基于机器视觉的短轴零件圆度检测

Roundness Detection of Short-axis Parts Based on Machine Vision

针对目前人工测量轴类零件圆度存在的效率低、综合误差大等问题,搭建了一个基于机器视觉的轴类零件圆度检测平台,基于单目相机对工件轮廓进行三维重构,并提出一种基于最小包容区域原则的自适应搜索逼近法对得到轮廓关键坐标点进行了圆度评定,实现了轴类零件圆度误差的非接触在线测量。其测量结果与文献中极点法所测结果对比精度更高,且与三坐标测量仪测量结果值平均有7.6μm的误差,较准确地测出了轴的圆度误差,证明该检测平台以及测量评定方法可以成功的对轴类零件进行圆度检测。

Aiming at the problems of low efficiency and large comprehensive error in the current manual measurement of the roundness of shaft parts,a machine vision-based roundness detection platform of shaft parts is built,and the contour of the workpiece is reconstructed in three dimensions based on a monocular camera.An adaptive search approximation method based on the principle of small containment area evaluates the roundness of the key coordinate points of the contour,and realizes the non-contact online measurement of the roundness error of the shaft parts.The measurement results are more accurate than those measured by the pole method in the literature,and have an average error of 7.6μm with the measurement results of the three-coordinate measuring instrument.The roundness error of the shaft is measured more accurately,which proves that the detection platform and measurement evaluation method can successfully detect the roundness of shaft parts.