FoodAllerGuard

Inspiration

One of our group members had an allergic reaction to food during an outing. The menu did not contain any listed allergens.

What it does

Scans text files (.txt, .pdf, etc), live scans, and photos of menus. Uses text recognition models to identify ingredients in food items. Cross-references with user-provided list of allergens and then recommends to user if food is safe to consume or not.

How we built it

Frontend: React, Tailwind CSS, Vite | Backend: Flask, OpenCV for training data, pytesseract for text recognition module.

Challenges we ran into

Lack of time, lack of experience, connection issues with camera, problems with integrating pytesseract.

Accomplishments that we're proud of

Furthered knowledge of AI models, finished despite lack of time, implementation of solution

What we learned

Text detection & recognition, image processing, computer vision, natural language processing.

What's next for FoodAllerGuard

Develop a mobile app, partner with restaurants to show allergens on their menus.

Built With

• css
• flask
• opencv
• pytesseract
• python
• react
• tailwind
• vite