LDN-DETR:轻量化降噪DETR实现高效输电线异物检测

doi:10.13474/j.cnki.11-2246.2026.0821

摘要/Abstract

摘要： [目的] 目标检测方法作为输电异物目标检测过程中的核心技术，为线路巡检带来了巨大的便利。然而，由于输电线路所处环境复杂多变，现有方法依旧存在推理速度慢、遮挡目标检测效果差及小目标检测精度低的问题。针对以上问题，本文对RT-DETR目标检测方法进行研究，最终提出LDN-DETR输电线异物目标检测方法。[方法] 首先，构建基于可变形卷积的轻量化特征提取网络，提升对不同形状、大小目标的检测效果并提高推理速度；然后，设计基于区域敏感注意力的跨尺度特征融合模块，提升遮挡目标的检测精度；最后，采用基于高斯调制的IoU感知查询降噪训练方法，解决小目标检测精度差的问题。[结果] 为证明该方法的有效性，本文在 RailFOD23、VisDrone2019上进行了试验。结果表明，LDN-DETR 在 RailFOD23 上的 mAP 较 RT-DETR 提升4%，在 VisDrone2019 上提升3%，同时推理速度达到 35帧/s(含后处理过程)。[结论] 本文方法优于多种主流目标检测方法，能够满足巡检异物检测巡检需要。

关键词: 输电巡检, 特征提取, 跨尺度特征融合, 查询降噪

Abstract: [Purposes] Object detection methods serve as the core technology in the process of detecting foreign objects on transmission lines, bringing significant convenience to line inspection. However, due to the complex and variable environment of transmission lines, existing methods still face issues such as slow inference speed, poor detection performance for occluded objects, and low precision for small object detection. To address these issues, research on the RT-DETR object detection method was conducted, leading to the proposal of LDN-DETR (lightweight denoising-DETR) for foreign object detection on transmission lines. [Methods] Firstly, a lightweight feature extraction network based on variable kernel convolution was constructed to enhance detection performance for objects of various shapes and sizes, while improving inference speed. Subsequently, a cross-scale feature fusion module based on region-sensitive attention was designed to improve detection accuracy for occluded objects. Finally, a Gaussian modulation-based IoU-aware query denoising training method was adopted to address the issue of poor detection precision for small objects. [Findings] To prove the effectiveness of the method, experiments were conducted in this paper on RailFOD23 and VisDrone 2019. The results show that LDN-DETR has a 4% improvement in mAP over RT-DETR on RailFOD23, a 3% improvement on VisDrone 2019, and an inference speed of 35 frames per second with post-processing. [Conclusions] This method outperforms various mainstream object detection methods, thereby meeting the needs for inspection of foreign objects.

Key words: transmission inspection, feature extraction, cross-scale feature fusion, query noise reduction

中图分类号:

P237

高书涵, 周超, 沈浩, 贾然, 刘辉, 刘传彬, 刘嵘. LDN-DETR:轻量化降噪DETR实现高效输电线异物检测[J]. 测绘通报, 2026, 0(8): 145-153.

Gao Shuhan, Zhou Chao, Shen Hao, Jia Ran, Liu Hui, Liu Chuanbin, Liu Rong. LDN-DETR:lightweight denoising DETR for efficient foreign object detection on power transmission lines[J]. Bulletin of Surveying and Mapping, 2026, 0(8): 145-153.

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://tb.chinasmp.com/CN/10.13474/j.cnki.11-2246.2026.0821

