基于机器视觉的柔性机械臂振动位移测量

 针对传统的振动位移测量方法中，间接测量法误差较大、准确度较低，而直接测量法虽能达到精度，但成本较高、实用性较差的问题，提出一种基于机器视觉的振动位移测量方法。选取电机驱动运动的柔性机械臂作为测量对象，通过CCD相机采集机械臂末端标记点运动图像，再经过阈值分割、质心检测等视觉处理，得到机械臂振动位移信息。通过搭建实验平台对该方法进行实验验证，并与压电材料测量法及仿真法得到的振动位移进行对比，结果表明，基于机器视觉的振动位移测量方法能够在不接触工况下，对柔性机械臂进行准确、实时的振动位移测量，相比于压电材料测量法，避免了公式误差等影响，具有明显的优越性。

Vibration displacement measurement of flexible manipulator using machine vision

Of traditional vibration displacement measurement methods, the indirect measurement method has its large error and low accuracy while the direct measurement method can achieve accuracy with high cost and poor utility. So a vibration displacement measurement method using machine vision is proposed. A motor-driven flexible arm is selected as the measurement object. A CCD camera is used to collect the motion image of the marker on the arm's manipulator. After the visual processing such as threshold segmentation and centroid detection, the vibration displacement information of the manipulator is obtained. An experimental platform is built to verify the method. The result is compared with the vibration displacements obtained by piezoelectric material measurement method and the simulation method. It is shown that the machine vision-based vibration displacement measurement method can be well applied to the flexible arm without any contacts. The accurate and real-time vibration displacement measurement of the arm is superior to the piezoelectric material measurement method, avoiding the influence of formula errors and other effects, and has obvious advantages.