摘要：基于图像处理技术提出一种非接触式高效且高精度的微小孔零件尺寸测量方式。通过对微小孔零件进行图像采样，依次进行图像灰度化、图像滤波、图像二值化处理和微小孔边缘轮廓提取。采用加权平均值法对图像进行灰度化处理，去除无用部分，减小采集到的原图大小；采用双边滤波去除图像中的噪音点，减小边缘提取时的误差，用最大类间方差法对图像进行二值化处理；采用改进型Canny算子进行图像轮廓提取，再用最小二乘法拟合像素尺寸；通过对标准件进行尺寸标定，进而获得零件测量尺寸。以铝板小孔为实验对象，通过实验对比和分析得到的测量结果误差达到微米级别。该方法具有高效且高精度的特性，具有较高的实用价值。

关键词：微小孔零件；图像处理；尺寸检测；高精度

中图分类号：TG937；TH124；TP391.9文献标志码：ADOI：10.3969/j.issn.1000-7008.2021.07.020

Research on Dimension Detection of Micro Hole Parts Based on Machine Vision

Jia Guobang，Song Lili，Cao Bowen，Xu Yi

Abstract:Based on image processing technology，a noncontact，high efficiency and high precision dimension measurement method for micro hole parts is proposed.Through the image sampling of micro hole parts，image graying，image filtering，image binarization and micro hole edge contour extraction are carried out successively.The weighted average method is used to gray the image to reduce the size of the original image and remove the useless parts；the bilateral filter is used to remove the noise points in the image to reduce the error of edge extraction，and then the maximum inter class variance method is used to binarize the image；the improved Canny operator is used to extract the image contour，and the least square method is used to fit the pixel size.Through the dimension calibration of standard parts，the measured dimensions of parts are obtained.Taking the aluminum plate hole as the experimental object，through the experimental comparison and analysis，the error of the measurement results reaches micron level.This method has the characteristics of high efficiency and high precision，and has high practical value.

Keywords:micro hole parts；image processing；dimension detection；high precision