基于车道线检测的移动机器人巡线方法研究  

Research on Mobile Robot Line Patrolling Based on Lane Line Detection Method

作　　者：薛源 张文典[2] 黄家才 王徐寅 汤文俊 汪涛 XUE Yuan;ZHANG Wendian;HUANG Jiacai;WANG Xuyin;TANG Wenjun;WANG Tao(School of Mechanical Engineering,Nanjing Institute of Technology,Nanjing 211167,China;Electrical Engineering Academy,Changchun Automobile Industry Institute,Changchun 130013,China)

机构地区：[1]南京工程学院机械工程学院,江苏南京211167 [2]长春汽车工业高等专科学校电气工程学院,吉林长春130013

出　　处：《南京工程学院学报（自然科学版）》2023年第4期71-77,共7页Journal of Nanjing Institute of Technology(Natural Science Edition)

摘　　要：SLAM技术在移动机器人巡检领域得到广泛应用,但对于环境特征单一的大型室内场景,SLAM技术的定位建图效果不能满足实际需求.针对大型仓库下移动机器人巡检问题提出基于车道线检测的移动机器人巡线方法.在HSV颜色空间下,首先采用多阈值叠加方法对采集的实时视频图像进行颜色阈值分割;然后通过霍夫变换与过滤得到与车道线唯一贴合的直线;最后通过该直线参数矫正移动机器人运动时的方向.试验结果表明:基于HSV阈值分割的车道线检测方法能有效检测车道线,正确率可达99.81%;该方法可使移动机器人正确调整运动方向,调整成功率可达93.33%.SLAM technology is widely employed in the domain of mobile robot inspection,however,its effectiveness in positioning and mapping falls short of practical requirements when dealing with large indoor environments characterized by a singular environmental feature.Addressing the challenges posed by mobile robot inspection in extensive warehouses,this paper explores a method for mobile robot line inspection based on lane line detection.In the HSV color space,the multi-threshold superposition method is first used to perform color threshold segmentation on the collected real-time video images,and the unique lane line is then determined through Hough transformation and filtering.Finally,the direction of mobile robot is corrected using the parameters of the identified straight line.Experimental results show that the lane line detection method based on HSV threshold segmentation can effectively detect lane lines,with an accuracy rate of 99.81%.This method facilitates the correct adjustment of the mobile robot's movement direction,achieving a success rate of 93.33%.

关 键 词：车道线检测 HSV颜色空间 阈值分割 霍夫变换 移动机器人 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]