基于机器视觉的亚像素精度法兰盘尺寸测量方法  

Flange size measurement method based on machine vision with sub-pixel precision

作　　者：焦博 刘国宁[1] 赵孟轩 马光岩 JIAO Bo;LIU Guoning;ZHAO Mengxuan;MA Guangyan(School of Mechanical and Power Engineering,Zhengzhou University,Zhengzhou 450001,China)

机构地区：[1]郑州大学机械与动力工程学院,郑州450001

出　　处：《现代制造工程》2022年第7期121-126,共6页Modern Manufacturing Engineering

基　　金：河南省产学研重点支持项目(172107000008)。

摘　　要：针对空调压缩机法兰盘尺寸人工测量方式存在效率低、精度难以保证等问题,提出一种基于机器视觉的法兰盘尺寸亚像素级别测量方法,并构建视觉测量系统。对法兰盘图像,先使用转灰度、中值滤波和二值化分割进行预处理,抑制图像噪声,同时提取目标法兰盘区域;再使用Canny算子获取法兰盘边缘像素坐标;然后根据Zernike算法对旋转、尺度等不敏感的特点,使用改进的Zernike矩亚像素边缘检测算法重新定位法兰盘边缘,获取法兰盘边缘的亚像素级别坐标;最后使用最小二乘拟合算法得到法兰盘外圆直径与内孔直径尺寸。实验证明,该检测方法与人工测量方式相比效率高、精度高,能够满足对法兰盘关键尺寸的在线测量需求。Aiming at the problems of low efficiency and difficulty in ensuring the accuracy of manual measurement of the flange size of the air conditioner compressor,a sub-pixel-level edge size measurement method of the flange based on machine vision was proposed,and a visual measurement system was constructed.The flange image was preprocessed by grayscale conversion,median filtering and binarization segmentation to suppress image noise and extract target flange region.Then Canny operator was used to obtain the pixel coordinates of the flange edge;Then according to the Zernike algorithm′s insensitivity to rotation,scale,etc.,the improved Zernike moment sub-pixel edge detection algorithm was used to relocate the flange edge to obtain the sub-pixel-level coordinates of the flange edge;finally,the least squares fitting algorithm was used obtain the flange diameter and inner hole diameter size.Experiments have proved that the detection method was more efficient and accurate than manual measurement,and can meet the requirements for online measurement of critical dimensions of flanges.

关 键 词：机器视觉 亚像素 法兰盘 尺寸检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]