摘要
重复定位精度是机器人精度评估的重要指标之一。提出一种基于光学图像非线性局部特征匹配的2-DOF机器人手臂重复定位精度测量方法,具有非接触、高精度和易测量等优点。在XY-Theta测试平台采用远心光学成像系统采集2种工况下机械臂定位点的视觉图像,并对其进行KAZE特征点提取和匹配,计算定位坐标点云并拟合其最小包围圆,进而得出机械臂的重复定位精度。对比实验结果表明:所提算法相比传统方法在精度、稳定性方面均具有显著优势,可推广到其他精度测量应用中。
Repeatability is one of the most important indicators for robot accuracy assessment. A repeatability measurement method for 2-DOF robot arm was proposed based on optical image local feature matching. The advantages of the proposed method include non-contact,high precision and easy measurement etc. The telecentric optical imaging system was used in the XY-Theta test platform to obtain the visual image of the robot arm positioning point under two working conditions. The nonlinear local features were extracted and were used for feature matching. The positioning coordinate point cloud was calculated according to the matching result. The minimum enclosing circle of point cloud was approximated,the radius of which denotes the repeatability of the manipulate arm. The experimental results show that the algorithm outperforms traditional methods in accuracy and stability,and can be extended to other measurement applications.
作者
张岩
李光旭
赵玉涛
郑洲洲
张森
ZHANG Yan;LI Guangxu;ZHAO Yutao;ZHENG Zhouzhou;ZHANG Sen(Electromechanical Engineering,Qingdao University of Science and Technology,Qingdao 266061,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2020年第5期196-203,244,共9页
Journal of Chongqing University of Technology：Natural Science
基金
山东省重大科技创新工程项目(2017CXGC0607)。
