单目相机与激光雷达点云融合的电力作业装备检测定位研究

打开文本图片集

中图分类号：TP391.4 文献标志码：B 文章编号：1671-5276（2025）03-0306-05

Monocular Camera and LiDAR Point Cloud Fusion for Power Operating Tool Detection

ZHAI Shixiong'，BAI Yuling1，ZHANG Liming¹，HU Yifei1，ZHANG Yuwei1，REN Shunan 2 （1.State Grid Tianjin Electric Power Company，Tianjin 3OoO1O，China； 2.Beijing Guodian Futong Technology Development Co.，Ltd.，Beijing 1OoO7O，China）

Abstract：Forthedetectingand positioningerorsofoperating tolsbyrobots during theprocessofpoweroperation，thispaper proposes a sensor fusion method that integrates monocular camera image and LiDAR point cloud data.ArUco markers on the operating toolaredetected，and theirposesunderthecameracordinatesystemareobtained.Basedonthejointcalibration betwencameraand LiDAR，theposes of marker planes under the LiDAR coordinate system areobtained through coordinate transformation.Withthenormalvectorof theplaneasthebaseline，andthepointcloudsofthemarkerplatesareobtained bythe filteringand clustering algorithm.Andthecharacteristicsof thetwomarkerplatepointcoudsare fusedtocalculatetheposeof theoperating tolundertheLiDARcoordinatesystem.Theexperimentunderoutdorhigh-intensitylightconditionsverifiesthat theproposedmethodcanefectivelyeliminatetheerrorcausedbythecameraintrinsiccallbrationandextrinsiccalibration between camera and LiDAR，accuratelydetect and locate the pose of the power operating tol in robot operation space.

Keywords：3D object detection；multiple sensor fusion；point cloud filter；point cloud cluster；electric robots

0 引言

近年来，随着人工智能与机器人技术的不断发展，带电作业机器人的发展取得了长足的进步，正在逐渐协助或取代电力工人完成各种配网带电作业任务[1-2]。（剩余7117字）