BerryHealthy

an intuitive, AI based app that combats childhood obesity by creating long-lasting healthy eating habits for children

Comment

Inspiration

Throughout the last decade, childhood obesity has been rising at an alarming rate and with the COVID-19 pandemic creating restrictions the rate of obesity among kids ranging from 2 to 19 has increased to 22.4 percent[1]. With almost a quarter of children being considered obese, we decided to develop a tool that will create a fun and healthy experience for kids to become more conscious about the foods they eat and help them create healthier habits.

What it does

Using a machine learning model programmed in Tensorflow BerryHealthy allows the end user to submit an image query of the food they will eat and the CNN will determine the food it is, the data is then sent to an API call that retrieves the nutrition information about the food. The login system gets the data from the front end React to the back end Node.js and the password is encoded using sha256 and the encrypted password and user credentials are stored in a MongoDB database. The user can log in and the backend will fetch the data from the database.

How we built it

The frontend was built using React JS, TailwindCSS, and HTML. The backend was made in Node.JS. For the database we used MongoDB which was used to store all the user's data. The machine learning model was made in Python using Tensorflow.

Challenges we ran into

  • Finding a working API that met our needs for the project.
  • Creating a color scheme that looked both good and provided proper contrast for those with visual impairments.
  • Making user authentication to work using Firebase and MongoDB, some of which was new to our members
  • Building the machine learning model and researching about Tensorflow, as this was new to some of our members

Accomplishments that we're proud of

  • User Authentication
  • Color accessibility
  • React frontend as a whole
  • MongoDB integration and user auth
  • Machine learning model

What we learned

  • Firebase auth
  • Tensorflow image processing

What's next for BerryHealthy

  • Creating a game within the application that allows users to spend their gems on app visuals such as different color schemes and icons etc.
  • Creating a mobile application to complement the web application.
  • Creating a community for users to connect and develop bonds, allowing them to encourage each other to eat healthier.
  • Connecting to information pages for parents, so they can learn about childhood obesity.
  • Improving the machine learning model to accommodate for even more foods.

Built With

Share this project:

Updates