Each day, 25,000 people, including more than 10,000 children, die from hunger and related causes. Around 854 million people worldwide are estimated to be undernourished, and high food prices might just drive another 100 million into poverty and hunger. All the food produced in the world but not eaten would be sufficient to feed two billion people. That's more than twice the number of undernourished people across the globe. So to address this issue and to solve the root cause, we present you frizzy

What it does

A product consisting of a Hardware box and an app to go along with it with this you'll always get to knows what's in your fridge and when is it going to expire be it a cooked food or a packaged food, with our computer vision model we'll identify the date of expiry and it'll be sent to the firebase from which we'll inculcate in the app, using the app you'll get to know all the food items that you've in your fridge and when are they going to expire, and don't worry we'll always remind well in hand to consume your food before it gets expired!

How we built it

We made a Hardware box containing Rpi, PiCam and two buttons, the PiCam captures the image and sends it to the google firebase, where we used computer vision model built on google collab, and using a database from kaggle (Food101) we trained the model on MobileNet using Vgg backbone and the expiry date an be extracted and it'll be sent to the app using firebase, The App consist of all the food items in your fridge their respective expire date and an option to set remainder according to your convince.

Challenges we ran into

Integration of Firebase was the major issue that we've faced.

Accomplishments that we're proud of

More than 90% implementation and an accuracy of 67% of the ML model although with the time and data constraints we had.

What we learned

Integration of hardware and software can be tedious and how to improve accuracy with limited dataset to name a few.

What's next for Frizzy

upscaling of this so that it can industry ready and can be used in multiple places like cold storages etc, and also giving the user the nutrition vales of the food they have in the fridge and give nutrition advices and suitable diet plans.

Built With

+ 24 more
Share this project: