CNN-based real-time classification of the availability of a car parking spot based on the video of the parking

A pre-trained CNN model is run on the Jetson Nano microcomputer to classify the availability of each parking spot in real-time based on the frame patches obtained from the video. The video was recorded using the Raspberry Pi Camera attached to the artificically made car parking zone.

The classified output is sent to the server and displayed on the host website. Here is a flow diagram of the project:

The project has been performed in a team of 4 people and was a finalist in the national competition for hardware prototypes across all the universities.