机器视觉在堆煤检测系统中的应用  

Application of Machine Vision in Coal Pile Detection System

作　　者：徐慧琼[1] 王庆福[2] XU Huiqiong;WANG Qingfu(Shanxi Polytechnic College,Taiyuan 030006,China;Liaoning Academy of Governance,Shenyang 110161,China)

机构地区：[1]山西职业技术学院,太原030006 [2]辽宁行政学院,沈阳110161

出　　处：《煤炭技术》2022年第7期163-165,共3页Coal Technology

基　　金：山西省自然科学基金项目(201901D111056)。

摘　　要：对堆煤检测总体系统进行研究,通过线阵CCD相机进行图像采集,并通过CAN总线等通信方式,完成与图像处理模块、上位机、报警器以及执行机构的通信;然后对获取的实时图像进行预处理后,通过Soble算法完成图像中像素点的灰度值计算,以及梯度方向的判断,最终完成图像边缘检测;最后通过阈值线识别法和面积比例法进行“与”结合判定方法,完成图像处理结果与堆煤报警的转换,并通过实验验证了堆煤检测系统的可靠性。The overall system of coal pile detection is studied.Images are collected by linear CCD camera,and then the communication with image processing module,upper computer,alarm and actuator is completed through CAN bus and other communication methods.Then,after the real-time image is preprocessed.Soble algorithm is used to calculate the gray value of pixels in the image and judge the gradient direction and image edge detection is completed.Finally,through the threshold line identification method and area ratio method,the combination of "and" judgment method,the conversion of image processing results and coal pile alarm is completed.At last,iability of coal pile detection system has been verified through experiments.

关 键 词：机器视觉 堆煤检测 CAN通信 Soble算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TD528.1[自动化与计算机技术—计算机科学与技术]