Inspiration
We wanted to create an app that would make recycling easier. Given that recycling guidelines can vary from counties to cities to even universities, it can be hard to figure out what can be thrown in the recycling bin. When things are incorrectly sorted, it can contaminate a large batch of waste to be recycled, causing the whole load to be thrown away. This app aims to reduce that problem and educate users. There are currently no apps that allow users to take a picture of an object and tell them if it's recyclable based on their location. We wanted to remedy that.
What it does
The user enters the location where they want to recycle and uploads a picture of the object they'd like to learn about. The app then uses machine learning to classify the material of the object and tell the user to either recycle it or throw it away based on their city's guidelines. The user can also change their city of interest by going back, which would update the recycling instructions.
Challenges we ran into
Pre-processing the image dataset, converting image files, routing pages
What's next for Recycle Buddy
While this is a good start, there are many ways to improve it:
- increase the training dataset to improve the accuracy
- be able to classify many more types of materials
- be able to classify multiple materials within a picture
- tell the user which bin to place the object in
- create a streamlined process that allows users to request to add a city
Built With
- android-studio
- firebase
- jupyter
- pyflask
- python
- react-native
- sk-learn
Log in or sign up for Devpost to join the conversation.