YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment
Real-time insulator defect detection is critical for ensuring the reliability and safety of power transmission systems. However, deploying deep learning models on edge devices presents significant challenges due to limited computational resources and strict latency constraints. To address these issu...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/18/7/1668 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849730423607787520 |
|---|---|
| author | Qinglong Wang Zhengyu Hu Entuo Li Guyu Wu Wengang Yang Yunjian Hu Wen Peng Jie Sun |
| author_facet | Qinglong Wang Zhengyu Hu Entuo Li Guyu Wu Wengang Yang Yunjian Hu Wen Peng Jie Sun |
| author_sort | Qinglong Wang |
| collection | DOAJ |
| description | Real-time insulator defect detection is critical for ensuring the reliability and safety of power transmission systems. However, deploying deep learning models on edge devices presents significant challenges due to limited computational resources and strict latency constraints. To address these issues, we propose YOLOLS, a lightweight and efficient detection model derived from YOLOv8n and optimized for real-time edge deployment. Specifically, YOLOLS integrates GhostConv to generate feature maps through stepwise convolution, reducing computational redundancy while preserving representational capacity. Moreover, the C2f module is restructured into a ResNet–RepConv architecture, in which convolution and Batch Normalization layers are fused during inference to reduce model complexity and enhance inference speed. To further optimize performance, a lightweight shared-convolution detection head significantly reduces parameter count and computational cost without compromising detection accuracy. Additionally, an auxiliary bounding box mechanism is incorporated into the CIoU loss function, improving both convergence speed and localization precision. Experimental validation on the CPLID dataset demonstrates that YOLOLS achieves a 42.4% reduction in parameters and a 48.1% decrease in FLOPs compared to YOLOv8n while maintaining a high mAP of 91%. Furthermore, when deployed on Jetson Orin NX, YOLOLS achieves 44.6 FPS, ensuring real-time processing capability. Compared to other lightweight YOLO variants, YOLOLS achieves a better balance between accuracy, computational efficiency, and inference speed, making it an optimal solution for real-time insulator defect detection in resource-constrained edge computing environments. |
| format | Article |
| id | doaj-art-2cf116dca3734986aa81f8638aed1a17 |
| institution | DOAJ |
| issn | 1996-1073 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj-art-2cf116dca3734986aa81f8638aed1a172025-08-20T03:08:52ZengMDPI AGEnergies1996-10732025-03-01187166810.3390/en18071668YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge DeploymentQinglong Wang0Zhengyu Hu1Entuo Li2Guyu Wu3Wengang Yang4Yunjian Hu5Wen Peng6Jie Sun7Department of Mechanical Engineering, North China Electric Power University, Baoding 071003, ChinaDepartment of Mechanical Engineering, North China Electric Power University, Baoding 071003, ChinaDepartment of Mechanical Engineering, North China Electric Power University, Baoding 071003, ChinaDepartment of Mechanical Engineering, North China Electric Power University, Baoding 071003, ChinaDepartment of Mechanical Engineering, North China Electric Power University, Baoding 071003, ChinaSchool of Mechanical Engineering, Shenyang Jianzhu University, Shenyang 110168, ChinaThe State Key Laboratory of Digital Steel, Northeastern University, Shenyang 110819, ChinaThe State Key Laboratory of Digital Steel, Northeastern University, Shenyang 110819, ChinaReal-time insulator defect detection is critical for ensuring the reliability and safety of power transmission systems. However, deploying deep learning models on edge devices presents significant challenges due to limited computational resources and strict latency constraints. To address these issues, we propose YOLOLS, a lightweight and efficient detection model derived from YOLOv8n and optimized for real-time edge deployment. Specifically, YOLOLS integrates GhostConv to generate feature maps through stepwise convolution, reducing computational redundancy while preserving representational capacity. Moreover, the C2f module is restructured into a ResNet–RepConv architecture, in which convolution and Batch Normalization layers are fused during inference to reduce model complexity and enhance inference speed. To further optimize performance, a lightweight shared-convolution detection head significantly reduces parameter count and computational cost without compromising detection accuracy. Additionally, an auxiliary bounding box mechanism is incorporated into the CIoU loss function, improving both convergence speed and localization precision. Experimental validation on the CPLID dataset demonstrates that YOLOLS achieves a 42.4% reduction in parameters and a 48.1% decrease in FLOPs compared to YOLOv8n while maintaining a high mAP of 91%. Furthermore, when deployed on Jetson Orin NX, YOLOLS achieves 44.6 FPS, ensuring real-time processing capability. Compared to other lightweight YOLO variants, YOLOLS achieves a better balance between accuracy, computational efficiency, and inference speed, making it an optimal solution for real-time insulator defect detection in resource-constrained edge computing environments.https://www.mdpi.com/1996-1073/18/7/1668lightweight detectiontransmission linepower insulator defectsedge computingreal-time detection |
| spellingShingle | Qinglong Wang Zhengyu Hu Entuo Li Guyu Wu Wengang Yang Yunjian Hu Wen Peng Jie Sun YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment Energies lightweight detection transmission line power insulator defects edge computing real-time detection |
| title | YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment |
| title_full | YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment |
| title_fullStr | YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment |
| title_full_unstemmed | YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment |
| title_short | YOLOLS: A Lightweight and High-Precision Power Insulator Defect Detection Network for Real-Time Edge Deployment |
| title_sort | yolols a lightweight and high precision power insulator defect detection network for real time edge deployment |
| topic | lightweight detection transmission line power insulator defects edge computing real-time detection |
| url | https://www.mdpi.com/1996-1073/18/7/1668 |
| work_keys_str_mv | AT qinglongwang yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT zhengyuhu yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT entuoli yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT guyuwu yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT wengangyang yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT yunjianhu yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT wenpeng yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment AT jiesun yololsalightweightandhighprecisionpowerinsulatordefectdetectionnetworkforrealtimeedgedeployment |