ballsports-rental

Computer Vision project to automate the dispense and retrieval of rental equipment for selected ball sports.

---

Installation

Repo preparation

Clone the repository via HTTPS:

git clone https://github.com/emberfox205/ballsports-rental.git 

Alternatively, if you have SSH installed:

git clone git@github.com:emberfox205/ballsports-rental.git

[Optional] Set up a virtual environment of your choice (e.g. venv, conda, etc.). Make sure the Python version is 3.12.0.

Install dependencies:

pip install -r requirements.txt

In the file check_database.json, set {"database_init": 0}.

Model installation

This repository only offers the web application interface for the ball classification and logo detection models.

To use the application, download the required models here: ballsports-rental-ai

In your local repo, setup a folder named models.

Move the 2 models into models.

Rename the ball classification model into model.keras.

Rename the logo detection model into logo.keras.

HTTPS Certification

As the application utilizes the device's integrated webcam, you are required to generate an https certificate and key.

Use OpenSSL:

• Windows:
  • Third-party releases
  • Usage guide
• Linux:
  • Guide

Note

If you follow the usage guide for Windows, the key and certificate should be found in C:\\Users\\USER. USER may differ on your device.

Now you should have a certificate (*.crt) and private key (*.key).

In your local repo, setup a folder named certs.

Move the certificate and key into certs.

Rename the certificate into certificate.crt.

Rename the key into private.key.

Running the application

While in your local repo, enter into the terminal:

python dashboard.py

Follow either of the addresses following Flask's warning in the terminal.

Upon browser's warning, click 'Advanced', then 'Continue/Proceed to ...'.

Allow for Camera permission if asked.