arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Computer Science · Computer Vision and Pattern Recognition

Insulator detection of Yolov5 fused with GBAM and α-IOU

Xin Luo, Junsun Chen, Guangling Sun, Ziqi Dai

See on the map

Abstract

Insulator fault of power line is easy to lead to transmission system fault, so insulator detection system is widely used. Based on this, an improved detection method for Yolov5 is proposed, this method constructs a new network framework through Yolov5. First, the Yolov5 loss function CIoU is improved, and the generalized power transformation is introduced into the existing CIoU loss to optimize the boundary frame regression and improve the accuracy of the boundary frame regression. Then, GBAM attention mechanisms are introduced in each C3 module of the Yolov5 backbone to reduce information dispersion and amplify global interaction features. Finally, create a custom damaged insulator data set. The results show that the network is more accurate and faster than the control group, with an accuracy of 97.42% for insulator defect detection and 4.01 images per second, which is good efficiency and practicality in insulator defect detection.

More like this