Inspired by the SeeFood app in the hit show Silicon Valley, we wanted to make an app that can detect types of food and display relevant information concerning that food. Unlike Jian-Yang, our app does not only differentiate between hotdogs and not hotdogs, but also other types of food and not foods.
What it does
FoodEye takes an image of food from your webcam or laptop camera. It then connects with the Watson Visual Recognition API, and using their AI technology, determines what type of food that image is. That data is then displayed through a nicely designed UI using ejs, Bootstrap, and HTML to inform the user. Users can click on a button to retake the image.
How we built it
We used node.js as the main backend of our project while using ejs, Bootstrap, and HTML for our frontend. A number of npm packages were used to assist with the construction of our project, including the jimp and node-webcam packages.
Challenges we ran into
One challenge was connecting to the FatSecure API in order to pull data about nutritional information. Though we would have liked that to have been implemented in our project, time constraints made it very difficult.
Accomplishments that we're proud of
Connecting to the Watson Visual Recognition API and getting a functional project that takes images and determines what food the image is.
What we learned
How to connect to the Watson Visual Recognition API and how to better use node.js.
What's next for FoodEye
Adding a display for nutritional information and perhaps moving onto a mobile platform or putting it online.