基于关键帧提取与图像识别的电网设施数字化重建

doi:10.11930/j.issn.1004-9649.202306018

中国电力 ›› 2024, Vol. 57 ›› Issue (9): 217-223.DOI: 10.11930/j.issn.1004-9649.202306018

基于关键帧提取与图像识别的电网设施数字化重建

陈志永¹(), 韩璟琳¹^,²(), 刘洋¹(), 李光毅¹, 蒋雨³, 蔡瑞天⁴

1. 国网河北省电力有限公司经济技术研究院，河北石家庄　050021
2. 国网河北省电力有限公司，河北石家庄　050000
3. 国网河北省电力有限公司雄安新区供电公司，河北保定　071000
4. 华北电力大学电气与电子工程学院，北京　102206

收稿日期:2023-06-05 接受日期:2024-08-10 出版日期:2024-09-28 发布日期:2024-09-23
作者简介:陈志永（1971—），男，高级工程师，从事电网规划、微电网运行优化研究，E-mail：18207750059@163.com
韩璟琳（1989—），男，通信作者，硕士，从事电网规划、微电网运行优化研究，E-mail：jinglinoliver@163.com
刘洋（1982—），男，硕士，高级工程师，从事电网规划管理研究，E-mail：2429109351@qq.com
基金资助:
国家电网有限公司科技项目（5100-202113564A-0-5-SF）。

3D Reconstruction of Power Grid Facilities Based on Keyframe Extraction and Image Recognition

Zhiyong CHEN¹(), Jinglin HAN¹^,²(), Yang LIU¹(), Guangyi LI¹, Yu JIANG³, Ruitian CAI⁴

1. State Grid Hebei Electric Power Co. Ltd., Economic and Technical Research Institute, Shijiazhuang 050021, China
2. State Grid Hebei Electric Power Co. Ltd., Shijiazhuang 050000, China
3. State Grid Hebei Electric Power Co., Ltd., Xiong'an New Area Power Supply Company, Baoding 071000, China
4. School of Electrical and Electronic Engineering, North China Electric Power University, Beijing 102206, China

Received:2023-06-05 Accepted:2024-08-10 Online:2024-09-28 Published:2024-09-23
Supported by:
This work is supported by Science and Technology Project of SGCC (No.5100-202113564A-0-5-SF).

摘要/Abstract

摘要：

提出一种基于视频关键帧提取与图像识别的电网设施数字化重建方案。首先，通过移动作业终端拍摄待重建设施的视频序列，提取视频的关键帧并利用YOLOv5网络识别并提取电力设施锚框去除环境背景，构造图像集；然后，以运动恢复结构为原理进行三维重建，提出改进定向快速旋转算法提高相机位姿计算精度，进而实现电网设施三维模型重建；最后，以变压器为例，验证所提方法可以有效完成电力设施三维模型的重建。

关键词: 数字孪生, 关键帧提取, 图像识别

Abstract:

A digital reconstruction scheme for power grid facilities based on video keyframe extraction and image recognition is proposed. Shoot a video sequence of the facility to be reconstructed through a mobile homework terminal, extract key frames from the video, and use YOLOv5 network to recognize and extract anchor frames of power facilities to remove environmental background, constructing an image set; Then, based on the principle of motion recovery structure, 3D reconstruction is carried out, and an improved directional fast rotation algorithm is proposed to improve the accuracy of camera pose calculation. Thereby achieving the reconstruction of 3D models of power grid facilities; Finally, taking common transformers in the power grid as an example, it is verified that the proposed method can effectively complete the 3D model reconstruction of power facilities.

Key words: digital twins, keyframe extraction, image recognition

陈志永, 韩璟琳, 刘洋, 李光毅, 蒋雨, 蔡瑞天. 基于关键帧提取与图像识别的电网设施数字化重建[J]. 中国电力, 2024, 57(9): 217-223.

