Deep Learning Oriented Smart Constructional Material and Face Recognition System for the Site Safety

Ensuring worker safety and compliance on construction sites is critical due to the high-risk nature of the environment and the need for efficient material handling. This paper presents a deep learning-based system for real-time detection and recognition of construction materials and equipment to enhance safety and optimize workflows. The model integrates pre-trained facial recognition networks (VGG Face) and object detection models (YOLO) to achieve accurate worker identification and material classification. A custom dataset was created, comprising 52 images for each of the 24 material/equipment classes and 20 worker profiles from 7 trades, including carpenters, metalworkers, and heavy machinery operators. The system successfully detects and classifies 24 distinct material classes, achieving 88.4 % precision and 63.7 % recall. By verifying workers' presence in designated environments and detecting materials in real-time, the proposed system ensures compliance with safety regulations and addresses critical challenges in construction site monitoring and safety enforcement.