Title-Detection method of PPE wearing and small target positioning for offshore operators: The improved YOLOv11 model and targets recognitionShort Title-Detection method of PPE wearing and small target positioning for offshore operators

Focusing on the practical challenges of insufficient samples, incomplete categories, and low detection accuracy (particularly for small targets) in Personal Protective Equipment (PPE) wearing condition monitoring for operators in offshore environments, this research investigates PPE targets detection for offshore operators using an optimized YOLOv11 model. The optimized model integrates the time-frequency features enhancement module (SFEAF) into the model's backbone network, employs a statistical-driven dynamic gating attention module (TSSA) to refine attention weight distribution in the original C2PSA module, and incorporates a Normalized Wasserstein Distance (NWD) loss function. These modifications collectively enhance the model's capability to detect PPE targets for offshore operators. To mitigate missed detection problem of small targets such as earplugs and gloves, a cascadednetwork of YOLOv11 and YOLOv11-Pose models is proposedfor small targetsdetection. The solution involves extracting human key points through YOLOv11-Posemodel, constructing spatial constraint regions via two-pointarea positioningmethod, enhancing small target features through localized region cropping and normalization, and performing secondary detection on refined regions using YOLOv11model. Ablation and comparative experiments validate the optimization method's efficacy, and the superiority of the cascaded network detection is verified through secondary detection experiments on small targets.