基于自注意力感知的配电网杆塔点云语义分割

doi:10.3969/j.issn.1008-0198.2024.06.011

湖南电力 ›› 2024, Vol. 44 ›› Issue (6): 76-82.doi: 10.3969/j.issn.1008-0198.2024.06.011

基于自注意力感知的配电网杆塔点云语义分割

黄志鸿^1,2, 刘宇³, 张辉³, 徐先勇¹, 彭金柱⁴

1.国网湖南省电力有限公司电力科学研究院,湖南长沙 410208;
2.湖南省湘电试验研究院有限公司,湖南长沙 410208;
3.湖南大学机器人学院,湖南长沙 410082;
4.郑州大学电气与信息工程学院,河南郑州 450001

收稿日期:2024-07-29 修回日期:2024-09-29 出版日期:2024-12-25 发布日期:2024-12-25
通信作者: 刘宇(2000),女,硕士,研究方向为电力场景三维点云语义分割。
作者简介:黄志鸿(1993),男,博士,高级工程师,研究方向为电力设备故障智能诊断。张辉(1983),男,博士,教授,博士生导师,研究方向为机器人视觉检测与控制、深度学习图像处理。徐先勇(1982),男,博士,高级工程师,研究方向为电力智能巡检、调频式谐振特高压电源等。彭金柱（1980）,男,博士,教授,博士生导师,研究方向为智能控制理论及其在电力控制与传动技术中的应用。
基金资助:
湖南省杰出青年科学基金（2021JJ10025); 国网湖南省电力有限公司科技项目(5216A522001Y); 湖南省科技人才托举工程-“小荷”科技人才项目（2023TJ-X48）

Point Cloud Semantic Segmentation of Distribution Net‍work Tower Based on Self-Attention Perception

HUANG Zhihong^1,2, LIU Yu³, ZHANG Hui³, XU Xianyong¹, PENG Jinzhu⁴

1. State Grid Hunan Electric Power Company Limited Research Institute, Changsha 410208, China;
2. Hunan Xiangdian Test and Research Institute Co., Ltd., Changsha 410208, China;
3. School of Robotics,Hunan University, Changsha 410082, China;
4. School of Electrical and Information Engineering, Zhengzhou University, Zhengzhou 450001, China

Received:2024-07-29 Revised:2024-09-29 Online:2024-12-25 Published:2024-12-25

摘要/Abstract

摘要： 从机载激光点云数据中自动、准确地提取配电场景中的电力杆塔是常规电力巡检的关键步骤。由于山区地形起伏大,植被茂密,特征难以区分,利用现有方法从山区电力场景中提取目标仍具有挑战。为了解决这个问题,提出一种基于自注意力感知的语义分割方法,主要包括颜色特征提取、局部位置编码和多层注意力感知模块。该方法在处理大规模配电网场景点云时,可将RGB特征作为点云空间特征的附着属性进行融合,从而提高语义分割精度。经实验结果表明,该方法在电力场景中具有显著的性能优势,点云分割精度超过了90%,其有效性在公共数据集S3DIS中得到验证。

关键词: 电力巡检, RGB, 激光点云, 语义分割, 深度学习

Abstract: Automatically and accurately extracting power towers from airborne laser point cloud data in power distribution scenes is a critical step in routine power inspections. Due to the undulating terrainand dense vegetationin mountainous areas,the features are difficult to distinguish, and it remains challenging to extract targets from mountainous power scenes using existing methods. To solve this problem, a semantic segmentation method based on self-attention perception is proposed,which mainly includes color feature extraction,local position encoding,and multi-layer attention perception modules. When processing large-scale point clouds of distribution network scenes, this method can integrate RGB features as attached attributes of the point cloud's spatial features,thereby improving the accuracy of semantic segmentation. Experimental results show that the proposed method has significant performance advantages in power scenarios, achieving a point cloud segmentation accuracy of over 90%, and its effectiveness has been validated on the public S3DIS dataset.

Key words: power inspection, RGB, laser point cloud, semantic segmentation, deep learning

中图分类号:

TM754

黄志鸿, 刘宇, 张辉, 徐先勇, 彭金柱. 基于自注意力感知的配电网杆塔点云语义分割[J]. 湖南电力, 2024, 44(6): 76-82.

HUANG Zhihong, LIU Yu, ZHANG Hui, XU Xianyong, PENG Jinzhu. Point Cloud Semantic Segmentation of Distribution Net‍work Tower Based on Self-Attention Perception[J]. Hunan Electric Power, 2024, 44(6): 76-82.

