Clean Vision Demo

clearvision.ravenjs.net

Inspiration

Introducing Clear Vision, a device that combines the power of computer vision and AI to detect and sort waste materials. With its state-of-the-art technology, Clear Vision uses image recognition to identify and classify various materials in real-time, making sorting easier and more efficient. The device is easy to install and can be integrated into existing waste management systems. By reducing human error and increasing accuracy, this Arduino device offers an eco-friendly solution to waste management, helping to create a cleaner, more sustainable future.

What it does

Clean Vision is an Arduino device that utilizes computer vision and AI to sort waste materials into the correct options
It also has an connected UI to showcase whether or not you had successfully put the item in the correct waste disposal options.

How we built it

It uses OpenCV and Tensorflow to detect what the current material in focus is.
Use python to perform the calculations to see what material the object should be.
Uses WebSockets to provide real-time updates to the user interface.

Challenges we ran into

Getting stuff to connect well together.
Troubleshooting computer vision, problems with overfitting.
Using ArduinoUno properly
Dealing with race conditions
Problems with playing audio in python

Accomplishments that we're proud of

Since this was our first time using OpenCV and Tensorflow completing the project that uses them to do perform a task that can benefit society was quite the experience.

What we learned

We learned that Tensorflow modelling was quite interesting to implement.
We combined a working website with Tensorflow and audio files.

What's next for Clean Vision

support for more types of bins
automatic system to handle many items at once sequentially because realistically, people have many things to throw away all at once: avoids needing to have people think about where each item belongs