摘要
研制了一种基于测量机器人的激光三维扫描仪,介绍了系统的测量原理,根据双三角法原理推导出物体空间三维点坐标,提出了具有一定泛化能力的神经网络系统整体标定方法,并对一茶杯盖进行了实际测量实验,测量误差小于0.1mm,实验结果验证了测量原理的有效性和系统的可靠性。
Based on the passive robot, a 3-D laser scanning system is put forward, the principle of the system is introduced,The 3-D coordinates are derived according to double-trigonometry method. The serf-learning algorithm of the network based on generalization ability has been developed to calibrate the 3-D laser scanner. A teacup cover is scanned with the measuring system,The measuring precision is less than 0. 1mm。 Experimental results prove the effectiveness of the measuring principle and the reliability of this measuring system。
出处
《激光技术》
CAS
CSCD
北大核心
2005年第4期366-369,共4页
Laser Technology
关键词
三维激光扫描仪
双三角法
神经网络
摄像机标定
3-D laser sanner;double-trigonometry method
neural network
camera calibration