HardHat-Object-Detection-YOLOv8 is a computer vision project designed to enhance safety measures in construction environments. By employing the state-of-the-art YOLOv8 (You Only Look Once version 8) algorithm, this project aims to accurately detect and identify whether construction workers are wearing their safety helmets or not. We will also take a look at YOLOv5, for comparison.
The primary goal of this initiative is to assist supervisors, safety officers, and automated systems in maintaining and enforcing safety regulations on site. The integration of AI in safety protocol monitoring opens the potential for real-time alerting and tracking of safety breaches, allowing for swift corrective actions.
This project leverages the YOLOv8 object detection algorithm, which is well-regarded for its detection accuracy and real-time performance. This algorithm enables the software to identify and differentiate between construction workers wearing hard hats and those who are not.
The software can be integrated into an existing video surveillance system to continuously monitor compliance with safety protocols. On detecting a worker without a helmet, the system can trigger alerts for immediate action. This way, the project aims to contribute significantly to accident prevention in the construction industry.
To use HardHat-Object-Detection-YOLOv8, simply clone this repository and execute the main Python notebook. (Since it is a personal project, i don't plan to get it production-ready) Ensure that the software has access to either a live video feed or a recorded video file from your construction site.
Please note: This project is for academic or personal use and should not replace manual supervision in critical safety situations. Its performance may vary based on numerous factors, including the quality of input video and the presence of obstructions in the scene.
