基于机器视觉的矿用胶带输送机胶带跑偏监测研究  

Research on machine vision based monitoring of belt deviation in mining belt conveyors

作　　者：杨杰 YANG Jie(Production Service Center,National Energy Group Shendong Coal Group,Erdos 017010,China)

机构地区：[1]国家能源集团神东煤炭集团生产服务中心,内蒙古鄂尔多斯017010

出　　处：《中国高新科技》2025年第1期35-37,共3页

摘　　要：矿用胶带输送机工作过程中,监测结果受到周围环境的影响。文章提出了基于机器视觉的矿用胶带输送机胶带跑偏监测方法。运用机器视觉,建立监测框架,通过双目视觉成像系统采集现场图像,观察输送机胶带跑偏情况。依托于非局部均值算法对输送机胶带图像进行去噪处理,并采用引力检测算法优化后的Canny(坎尼)算子,检测图像包含的边缘信息。利用Hough(霍夫)直线变换思想获取图像边缘直线,识别输送机胶带的扭曲和偏移状态,完成胶带跑偏监测。实验结果表明,新研究方法监测结果的AUC值为0.87,可准确反映矿用胶带输送机胶带的工作状态。During the operation of the mining belt conveyor,the monitoring results are affected by the surrounding environment.This paper proposes a monitoring method of belt deviation of mining belt conveyor based on machine vision.Machine vision is used to establish a monitoring frame,and the binocular vision imaging system is used to collect field images and observe the deviation of conveyor belt.Based on the non-local mean algorithm,the conveyor belt image is denoised,and the Canny operator optimized by the gravity detection algorithm is used to detect the edge information contained in the image.The image edge line is obtained by using Hough linear transformation idea,and the distortion and offset state of conveyor belt are identified to complete the belt deviation monitoring.The experimental results show that the AUC value of the monitoring result of the new method is 0.87,which can accurately reflect the working state of the belt conveyor.

关 键 词：机器视觉 图像处理 边缘检测 中心线 输送机 胶带跑偏 

分 类 号：TD528[矿业工程—矿山机电]