Automated 3D scenes reconstruction for mobile robots using laser scanning

Authors:
Zheng Keqiang;Zhuang Yan;Wang Wei
Affiliations:
Research Center of Information and Control, Dalian University of Technology, Dalian;Research Center of Information and Control, Dalian University of Technology, Dalian;Research Center of Information and Control, Dalian University of Technology, Dalian
Venue:
CCDC'09 Proceedings of the 21st annual international conference on Chinese control and decision conference
Year:
2009

Citing 4
Cited 1

Least-Squares Fitting of Two 3-D Point Sets

IEEE Transactions on Pattern Analysis and Machine Intelligence
A Method for Registration of 3-D Shapes

IEEE Transactions on Pattern Analysis and Machine Intelligence - Special issue on interpretation of 3-D scenes—part II
Fast Acquiring and Analysis of Three Dimensional Laser Range Data

VMV '01 Proceedings of the Vision Modeling and Visualization Conference 2001
Robust meshes from multiple range maps

NRC '97 Proceedings of the International Conference on Recent Advances in 3-D Digital Imaging and Modeling

Self-calibration for a 3D laser

International Journal of Robotics Research

Quantified Score

Hi-index	0.00

Visualization

Abstract

The fast reconstruction for 3D environment is a challenging work for autonomous mobile robots. This paper presents an automated 3D scenes reconstruction method based on ICP algorithm. According to the order relations of 3D laser scanning, an effective edge feature extraction method is proposed to extract the edge points in the 3D scenes. The edge points instead of the initial points are used in pairs matching in the course of ICP registration, which can reduce the computational burden significantly. A 3D-grid-division method is adopted to reduce the redundant data points in the overlapping areas of multiple scans. The data points are distributed into small voxels and the point nearest to the center is extracted to reconstruct a new scene. According to the reduced points in the reconstructed scene, digitalization of 3D environment is implemented by using 3D grids and surface meshes. A series of experiment results and data analysis show the method's validity and practicability.