摘要
输电线异物检测普遍采用无人机拍照,但需要人工复检,导致电力巡检效率偏低,为解决这一问题,提出输电通道交叉跨越段导地线异物三维信息检测算法。通过智能无人机拍摄相关区域,并采用双目立体视觉理论,获取异物图像的三维深度信息;引入双边滤波算法消除图像中的噪声及其他因素的干扰,对于动态异物,算法可分析其特征并完成连续跟踪;通过计算异物与导地线之间的安全距离,完成异物检测。结果表明,该方法能够精准地检测出多幅输电通道交叉跨越段导地线图像中的各类异物,且检测精度较高。
For foreign matter detection of power transmission lines,unmanned aerial vehicles are generally used,but manual reinspection is required,resulting in low efficiency of power inspection.In order to solve this problem,three-dimensional information detection of foreign matters for conductor and ground wires in crossing section of transmission channel was proposed.Through the intelligent unmanned aerial vehicle(UAV)shooting the relevant area,the three-dimensional depth information of foreign object image was obtained by using the binocular stereo vision theory.The bilateral filtering algorithm was introduced to eliminate the noise and other interfering factors in the image.For the dynamic foreign matter,the characteristics were analyzed and the continuous tracking was completed,the safe distance between the foreign object and the grounding wire was calculated to finish foreign matter detection.The results show that the as-proposed method can accurately detect all kinds of foreign matters in multiple images of the conductor and ground wires in the crossing section of transmission channel with relatively higher detection accuracy.
作者
朱登杰
李昊
李锐海
廖永力
张志强
何锦强
ZHU Dengjie;LI Hao;LI Ruihai;LIAO Yongli;ZHANG Zhiqiang;HE Jinqiang(School of Energy Power and Mechanical Engineering,North China Electric Power University,Baoding 071003,Hebei,China;Production Technology Support Center,Southern Power Grid Research Institute Co.,Ltd.,Guangzhou 510000,Guangdong,China)
出处
《沈阳工业大学学报》
CAS
北大核心
2023年第6期680-685,共6页
Journal of Shenyang University of Technology
基金
国家重点研发计划项目(2021YEE0204200)
广西电网公司科技项目(GXKJXM20190616)。
关键词
输电通道
交叉跨越段导地线
异物检测
三维信息
双目视觉系统
安全距离
双边滤波算法
动态跟踪
transmission channel
conductor and ground wire in crossing section
foreign matter detection
three-dimensional information
binocular vision system
safe distance
bilateral filtering algorithm
dynamic tracking
