UAV Low-Altitude Remote Sensing Inspection System Using a Small Target Detection Network for Helmet Wear Detection

Automated construction site supervision systems are critical for reducing accident risks. We propose a helmet detection system with low-altitude remote sensing by UAVs in this paper to automate the of helmet-wearing workers overcome limitations most efforts that rely on ground surveillance cameras and improve efficiency safety supervision. The proposed has following key aspects. (1) an approach speedy automatic at sites regularly leveraging flexibility mobility UAVs. (2) A single-stage high-precision attention-weighted fusion network is proposed, allowing AP small-sized targets be enhanced 88.7%, considerably improving network’s performance targets. (3) Our method can accurately categorize helmets based whether they worn or not type color, mAP 92.87% maximum accuracy each category.