Wydoinn/Pothole-Detection

This repository contains a pothole detection system using YOLOv8 for instance segmentation.

Pothole Detection

Overview

This project implements an advanced pothole detection system using computer vision and deep learning techniques. By leveraging the YOLOv8-small model, we've created a robust and efficient solution for identifying and localizing potholes in road images and videos.

Demo

demo.mp4

Key Features

• YOLOv8-small Model: Utilizes the compact yet powerful YOLOv8-small architecture for real-time object detection and segmentation.
• Multi-format Input: Processes both images and videos for versatile application.
• Real-time Detection: Achieves fast inference times, suitable for mobile and edge devices.
• User-friendly Interface: Implemented with Streamlit for easy interaction.

Technology Stack

• Deep Learning Framework: YOLO (You Only Look Once) v8
• Computer Vision: OpenCV and Supervision
• Data Processing: NumPy
• UI: Streamlit

Setup and Installation

1. Clone the repository:

   git clone https://github.com/Wydoinn/Pothole-Detection.git
   cd pothole-detection
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Run the Streamlit app:

   streamlit run app.py
   

Usage

• Use the Streamlit app to upload images or videos for pothole detection.
• Adjust confidence thresholds and other parameters as needed.