Zhiyong CHEN, Jinglin HAN, Yang LIU, Guangyi LI, Yu JIANG, Ruitian CAI. 3D Reconstruction of Power Grid Facilities Based on Keyframe Extraction and Image Recognition[J]. Electric Power, 2024, 57(9): 217-223.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202306018

https://www.electricpower.com.cn/CN/Y2024/V57/I9/217

图/表 11

图 1 基于视频关键帧提取与图像识别的三维重建方法架构

Fig.1 Architecture of the 3D reconstruction method based on video keyframe extraction and image recognition

图 2 基于视频关键帧提取与图像识别的三维重建方法

Fig.2 The three-dimensional reconstruction method based on video keyframe extraction and image recognition

图 3 清晰度评估值

Fig.3 Schematic diagram of the clarity assessment

图 4 识别检测结果

Fig.4 Transformer recognition test results

图 5 稀疏特征点提取结果

Fig.5 Sparse feature point extraction results

表 1 特征检测算法的比对结果

Table 1 Comparison of feature detection algorithms

算法	提取特征数/个	正确匹配特征/对	匹配准确率/%
SIFT	327	233	71.3
SURF	286	209	73.1
原始ORB	397	304	76.6
本文改进ORB	348	286	82.2

图 6 有载调压电力变压器重建结果

Fig.6 Reconstruction of the on-load tap changer transformer

图 7 普通电力变压器重建结果

Fig.7 Reconstruction of the common transformer

图 8 使用原始图像重建结果示意

Fig.8 Reconstruction based on the original image

图 9 固定重投影误差点云对比结果

Fig.9 Comparison of reprojection error point cloud

图 10 点云平均重投影误差对比结果

Fig.10 Comparison of point cloud average reprojection error

参考文献 25

1	苏寅生, 周挺辉, 郑外生, 等. 基于云计算的电力系统计算分析平台构建[J]. 南方电网技术, 2022, 16 (7): 67- 75.
	SU Yinsheng, ZHOU Tinghui, ZHENG Waisheng, et al. Construction of power system computing and analysis platform based on cloud computing[J]. Southern Power System Technology, 2022, 16 (7): 67- 75.
2	孙超, 夏晨, 吕志强, 等. 省地一体化稳态分析服务弹性调度系统设计[J]. 电力系统保护与控制, 2022, 50 (5): 136- 145.
	SUN Chao, XIA Chen, LÜ Zhiqiang, et al. Design of a service elastic scheduling system for steady state analysis across provincial and regional dispatching centers[J]. Power System Protection and Control, 2022, 50 (5): 136- 145.
3	何永胜, 罗丹, 鲁宗相. 特高压阀侧套管数字孪生系统集成设计方法[J]. 中国电力, 2024, 57 (6): 165- 173.
	HE Yongsheng, LUO Dan, LU Zongxiang. Integration of digital twin model system for ultra-high voltage converter transformer valve-side bushing[J]. Electric Power, 2024, 57 (6): 165- 173.
4	孙毅, 陈恺, 左强, 等. 考虑5G通信负荷协同优化的云边计算网络能量管理模型[J]. 中国电机工程学报, 2023, 43 (23): 9020- 9033.
	SUN Yi, CHEN Kai, ZUO Qiang, et al. Energy management model of cloud-edge computing network considering the coordinated optimization of 5G communication load[J]. Proceedings of the CSEE, 2023, 43 (23): 9020- 9033.
5	梁珩, 黄耕, 侯宾, 等. 工业用户连续参与需求响应的用户基线负荷精准计算方法[J]. 中国电力, 2024, 57 (3): 34- 42.
	LIANG Heng, HUANG Geng, HOU Bin, et al. Accurate estimation method of customer baseline load for continuous participation of industrial users in demand response[J]. Electric Power, 2024, 57 (3): 34- 42.
6	李博, 方彤. 北斗卫星导航系统(BDS)在智能电网的应用与展望[J]. 中国电力, 2020, 53 (8): 107- 116.
	LI Bo, FANG Tong. Application and prospect of BeiDou navigation satellite system(BDS) in smart grid[J]. Electric Power, 2020, 53 (8): 107- 116.
7	朱睿, 欧乙丁, 李筱天, 等. 基于外特性等值的虚拟电厂灵活性资源价值评估[J]. 中国电力, 2024, 57 (1): 30- 39.
	ZHU Rui, OU Yiding, LI Xiaotian, et al. Evaluation of virtual power plant flexibility resources based on external characteristic equivalence[J]. Electric Power, 2024, 57 (1): 30- 39.
8	刘林, 王大龙, 綦晓, 等. 基于双锁相环的海上风场综合惯量调频策略研究[J]. 发电技术, 2024, 45 (2): 282- 290. DOI
	LIU Lin, WANG Dalong, QI Xiao, et al. Study on double phase-locked loop on the synthetic inertia control of offshore wind farm frequency regulation[J]. Power Generation Technology, 2024, 45 (2): 282- 290. DOI
9	钱仲豪, 胡骏, 沈思辰, 等. 考虑条件风险价值的多源协调优化运行策略[J]. 发电技术, 2023, 44 (6): 781- 789. DOI
	QIAN Zhonghao, HU Jun, SHEN Sichen, et al. Multi-power coordinated optimization operation strategy considering conditional value at risk[J]. Power Generation Technology, 2023, 44 (6): 781- 789. DOI
10	杜夏恒, 赫玉莹, 邹文, 等. 跟网型和构网型逆变器的阻抗无源化方法综述[J]. 东北电力大学学报, 2024, 44 (2): 12- 20.
