基于机器视觉的软包装边缘测量及纠偏系统

Soft Package Edge Measurement and Correction System Based on Machine Vision [JZ)]

摘要:

针对软包装带材在运动过程中易发生跑偏现象的工程背景,设计了一种自动测量与纠偏系统。介绍了系统组成和纠偏原理,系统采用机器视觉实时采集带材边缘图像,用直方图修正方法修正图象灰度,通过增强对比度突出图像中的带材边缘特征,用基于一阶微分的边缘检测方法实现边缘识别及判定边缘位置,运用FUZZY+PID相结合的控制模型实施纠偏控制；实验结果表明,采用LED光源和Microvision数字摄像机,在200 m/min中等车速和纠偏机构速度10 mm/s条件下,系统纠偏精度在±0～±1 mm之间,稳态误差小,相对误差小,能够满足软包装带材生产过程中的纠偏要求。

Abstract:

A method is proposed to solve the problem that soft packaging strip occurrence deviation in movement.The working principle and composition of the system were introduced, this system uses machine vision real-time capture strip edge, uses histogram modified method correction image gray, highlights the strip edge characteristic in the image through enhanced contrast,uses edge detection method based on first-order differential identify edge and determine the edge location, uses Fuzzy+PID combining model implement deviation control. The experiment results show that, the LED light source and Microvision digital camera, in 200 m/min medium speed and in 10 mm/s correcting mechanism speed, correction precision between ±0~±1 mm, steady-state error is small, the relative error is small. This system satisfies the needs of soft packaging strip correction.