A project for Dubhacks 2018 using OCR and Algolia to help people with dietary restrictions classify food that's good and bad for them based on a menu image file.
The pipeline is as follows:
- You go to a restaurant, but realizes that you have a few food restrictions you need to adhere to.
- There are tons of things on the menu and you don't know what you can and cannot get.
- You take a picture using glyco.me
- Glyco.me parses the menu into list of dishes (In the future will use Algolia to support typos and synonyms, and ideally have a feature for users to add unique menu items as a keyword to its original dish name and improve search accuracy)
- Glyco.me searches the ingredients of said dish
- Glyco.me searches Algolia for the glycemic index/nutrition contents of each ingredient
- Glyco.me presents the menu to you in a whole new perspective!
- Limited OCR libraries for react-native
- Many different glycemic index databases (difficult to standardize/choose a plan of attack)
- Finish integrating OCR
- Better UI/UX