摘要
现有厚度测量方法(激光法、超声法、电容法)无法实现在线监控载带厚度。针对此问题提出了一种双工位的视觉测厚成像机构,低成本实现在线高速测量载带平面尺寸和厚度。针对载带高速运动下,型腔离焦成像、视差等因素导致的测量重复精度低的问题,提出了离焦补偿校正方法。首先引入基于局部区域面积效应的边缘定位算法实现高精度边缘检测,然后由所得关键点的几何关系计算出尺寸,最后基于离焦成像环境测量的F值和P2值的补偿校正方法,实现载带尺寸高精度测量。通过在实际载带生产线运行验证,该检测系统允许载带最大生产速度1800 m/h,重复精度为5μm;以人工测量均值为准确值时,测量系统精度为0.03 mm。
The existing thickness measurement methods(laser method,ultrasonic method,capacitance method)can’t monitor the thickness of carrier tape online.Aiming at this problem,a visual thickness measurement imaging mechanism was proposed to realize a low-cost online highspeed measurement of carrier tape thickness.Aiming at the problem of low measurement repetition accuracy caused by factors such as defocus imaging and parallax under the high-speed movement of carrier tape,a correction method was proposed.In this paper,a double-position vision measuring mechanism was designed,which can measure the plane size and thickness.Firstly,an edge location algorithm based on local area effect was introduced to achieve high-precision edge detection.Then,each size was calculated from the obtained geometric relationship of key points.Finally,based on the compensation and correction method of F value and P2 value measured in defocused imaging environment,the high-precision measurement of carrier tape size was realized.Through the actual tape production line operation verification,the detection system allows the maximum tape production speed of 1800 m/h,and the repetition accuracy is 5μm;When the average value of manual measurement is taken as the accurate value,the accuracy of the measurement system is 0.03 mm.
作者
朱文鹏
汪志成
周显恩
Zhu Wenpeng;Wang Zhicheng;Zhou Xian’en(School of Mechanical and Electronic Engineering,East China University of Technology,Nanchang 330013,China;Ji’an Electronic Information Research Institute,Ji’an,Jiangxi 343099,China)
出处
《机电工程技术》
2023年第5期29-34,共6页
Mechanical & Electrical Engineering Technology
基金
国家自然科学基金青年基金(62103142)。
关键词
载带厚度测量
机器视觉
亚像素
离焦校正
thickness measurement
machine vision
sub-pixel
defocus correction
作者简介
朱文鹏(1994-),男,硕士研究生,研究领域为机器视觉;汪志成(1982-),男,博士,副教授,研究领域为人工智能应用。
