LW-YOLO: A Lightweight Network for Defect Detection in Photovoltaic Modules

With the swift advancement of photovoltaic power generation, defect detection in photovoltaic modules has become a critical aspect of ensuring their stable operation. Currently, detecting defects in photovoltaic modules with limited computational resources presents a significant challenge. To address this issue, this paper presents a lightweight model for photovoltaic modules detection, called LW-YOLO. First, a lightweight backbone, LDNet, is designed based on YOLOv10n. It utilizes re-parameterized convolutions and attention mechanisms to effectively extract features while reducing computational complexity. Then, the GSConv is introduced to optimize the feature fusion network. Finally, the shared-parameter detection head, LSDECD, is proposed, which significantly improves both the detection accuracy and efficiency of the model. Experimental results show that compared to the baseline model, the number of parameters and computational complexity of LW-YOLO are reduced by 23% and 43%. Moreover, the mAP@0.5 and mAP@0.5:0.95 are improved by 3.3 and 0.8.