This program is used to count the calories of different kinds of food. It applies
Google Cloud's Vision API so that the user need simply take a photo to identify
their food, as opposed to typing in the name of the food. The program then
passes the identifying keywords to a database of food nutrition values,
managed by the U.S. Department of Agriculture (USDA). The database lookup provides an
estimate of the calorie count of your food.
Food data has been scraped from college dining halls (namely, UCSD) and locally stored. College students can compare what they eat to something they may be eating quite frequently, and perhaps realize they need to remedy their eating habits.
[comment]: # (All you need to do is to tap "Take Photo" to take a photo of that food. Then our program will identify that food and provide both estimated calories and actual calories. Estimated calories come from USDA with the API. Actual calories come from the menu of each dinning hall. If you think the result is not accurate, you can add more keywords to the picture to get a more accurate result. Throughout the comparison between esmitated and actual calories, you can know what you eat may not be healthy.)
We have finished integrating the USDA database and Google Cloud Vision API. In future installations of this program, we'll need to ensure that the calorie counts are as accurate by more closely examining the USDA database. Also, we could include other data about college dining halls, as we made but did not apply an interface to the Foursquare API. Finally, to put this program online, we'll have to find a web framework compatible with Google Cloud's App Engine that will allow us to run Python code in an HTML document. (We ran into insurmountable difficulties with Flask.)
Log in or sign up for Devpost to join the conversation.