PPE Wearability Detection of Offshore Platform Operators based on Improved RT-DETR

Correctly wearing PPE serves as the last barrier to safety protection for offshore platform operators, necessitating real-time, accurate, and large-scale monitoring of personnel PPE compliance. Based on RT-DETR, a model optimization method incorporating the Time-Frequency Feature Enhancement Module (SFEAF) and the Dual-Branch Attention Mechanism (HiLo) into the backbone network was studied to improve the model's detection accuracy regarding PPE compliance among offshore platform operators. Experiments show that the optimized model exhibits significant advantages in PPE target detection accuracy, speed, and computational efficiency.