基于改进RetinaNet算法的输电线路电力器件及异常目标检测

doi:10.3969/j.issn.1008- 0198.2023.05.012

湖南电力 ›› 2023, Vol. 43 ›› Issue (5): 79-84.doi: 10.3969/j.issn.1008- 0198.2023.05.012

基于改进RetinaNet算法的输电线路电力器件及异常目标检测

彭紫扬¹, 陈诺天¹, 易俊飞¹, 陶梓铭¹, 毛建旭¹, 谢锦莹²

1.湖南大学电气与信息工程学院,湖南长沙 410082;
2.国网湖南省电力有限公司长沙供电分公司,湖南长沙 410015

收稿日期:2023-06-14 修回日期:2023-07-24 出版日期:2023-10-25 发布日期:2023-11-03
作者简介:毛建旭(1974),男,教授,通信作者,从事机器视觉与图像处理、智能机器人系统工作。
基金资助:
国家自然科学基金项目 (62133005)

Power Devices and Abnormal Object Detection of Transmission Lines Based on Improved RetinaNet Algorithm

PENG Ziyang¹, CHEN Nuotian¹, YI Junfei¹, TAO Ziming¹, MAO Jianxu¹, XIE Jinying²

1. College of Electrical and Information Engineering, Hunan University, Changsha 410082, China;
2. State Grid Changsha Power Supply Company,Changsha 410015, China

Received:2023-06-14 Revised:2023-07-24 Online:2023-10-25 Published:2023-11-03

摘要/Abstract

摘要： 根据电力作业场景的实际需求,针对电力设备可能存在遮挡和检测目标较小、难以识别等问题,采用自适应样本选择方法中的样本分配策略来解决设备遮挡问题。同时,采用CIoU损失函数克服小目标识别困难问题,进一步提高检测准确性。此外,通过数据增强方法增加算法的鲁棒性,优化检测效果。理论和实践结果均显示,该优化后的算法能够显著提高输电线路电力器件及异常目标的检测准确性,从而实现对输电线路电力器件及异常目标的有效检测。

关键词: 电力作业场景, CIoU损失函数, 数据增强, 输电线路电力器件, 异常目标检测

Abstract: In this paper,aiming at the problems that power equipment may be occluded and the detection target is small and difficult to identify, adaptive training sample selection strategy is adopted to solve the equipment occlusion problem. Additionally, the CIoU loss function is utilized to overcome difficulties in detecting small targets, further enhancing detection accuracy. Furthermore, data augmentation methods are employed to increase algorithm robustness and optimize detection performance.Both theoretical analysis and practical results demonstrate that the optimized algorithm significantly improves the accuracy of detecting power equipment and abnormal targets in transmission lines, thus achieving effective dectection of power components and abnormal targets on transmission lines.

Key words: power operation scenarios, CIoU loss function, data augmentation, transmission line power devices, abnormal target detection

中图分类号:

TM755

彭紫扬, 陈诺天, 易俊飞, 陶梓铭, 毛建旭, 谢锦莹. 基于改进RetinaNet算法的输电线路电力器件及异常目标检测[J]. 湖南电力, 2023, 43(5): 79-84.

PENG Ziyang, CHEN Nuotian, YI Junfei, TAO Ziming, MAO Jianxu, XIE Jinying. Power Devices and Abnormal Object Detection of Transmission Lines Based on Improved RetinaNet Algorithm[J]. Hunan Electric Power, 2023, 43(5): 79-84.

参考文献

[1] SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions[C]//IEEE Conference on Computer Vision and Pattern Recognition:Boston,MA,USA:IEEE,2015:1-9.
[2] 龙昊,何应强,邢博为,等.基于Faster R-CNN的电力系统故障的图像检测方法[J].科技与创新,2023(11):135-138.
[3] 方汀,刘艺超,唐哲,等.基于高效通道注意力模块(ECA)和YOLOv5的图像检测方法研究[J].科学技术创新,2023(8):88-91.
[4] HINTON G,DENG L,YU D,et al.Deep neural networks for acoustic modeling in speech recognition: The shared views of four research groups[J].IEEE Signal Processing Magazine,2012,29(6):82-97.
[5] ONG B T,SUGIURA K,ZETTSU K.Dynamically pre-trained deep recurrent neural networks using environmental monitoring data for predicting PM2.5[J].Neural Computing and Applications,2016,27(6).
[6] 王万国,田兵,刘越,等.基于RCNN的无人机巡检图像电力小部件识别研究[J].地球信息科学学报,2017,19(2):256-263.
[7] 王子昊. 深度学习在输电铁塔关键部件缺陷检测中的应用研究[D].天津:中国民航大学,2018.
[8] 翁智,程曦,郑志强.基于改进YOLOv3的高压输电线路关键部件检测方法[J].计算机应用,2020,40(S2):183-187.
[9] 赵永强,饶元,董世鹏,等.深度学习目标检测方法综述[J].中国图象图形学报,2020,25(4):629-654.
[10] 邓姗姗,黄慧,马燕.基于改进Faster R-CNN的小目标检测算法[J].计算机工程与科学,2023,45(5):869-877.
[11] 杨辉羽,李海明.单聚合YOLO航拍小目标检测算法[J].国外电子测量技术,2023,42(4):131-140.
[12] 史建柯,乔美英,李冰锋,等.基于注意力机制的水下遮挡目标检测算法[J].电子科技,2023,36(5):62-70.
[13] 孙方伟,李承阳,谢永强,等.深度学习应用于遮挡目标检测算法综述[J].计算机科学与探索,2022,16(6):1243-1259.
[14] 阳珊,王建,胡莉,等.改进RetinaNet的遮挡目标检测算法研究[J].计算机工程与应用,2022,58(11):209-214.
[15] TSUNG Y L, PRIYA G, ROSS G,et al. Focal loss for dense object detection[C].Proceedings of the IEEE International Conference on ComputerVision,2017:2980-2988.
[16] ZHANG S F,CHI C,YAO Y Q,et al.Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2020:9759-9768.
[17] KIM K, LEE H S. Probabilistic anchor assignment with iou prediction for object detection[C]//Computer Vision-ECCV 2020: 16th European Conference.Glasgow, UK, August 23-28, 2020, Proceedings, Part XXV 16. Springer International Publishing, 2020: 355-371.
[18] GE Z, LIU S, WANG F, et al. Yolox: Exceeding yolo series in 2021[J]. arXiv preprint arXiv:2107.08430, 2021.
[19] FENG C, ZHONG Y, GAO Y, et al. Tood: Task-aligned one-stage object detection[C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV). IEEE Computer Society, 2021: 3490-3499.
[20] 翟永杰,胡哲东,白云山,等.融合迁移学习的绝缘子缺陷分级检测方法[J].电子测量技术,2023,46(6):23-30.

基于改进RetinaNet算法的输电线路电力器件及异常目标检测

Power Devices and Abnormal Object Detection of Transmission Lines Based on Improved RetinaNet Algorithm

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

编辑推荐

Metrics

本文评价

[1]	刘道生, 赖勇耀, 陈星蓉, 刘晖. 输电线路导地线图像识别算法综述[J]. 湖南电力, 2023, 43(3): 103-108.
[2]	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.
[3]	曹先慧, 欧阳克俭, 陈军君, 刘维可, 冯超, 杨涛. 220 kV三跨输电线路子导线间隔棒脱落原因分析[J]. 湖南电力, 2020, 40(4): 39-42.