Fault Detection Method of Glass Insulator Aerial Image Based on the Improved YOLOv5

Zhou, Ming; Li, Bo; Wang, Jue; He, Shi

doi:10.1109/tim.2023.3269099

Cited by 11 publications

(5 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although YOLO decreased slightly by 0.2% in the variable 𝐴𝑃𝑀 21 compared to the Libra R-CNN model, the network structure was improved with the gradual introduction of YOLO9000 22 and YOLOv5 in the YOLO series. Compared with the traditional depth model, the evaluation indexes of YOLOv5 model are improved, 16,23 AP50 is increased by 18.1%, 24 the experimental results show that the dense target identification is more accurate. The recognition accuracy was 97.2%, 96.2%, and 97.9%, respectively.…”

Section: Deep Learning Modelsmentioning

confidence: 97%

“…In the adaptive aurora synthesis, the aurora effect is adjusted through the frequency domain based on the local brightness and contrast, as shown in Equation (16).…”

Section: Adaptive Augmentation Strategy Based On Brightness and Contrastmentioning

confidence: 99%

“…𝐹(𝑖, 𝑗) = 𝐼(𝑖, 𝑗) × (1 + 𝛼𝐵(𝑖, 𝑗)) × (1 + 𝛽𝐶(𝑖, 𝑗)) (16) In this expression, 𝐹(𝑖, 𝑗) represents the image in the frequency domain. 𝐵(𝑖, 𝑗) and 𝐶(𝑖, 𝑗) correspond to the representations of local brightness and contrast in the frequency domain.…”

Section: Adaptive Augmentation Strategy Based On Brightness and Contrastmentioning

confidence: 99%

“…Extensive experiments have been performed based on the YOLOv5 framework. [15][16][17] The proposed method significantly improves the accuracy and generalization ability of the ice condition detection model, which can better guide polar ships to formulate the best route for reliable obstacle avoidance and safe navigation.…”

Section: Introductionmentioning

confidence: 99%

“…Finally, an adaptive augmentation strategy based on brightness and contrast is proposed to adaptively adjust the snowflake texture density and aurora intensity, which improves the distinguishability of ice boundaries in the augmented dataset, and improves the accuracy of the trained model for ice condition detection. Extensive experiments have been performed based on the YOLOv5 framework 15–17 . The proposed method significantly improves the accuracy and generalization ability of the ice condition detection model, which can better guide polar ships to formulate the best route for reliable obstacle avoidance and safe navigation.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

An image augmentation‐based ice monitoring method for safe navigation of polar ships

Liang,

Wang,

Zhao

et al. 2024

Quality & Reliability Eng

View full text Add to dashboard Cite

The navigation safety of polar ships depends on accurate monitoring of ice conditions along the route. Complex polar environments but scarce ice datasets pose challenges for deep learning‐based object detection methods. To address this issue, we propose an ice monitoring method based on polar environment adaptive image enhancement. This approach adaptively augments the ice dataset by simulating polar environments such as snowflakes and aurora to improve the training precision of ice detection models in harsh polar environments and meet the needs of safe navigation of polar ships. First, an image augmentation method based on randomly generated snowflake textures is proposed to address the difficulty of covering polar storm scenarios in existing ice datasets, which leads to weak generalization of ice monitoring models. Second, targeting the problem that the boundary between ice and background is confused by aurora phenomena, which severely reduces the ice monitoring accuracy, we propose an image augmentation method based on aurora simulations to improve the ice monitoring performance of trained models under aurora. Finally, an adaptive image augmentation strategy based on the brightness‐contrast threshold is proposed to optimize the snowflake density and aurora intensity in the augmented images to improve the discrimination of ice boundaries. Experimental results demonstrate the effectiveness of the proposed method, which improves the ice state detection accuracy from 0.640 to 0.950, thus providing a reliable guarantee for the safe navigation of polar ships.

show abstract

Section: Deep Learning Modelsmentioning

confidence: 97%

“…In the adaptive aurora synthesis, the aurora effect is adjusted through the frequency domain based on the local brightness and contrast, as shown in Equation (16).…”

Section: Adaptive Augmentation Strategy Based On Brightness and Contrastmentioning

confidence: 99%

Section: Adaptive Augmentation Strategy Based On Brightness and Contrastmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

An image augmentation‐based ice monitoring method for safe navigation of polar ships

Liang,

Wang,

Zhao

et al. 2024

Quality & Reliability Eng

View full text Add to dashboard Cite

show abstract

Dual Constraint Parallel Multi-scale Attention Network for Insulator Detection in Foggy Scene

Sun,

Huang,

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Insulator Defect Recognition Based on Vision Big‐Model Transfer Learning and Stochastic Configuration Network

Liu,

Ma,

Zheng

et al. 2024

IET Signal Processing

View full text Add to dashboard Cite

Insulator faults are an important factor in causing outages and accidents in power transmission lines. In response to problems related to inefficient insulator positioning, limited robustness of insulator defect feature extraction methods, and the scarcity of defective insulator samples leading to poor classifier generalization, a method for insulator defect detection and recognition based on vision big‐model transfer learning and a stochastic configuration network (SCN) is proposed. First, data augmentation methods, such as Mosaic and Mixup, are employed to mitigate overfitting in the YOLOv7 network. Second, StyleGanv3 adversarial generative networks are used to augment the dataset of defective insulators, which enhances dataset diversity. Third, a vision big‐model transfer learning method based on DINOv2 is introduced to extract features from insulator images. Finally, an SCN classifier is used to determine the status of insulators. Experimental results demonstrate that the applied data augmentation methods effectively mitigate overfitting. YOLOv7 accurately detects insulator positions, and the use of the DINOv2 feature extraction method increases the accuracy of insulator defect recognition by 28.6%. Compared with machine learning classification methods, the SCN classifier achieves the highest accuracy improvement of 17.4%. The proposed method effectively detects insulator positions and recognizes insulator defects.

show abstract

Fault Detection Method of Glass Insulator Aerial Image Based on the Improved YOLOv5

Cited by 11 publications

References 27 publications

An image augmentation‐based ice monitoring method for safe navigation of polar ships

An image augmentation‐based ice monitoring method for safe navigation of polar ships

Dual Constraint Parallel Multi-scale Attention Network for Insulator Detection in Foggy Scene

Insulator Defect Recognition Based on Vision Big‐Model Transfer Learning and Stochastic Configuration Network

Contact Info

Product

Resources

About