3D Perception and Reconstruction System Based on 2D Laser Scanner

Abstract

3D perception and reconstruction is an important research in the field of robot motion control and 3D manufacturing. A system for surface perception and reconstruction of the 3D area based on 2D laser scanner is outlined in this paper. Firstly, aiming at the limited work area of a single device, the “S-shaped” control strategy for scanning device is proposed to extend the effective operating range of the detection system so that the surface could be completely measured, analyzed and processed. The scanning point cloud could be registered to obtain a complete three-dimensional point cloud image. Secondly, due to the problem of the impulse noise, a weighted median filter based on the Euclidean distance is proposed to recover the correct depth information. The performance of the proposed system shows that the designed Euclidean distance-based median filtering method effectively removes impulse noise and preserves original depth information. The system can accurately get the 3D point cloud information, effectively process the scan data, and realize 3D perception and reconstruction for object.