With the rapid advancement of deep learning technology, deep learning-based methods have become the mainstream approach for detecting potential safety hazards in transmission lines, playing a crucial role in power grid safety monitoring. However, existing models are often overly complex and struggle with detecting small or occluded targets, limiting their effectiveness in edge-device deployment and real-time detection scenarios enhanced the YOLOv11 model by integrating it with the ConvNeXt network, a multi-level cross-domain analysis detection model (ConvNeXt-You Only Look Once) is proposed. Additionally, Bayesian optimization was employed to fine-tune the model’s hyperparameters and accelerate convergence. Experimental results demonstrate that CO-YOLO mAP@0.5 reached 98.4%, mAP@0.5:0.95 reached 66.1%, and FPS was 303, outperforming YOLOv11 and ETLSH-YOLO, in both accuracy and efficiency. Compared with the original model, CO-YOLO model improved by 1.9% in mAP@0.5 and 2.2% in mAP@0.5:0.95.
Detecting transformer oil leakage from infrared thermal images is challenging due to complex thermal backgrounds, low contrast, and sensor noise. To address these issues, this study proposes IR-OilNet, a lightweight CNN–Transformer fusion network designed for real-time infrared leakage detection in inspection robots. This research contributes to SDG 6 by preventing industrial water pollution resulting from transformer oil runoff, thereby protecting vital water sources in remote environments. The proposed framework integrates three key components: (1) a parameter-free Thermal Noise Suppression (TNS) module that reduces high-frequency thermal noise while preserving structural information; (2) a compact CNN backbone for local thermal texture extraction; and (3) a lightweight Transformer branch for global contextual modeling. To further enhance feature integration, a Dynamic Weight Fusion (DWF) module is introduced to adaptively balance local and global representations. Experiments on an expanded dataset of 2700 infrared thermal images demonstrate that IR-OilNet achieves an accuracy of 0.9992, F1-score of 0.9992, and AUC of 0.9993, while maintaining only 0.20 M parameters and 0.19 GFLOPs with real-time inference at 433 FPS. Compared with representative CNN and Transformer models, the proposed method achieves a superior trade-off between accuracy and computational efficiency, making it suitable for deployment in resource-constrained inspection systems.