Recipe suggestion view
The origin of our idea was the need of having an overview of the different products in the fridge. But, leftovers and food waste are a more general problem. Having an complete overview over his goods in the fridge and/or in the store room are essential to consume them before it's too late. We provide a intuitive and easy accessible way to track your leftovers and goods. However, knowing the available items doesn't make themselves a nice meal. We help you to become a even better chief and and provide provide you great recipes. Get out the most of your food!
What it does
Keeps track of leftovers that are conducted by barcode scanning, image recognition and manual recording. Shows an overview of the leftovers and suggests recipes for a user defined selection of them. In case of no preferences, random menus are displayed.
How I built it
We use react-native to provide cross-platform availability and support Android as well as iOS users. To record the goods and leftovers we use a manual input form, image recognition and barcode scanning. Our barcode scanning interface is built on the scandit api. We use Microsofts Vision services on Azure to do image recognition and categorise the items. In addition to these technologies we provide the users with a simple input form to manually record goods. We use the food2fork API to get access to a huge database of recipes. We use http requests directly from our client software.
Challenges I ran into
Cross-platform developing on Android and iOS. Using completely different API's of various services directly from the client-side.
Accomplishments that I'm proud of
Running image recognition of food. Easy accessible and clean user interface providing real-time barcode scanning. Completely independent client software without any server backend. Integration of recipe data provider with a huge amount of different meals.
What I learned
Cross-platform development sounds nice at the first glance, but isn't that easy to do in the end. It's possible to develop a full working app without a heavyweight server in the backend.
What's next for LeftLovers
- Use more complete barcode provider
- Enable sharing of goods for shared flats and families (including a reservation system for leftovers)
- Text recognition to add whole shopping receipts
- IOT integration into fridge and shelves to make make out checking/removing items easier
- Nutrition facts for menus and single items.
- Rating and sharing of the best recipes with friends and family via social-media