Eating healthy is a major concern of people who would like to pursue a healthy life. Nevertheless, when dining out, it would be hard to know exactly how many calories we have consumed. We hope that there is an easy way to figure out the number, and we are inspired by how the banks enable people to use their phone to scan a check, or how apple can recognize a credit card for Apple pay.
What it does
we decide to use some vision technique/ OCR/ Machine Learning to get the calories number from the checks people receive when dining out.
How we built it
We use google vision API to do the text detect so we know what food the users have consumed. And then we use the food name as a key to find the calories number in the dictionary (hopefully we can get more data from the restaurants and build a database). We use flask to build a web app so users can upload their check/ receipt easily and get the number. They can also record the data so they have a monthly/ weekly record of it.
Challenges we ran into
Set up the cloud environment correctly. Parse the text detect API's response correctly and filter the result correctly. Get the data from the restaurants...
Accomplishments that we're proud of
We completed a web app that used advanced machine learning/ cloud technology that targets a healthcare problem! And this can promote people's awareness of a healthy lifestyle, also hope to make the restaurants be more transparent with the nutrition data of the food they provide!
What we learned
Web framework, ML API
What's next for InstaCal
Visualize the calories result, make them more aesthetic, which means there can be some graphs to help users analyze the consumption better.
Build a database to collect restaurants' food nutrition data. Support more records like sodium, calcium intake...