基于机器视觉的电梯导轨垂直度检测

Elevator Guide Rail Verticality Detection Based on Machine Vision

为更有效地测量电梯导轨的垂直度,设计一种基于机器视觉的自动检测系统。利用电梯导轨垂直度检测机器人携带垂直安装在导轨上的发光圆环,使其在导轨上自主运行。通过Canny变换提取摄像机采集的成像椭圆图像边缘,拟合椭圆参数,推导发光圆环几何参数与椭圆参数的关系式,确定其圆心和法向向量。发光圆环圆心的变化反映导轨垂直度的变化,法向向量与竖直方向夹角反映导轨的扭曲方向。实验结果表明,该系统能方便准确地测量出导轨形变。

An automatic detection system based on machine vision is designed to measure the verticality of elevator guide rail more effectively. A verticality measuring robot for elevator guide rail moves along the rail with a vertically fixed lighten ring. Canny transform is used to extract the elliptical edge points acquired by a video camera. Then ellipse parameters are fitted. The relationships between lighten ring geometrical parameters and ellipse parameters are derived. So its geometric center and the normal vector are determined. The changes in the center of the lighten ring reflect those in the verticality of elevator guide rail, and the changes of the angle between normal vector and the vertical reflect the distortion direction of the guide rail. Experimental results show that the proposed system can measure the deformation of the guide rail easily and accurately.