Electrical Insulator Defects Detection Method Based on YOLOv5

Feng, Zhiqiang; Guo, Li; Huang, Darong; Li, Runze

doi:10.1109/ddcls52934.2021.9455519

Cited by 48 publications

(34 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The latest work shows that Cascade R-CNN has been used in insulator defect detection (Wen et al, 2021). Feng et al (2021) have proved that YOLOv5 can achieve the highest accuracy at 86.8%, and mAP is 95.5%. In order to evaluate the advantage of the proposed model compared with other models, we have Cascade Mask R-CNN (Cai and Vasconcelos, 2019) and YOLOv5 (Jocher et al, 2022) as strong competitors.…”

Section: Comparison With Other Object Detection Modelsmentioning

confidence: 98%

End-to-End Insulator String Defect Detection in a Complex Background Based on a Deep Learning Model

Xu¹,

Zhong²,

Luo³

et al. 2022

Front. Energy Res.

View full text Add to dashboard Cite

Normal power line insulators ensure the safe transmission of electricity. The defects of the insulator reduce the insulation, which may lead to the failure of power transmission systems. As unmanned aerial vehicles (UAVs) have developed rapidly, it is possible for workers to take and upload aerial images of insulators. Proposing a technology to detect insulator defects with high accuracy in a short time can be of great value. The existing methods suffer from complex backgrounds so that they have to locate and extract the insulators at first. Some of them make detection relative to some specific conditions such as angle, brightness, and object scale. This study aims to make end-to-end detections using aerial images of insulators, giving the locations of insulators and defects at the same time while overcoming the disadvantages mentioned above. A DEtection TRansformer (DETR) having an encoder–decoder architecture adopts convolutional neural network (CNN) as the backbone network, applies a self-attention mechanism for computing, and utilizes object queries instead of a hand-crafted process to give the direct predictions. We modified this for insulator detection in complex aerial images. Based on the dataset we constructed, our model can get 97.97 in mean average precision when setting the threshold of intersection over union at 0.5, which is better than Cascade R-CNN and YOLOv5. The inference speed of our model can reach 25 frames per second, which is qualified for actual use. Experimental results demonstrate that our model meets the robustness and accuracy requirements for insulator defect detection.

show abstract

Section: Comparison With Other Object Detection Modelsmentioning

confidence: 98%

End-to-End Insulator String Defect Detection in a Complex Background Based on a Deep Learning Model

Xu¹,

Zhong²,

Luo³

et al. 2022

Front. Energy Res.

View full text Add to dashboard Cite

show abstract

“…Liu et al [37] had a F1-score of 0.9499 using YOLOv3 and Feng et al [70] had a F1-score of 0.9293 using YOLOv5, which is the most current model for object detection nowadays. The applications of Jiang et al [46] and Miao et al [34] using single shot multibox detector had a F1-score of 0.9244 and 0.9184, respectively.…”

Section: State-of-the-art Approachesmentioning

confidence: 99%

Semi-ProtoPNet Deep Neural Network for the Classification of Defective Power Grid Distribution Structures

Stefenon

Singh

Yow

et al. 2022

Sensors

View full text Add to dashboard Cite

Power distribution grids are typically installed outdoors and are exposed to environmental conditions. When contamination accumulates in the structures of the network, there may be shutdowns caused by electrical arcs. To improve the reliability of the network, visual inspections of the electrical power system can be carried out; these inspections can be automated using computer vision techniques based on deep neural networks. Based on this need, this paper proposes the Semi-ProtoPNet deep learning model to classify defective structures in the power distribution networks. The Semi-ProtoPNet deep neural network does not perform convex optimization of its last dense layer to maintain the impact of the negative reasoning process on image classification. The negative reasoning process rejects the incorrect classes of an input image; for this reason, it is possible to carry out an analysis with a low number of images that have different backgrounds, which is one of the challenges of this type of analysis. Semi-ProtoPNet achieves an accuracy of 97.22%, being superior to VGG-13, VGG-16, VGG-19, ResNet-34, ResNet-50, ResNet-152, DenseNet-121, DenseNet-161, DenseNet-201, and also models of the same class such as ProtoPNet, NP-ProtoPNet, Gen-ProtoPNet, and Ps-ProtoPNet.

show abstract

“…This work finished 25 images per second, which just reaches the minimum standard for real-time detection, while the final accuracy only reached 88%. In the work by Feng, Guo et al (2021), YOLOv3 was applied to detect insulators defect. This method realized a fast detection speed with 100 images per second.…”

Section: Insulator Detectionmentioning

confidence: 99%

High Accuracy Real-Time Insulator String Defect Detection Method Based on Improved YOLOv5

et al. 2022

View full text Add to dashboard Cite

Insulator string is a special insulation component which plays an important role in overhead transmission lines. However, working outdoors for a long time, insulators often have defects because of various environmental and weather conditions, which affect the normal operation of transmission lines and even cause huge economic losses. Therefore, insulator defect recognition is a crucial issue. Traditional insulator defect identification relies on manual work, which is time-consuming and inefficient. Therefore, the use of artificial intelligence to detect the defect location and recognize its class has become a key research field. By improving the classical YOLOv5 (you only look once) model, this article proposes a new method to enable high accuracy and real-time detection. Our method has three advantages: 1) Efficient-IoU (EIoU) replaces intersection over union (IoU) to calculate the loss of box regression, which overcomes that the detection is sensitive to various scale insulators in aerial images. 2) Since YOLOv5 itself detects some natural scenes in the real world, some anchors setting by default are not suitable for defect detection, this article introduces Assumption-free K-MC2 (AFK-MC2) algorithm into YOLOv5 to modify the K-means algorithm to improve accuracy and speed. 3) The cluster non-maximum suppression (Cluster-NMS) algorithm is introduced to avoid missing detection of insulators because of mutual occlusion in images and improve the computation speed at the same time. The experiments’ results show that this model can improve detection accuracy compared with YOLOv5 and realize real-time detection.

show abstract

Electrical Insulator Defects Detection Method Based on YOLOv5

Cited by 48 publications

References 7 publications

End-to-End Insulator String Defect Detection in a Complex Background Based on a Deep Learning Model

End-to-End Insulator String Defect Detection in a Complex Background Based on a Deep Learning Model

Semi-ProtoPNet Deep Neural Network for the Classification of Defective Power Grid Distribution Structures

High Accuracy Real-Time Insulator String Defect Detection Method Based on Improved YOLOv5

Contact Info

Product

Resources

About