Inspiration

Studies have shown for years that going outside drastically improves people’s overall mental and physical wellbeing. However, in the modern world, an increasing number of individuals are not spending enough time in nature. We wanted to make an app that would incentivize people to connect with their local environment. Fleurish is an app that allows you to identify, collect, and catalogue images of plants in your local area.

What it does

Fleurish is a mobile application that works both on Android and iOS and allows users to capture images of plants in their local area. After an image is captured, our machine learning model analyzes what the plant is and identifies it! Capturing an image will also allow our user to gain points or level up!

How we built it

For the frontend of our mobile application we used the Expo framework with React Native and JavaScript. The backend was built using Django to be able to integrate with the machine learning model that was trained using Python and TensorFlow, and we used a PostgreSQL database to house the data of all our plans.

Challenges we ran into

It was some of our first times’ working with mobile applications (and most of us have never used React Native!), so it was a bit tricky to get Expo and to transition from building websites to mobile applications.

Accomplishments that we're proud of

  • Training our own machine learning model to recognize plants: we were able to reach a validation accuracy of 96%!
  • Learned how to use react-native and Expo to build mobile applications!

What we learned

Through working on Fleurish, our team learned

  • A lot about mobile development
  • Training machine learning models

What's next for Fleurish

For future developments of Fleurish, we hope to implement geographic plant data to correspond to a user’s location, allowing them to be able to see the plants available in their local area. In addition, we hope to further develop our plant point system for each new plant found to improve the entertainment quality of our application.

Share this project:

Updates