杨坚,李剑,徐硕
YANG Jian,LI Jian,XU Shuo

• 1:国网浙江省电力有限公司台州供电公司

摘要(Abstract)：

无人机自动巡检是输电线路智能化巡检的重要环节。针对巡检过程中无人机拍摄的视频图像背景复杂、电力线检测精度差、检测速率较低等问题，提出一种基于改进SegNet模型的电力线检测算法。首先，在编码器中引入残差模块和非对称卷积，减小网络计算负担；其次，减少解码层网络层数，并对编码器与解码器进行特征融合，提高检测精度；最后，利用改进SegNet在构建的电力线数据集中进行训练，准确率和交并比均值分别达到了89.4%和86.62%，单张检测时间仅46 ms。实验结果表明，基于改进SegNet模型的电力线检测算法可实现较高精度的实时检测。
UAVs(unmanned aerial vehicles) are now involved in intelligent transmission line inspection. Given the complex image backgrounds captured by UAVs and poor line inspection accuracy and low detection speed of UAVs, the paper proposes a power line inspection algorithm based on an improved SegNet model. Firstly, residual modules and asymmetric convolutions are introduced into the encoder to reduce the computational burden on the network.Secondly, the network layers of the decoding layer are reduced, and the features of the encoder and decoder are fused to improve inspection accuracy. Finally, the improved SegNet algorithm is used to train the power line dataset.The accuracy and mean intersection over union reach up to 89.4% and 86.62% respectively, and the single detection time is 46 ms. The experimental results show that the algorithm based on the improved SegNet model can achieve high-precision and real-time power line detection.

关键词(KeyWords)： 无人机巡检;深度学习;改进SegNet;残差模块;非对称卷积
UAV inspection;deep learning;improved SegNet;residual module;asymmetric convolution

Abstract:

Keywords:

基金项目(Foundation): 国网浙江省电力有限公司科技项目（5211TZ2000M7）

作者(Author): 杨坚,李剑,徐硕
YANG Jian,LI Jian,XU Shuo

DOI: 10.19585/j.zjdl.202306013

参考文献(References)：

• [1] BIAN J A,HUI X L,ZHAO X G,et al. A monocular vision-based perception approach for unmanned aerial vehicle close proximity transmission tower inspection[J].International Journal of Advanced Robotic Systems,2019,16(1):172988141882022.
• [2] TAN J X,ZHAO H J,YANG R H,et al. An entropyweighting method for efficient power-line feature evaluation and extraction from LiDAR point clouds[J].Remote Sensing,2021,13(17):3446.
• [3] WU J A,SHAO S A,JIANG D P,et al. Research on adaptive power transmission line fault inspection[J].IOP Conference Series:Earth and Environmental Science,2021,787(1):012050.
• [4]韩冰，尚方.面向无人机输电线路巡检的电力杆塔检测框架模型[J].浙江电力，2016,35(4):6-11.HAN Bing,SHANG Fang.A frame model of power pylon detection for UAV-based power transmission line inspection[J].Zhejiang Electric Power,2016,35(4):6-11.
• [5]陈舫明，杜雅慧，陈弘.图像处理技术在输电线路直升机智能巡检中的应用[J].浙江电力，2012,31(9):63-66.CHEN Fangming,DU Yahui,CHEN Hong. Application of image processing technology in intelligent patrol inspection of transmission lines by helicopter[J].Zhejiang Electric Power,2012,31(9):63-66.
• [6]刘哲，闫佳文，陈长金，等.电力线图像检测方法研究[J].机械与电子，2022,40(3):54-57.LIU Zhe,YAN Jiawen,CHEN Changjin,et al.Research on power line image detection method[J]. Machinery&Electronics,2022,40(3):54-57.
• [7]王植，贺赛先.一种基于Canny理论的自适应边缘检测方法[J].中国图象图形学报，2004,9(8):957-962.WANG Zhi,HE Saixian. An adaptive edge-detection method based on canny algorithm[J].Journal of Image and Graphics,2004,9(8):957-962.
• [8]杨坚，徐硕，林中圣.基于图像处理的无人机电力巡线航向偏差检测[J].浙江电力，2021,40(9):105-110.YANG Jian,XU Shuo,LIN Zhongsheng.The course deviation detection of power line tracking UAV based on image processing[J].Zhejiang Electric Power,2021,40(9):105-110.
• [9] REN S Q,HE K M,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.
• [10] REDMON J,DIVVALA S,GIRSHICK R,et al. You only look once:unified, real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).June 27-30,2016,Las Vegas,NV,USA:IEEE,2016:779-788.
• [11] WEI Y C,LIANG X D,CHEN Y P,et al. Learning to segment with image-level annotations[J].Pattern Recognition,2016,59:234-244.
• [12] SHAN J C,LI X Z,JIA S M,et al.Semantic segmentation based on deep convolution neural network[J]. Journal of Physics:Conference Series,2018,1069:012169.
• [13] HUI X L,BIAN J A,ZHAO X G,et al.Vision-based autonomous navigation approach for unmanned aerial vehicle transmission-line inspection[J]. International Journal of Advanced Robotic Systems,2018,15(1):172988141775282.
• [14] SAURAV S,GIDDE P,SINGH S,et al.Power line segmentation in aerial images using convolutional neural networks[M]//Lecture Notes in Computer Science. Cham:Springer International Publishing,2019:623-632.
• [15] BADRINARAYANAN V,KENDALL A,CIPOLLA R.SegNet:a deep convolutional encoder-decoder architecture for image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(12):2481-2495.
• [16] HE K M,ZHANG X Y,REN S Q,et al.Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).June 27-30,2016,Las Vegas,NV,USA:IEEE,2016:770-778.