This repository contains the scripts and code for the Reverse Vending Machine project. This is a Data Science project where we used a RaspberryPI 3, and a Deep Learning model to create a reverse vending machine targeted for the tunisian market, specializing in the classification of plastic bottles and aluminum/metal cans. The users have the chance to receive points/vouchers depending on the object they insert into the vending machine.
For more information on how a reverse vending machine works :
The "Vegeta Reverse Vending Machine" does not rely on expensive sensors to detect the type of the object inserted, it uses AI, A Deep Learning CNN model to perform object classification, from a cheap 1080p HD IR camera.
The types of the objects which the deep learning model has been trained on :
- Plastic bottles, ranging from 25cl to 2 liters.
- Aluminum/Metal cans.
A Raspberry PI 3 Model B running on Raspian OS has been used for this project. The main code that has been used capture the bottle pictures can be found in the Vending_Machine_Python_Script.py file.
Here's how we've setup the card and the components :
For the Demo Videos :