药卷传输姿态快速识别与诊断方法研究

工业炸药包装过程中,药卷传输姿态异常会影响后续工序;针对药卷传送皮带快速运动的特点,提出基于机器视觉的药卷传输姿态快速识别与诊断方法;该方法经过图像预处理,提取与药卷中轴线平行的药卷边缘直线,计算出药卷传输姿态的夹角θ,再由θ诊断药卷传输姿态;在搭建的仿真平台上,利用OpenCV开发仿真实验程序;实验结果表明,文中的方法每帧图像检测速度比药卷传输速度快4.8倍,可满足在线检测药卷传输姿态的要求。

Research on Method of Fast Identification and Diagnosis for Transmission Attitude of Explosive Roll

In the industrial explosive packaging process,the abnormal posture of explosive roll transmission attitude will affect the follow-up process.For the transmission attitude of explosive roll on the fast movement conveyer belt,a fast method of identification and diagnosis based on machine vision is proposed.In our proposed method,after an image preprocessed,the edge lines of explosive roll which parallel to its central axis is extracted from the image.Then the angle θ of roll transmission attitude is calculated,and its transmission attitude is identified.In our simulation platform,we use OpenCV for a experiment simulation program to come the proposed method true.The experiment results show that the detection speed of each frame image is 4.8 times faster than the actual explosive roll transfers using this method.This situation can meet the requirements of on-line detection for the roll transmission attitude.