Detection of Self-explosive Insulators in Aerial Images Based on Improved YOLO v4

Yi, Li; Li, Jinguo; Zhai, Yongjie; Meng, Ping

doi:10.1088/1742-6596/2320/1/012025

Cited by 7 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different versions of YOLOv5 have the added depth of the CSP module 30 . In addition, some other methods based on YOLO have been proposed 31,32 . The structure of the YOLOv5s network is shown in Figure 1.…”

Section: Yolov5mentioning

confidence: 99%

Overhead line insulator defect detection method based on improved YOLOv5s

Zhang

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

Target defect detection on overhead lines is critical for safe and reliable grid operation. When targeting line images with complex backgrounds, there are significant challenges in detecting defective insulators. A detection model of overhead line insulators with an improved YOLOv5s network is proposed to improve the detection efficiency of target defects. First, MobileNetV3 is used to lighten the YOLOv5s backbone network and reduce the network parameters; second, an improved ASFF adaptive feature fusion structure is used to equalize the inconsistency of different feature layers; then a CBAM attention mechanism is added to the head network to increase the weight of the target information to improve the model accuracy; finally, the edge loss function is adopted as EIOU_Loss loss function, to accelerate the convergence of the network. The experimental results show that the GFLOPs of the improved YOLOv5s algorithm is 11.6G, which is 1.3% higher compared with the YOLOv5s algorithm, and the Map_0.5 is 97%, which is 0.2 percentage point higher than the YOLOv5s algorithm, so the proposed improved YOLOv5s algorithm has higher detection speed while ensuring the model detection accuracy, and can meet the requirements of detection accuracy in light-weight target detection scenarios.

show abstract

Section: Yolov5mentioning

confidence: 99%

Overhead line insulator defect detection method based on improved YOLOv5s

Zhang

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…By using cross-entropy loss function and considering classification error and boundary frame coordinate error, the network weight is updated by backpropagation algorithm. In the test phase, the network propagates forward to the input image, obtains the category probability and confidence score of each boundary box, and finally outputs the prediction box [8] .…”

Section: Basic Algorithmmentioning

confidence: 99%

Research on mechanical parts classification and recognition technology based on improved convolutional neural network YOLOv5

Xu,

Guo

2023

Third International Conference on Computer Graphics, Image, and Virtualization (ICCGIV 2023)

View full text Add to dashboard Cite

In the field of classification and recognition of mechanical parts, in order to solve the difficulties of small target detection, adhesion and overlapping parts classification and recognition, this paper proposes an automatic classification and recognition algorithm of mechanical parts based on improved convolutional neural network YOLOv5. The feature pyramid module PANet is added to the traditional YOLOv5 network structure to strengthen the characterization ability of the network. At the same time, NMS algorithm and image segmentation algorithm are used to improve the detection ability of network boundary segmentation, to improve the detection accuracy and recall rate. The experimental results show that both the traditional YOLOv5 and the improved YOLOv5 in this paper can realize the classification and identification of mechanical parts, and the improved YOLOv5 has better detection precision and accuracy, faster detection speed, and better network robustness.

show abstract