摘要
人工测量医药罐体瓶口螺纹尺寸的精度与重复性较差,使用三坐标机虽能提高精度但检测速度慢,针对以上缺点,本文设计了基于点激光和视觉引导的瓶口螺纹测量系统。该系统能实现罐体的夹取、视觉引导定位以及瓶口螺纹的高精度测量。首先,针对上料位置的动态变化,使用机器视觉算法完成罐体的精确定位;其次,针对螺纹尺寸测量的高精度要求,使用激光测距传感器完成螺纹轮廓的数据采集;最后,针对螺纹的复杂外形,使用最小二乘法拟合螺纹轮廓,分别计算螺纹相关尺寸。试验结果表明,该系统能够准确地测量出不同规格的瓶口螺纹尺寸,绝对精度、重复精度分别能够达到0.01 mm和0.003 mm,基本满足医药罐体瓶口螺纹的测量要求,在实际生产中具有良好的可行性。
The accuracy and repeatability of manually measuring the medical tank’s bottle-mouth thread size are poor. Although the accuracy can be improved by using three coordinate machine, the detection speed is slow. In view of the above shortcomings, a measurement system of bottle thread based on point laser and vision guidance is designed in this paper. The system can realize the clamping of can body, visual guidance and positioning, and high-precision measurement of bottle mouth thread. Firstly, aiming at the dynamic change of feeding position, the machine vision algorithm is used to complete the accurate positioning of can body. Secondly, according to the high precision requirements of thread size measurement, the laser ranging sensor is used to complete the data acquisition of thread profile. Finally, for the complex shape of the thread, the least square method is used to fit the thread contour, and then the relevant dimensions of the thread are calculated respectively. The results show that the system can accurately measure the size of bottle mouth thread of different specifications, and the absolute accuracy and repetition accuracy can reach 0.01 mm and 0.003 mm respectively, which basically meets the measurement requirements of bottle mouth thread of medical tank, and has good feasibility in practical production.
作者
张爱云
王吉华
高崴
ZHANG Aiyun;WANG Jihua;GAO Wei(School of Automobile and Transportation Wuxi Instiute of Technology,Wuxi Jiangsu 214121,China;Wuxi Fuel Injection Equipment Instiute,Faw.Wuxi Jiangsu 214063,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2021年第12期1697-1704,共8页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(61773182)。
关键词
机器视觉
螺纹测量
视觉引导
梯度特征
最小二乘法
machine vision
thread measurement
vision guidance
gradient characteristics
least square method