Helmet and Safety Vest Detection System with Integrated Database and Workplace Safety Dashboard
Specifications
- Code
- RA2611
- Department
- DKR
- Technology
- Raspberry Pi · Node-RED · MySQL · YOLOv8
- Theme
- Safety & Security · AI & Computer Vision
- Student
- Meor Iskandar bin Meor Asri
- Supervisor
- Khairunnisa binti Zainal Azhar
Objectives
To implement a camera sensor module to capture worker images and detect helmet and vest compliance in real time. Notification & Dashboard System To develop a Node-RED dashboard to display compliance status and send alerts to supervisors instantly. Database Management To create a MySQL database to store worker profiles and log PPE compliance automatically.
Scope of project
Data Logging Save data into a local file/database for monitoring and analysis. Environment System operates in a controlled environment. Detection Implement detection using OpenCV and a trained deep learning model (YOLO). Live Input Use a camera module to capture images/video for real-time processing.
Project description
This project develops a real-time workplace safety system that combines PPE detection with facial recognition. A camera feeds video to a Raspberry Pi running a YOLOv8 model with OpenCV to check whether workers are wearing safety helmets and vests, while facial recognition identifies who each worker is. Each result, including worker identity, PPE status, and timestamp, is logged in a MySQL database and shown on a Node-RED dashboard. Supervisors can then see exactly who is not compliant and monitor safety remotely, replacing inconsistent manual checks with a low-cost, automated solution. Not Complied: helmet or vest missing. The detection box is red and the dashboard logs it as non-compliant. Complied: worker wearing both helmet and safety vest. The detection box is green and the dashboard logs it as compliant. Unknown: The system can’t match the face to a registered worker. The dashboard logs it as unknown worker. No matter the compliance.