belongs to each coordinate. Then, all of the point clouds must be aligned and combined using the ICP algorithm. The experimental results show that the quality of the method is significantly improved.