基于视觉的六自由度机械臂运动学参数辨识

Design of a Vision Calibration Method of Kinematic Parameters for 6-DoF Manipulator

提出一种新型低成本的基于单目视觉的机器人末端位姿测量方法,设计并实现了六自由度机械臂的运动学参数辨识。采用分级测量方法和标定板绝对编码方法,解决了当前视觉测量过程中测量范围小、测量精度受相机畸变影响大等问题;使用基于位置误差的运动学参数辨识模型和单目视觉测量系统,简化了机器人的标定过程。最后,通过实验验证了方案的实用性和有效性。

A novel low-cost monocular vision based robot end pose measurement method was proposed,and the kinematic parameters identification of a 6-DOF manipulator was designed and implemented.By using the hierarchical measurement method and the absolute coding method with calibration plate,the problems such as the small measurement range and the large influence of the camera distortion on the measurement accuracy in the current visual measurement process were solved.The robot calibration process was simplified by using the kinematic parameter identification model based on position error and monocular vision measurement system.Finally,the practicability and effectiveness of the proposed scheme were verified by experiments.