Year: 2026
Type: article
Abstract: To accurately and quickly diagnose overheating defects in dry-type flexible cable terminals to reduce failure rates, hotspots must be detected in real time using aerial infrared images of the cable terminals. To address the challenges of the small target size, complex background, and real-time performance in the task, I defined four types of hotspots in dry-type flexible terminals and propose an improved model based on You Only Look Once model. This model embeds Diverse Branch Blocks to enhance the representation capabilities, introduces the Multi-Scale Convolutional Attention to better focus on key areas and detect small targets, and reduces the model's complexity by introducing the FasterNet Block. The proposed model performs significantly better than the baseline model and other mainstream You Only Look Once series models, satisfying the real-time detection demand for hotspots in dry-type flexible cable terminals using aerial infrared images. The source code can be found at https://gitee.com/yao-shunyu1/real-time-detection-of-hotspots-in-dry-type-flexible-cable-terminals.git .
Source: Scientific Reports
Author: Shunyu Yao
Institution: State Grid Henan Electric Power Company (China)
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7160711248
FWCI:
Cites:
Cited by:
Related to: