We wanted to create something that would be useful and also use various technologies we were interested in.

What it does

intelliFu allows people to scan their grocery receipts, and uses clever algorithms to maintain an accurate database of items in your fridge and pantry. Then, intelliFu suggests recipes based on available ingredients.

How we built it

We used optical character recognition (OCR) methods in Python (tesseract) to scan the receipts, and the dropbox API to store the images for easy access from a web server. When the user accesses the intelliFu web-app, the most recent dropbox scans are processed by OCR and added to a SQLite database. Finally, we use Spoonacular's Food API to suggest recipes based on matches to the pantry database.

Challenges we ran into

We spent the first night troubleshooting a faulty ssh connection to our raspberry pi, which we were going to originally use to scan (pi camera) the receipts. More significant challenges arose when attempting to access our SQLite database from javascript, and when attempting to run the python code which analyzed our images from a .php file.

Accomplishments that we're proud of

Merging multiple python libraries, powering through the many SQLite and php errors, and running a 5k.

What we learned

We experimented with methods for optical character recognition, gained proficiency in jQuery and SQLite, and learned about APIs that are made just for sending requests to use OTHER APIs!

What's next for intelliFu

First, a simple mobile app to quickly scan receipts and see recipe ideas (currently, the user takes a photo and uploads it to a pre-named folder in dropbox). We also want to migrate all of the code and databases to a web server, and incorporate a secure login screen for people to access their own pantries.

Share this project: