Abstract
A hand-held 3D laser scanning sensor has been widely used as an advanced solution to digitise the 3D shape for a wide variety of objects. To improve the reliability and efficiency of computing the sensor's position and orientation, this paper proposes a novel matching algorithm that forms and matches triangles from the two lists of reference points of the global world frame and the current local stereo frame. Delaunay triangulation is used to form triangles in these frames. Theoretical analysis and experimental results show that the algorithm can match the corresponding reference points more precisely and efficiently and therefore meet the real-time requirement of the hand-held sensor applications compared with previous algorithms.
Acknowledgements
This work was supported under an Eleventh Five-Year Pre-Research Project, which was undertaken by the State Key Laboratory of Material Process and Die and Mould Technology, HuaZhong University of Science and Technology. The authors would like to take this opportunity to express their sincere gratitude to Mr. YuSheng Shi and Mr. Ho Simon Wang for their valuable comments and help in improving the manuscript.