基于机器视觉的带式输送机监控系统中输送带积水检测  

Water detection in belt conveyor monitoring system based on computer vision

作　　者：苗长云[1] 王春华 李现国[1] MIAO Chang-yun;WANG Chun-hua;LI Xian-guo(School of Electronics and Information Engineering,Tiangong University,Tianjin 300387,China)

机构地区：[1]天津工业大学电子与信息工程学院

出　　处：《天津工业大学学报》2019年第6期52-57,共6页Journal of Tiangong University

基　　金：国家自然科学基金项目(51274150);天津市应用基础及前沿技术研究计划重点项目(17JCZDJC31600);天津市重点研发计划科技支撑重点项目(18YFZCGX00930)

摘　　要：为了解决带式输送机需要人工检测输送带积水的问题,提出了一种基于机器视觉的带式输送机监控系统中输送带积水检测方案和基于自适应阈值的输送带表面积水检测算法,采用工业面阵相机采集输送带上表面图像,通过以太网传输给带式输送机监控系统,其上位机对图像进行处理;积水检测算法首先计算自适应阈值,并通过阈值化得到积水区域轮廓,再根据积水区域轮廓计算积水面积。实验结果表明:该积水检测方案和算法实现了带式输送机中输送带积水的实时检测和面积的计算,积水检测准确率达到98.4%,解决了带式输送机需要人工检测输送带积水的问题,提高了带式输送机的自动化水平,能够集成在带式输送机监控系统中,在港口、煤炭、矿山、电力等领域具有广泛应用前景。In order to solve the problem that the belt conveyor needs to manually detect the water in the conveyor belt,a water detection method in conveyor belt based on machine vision and a water detection algorithm based on adaptive threshold are proposed.The industrial surface array camera is used to collect the image of the upper conveyor belt,and transmitted to the belt conveyor monitoring system through Ethernet.The host computer of the monitoring system processes the image;the water accumulation detection algorithm first calculates the threshold adaptively.The contour of the water accumulation area is calculated according to the threshold and the area is calculated based on contour.The experimental results show that the water accumulation detection scheme and algorithm realize the real-time detection and area calculation of the water in the conveyor belt and the detection accuracy of the accumulated water reaches 98.4%,which solves the problem that the belt conveyor needs to manually detect the water accumulated in the conveyor belt,improves the automation level of the belt conveyor,and can be integrated in the belt conveyor monitoring.The system has broad application prospects in the fields of ports,coal,mining,and electric power.fields.

关 键 词：带式输送机 积水检测 偏振镜 机器视觉 监控系统 

分 类 号：TP274.5[自动化与计算机技术—检测技术与自动化装置] TD67[自动化与计算机技术—控制科学与工程]