A novel matching algorithm for a Hand-Held 3D Laser Scanning Sensor
A hand-held 3D laser scanning sensor has been gradually used as a novel solution to digitize the 3D shape for a wide variety of objects, as its maneuverability and portability. It consists of two synchronized CCD cameras and a laser projector. The current research challenge of a hand-held 3D laser scanning sensor is how to reliably and rapidly find its position and orientation at all times in a global world frame. It covers a key issue how to reliably and quickly match the list of observed reference points with the sensors list of scattered reference points. To work out this issue, this paper proposes a novel matching algorithm based on Delaunay triangulation. It is assumed that all reference points are accurately measured and stored in the sensors list beforehand. Above all, all reference points are projected to the X-Y plane and the mapping list is gained. The usage of Delaunay triangulation in the mapping list can determine which triplets of reference points are grouped together to form a set of triangles in the sensors list of scattered reference points. At last,in order to accelerate matching the triangles in the current stereo frame with the corresponding triangles in the global world frame, the radius of inscribed circle within the triangle is regarded as index and array the set of triangles according to indexs ascending order. Practical experiments show that the algorithm can precisely match reference points and furthermore meet real time requirement of the hand-held sensor application.
3D shape measurement Self-Referencing Matching algorithm Delaunay triangulation
Kui Huang Zhongwei Li Jianhua Mo Dahui Qing Congjun Wang
State Key Laboratory of Material Processing and Die & Mould Technology, Huazhong University of Science and Technology, Wuhan, 430074, China.
国际会议
北京
英文
172-179
2008-11-06（万方平台首次上网日期，不代表论文的发表时间）