Abstract: In order to reduce the risks of accidents on offshore platforms caused by unsafe behaviors of personnel, this study investigates a monitoring algorithm for the wearing of personal protective equipment (PPE) to reduce the risk of unsafe human behaviors. Traditional machine vision algorithms often struggle to detect PPE wearing behavior under challenges such as multi-scale variation, severe occlusion, and small target size. To address these issues, three improved modules of C3K2_WT, C3K2_ADV and C2PSA_SIM are proposed respectively based on the conventional YOLOv11 framework, forming an enhanced YOLOv11-based algorithm for monitoring PPE wearing behavior on offshore platforms. Experiment results show that compared to the original model, the improved YOLOv11 model achieves an increase of 2.7% in average precision mean (mAP@0.5), 3% in recall, 1% in precision and 1.7% in mean average precision (reaching 88.7%) for offshore platform PPE wearing behavior monitoring. The findings demonstrate that the proposed method outperforms other conventional approaches and can effectively monitor PPE wearing behaviors on offshore platforms.