This project leverages the powerful YOLOv8 object detection model, implemented in PyTorch, to accurately identify potholes in road images. The goal is to enhance road safety and maintenance efficiency through advanced computer vision techniques.
Potholes pose significant risks to road safety, contributing to accidents and vehicle damage. This project addresses this issue by employing YOLOv8, a state-of-the-art object detection model, to detect and localize potholes in various road conditions.
Ensure you have the following prerequisites installed:
- Python 3.6+
- PyTorch 1.8.1+
- CUDA Toolkit (for GPU acceleration)
The training dataset can be downloaded from Kaggle at https://www.kaggle.com/datasets/anggadwisunarto/potholes-detection-yolov8. After downloading, extract the dataset into a directory within the project folder.
The project includes a Jupyter Notebook (potholes-detection-yolov8-notebook-01.ipynb) that guides you through the model training process. You can run this notebook on kaggle.