ParkSmart Analytics

A research project using computer vision to detect parking violations from parking lot images. Built with Streamlit and Python, it identifies improperly parked cars, logs violations, and provides a dashboard for review and management. Ideal for exploring AI in everyday scenarios.

Overview

Parking management can be challenging, especially in busy areas where improper parking can lead to reduced capacity and congestion. ParkSmart Analytics aims to provide a simple tool to analyze parking lot images, detect improperly parked cars, and log these incidents for review.

This project is purely for R&D purposes and aims to explore the capabilities of image analysis and machine learning in addressing everyday problems like parking management.

Features

• Parking Violation Detection: Identifies cars that are not parked correctly within the designated parking lines.
• Image Analysis: Uses computer vision models such as YOLOv5 to detect cars and parking lines.
• Logging and Record-Keeping: Stores records of violations, including images and status updates.
• User Dashboard: Allows users to upload images and view past violation records.
• Admin Dashboard: Provides basic admin controls to manage parking violations (issue warnings, etc.).

Project Structure

parking_violation_detection/
│
├── park_smart/                     
│   ├── __init__.py
│   ├── config/
│   │   ├── __init__.py
│   │   ├── logging_config.py       
│   │   ├── model_config.py         
│   ├── utils/
│   │   ├── __init__.py
│   │   ├── image_processing.py             
│   │   ├── database.py             
│   │   ├── number_plate_recognition.py 
│   │   ├── parking_validations.py  
├── app.py                          
├── requirements.txt                
└── data/                           
    └── park_smart.db

Getting Started

Prerequisites

• Python 3.7 or higher
• Streamlit
• OpenCV
• NumPy
• Pillow
• PyTorch

Installation

1. Clone the Repository

   git clone https://github.com/prblsing/ParkSmart.git
   cd ParkSmart

2. Install Dependencies

   Use pip to install the necessary packages:

   pip install -r requirements.txt

3. Run the Application

   Start the Streamlit app:

   streamlit run app.py

4. Open in Browser

   Visit http://localhost:8501 to access the app.

Usage

• User Dashboard: Upload an image to analyze for parking violations or view past violation records.
• Admin Dashboard: Log in as an admin to manage violations (issue warnings, etc.).

Future Work

• Enhance Car Detection Accuracy: Improve detection accuracy by refining models and training data.
• License Plate Recognition: Add functionality for license plate recognition.
• Advanced Analytics: Incorporate more detailed parking analytics, such as space utilization rates.

Contributions and License

This project is licensed under the MIT License. Contributions, suggestions, and collaboration are welcomed! Feel free to fork the repository and submit a pull request.

Disclaimer

This project is intended for educational and research purposes only.