Fridgify

Inspiration

The creation of Fridgify came from concern for an issue that we had seen quite frequently during the winter. That common issue for many individuals is seasonal depression, which tends to peak during these times. Seasonal depression can inhibit many daily functions and routines, because of symptoms such as changing appetite, low energy and difficulty going outside. Our solution for this issue came to us through an idea that could help make life a lot more accessible and convenient for those struggling.

What it does

Fridgify is a web application that is meant to deliver innovative and accessible recipes to users from the comfort of their own home!

It allows the user to show a camera ingredients that they have in their own fridge, which is scanned through a database to find recipes that contain those ingredients. A variety of recipes will be displayed to the user, who will be able to access the ingredients and instructions of the recipe at ease, with the ability to have instructions read out to them.

How we built it

Fridgify utilizes Image Processing Software with TensorFlow, OpenCV-python, CVLIB to identify ingredients. The front-end is implemented using HTML and CSS with the back-end using Python Flask. The database of recipes is sourced from MongoDB Atlas.

Challenges we ran into

One of the main challenges we faced was utilizing some of the technologies that we were unfamiliar with, such as the Image Processing and using unfamiliar Python libraries. There was a lot of work put into finding some type of technology that could adequately identify ingredients, and just discovering something that was viable for our ideas. Another challenge that we faced was developing the user interface for this project. We wanted to create a very welcoming and accessible interface, as this would mainly be used to support people in the best way possible. Achieving such a task required a lot of testing and experimentation to find the right theme for our vision of the project.

Accomplishments that we're proud of

Something we are proud of is that we were able to develop a complete project similar to what we imagined. At first, it seemed insurmountable to develop all these features into one working project. However, looking back on it, it was a very rewarding and accomplishing experience to work together and create a tool that can be useful for others.

What we learned

We learned a lot about the importance of both front-end and back-end. As this project required the prioritization of both components, there was a necessary balance we needed to find. We learned how to utilize an object processing model as well as how to implement sound libraries.

What's next for Fridgify

For the future, we would ideally like to train the model to perform with higher accuracy, and be able to expand its scope of recognition. We additionally would like to improve the object processing.