基于单目机器视觉的高压输电线路障碍物定位研究
|
摘要
高压输电线巡线机器人运行线路上的障碍物识别定位,是实现巡线机器人自主导航的关键技术之一。针对110k V高压输电线路地线结构的特点,提出了基于单目机器视觉的高压输电线路障碍物定位研究方法。通过在巡检机器人本体上安装单目摄像机,利用单目摄像机采集前方线路图像,对图像中的障碍物进行识别,利用识别出来的障碍物位置中心与摄像机的位置关系建立测距几何模型,来对巡线机器人前方障碍物进行定位。该方法可以较精确地对巡线机器人前方约(1-2)m内的障碍物进行定位并测距,且误差小、识别率高、响应快。最后,通过实验验证该方法有较高的可行性和有效性,能够极大提高巡线机器人自主导航能力。
Obstacle recognition and orientation of high voltage transmission line inspection robot on the line, is one of the key techniques for implementation of inspection robot autonomous navigation. In light of the wire structure of 110 k V high voltage transmission line, a method of obstacle recognition and orientation based on monocular machine vision is put forward. By installing a monocular camera on a patrol robot body, using monocular camera acquisition front line image, identifies obstacles of the image, establishes distance geometry model by using the relationship between the identified obstacles central location and the camera position, to locate obstacles in front of the inspection robot. This method can be more accurate to position and range obstacles within about( 1-2) meters in front of the robot, and it has small error, high recognition rate, and fast response. Finally, experiments results verify that this method has high feasibility and effectiveness,can greatly improve the ability for autonomous navigation of inspection robot.
Obstacle recognition and orientation of high voltage transmission line inspection robot on the line, is one of the key techniques for implementation of inspection robot autonomous navigation. In light of the wire structure of 110 k V high voltage transmission line, a method of obstacle recognition and orientation based on monocular machine vision is put forward. By installing a monocular camera on a patrol robot body, using monocular camera acquisition front line image, identifies obstacles of the image, establishes distance geometry model by using the relationship between the identified obstacles central location and the camera position, to locate obstacles in front of the inspection robot. This method can be more accurate to position and range obstacles within about( 1-2) meters in front of the robot, and it has small error, high recognition rate, and fast response. Finally, experiments results verify that this method has high feasibility and effectiveness,can greatly improve the ability for autonomous navigation of inspection robot.
引文
[1]SAWADA jun,KUSUMOTO kazuyuki,MAIKAWA yasuhisa.A mobile robot for inspection of power transmission lines[J].IEEE Transaction on Power Delivery,1991,6(1):309-315.
[2]MONTAMBAULT S,POULIOT N.Robots repair and examine live lines in sever condition[J].Electrical World,1989(5):71-72.
[3]郭磊,徐友春,李克强.基于单目视觉的实时测距方法研究[J].中国图象图形学报,2006,11(1):74-81.(Guo Lei,Xu You-chun,Li Ke-qiang.Study on real-time distance detection based on monocular vision technique[J].Journal of Image and Graphics,2006,11(1):74-81.)
[4]王力超,熊超,王晨毅.基于竞争机制的简化双目立体视觉测距算法及系统设计[J].传感技术学报,2007,20(1):150-153.(Wang Li-chao,Xiong Chao,Wang Chen-yi.Simplified binocular stereo distance-measurement algorithm and a system based on competition[J],Chinese Journal of Sensors and Actuators,2007,20(1):150-153.)
[5]施小成,王晓娟.一种面向AUV水下对接的双目视觉测距方法[J].计算机测量与控制,2008(10):1460-1462+1488.(Shi Xiao-cheng,Wang Xiao-juan.A ranging method in AUV underwater docking by binocular vision[J].Computer Measurement&Control,2008(10):1460-1462+1488.)
[6]胡彩石,吴功平,曹珩.高压输电线路巡线机器人障碍物视觉检测识别研究[J].传感技术学报,2008,21(12):2092-2096.(Hu Cai-shi,Wu Gong-ping,Cao Hang.Research of obstacle recognition based on vision for high voltage transmission line inspection robot[J].Chinese Journal of Sensors and Actuators,2008,21(12):2092-2096.)
[7]章毓晋.图像理解与计算机视觉[M].北京:清华大学出版社,2000(8).(Zhang Yu-jin.Image Understanding and Computer Vision[M].Beijing:Tsinghua University Press,2000(8).)
[2]MONTAMBAULT S,POULIOT N.Robots repair and examine live lines in sever condition[J].Electrical World,1989(5):71-72.
[3]郭磊,徐友春,李克强.基于单目视觉的实时测距方法研究[J].中国图象图形学报,2006,11(1):74-81.(Guo Lei,Xu You-chun,Li Ke-qiang.Study on real-time distance detection based on monocular vision technique[J].Journal of Image and Graphics,2006,11(1):74-81.)
[4]王力超,熊超,王晨毅.基于竞争机制的简化双目立体视觉测距算法及系统设计[J].传感技术学报,2007,20(1):150-153.(Wang Li-chao,Xiong Chao,Wang Chen-yi.Simplified binocular stereo distance-measurement algorithm and a system based on competition[J],Chinese Journal of Sensors and Actuators,2007,20(1):150-153.)
[5]施小成,王晓娟.一种面向AUV水下对接的双目视觉测距方法[J].计算机测量与控制,2008(10):1460-1462+1488.(Shi Xiao-cheng,Wang Xiao-juan.A ranging method in AUV underwater docking by binocular vision[J].Computer Measurement&Control,2008(10):1460-1462+1488.)
[6]胡彩石,吴功平,曹珩.高压输电线路巡线机器人障碍物视觉检测识别研究[J].传感技术学报,2008,21(12):2092-2096.(Hu Cai-shi,Wu Gong-ping,Cao Hang.Research of obstacle recognition based on vision for high voltage transmission line inspection robot[J].Chinese Journal of Sensors and Actuators,2008,21(12):2092-2096.)
[7]章毓晋.图像理解与计算机视觉[M].北京:清华大学出版社,2000(8).(Zhang Yu-jin.Image Understanding and Computer Vision[M].Beijing:Tsinghua University Press,2000(8).)