Overview:
This project focuses on helmet detection and segmentation using computer vision. It uses the YOLOv8l model trained on 145 images to detect whether helmets are being worn in real-time. The system aims to improve workplace safety by automatically checking for helmet compliance.
Key Features:
1. YOLOv8l for Helmet Detection:
The system uses YOLOv8l for detecting helmets in real-time, ensuring quick identification of helmet use in work environments.
2, Dataset:
The model was trained on 145 images, focusing on detecting the presence of helmets to ensure compliance with safety regulations.
3. Performance:
Achieved a mean average precision (mAP) of 0.774, precision of 0.632, and recall of 0.825, demonstrating its reliability in detecting helmets.
4.No Data Augmentation:
Trained over 40 epochs without any data augmentation, indicating good performance even with a limited dataset.
Category:
Tags:
Links
Portfolio 
Leave a comment