参考文献

[1] LIU X,SHUANG F,LI Y,et al.SS-IPLE:semantic segmentation of electric power corridor scene and individual power line extraction from uav-based lidar point cloud[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2023,36:38-50.
[2] MILLS S J,CASTRO M P G,LI Z R,et al. Evaluation of aerial remote sensing techniques for vegetation management in power-line corridors[J]. IEEE Transactions on Geoscience and Remote Sensing,2010,48(9):3379-3390.
[3] 杨洪朝,席燕辉,项胜,等. 基于深度学习的输电线路巡检无人机图像分类方法[J]. 湖南电力,2024,44(1):53-58.
[4] LI W,LUO Z,XIAO Z,et al.A GCN-based method for extracting power lines and pylons from airborne LiDAR data[J]. IEEE Transactions on Geoscience and Remote Sensing,2022,60:1-14.
[5] HE T,ZENG Y,HU Z.Research of multi-rotor UAVs detailed autonomous inspection technology of transmission lines based on route planning[J]. IEEE Access,2019,7:114955-114965.
[6] AZEVEDO F,DIAS A,ALMEIDA J,et al.Lidar-based real-time detection and modeling of power lines for unmanned aerial vehicles[J]. Sensors,2019,19(8):1812.
[7] SHEN X J,QIN C,DU Y,et al.An automatic extraction algorithm of high voltage transmission lines from airborne LIDAR point cloud data[J]. Turkish Journal of Electrical Engineering and Computer Sciences,2018,26(4):2043-2055.
[8] WANG Y J,CHEN Q,LIU L,et al.Supervised classification of power lines from airborne LiDAR data in urban areas[J]. Remote Sensing,2017,9(8):771.
[9] YANG J,KANG Z.Voxel-based extraction of transmission lines from airborne LiDAR point cloud data[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2018,11(10):3892-3904.
[10] YADAV M,CHOUSALKAR C G.Extraction of power lines using mobile LiDAR data of roadway environment[J]. Remote Sensing Applications:Society and Environment,2017,8:258-265.
[11] GUO B,LI Q Q,HUANG X F,et al.An improved method for power-line reconstruction from point cloud data[J]. Remote sensing,2016,8(1):36.
[12] LECUN Y,BENGIO Y,HINTON G.Deep learning[J]. Nature,2015,521(7553):436-444.
[13] LU X,ZHONG Y,ZHENG Z,et al.Multi-scale and multi-task deep learning framework for automatic road extraction[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019,57(11):9362-9377.
[14] HUANG J J,SHEN Y Q,WANG J G,et al.Automatic pylon extraction using color-aided classification from UAV LiDAR point cloud data[J].IEEE Transactions on Instrumentation and Measurement, 2023,72:3293543.
[15] SU H,MAJI S,KALOGERAKIS E,et al.Multi-view convolutional neural networks for 3D shape recognition[C]//2015 IEEE International Conference on Computer Vision (ICCV).Santiago,Chile.2015:945-953.
[16] MATURANA D,SCHERER S.VoxNet:A 3D convolutional neural network for real-time object recognition[C]//2015 IEEE/RSJ International Conference on Intelligent Robots and Systems(IROS). IEEE,2015.
[17] WANG Y,SUN Y,LIU Z,et al.Dynamic graph CNN for learning on point clouds[J].ACM Transactions on Graphics(tog),2019,38(5):1-12.
[18] QI C R,SU H,MO K,et al.PointNet:Deep learning on point sets for 3d classification and segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Honolulu,HI.IEEE, 2017: 652-660.
[19] QI C R,YI L,SU H,et al.PointNet++:Deep hierarchical feature learning on point sets in a metric space[J]. Neural Information Processing Systems,2017,30.
[20] LI G H,MULLER M,THABET A,et al.Deep GCNs:Can GCNs go as Deep as CNNs?[C]//2019 IEEE/CVF International Conference on Computer Vision.2019:9267-9276.
[21] LIN H,ZHENG X,LI L,et al.Meta architecture for point cloud analysis[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023:17682-17691.

基于自注意力感知的配电网杆塔点云语义分割

Point Cloud Semantic Segmentation of Distribution Net‍work Tower Based on Self-Attention Perception

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 6

编辑推荐

Metrics

本文评价

[1]	黄志鸿, 颜星雨, 陶岩, 张辉, 徐先勇. 基于多模态图像信息的配电网部件定位方法[J]. 湖南电力, 2024, 44(6): 83-89.
[2]	瞿旭, 龙崦平, 张伟, 袁超雄. 基于位图和矢量图联合识别的电气二次图纸解析方法研究[J]. 湖南电力, 2024, 44(6): 97-103.
[3]	刘云飞, 周光远, 尹凤梅, 刘聪, 郭清玥, 黄照阳. 基于改进小波阈值的无人机巡检图像去噪方法[J]. 湖南电力, 2024, 44(6): 113-119.
[4]	张家盛, 梁进兴. 基于深度学习的无人机巡检架空输电线路金具锈蚀缺陷检测方法[J]. 湖南电力, 2022, 42(5): 75-78.
[5]	李振华, 李浩, 黄景光, 张磊, 吴琳. 基于声信号和一维卷积神经网络的绝缘子污秽度监测方法研究[J]. 湖南电力, 2022, 42(4): 18-22.
[6]	周可慧, 肖剑, 张可人, 徐先勇, 唐海国, 王珂, 朱吉然. 变压器套管运行状态自动检测技术现状分析[J]. 湖南电力, 2020, 40(6): 40-.