ZHANG Jiasheng, LIANG Jinxing. Detection Method of Metal Fitting Rust Defects for Overhead Transmission Lines Based on UAV Patrol of Deep Learning[J]. 2022, 42(5): 75-78.
ZHANG Jiasheng, LIANG Jinxing. Detection Method of Metal Fitting Rust Defects for Overhead Transmission Lines Based on UAV Patrol of Deep Learning[J]. 2022, 42(5): 75-78. DOI: 10.3969/j.issn.1008-0198.2022.05.013.
In order to improve the efficiency of UAV inspection of overhead lines and the intelligent detection efficiency of rust defects of overhead line metal fitting
a deep learning-based inspection method for rust defects of overhead lines metal fitting is proposed. Due to the characteristics of large environmental background
small target
and large difference in shooting angle and shooting light in the intelligent detection of metal fittings rust defects of overhead transmission lines
this paper uses image preprocessing algorithm to expand the data set. And YOLO is replaced by MobileNet′s backbone feature extraction network to improve the generalization ability and robustness of the
algorithm
and the actual inspection images are used for experimental testing. In the test set validation
when the confidence threshold is 0.5
the P value is 0.92
the R value is 0.84
and the A
P
value is 91.34%. The results show that this method has a good detection effect on the rust defects of overhead line metal fittings
and can provide reference for the assessment of equipment health status.
WANG C-Y, BOCHKOVSKIY A, LIAO H-Y M, et al. Scaled-YOLOv4: Scaling Cross Stage Partial Network[C]//proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Electr Network, 2021.
REDMON J, DIVVALA S, GIRSHICK R, et al. You Only Look Once: Unified, Real-Time Object Detection[C]//proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA, 2016.
Andrew G.Howard,Menglong zhu,Bo Chen, et al. MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications [J]. arxiv:1704.04861[CS.CV], 2017.