Skip to content

Latest commit

 

History

History
20 lines (13 loc) · 920 Bytes

README.md

File metadata and controls

20 lines (13 loc) · 920 Bytes

toucan-hardware

This repository contains all the necessary hardware interfacing functionality to support our application, parkn, which uses Computer Vision and Deep Learning to detect vacant parking spaces in crowded urban areas.

The project includes an embedded application which utilizes a Raspberry Pi along with a servo motor to give your camera a better field of view of the road.

Get started

Required hardware:

  • A camera
  • Raspberry Pi (optional)
  • Servo Motor (optional)

Setup

Run setup.sh to setup the virtual environment and install the requirements. Please note this project requires venv to be installed.

If you wish to use our embedded solution, carry out this process on your Raspberry Pi and add the following argument setup.sh -rpi. You will also have to set RPI=True in main.py to import the correct modules.

Running Toucan

Run run.sh to activate venv and start script