http://tb.chinasmp.com/CN/Y2026/V0/I8/145

参考文献

[1] Dalal N,Triggs B.Histograms of oriented gradients for human detection[C]//Proceedings of 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).San Diego:IEEE,2005:886-893.
[2] Suthaharan S.Support vector machine[M]//Machine Learning Models and Algorithms for Big Data Classification:Thinking with Examples for Effective Learning.Boston:Springer US,2016:207-235.
[3] Girshick R,Donahue J,Darrell T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of 2014 IEEE Conference on Computer Vision and Pattern Recognition.Columbus:IEEE,2014:580-587.
[4] Yang Xue,Sun Hao,Sun Xian,et al.Position detection and direction prediction for arbitrary-oriented ships via multitask rotation region convolutional neural network[J].IEEE Access,2018,6:50839-50849.
[5] Zhang Hongkai,Chang Hong,Ma Bingpeng,et al.Dynamic R-CNN:towards high quality object detection via dynamic training[M]//Computer Vision-ECCV 2020.Cham:Springer International Publishing,2020:260-275.
[6] Ren Shaoqing,He Kaiming,Girshick R,et al.Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149.
[7] Redmon J,Divvala S,Girshick R,et al.You only look once:unified,real-time object detection[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Las Vegas:IEEE,2016:779-788.
[8] Liu Wei,Anguelov D,Erhan D,et al.SSD:single shot MultiBox detector[M]//Computer Vision-ECCV 2016.Cham:Springer International Publishing,2016:21-37.
[9] 刘桂超,王怀光,任国全,等.基于深度学习的单目视觉目标检测综述[J].计算机工程与应用,2026,62(1):1-19.
[10] Beal J,Kim E,Tzeng E,et al.Toward transformer-based object detection[EB/OL].[2025-11-10].https://doi.org/10.48550/arXiv.2012.09958.
[11] Li Feng,Zhang Hao,Liu Shilong,et al.DN-DETR:accelerate DETR training by introducing query DeNoising[C]//Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).New Orleans:IEEE,2022:13609-13617.
[12] Zhao Y,Lü W,Xu S,et al.DETRs Beat YOLOs on real-time object detection[EB/OL].[2025-11-10].https://doi.org/10.48550/arXiv.2304.08069.
[13] Zhu Xizhou,Su Weijie,Lu Lewei,et al.Deformable DETR:deformable transformers for end-to-end object detection[EB/OL].[2025-11-11].https://doi.org/10.48550/arXiv.2010.04159.
[14] Zhao Yian,Lv Wenyu,Xu Shangliang,et al.DETRs beat YOLOs on real-time object detection[C]//Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle:IEEE,2024:16965-16974.
[15] 盛伟,刘明剑,刘殿臣.面向人员密集与遮挡的实时目标检测方法[J/OL].计算机工程与科学,1-10[2025-11-12].https://link.cnki.net/urlid/43.1258.tp.20250103.0853.002.
[16] 王坤,冯康威.FFE-YOLOv8:交通场景遮挡目标检测算法[J].计算机工程与应用,2025,61(15):156-166.
[17] 田艳萍,金淼,陈习文,等.基于小目标遮挡感知的烟头检测算法[J].计算机工程与应用,2025,61(20):270-280.
[18] Yang Zhiqiang,Guan Qiu,Zhao Keer,et al.Multi-branch auxiliary fusion YOLO with re-parameterization heterogeneous convolutional for accurate object detection[M]//Pattern Recognition and Computer Vision.Singapore:Springer Nature Singapore,2024:492-505.
[19] Liu Shilong,Li Feng,Zhang Hao,et al.DAB-DETR:dynamic anchor boxes are better queries for DETR[EB/OL].[2025-11-12].https://arxiv.org/abs/2201.12329.
[20] Zhang Shilong,Wang Xinjiang,Wang Jiaqi,et al.Dense distinct query for end-to-end object detection[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Vancouver:IEEE,2023:7329-7338.
[21] 左露,牛晓伟,朱春惠.基于改进YOLOX的航拍遥感图像检测模型[J].电子测量技术,2023,46(16):179-186.
[22] 沈涛,张秀再,许岱.改进RT-DETR的遥感图像小目标检测算法[J].计算机科学,2025,52(8):214-221.
[23] Teng Xiaoxi,Zhang Wendong,Liu Tao,et al.Stff-rtdetr:a small object detection algorithm based on drone aerial photography[J].The Journal of Supercomputing,2025,81(8):928.
[24] Chattopadhay A,Sarkar A,Howlader P,et al.Grad- CAM++:generalized gradient-based visual explanations for deep convolutional networks[C]//Proceedings of 2018 IEEE Winter Conference on Applications of Computer Vision (WACV).Lake Tahoe:IEEE,2018:839-847.
[25] Khanam R,Hussain M.What is YOLOv5:a deep look into the internal features of the popular object detector[EB/OL].[2025-11-12].https://arxiv.org/abs/2407.20892.
[26] Varghese R,M S.YOLOv8:a novel object detection algorithm with enhanced performance and robustness[C]//Proceedings of 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS).Chennai:IEEE,2024:1-6.
[27] Khanam R,Hussain M.YOLOv11:an overview of the key architectural enhancements[EB/OL].[2025-11-12].https://arxiv.org/abs/2410.17725.
[28] Tian Yunjie,Ye Qixiang,Doermann D.YOLOv12:attention-centric real-time object detectors[EB/OL].[2025-11-12].https://arxiv.org/abs/2502.12524.
[29] Zhang Hao,Li Feng,Liu Shilong,et al.DINO:DETR with improved DeNoising anchor boxes for end-to-end object detection[EB/OL].[2025-11-12].https://arxiv.org/abs/2203.03605.