Inspiration
An app that tracks pantry items and suggests recipes based on available ingredients offers several benefits. It helps reduce food waste by utilizing what's already on hand, saves time and encourages cost-effective meal planning. Additionally, it can promote healthier eating, introduce variety to meals, and cater to dietary preferences, making it a valuable tool for simplifying cooking and reducing environmental impact.
What it does
Pantry Pal is a kitchen-assistant that aims to reduce food waste by suggesting recipes based on what’s in your digital pantry. Users can add to their pantries manually or by using computer vision to scan their receipts. Pantry Pal can also be used to search for specific recipes.
How we built it
The web app itself was built in React.js. The receipt-scanning computer vision scripts were built using the Tesseract.js library. After collecting the text data, we used regular expressions to parse for ingredients. These ingredients form the pantry and can be used with the Spoontacular API to look for recipes.
Challenges we ran into
React was unfamiliar to us, so a few of our original plans for implementing features did not work out due to incompatibility issues (such as computer vision through the Google Cloud Vision API). We had to look in different directions and compromise. Debugging was very difficult in such a new environment.
Accomplishments that we're proud of
This is our first project working in React. For two of our members, this is also their first ever Hackathon. We were incredibly proud that we were able to create a finished product that we believe can create a positive impact on others.
What we learned
We have largely only worked with object-oriented programming, so it was a great experience to learn a new function-based approach that React requires. We also learned a lot about computer vision, node.js, and several other unique libraries and APIs.
What's next for Pantry Pal
We would love to expand Pantry Pal to mobile devices and add more features to become an all-in-one kitchen app. Some ideas we worked on but didn’t have time to fully implement include graph visualizations for grocery store spending and using a webcam to take live photos of receipts. We would also like to implement an AI training model as a better alternative for ingredient recognition.
Log in or sign up for Devpost to join the conversation.