Recently, I read an article about the mountains of trash accumulating around the world. It boggled my mind that so much trash is produced every single day. However, that comes to say, the amount of waste around the world can be decreased by thousands of tons if people were able to throw away our waste correctly. If we were able to recycle what's suppose to be recycled, throw away what's suppose to be thrown away, and place bio-degradable products where they belong, I believe the waste problem in this world would be minimized. From there, we found the inspiration to create LitterSmart.

What it does

LitterSmart is a program that allows users to analyze items through video and image analysis to see whether it should be considered compost, trash, or recyclable.

How we built it

Using opencv2 and the Google Cloud Vision API through Python, we were able to create a program that pulled in a live video feed, convert it into a series of frame images, and identify different categories for items. We first used opencv2 library to access the webcam of our computer and saved the video into a .mov file. Using this file we were able to capture still frames to analyze each component and then using the Google Cloud Vision API, analyze key features of each item. Using keywords, we were able to categorize items, whether they were considered 'Trash', 'Compost', or 'Recyclable'. We then used pygame to pop up a GUI for users to know what their item scanned to be.

Challenges we ran into

The main challenge we ran into was figuring out which keywords were needed to organize items. This was the most time consuming part of the entire program. We also had difficulty converting the .mov file into multiple still frames.

Accomplishments that we're proud of

As first-years with little to no experience with creating a program from scratch, we believe we created something very unique that challenged our coding abilities as well as sparked the creativity inside of us. We were proud that we implemented the API's correctly and were able to integrate them within our LitterSmart program.

What we learned

Through this entire Hackathon, we learned the valuable technical skills to handle API's. We learned the powerful tools that the Google Cloud APIs can offer and this sparked our curiosity for future projects down the line.

What's next for LitterSmart

In practice, LitterSmart can have the capability to change our waste in a massive way. Hopefully, through refining the code to make the Google Cloud Vision API to analyze images more efficiently, we would create a faster and more accurate program that analyzes items. We could eventually implement hardware into this, creating trash bins that only open if it scans an item as 'trash'.

Built With

Share this project: