This GitHub repository contains the code and instructions to create a custom object detection model for helmet detection using YOLOv8 and Roboflow. This project uses Roboflow for data annotation and preparing the custom dataset.
In this project, we have used YOLOv8, a single stage object detection algorithm that has shown great results in detecting various objects. We have trained our own custom model to detect helmets in images as well as videos.
Roboflow is an end-to-end computer vision platform that helps you to
- Annotate, and create datasets quickly and easily with bounding boxes or other object detection labels, and export the dataset in a format that can be used to train your model.
- Collaborate with your team on computer vision projects.
- Export, train, and deploy computer vision models.
To get started, you will need to create an account on Roboflow and have the following dependencies installed on your system:
- Helmet Detection Dataset : https://app.roboflow.com/helmet-dataset-aho0e/sighted/3
- Number Plate Dataset : https://app.roboflow.com/numberplateocr/ocr-qdbjr/1