基于机器视觉的异形零件尺寸精密测量方法

Precision Measurement of Special-shaped Parts Based on Machine Vision

针对目前异形零件尺寸测量中存在的精度低、成本高、难以实现在线测量等问题,提出一种基于Halcon的异形零件尺寸精密测量方案,包括硬件和软件算法两部分。硬件部分主要对工业相机、工业镜头和光源进行选型,并设计了测量装置。软件算法部分,首先利用Halcon对相机进行标定获得其内外参数,然后通过灰度处理、阈值分割等算子得到零件区域,接着通过区域特征筛选、最小二乘算法拟合区域最大内接圆和最小外接圆、计算两线段距离、计算两直线夹角、计算区域个数等方法实现对异形零件的齿高、外槽宽度、外槽角度、圆孔角度、外槽深度、齿数等参数的测量。最后,与高精度三坐标测量机的测量结果对比,结果表明采用本方法测量异形零件的齿高、外槽宽度和外槽深度的精度达到0.04mm,测量外槽角度和圆孔角度的精度达到0.03°。

To address the challenges of low precision,high cost,and difficulties in on-line measurement encountered in dimension measurement of special-shaped parts,a scheme of precise dimension measurement of special-shaped parts based on Halcon is proposed,including two parts of hardware and software algorithm.The hardware part mainly selects the industrial camera,industrial lens and light source,and designs the measuring device.In the software algorithm part,the Halcon is utilized for camera calibration to obtain its internal and external parameters.Subsequently,the parts region is obtained through gray processing,threshold segmentation,and other operators.Then,the parameters such as tooth height,outer groove width,outer groove angle,circular hole angle,outer groove depth and tooth number of the special-shaped parts are measured by the methods.This involves selecting the region features,fitting the maximum inner circle and minimum outer circle using least square algorithm,calculating the distance between two lines,determining the angle between two lines,calculating the number of regions and so on.Finally,by comparing the measurement results with high-precision coordinate measuring machines,it is shown that the accuracy in measuring tooth height,outer groove width and outer groove depth for special-shaped parts by our proposed method reaches 0.04 mm while measuring outer groove angle and circular hole angle achieve an accuracy up to 0.03°.