🅿️ Parking Space Recognition System

An intelligent computer vision system that detects and classifies parking spaces in real-time using machine learning algorithms.

📋 Table of Contents

• Overview
• Features
• Models
• Installation
• Usage
• Project Structure
• Documentation
• Demo
• Performance
• Technical Stack
• Contributing
• Authors
• Contact

🎯 Overview

This project addresses the growing challenge of urban parking management by leveraging advanced computer vision and machine learning techniques. The system can automatically detect and classify parking spaces as "empty" or "occupied" from video streams or images.

Academic Project Details:

• Course: Modélisation et Simulation en IA (4th Year)
• Academic Year: 2024-2025
• Institution: ENSAM Meknès
• Supervisor: Mr. Tawfik Masrour

✨ Features

• Real-time Detection: Process video streams for live parking space monitoring
• Dual Model Support: Choose between SVM and YOLOv8 models based on your requirements
• Web Interface: User-friendly Streamlit application for easy interaction
• High Accuracy: Achieves ~85% accuracy with SVM and ~95% with YOLOv8
• Scalable Architecture: Modular design for easy integration and expansion
• Visual Analytics: Comprehensive performance metrics and confusion matrices

🧠 Models

Support Vector Machine (SVM)

• File: models/model.p
• Type: Classification model
• Accuracy: ~85%
• Use Case: Suitable for smaller datasets and simpler setups
• Training: Grid search optimization for hyperparameters

YOLOv8 (You Only Look Once v8)

• File: models/best.pt
• Type: Object detection model
• Accuracy: ~95%
• Use Case: Real-time applications with high accuracy requirements
• Training: 100 epochs on 1700 labeled images

🚀 Installation

Prerequisites

• Python 3.8 or higher
• pip package manager
• Git (for cloning the repository)

Clone the Repository

git clone https://github.com/yourusername/parking-space-recognition.git
cd parking-space-recognition

Install Dependencies

pip install -r requirements.txt

Additional Requirements (if needed)

pip install -r requirementProject.txt

💻 Usage

Running the Streamlit Application

streamlit run src/app.py

Training Models

For detailed training steps and model development, refer to:

• Jupyter Notebook: Notebook/ParkingDetection_Bellmir_Chegdati.ipynb
• Training Script: src/ParkingSpaceRecognition.py

Using Individual Components

# Example usage of the SVM model
from src.model import load_model
from src.util import empty_or_not

# Load the trained model
model = load_model('models/model.p')

# Classify a parking spot
result = empty_or_not(model, image_crop)

📁 Project Structure

parking-space-recognition/
├── 📁 src/                          # Source code
│   ├── app.py                       # Streamlit web application
│   ├── main.py                      # Main execution script
│   ├── model.py                     # Model utilities
│   ├── ParkingSpaceRecognition.py   # SVM training and evaluation
│   ├── util.py                      # Utility functions
│   └── yolo_page.py                 # YOLOv8 integration
├── 📁 models/                       # Trained models
│   ├── model.p                      # SVM model (pickle)
│   └── best.pt                      # YOLOv8 model (PyTorch)
├── 📁 data/                         # Dataset and videos
│   └── parking_1920_1080.mp4       # Sample parking video
├── 📁 mask/                         # Parking spot masks
│   ├── mask_1920_1080.png          # Main mask file
│   └── mask_crop.png               # Cropped mask
├── 📁 clf-data/                     # Classification dataset
│   ├── empty/                       # Empty parking spots
│   └── not_empty/                   # Occupied parking spots
├── 📁 Notebook/                     # Jupyter notebooks
│   └── ParkingDetection_Bellmir_Chegdati.ipynb
├── 📁 Readthedocs/                  # Documentation
├── requirements.txt                 # Python dependencies
├── requirementProject.txt           # Additional requirements
└── README.md                        # This file

📚 Documentation

Comprehensive documentation is available on ReadTheDocs: 📖 View Documentation

Key Documentation Sections:

• Technical Implementation: Detailed code explanations
• Model Comparison: SVM vs YOLOv8 performance analysis
• API Reference: Function and class documentation
• Training Guide: Step-by-step model training process

🌐 Demo

Live Demo

Experience the system in action: 🚀 Try the Live Demo

Note: Some features may be limited on Streamlit Cloud. For the best experience, run the application locally.

Project Report

For comprehensive project details and research methodology: 📄 View Full Report

📊 Performance

Feature	SVM	YOLOv8
Model Type	Classifier	Object Detector
Accuracy	~85%	~95%
Real-time Capability	Limited	Excellent
Implementation Effort	Medium	High
Dataset Size	Smaller datasets	Large datasets
Use Case	Simple setups	Production environments

🛠️ Technical Stack

• Computer Vision: OpenCV
• Machine Learning: scikit-learn
• Deep Learning: YOLOv8 (Ultralytics)
• Web Framework: Streamlit
• Image Processing: scikit-image
• Data Visualization: Matplotlib, Seaborn
• Data Handling: NumPy, Pandas

🤝 Contributing

We welcome contributions to improve the system! Here's how you can help:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

👥 Authors

Students:

• Bellmir Yahya - Github | LinkedIn
• Chegdati Chouaib - Github | LinkedIn

Supervisor:

• Mr. Tawfik Masrour - Academic Supervisor

📞 Contact

For questions, suggestions, or collaboration opportunities:

• Email: [yahyabellmir@gmail.com]
• Project Link: [https://github.com/yourusername/parking-space-recognition]
• Documentation: [https://parking-recognition-system-2.readthedocs.io/en/latest/]

---

⭐ Star this repository if you found it helpful! ⭐