Parking Control

This computer vision system is designed to detect the presence of vehicles in the regions of interest by monitoring and showing vacant parking places.

Objective

This project is intended to demonstrate the power of the DragonBoard™ 410c from Arrow Electronics in a Smart Cities solution. Rather than monitor the parking lot with a lot of sensors, you can simply use this board and a camera to cover a huge area of parking slots, utilizing the capacity of edge computing.

Materials Required / Parts List / Tools

• DragonBoard 410c boards
• 12V wall adapter
• Micro SD Card (8GB)
• Fanshine Eletronics IP Camera

Source Code / Source Examples / Application Executable

Build / Assembly Instructions

The build instructions can be found in the README file in the GitLab link. Basically, you have to install all the libraries and the application onto the board. You will need to physically install the camera pointed towards parking slots and know how to access the camera stream. The code on the repository uses an IP camera. When using an IP camera, change the IP and login info. If you use another model, you will need some OpenCV knowledge, but it's simple.

Usage Instructions

With the camera set up and the desired regions marked, run the application on the DragonBoard 410c. The app will show the image with the result and update a string with the status of each parking slot. You can send that string to the cloud or to a front-end platform like a mobile app, because all the hard work is done in the edge.