It’s clear that textile waste is a major problem harming the environment. On average, Americans throw away 14 million tons of clothing each year, even if it’s still wearable. In fact, one in three women consider an article of clothing worn once or twice to be old, contributing to their decision to dispose of clothes that are in good condition. And because people mostly throw away their clothes rather than recycle them, textiles may sit in landfills for over 200 years, leaching toxic chemicals into the environment or releasing harmful greenhouse gases.
What it does
TerraFit aims to connect people to each other through clothes. Users can register for an account and find nearby places to donate the clothes they no longer want by uploading their zip code. They can also choose to upload a picture of their clothes and donate it to other members of the TerraFit community. On top of that, with tensorflow, a library for machine learning, the website suggests certain types of clothing for the user based on the pictures they scan. Then, for each article of clothing they donate, they receive a flower to plant in their virtual garden! As an added bonus, TerraFit also includes guides for members to read relating to sustainable fashion.
How I built it
We used Python to web scrape for locations of donation centers as well as to implement the tensorflow model into the code. We primarily used Python to code the functions of the website, and we then utilized html and css to create each webpage. Then, we used Flask to connect the backend Python programs to the html pages.
Challenges I ran into
We struggled to implement the tensorflow model into the website and connect that to the webcam in the browser. However, we were able to add a webcam in the browser that allowed users to scan items, which would then recommend clothes based on the scanned item. Also, none of us had experience designing a website, so we had to quickly learn html and css.
Accomplishments that I'm proud of
We are proud of creating an organization that promotes sustainable fashion without sacrificing fashion. We also love how we branded TerraFit; its brand colors are continuous throughout the website and the info guides are clean and easy to read.
What I learned
This project was the first time we implemented Flask for a webpage, so we had to quickly learn how to use Flask. Also, we incorporated tensorflow in order to identify the article or clothing scanned and recommend clothes based on that, and we had never used tensorflow in Python before.
What's next for TerraFit
Sometimes, clothes can’t be reused. So, we want to expand TerraFit to aid with recycling clothes too by allowing users to find local recycling centers or providing guides on how to recycle clothes. We also hope to add more data for the tensorflow to learn so that it provides more accurate suggestions of clothing for each user.