江西省电机工程学会网站

Abstract. In transmission line corridors，there are numerous abnormal conditions，such as bird nests，suspended foreign objects，wildfires，and smog，which always affect the normal operation of transmission systems. Considering the various types of foreign objects，complex backgrounds，occlusion，wildfires，irregular smoke shapes，and the real-time require- ments of drone detection，this paper introduces an information aggregation algorithm based on YOLOX-S. This algorithm enhances relevant features and suppresses irrelevant features by aggregating spatial information and channel information in the feature map，which raises the overall learning ability of the network to improve the detection accuracy. Finally，the test on the transmission line environmental data set shows that the detection accuracy of this method in abnormal conditions such as bird/s nest，hanging foreign body，wildfire and smog is 88.5%，which is 13.9% higher than that before the im- provement，and the increase of reasoning time is very small. The experimental results show that this method has made sig- nificant progress in the abnormal environmental condition detection of power transmission line corridors.

[1] S.Ren，K.He，and R. Girshick，“Faster R-CNN：toward real-time object detection with region proposal networks，”Internation- al Conference on Neural Information Processing Systems， 2015，pp. 91-99. [2] J. Dai，Y. Li，K. He，and J. Sun，“R-FCN：object detection via re- gion-based fully convolutional networks，”Conference on Neu- ral Information Processing Systems. Barcelona，Spain，2016， pp. 379-387. [3] Y.Xue，H.Wu，N.Zhang，Z.Yu，X.Ye，and Hua X，“Detectionof insulation piercing connectors and bolts on the transmis- sion line using improved faster R-CNN，”Laser and Opto- electronics Progress，2020，pp.84-91. [4] Z.Zhao，H. Qi，Y Qi，K.Zhang，Y.Zhai，and W.Zhao，“Detec- tion method based on automatic visual shape clustering for pinmissing defect in transmission lines，”IEEE Transactions on Instrumentation and Measurement，2020，pp.6080-6091 [5] W. Liu，D.Anguelov，D. Erhan，and C. Szegedy，“SSD：single shot multibox detector，”Proceedings of the 2016 European Conference on Computer Vision，2016，pp.21-37. [6] J.Redmon，S.Divvala，R.Girshick，and A.Farhadi，“You Only Look Once：Unified Real-Time Object Detection，”Proceed ings of IEEE Conference on Computer Vision and PatternRecognition，Las V egas，United States，2016，pp.779-788. [7] T.Wu，T.Wang，and Y. Liu，“Real-Time Vehicle and Distance Detection Based on Improved Yolov5 Network，”IEEE World Symposium on Artificial Intelligence（WSAI），2021，pp.24-28. [8] J.Chen，Z.Liu，H.Wang，A.Núñez，and Z.Han，“Automatic de- fect detection of fasteners on the catenary support device us- ing deep convolutional neural network，”IEEE Transactions on Instrumentation and Measurement，2018，pp.257-269. [9]Z.Ge，S.Liu，F. Wang，and Z. Li，“YOLOX：ExceedingYOLO Se- ries in 2021，”Computer Vision and Pattern Recognition，2021. [10] Q. Houn，D.Zhou，and J.Feng，“Coordinate Attention for Ef- ficient Mobile Network Design，”IEEE Conference on Computer Vision and Pattern Recognition，2021.

优秀论文集模块列表

会员频道