机器视觉在圆锥滚子轴承内圈外表面缺陷检测中的应用

Using Machine Vision to Complete Tapered Roller Bearing out of the Loop Detection of Surface Defects

针对圆锥滚子轴承内圈外表面位置的特殊性,提出了一种新的缺陷检测方法,并设计了一套完整的缺陷自动检测系统。利用机器代替人眼完成轴承内圈外表面缺陷的检测,该系统光源恒定,可获取高精度的图像,处理速度快,可以长时间不间断的工作。首先对获取的图像进行图像预处理,去除图像噪声。利用MATLAB图像处理工具箱中的函数对预处理后的图像进行图像分析,统计图像中灰度值的分布情况,即计算出灰度图像的均值,标准差,利用图像灰度值的标准差判别出图像的表面质量是否合格。经过实验测试取得了96%的检测成功率,证明了其可行性。

Considering the particularity of tapered roller bearing inner ring outer surface position,put forward a new method of defect detection and designed a set of complete defects automatic detection system. Used machines instead of human eyes for the surface defect detection and bearing inner ring. Constant light source,the system can obtain the high precision image,and processing speed was fast,being able to work continuously for a long time. First of all it acquired images for image processing,and removed the image noise. By using the MATLAB image processing toolbox function image analysis was carried out on the image after treatment. Statistics in the image gray value distribution,that was,to calculate the average gray level image,standard deviation,using the standard deviation of the image gray value criterion in the surface of the image quality was qualified.