Acessibility Parking Finder

About The Project

This project aims to improve accessibility parking for users with disabilities. Using AI and computer vision, we've developed a system that detects and tracks the availability of accessible parking spaces in real-time.

Key Features:

• Real-time detection of cars and accessibility symbols
• Dynamic routing to available accessible parking spots
• User-friendly interface inspired by ParkHub

Folder Structure

• .archive - This contains old files not required for the presentation.
• CarDetection - This contains tests we conducted utilizing a public general YOLOv5 model. This model could detect cars, but could not detect the accessibility sign. This model appeared to be too slow, prompting us to train our own car model.
• DatasetManipulation - Code for manipulating images to create larger datasets. Actual dataset not included. Resizes, colors, and distorts images.
• FlaskWebsite - The current complete website. Contains front end and backend, including our best model.
• OpenCV - Reference code for manipulating videos frame by frame which was used when writing the FlaskWebsite code

Tech Stack

• AI Model: YOLO-v5
• Backend: OpenCV, Flask
• Frontend: Flask templates, JavaScript

Performance

• Confidence Score: 50-70%
• FPS Achieved: 6-15 fps
• Response Time: 35-60ms
• Dataset size: 800 images

Installation

Prerequisites

• Python 3.7+
• pip

Required Libraries

opencv-python
numpy
torch
flask

Installation Steps

1. Clone the repository

   git clone https://github.com/SSKYAJI/Acessibility.git
   cd your-repo-name
   

2. Create a virtual environment (optional but recommended)

   python -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`
   

3. Install required libraries

   pip install -r requirements.txt
   

4. Run the application

   python app.py
   

5. Open a web browser and go to http://localhost:5000

Future Plans

1. Expand dataset and improve model accuracy
2. Develop a user-friendly mobile app
3. Partner with local businesses for wider coverage

Acknowledgments

• SMU Campus for data collection
• ParkHub for UI inspiration