基于机器视觉的轴承滚动体缺失检测

Detection of Bearing Rolling Element Missing Based on Machine Vision

为了更简单高效的检测轴承滚动体是否缺失,设计了单独背光源、单独同轴光源和背光源、同轴光源组合光源等三种光学系统,然后由实验效果选择组合光源的方案。利用机器视觉Halcon软件,根据检测需求,对实验图像进行预处理、区域筛选、特征分析、极坐标变换、截取矩形区域和灰度投影等处理流程,得到轴承的内、外圈的同心度,并得到灰度函数图及一阶导数图,最终自动检测轴承滚动体个数及判断滚动体是否缺失,达到预期的检测目的。

In order to more easily and efficiently detect whether the bearing rolling elements are missing,three optical system are designed which include a separate backlight source,a single coaxial light source,a backlight source and a coaxial light source combined light source,and then a combination light source scheme is chosen according to the experimental results.Using machine vision—Halcon software,according to the detection requirements,the experimental image is processed by image preprocessing,area screening,feature analysis,polar coordinate transformation,interception of rectangular areas,and gray-scale projection are performed on the experimental images to obtain the concentricity of the inner and outer rings of the bearing,and also to obtain the gray-scale function diagram and the first derivative chart,which automatically detect the number of rolling element in the bearing is automatically detected and whether the rolling element is missing or not,which achieves the expected detection goal.This design provides a new solution method for non-contact automatic detection of bearings.