EcoEats
This project was inspired by experiences with food waste.
University students often face busy schedules and may forget
the food in their fridge or buy whatever is available during
exam season. Reflecting on these decisions highlights the
significant impact of food consumption and interaction on
the environment. It is important for people to take every
purchasing decision seriously, as every bite wasted could
have gone to someone in need.
EcoScore is accessible via browser on a laptop. Upon opening,
users can choose to start scanning or add an item to inventory.
Selecting scan mode opens the camera, allowing users to place a
food item in view. A picture is automatically taken, processed,
and the food item is identified, displaying a corresponding
"Eco-Score" out of 5. The Eco-Score considers carbon emissions
and water consumption during production, comparing it with data
from other items.
The basic user interface (front end) was developed using HTML,
CSS, and JavaScript. The back end involves a machine learning
model built with TensorFlow. Hundreds of pictures of different
food items were used to train the model for accurate prediction.
An algorithm for developing the Eco-Score was also created.