摘要
针对传统的零件尺寸测量系统效率低、精度差、价格昂贵且操作难度大的问题,课题组提出了一种基于机器视觉的齿轮尺寸非接触式在线测量系统。该系统采用一种新的直齿圆柱齿轮测量方案:首先对CCD相机采集的图像进行灰度处理与滤波去噪;然后进行阈值分割,采用Canny算子来获得单个像素的图像边缘;针对齿轮的边缘轮廓形状特征,应用霍夫变换、改进的最小二乘拟合和改进的中心画圆法等原理,建立了齿轮中心孔直径、齿根圆直径、齿顶圆直径和齿数等齿轮参数的测量算法。研究结果表明:使用2560 dpi*1920 dpi分辨率的CCD相机测量20 mm以下的直齿圆柱齿轮时,测量精度可达到10μm,绝对误差小于1个像素0.0209 mm,系统响应时间为2.8 s。该系统满足工业测量要求,解决了传统系统测量精度低、价格高的问题。
In order to solve the problems of low efficiency,poor accuracy,high cost and high difficulty in operation of the current detection system,a non-contact gear size measurement system based on machine vision was proposed.A new measurement scheme of spur gear was adopted.Firstly,the image captured by CCD camera was processed by gray scale and denoised by filtering,then threshold segmentation was carried out,and Canny operator was used to obtain the image edge of single pixel.According to the shape characteristics of gear edge,the measurement algorithms of gear parameters such as center hole diameter,tooth root circle diameter,tooth tip circle diameter and tooth number were established by applying the principles of Hough transform,improved least square fitting and improved midpoint center circle drawing method.The research results show that when using the CCD camera with 2560 dpi*1920 dpi resolution to measure the spur gear below 20 mm,the measuring accuracy can be 10μm,the absolute error is less than one pixel 0.0209 mm and the response time of the system is 2.8 s.The proposed system meets the requirements of industrial measurement,which solves the problems of low measurement accuracy and high system price.
作者
王德祥
张周强
管声启
WANG Dexiang;ZHANG Zhouqiang;GUAN Shengqi(School of Mechanical and Electrical Engineering,Xi’an Polytechnic University,Xi'an 714245,China)
出处
《轻工机械》
CAS
2021年第2期67-73,共7页
Light Industry Machinery
基金
国家自然科学基金青年科学基金资助项目(61701384)。