	DU Xiaheng, HE Yuying, ZOU Wen, et al. An overview of impedance passivation methods for grid-following and grid-forming inverters[J]. Journal of Northeast Electric Power University, 2024, 44 (2): 12- 20.
11	陈继开, 侯文宇, 李浩茹, 等. 一种基于虚拟惯量控制的非故障雷击多端柔直低频振荡平抑方法[J]. 东北电力大学学报, 2023, 43 (4): 10- 19.
	CHEN Jikai, HOU Wenyu, LI Haoru, et al. A non-fault lightning strike multi-terminal flexible low-frequency oscillation suppression method based on virtual inertia control[J]. Journal of Northeast Electric Power University, 2023, 43 (4): 10- 19.
12	赵辉, 韩璟琳, 李光毅, 等. 基于运动恢复结构原理的电网设施三维静态重建方法[J/OL]. 现代电力: 1–9[2023-06-05]. https://doi.org/10.19725/j.cnki.1007-2322.2022.0402.
	ZHAO Hui, HAN Jinglin, LI Guangyi, et al. 3D static reconstruction method of power grid facilities based on structure from motion principle[J/OL]. Modern Electric Power: 1–9[2023-06-05]. https://doi.org/10.19725/j.cnki.1007-2322.2022.0402.
13	ZHANG K, YAN M Z, HUANG T Y, et al. 3D reconstruction of complex spatial weld seam for autonomous welding by laser structured light scanning[J]. Journal of Manufacturing Processes, 2019, 39, 200- 207. DOI
14	冯欣楠, 鲁志伟, 梁记云, 等. 基于三维模型法500 kV交流输电线路绕击耐雷性能分析[J]. 东北电力大学学报, 2022, 42 (1): 87- 95.
	FENG Xinnan, LU Zhiwei, LIANG Jiyun, et al. Analysis of lightning protection performance of 500 kV AC transmission lines based on three-dimensional model method[J]. Journal of Northeast Electric Power University, 2022, 42 (1): 87- 95.
15	李智韬. 基于无人机摄影测量的架空输电线路三维实景建模研究与应用[D]. 广州: 华南理工大学, 2019.
	LI Zhitao. Studies and apllications of overhead transmission lines' 3D real scene modeling based on drones equiped with imaging and measuring devices[D]. Guangzhou: South China University of Technology, 2019.
16	郑恩. 基于无人机航拍输电线路杆塔的三维重建[D]. 南宁: 广西大学, 2017.
	ZHENG En. 3D reconstruction of tower of aerial transmission line based on UAV[D]. Nanning: Guangxi University, 2017.
17	刘磊, 孙敏, 任翔, 等. 基于无人机影像序列的三维重建方法综述[J]. 北京大学学报(自然科学版), 2017, 53 (6): 1165- 1178.
	LIU Lei, SUN Min, REN Xiang, et al. Review on methods of 3D reconstruction from UAV image sequences[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2017, 53 (6): 1165- 1178.
18	张童, 谭南林, 包辰铭. 应用于嵌入式平台的实时红外行人检测方法[J]. 激光与红外, 2020, 50 (2): 239- 245.
	ZHANG Tong, TAN Nanlin, BAO Chenming. Real-time infrared pedestrian detection method applied to embedded platform[J]. Laser & Infrared, 2020, 50 (2): 239- 245.
19	ZHOU H Y, ZHANG T. The combination of SfM and monocular SLAM[C]//The 26th Chinese Control and Decision Conference (2014 CCDC). Changsha, China. IEEE, 2014: 5282–5286.
20	苏凯第, 赵巧娥. 基于YOLOv5算法的无人机电力巡检快速图像识别[J]. 电力科学与工程, 2022, 38 (4): 43- 48. DOI
	SU Kaidi, ZHAO Qiaoe. Fast image recognition of UAV power inspection based on YOLOv5 algorithm[J]. Electric Power Science and Engineering, 2022, 38 (4): 43- 48. DOI
21	庞博, 鲍志远, 杨明坤, 等. 基于YOLOv5的电力巡检图像缺陷识别研究[J]. 四川电力技术, 2022, 45 (5): 48- 53, 94.
	PANG Bo, BAO Zhiyuan, YANG Mingkun, et al. Research on defect recognition of power patrol images based on YOLOv5[J]. Sichuan Electric Power Technology, 2022, 45 (5): 48- 53, 94.
22	王文斌, 刘桂华, 刘先勇, 等. 本质矩阵五点算法伪解的两种剔除策略[J]. 光电工程, 2010, 37 (8): 46- 52.
	WANG Wenbin, LIU Guihua, LIU Xianyong, et al. Two removal tacties of pseudo solutions for essential matrix five-point algorithm[J]. Opto-Electronic Engineering, 2010, 37 (8): 46- 52.
23	姚依妮, 王玮. Harris角点检测算法的应用研究[J]. 智能计算机与应用, 2022, 12 (8): 148- 151. DOI
	YAO Yini, WANG Wei. Application research on Harris corner detection algorithms[J]. Intelligent Computer and Applications, 2022, 12 (8): 148- 151. DOI
24	张炎, 野莹莹, 马永轩, 等. KLT跟踪算法[J]. 信息技术与信息化, 2021, (5): 21- 23. DOI
	ZHANG Yan, YE Yingying, MA Yongxuan, et al. KLT tracking algorithm[J]. Information Technology and Informatization, 2021, (5): 21- 23. DOI
25	别治峰, 刘守山, 黄春凤. 基于SURF-RANSAC配准的三维重建[J]. 科学技术与工程, 2019, 19 (28): 239- 244. DOI
	BIE Zhifeng, LIU Shoushan, HUANG Chunfeng. 3D reconstruction based on SURF-RANSAC matching[J]. Science Technology and Engineering, 2019, 19 (28): 239- 244. DOI

