轴套零件的机器视觉检测  

Shaft Sleeve Parts Detection with Machine Vision

作　　者：陈向伟[1] 王海月[1] 

机构地区：[1]东北电力大学机械学院,吉林吉林132012

出　　处：《拖拉机与农用运输车》2015年第1期29-32,35,共5页Tractor & Farm Transporter

摘　　要：针对轴套零件的几何尺寸测量给出了一种基于机器视觉的检测方法。阐述了在Visual C++平台下,采用数字图像处理技术非接触测量轴套几何参数的方法,包括图像的预处理、改进Canny算子边缘检测、圆孔检测技术等。并对系统进行标定,得到实际尺寸。通过试验得到测量数据,分析了测量误差的原因。分析及试验表明,用该方法对轴套的几何尺寸进行测量及分析评定在实际应用中是可行的,并具有高效性和实用性。A new measurement system of shaft sleeve parts size is put forward based on machine vision. On theplatform of Visual C ＋ ＋, the non-contact measurement principle for geometry of part is introduced by using image measurement technology. The main measurement approaches include：image collecting,image pre-processing, improved Canny edge detection, contour extraction etc. Measurement data is obtained by experiment, and the reasons of measurement error are analyzed. Theory analysis and experiment show that it is feasible to measure and analyze the geometrical size of shaft sleeve parts with this measurement method, and the method is proved to be efficient and practical.

关 键 词：机器视觉 改进Canny 最小二乘法 图像测量 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]