基于FCIS模型的输电线路部件同时检测与分割方法
CSTR:
作者:
作者单位:

(1.华北电力大学电气与电子工程学院,河北 保定071003;2.华北电力大学河北省电力物联网技术重点实验室,河北 保定071003)

通讯作者:

耿劭锋(1996—),男,硕士研究生,主要从事电力图像超分辨率处理研究;E?mail:gsf13633220753@163.com

中图分类号:

TM726

基金项目:

国家自然科学基金(61871182);河北省省级科技计划(SZX2020034)


Simultaneous detection and segmentation method of transmission line components based on improved FCIS model
Author:
Affiliation:

(1.School of Electrical and Electronic Engineering, North China Electric Power University, Baoding 071003, China;2.Hebei Key Laboratory of Power Internet of Things Technology, North China Electric Power University, Baoding 071003, China)

  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献 [17]
  • | | | |
  • 文章评论
    摘要:

    实时监测并及时诊断输电线路故障是输电线路安全运行的前提。由于输电线路图像拍摄环境复杂,单独的检测或分割不能满足实时性要求,且图中小部件和遮挡部件难以提取,因此为了更精准地定位目标位置、检测并分割出图中小部件和遮挡部件,提出一种改进全卷积实例感知语义分割(FCIS)的输电线路部件同时检测与分割方法。将兴趣区对准(ROI Align)算法的思想引进到FCIS模型中,提出位置敏感区域前景/背景—感兴趣区域对准(PS2?ROI Align)方法,使用双线性插值法可以有效地解决输入图像特征图中的ROI与原图中位置信息不匹配的问题,并且引用梯度回传算法解决图像中小金具及遮挡金具特征难提取导致检测与分割精度差的问题。在本次构建的输电线路检测与分割数据集中进行检测分割实验,结果表明,改进前图中未能检测分割的小目标及遮挡目标得到了有效的检测分割,对比其他检测模型,改进后FCIS模型平均准确率(mAP)最高,并且相比改进前提升了1.73%。

    Abstract:

    Real?time monitoring and timely diagnosis of transmission line faults are the prerequisite for the safe operation of transmission lines. Due to the complex shooting environment of transmission line images, individual detection or segmentation can not meet the real?time requirements, and it is difficult to extract small parts and occluded parts in the picture. In order to more accurately locate the target position, detect and segment small parts and occluded parts in the picture, an improved fully convolutional instance?aware semantic segmentation (FCIS) simultaneous detection and segmentation method for transmission line components is proposed. This method introduces the idea of region of interest (ROI) Align algorithm into the FCIS model, and proposes position sensitive inside/outside?region of interest (PS2?ROI) Align, which uses bilinear interpolation method to effectively solve the problem that the ROI in the input image feature map does not match the position information in the original image. And the gradient backpropagation algorithm is used to solve the problem of poor detection and segmentation accuracy due to the difficulty in extracting the features of small fittings and occluded fittings in the image. The detection and segmentation experiment was carried out on the transmission line detection and segmentation data set of this structure. The results showed that the small targets that could not be detected and segmented in the modified figure had indicators and masked detection segmentation. Compared with other detection models, the FCIS model has the highest mean average precision (mAP), which is 1.73% higher than before improvement.

    参考文献
    [1] 仝卫国,苑津莎,李宝树.图像处理技术在直升机巡检输电线路中的应用综述[J].电网技术,2010,34(12): 204?208.TONG Weiguo,YUAN Jinsha,LI Baoshu.Application of image processing in patrol inspection of overhead transmission line by helicopter[J].Power System Technology,2010,34(12):204?208.
    [2] 黄欢,雷加智,曾华荣,等.极端外部环境下输电线路的综合风险评估方法[J].电力科学与技术学报,2019,34(2):119?127. HUANG Huan,LEI Jiazhi,ZENG Huarong,et al.Integrated risk assessment system of transmission line under extreme external environment[J].Journal of Eiectric Power Science and Technology,2019,34(2):119?127.
    [3] 赵振兵,齐鸿雨,聂礼强.基于深度学习的输电线路视觉检测研究综述[J].广东电力,2019,32(9):11?23. ZHAO Zhenbing,QI Hongyu,NIE Liqiang.Research overview on visual detection of transmission lines based on deep learning[J].Guangdong Electric Power,2019,32(9):11?23.
    [4] 廖瑞金,王有元,刘航,等.输变电设备状态评估方法的研究现状[J].高电压技术,2018,44(11):3454?3464.LIAO Ruijin,WANG Youyuan,LIU Hang,et al.Research status of condition assessment method for power equipment[J].High Voltage Engineering,2018,44(11):3454?3464.
    [5] EVERINGHAM M,GOOL L V,WILLIAMS C K I,et al.The pascal visual object classes (VOC) challenge[J].International Journal of Computer Vision,2010,88(2):303?338.
    [6] GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR),Columbus,OH,USA,2014.
    [7] VAN DE SANDE K E A,UIJLINGS J R R,GEVERS T,et al.Segmentation as selective search for object recognition[C]//2011 IEEE International Conference on Computer Vision (ICCV),Barcelona,Spain,2011.
    [8] KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[J].Communications of the ACM,2017,60(6):84?90.
    [9] GIRSHICK R.Fast R?CNN[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV),Santiago,Chile,2015.
    [10] REN S Q,HE K M,GIRSHICK R,et al.Faster R?CNN:towards real?time object detection with region proposal networks[C]//International Conference on Neural Information Processing Systems (NIPS),Montreal,Canada,2015.
    [11] DAI J F,LI Y,HE K M,et al.R?FCN:object detection via region?based fully convolutional networks[C]//Proceedings of the 30th International Conference on Neural Information Processing Systems,Barcelona,Spain,2016.
    [12] ZITNICK C L,DOLLAR P.Edge Boxes:locating object proposals from edges[C]//European Conference on Computer Vision (ECCV),Zurich,Switzerland,2014.
    [13] ZHAO H S,SHI J P,QI X J,et al.Pyramid scene parsing network[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition,Hawaii,USA,2017.
    [14] SHELHAMER E,LONG J,DARRELL T.Fully convolutional networks for semantic segmentation[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2017,39(4):640?651.
    [15] YANG H,ZHOU J T,CAI J,et al.Miml?fcn:Multi?instance multi?label learning via fully convolutional networks with privileged information[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition,Hawaii,USA,2017.
    [16] LI Y,QI H,DAI J F,et al.Fully convolutional instance?aware semantic segmentation[C]//Proceedings of 2017 IEEE Computer Vision and Pattern Recognition,Hawaii,USA,2017.
    [17] HE K M,GKIOXARI G,DOLLAR P,et al.Mask R?CNN[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,42(2):386?397.
    相似文献
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

耿劭锋,戚银城,史博强,等.基于FCIS模型的输电线路部件同时检测与分割方法[J].电力科学与技术学报,2023,38(2):124-132.
GENG Shaofeng, QI Yincheng, SHI Boqiang, et al. Simultaneous detection and segmentation method of transmission line components based on improved FCIS model[J]. Journal of Electric Power Science and Technology,2023,38(2):124-132.

复制
分享
文章指标
  • 点击次数:209
  • 下载次数: 844
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 在线发布日期: 2023-06-29
文章二维码