[1]	何永胜, 罗丹, 鲁宗相. 特高压阀侧套管数字孪生系统集成设计方法[J]. 中国电力, 2024, 57(6): 165-173.
[2]	刘曌, 孙庆凯, 许泽凯, 吴潇雨, 王小君, 吕金玲. 能源互联网中的数字孪生技术体系、应用与挑战[J]. 中国电力, 2024, 57(1): 230-243.
[3]	韩伟, 段文岩, 杜兴伟, 姚峰, 马伟东, 刘磊. 基于数字孪生的在运安控系统故障诊断方法[J]. 中国电力, 2023, 56(11): 121-127.
[4]	赵永良, 付鑫, 郭阳, 边迎迎, 王思宁. 基于深度学习和图像识别的电力配件智能出入库[J]. 中国电力, 2021, 54(3): 55-60.
[5]	周俊煌, 黄廷城, 谢小瑜, 范纹郡, 易婷婷, 张勇军. 视频图像智能识别技术在输变电系统中的应用研究综述[J]. 中国电力, 2021, 54(1): 124-134,166.
[6]	蒋晶, 赵洋洋, 樊宏伟, 周振宇, 董朝阳, 杨青波, 王晓丽. 图像识别在特高压换流阀元件故障在线监测系统的应用[J]. 中国电力, 2020, 53(11): 126-132.
[7]	赖秋频, 杨军, 谭本东, 王亮, 傅思遥, 韩立伟. 基于YOLOv2网络的绝缘子自动识别与缺陷诊断模型[J]. 中国电力, 2019, 52(7): 31-39.
[8]	贾思棋, 李军辉, 杜冬梅, 刘彬, 何青. 基于随机Hough变换的线路覆冰厚度图像识别技术[J]. 中国电力, 2019, 52(12): 39-45,53.
[9]	张宇娇, 杨骐, 姜岚, 徐天勇, 董晓虎. 一种基于视频图像识别的输电线路屏蔽环振动频率监测方法[J]. 中国电力, 2018, 51(3): 35-41.
[10]	董晓峰, 顾晓文, 王业, 朱国防. 电网倒闸操作票智能监护系统设计[J]. 中国电力, 2015, 48(12): 16-22.

基于关键帧提取与图像识别的电网设施数字化重建

3D Reconstruction of Power Grid Facilities Based on Keyframe Extraction and Image Recognition

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 25

相关文章 10

编辑推荐

Metrics

模态框（Modal）标题

基于关键帧提取与图像识别的电网设施数字化重建

3D Reconstruction of Power Grid Facilities Based on Keyframe Extraction and Image Recognition

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 25

相关文章 10

编辑推荐

Metrics