OpenLabel

This project lets users scan items in a grocery store and determines if its suitable based on a user's preferences.

Comment

Inspiration

According to food allergy research and education, 3.4 million patients have food allergy related emergency room visits

Most food products found in your average grocery store are not explicitly labeled and more often than not, the nutrition and ingredient labels on them are difficult to interpret for the average consumer.

Furthermore, some food products that are marketed as healthy products may contain potential harmful chemicals and allergens that can lead to drastic and unexpected outcomes.

We wanted to help people understand exactly what they are putting into their bodies in the context of their own health - not what the flashy packaging tends to lead you to believe.

We believe everyone deserves honest food.

What it does

OpenLabel snaps a picture of any food item that you can find in your grocery store and delivers:

  • Simplified ingredient breakdowns that anyone can understand
  • Personalized nutrition analysis tailored towards your own dietary needs
  • Healthier alternatives with better choices
  • A verdict on whether an item is actually appropriate for you

No more hidden chemicals. No more misleading words. Just the truth.

How we built it

  • CustomTkinter for a clean, interactive UI
  • OpenCV for live webcam integration
  • Google Gemini API for personalized analysis and brand detection
  • Python to tie it all together into one seamless experience

Challenges we ran into

Our biggest challenge by far was trying to find out how to accurately detect the item and its nutrition information just by a picture of its packaging.

We did not want to use barcodes because we didn't want OpenLabel to be limited to just pre-packaged products, we wanted OpenLabel to be able to identify all kinds of foods just by the visual itself.

Thankfully, the power of Google Gemini combined with OpenCV bounding was capable of providing extremely accurate product lookups and breakdowns.

Accomplishments that we're proud of

  • We developed a product that we, ourselves, are actually excited about and want to use in our daily lives
  • We gave people a way to understand their food without a science degree
  • We gave people reassurance in the fact that the products they consume are right for them
  • We were able to effectively leverage Google Gemini API and develop an application with complexity while maintaining simplicity.

What we learned

  • Simplicity is powerful
  • Dieting can be intimidating; especially when people don't know what 80% of the ingredients are
  • Many people feel helpless and misled when shopping - OpenLabel can be a lifeline

We constantly thought of ways that we could add as much technical complexity into our project as possible. However, we eventually settled on an idea that, while seemed simple at first, actually turned out far better than we would have ever expected.

We learned that you don't need to make a project extremely lucrative with complex algorithms to solve a real world problem. Sometimes, simplicity is key.

What's next for OpenLabel

  • Mobile integration for on-the-go scanning
  • User profiles to tailor feedback based on dietary needs
  • A custom trained model that is specifically designed to identify food products
  • Making OpenLabel a companion for every grocery trip

OpenLabel doesn't just tell you waht's in your food - it tells you what that means.

Because you never know when one bad product can change your life forever - your health deserves honesty

Built With

Share this project:

Updates