Alexander Kozhevin

Food Additive Scanner - Take a photo of food components and learn more about it

Make a photos of food components and get a full list of
additives with description and recommendation.
All operations are done on device. NO cloud is used.

Add a comment

Replies

Best
Alexander Kozhevin
I just decided to make something useful for myself.
MacBriC
@alex_polymath Very nice! What database do you use for descriptions and recommendations?
Alexander Kozhevin
@macbric Actually there is no clear database. We used this table to Australian government https://www.foodstandards.gov.au... And some info from wikipedia.
Alexander Kozhevin
@macbric Everything is stored on device. NO cloud
Beenish Vafaqani
@alex_polymath @macbric Yeah you can do this because I've just read that technology transforming the Health care industry through many ways like Artificial Intelligence. https://www.mmcdubai.ae/blog/how...
Boris Dedejski
Great work !
Beenish Vafaqani
Great
Alexander Kozhevin
@vafaqani Thank you :)
Ben Patton
Great idea. Couldn't you do it using the barcode?
Alexander Kozhevin
@ben_patton1 Nope, not really. We just make OCR and search for E codes based on numbers.
Nicolas Roulis
I'm curious about your app because some powerful app like Yuka or Foodvisor do the same and more, but maybe it's only available in my country (France), if not I don't get your strengh :/
Alexander Kozhevin
@drdou well, my app is not unique. And it's worldwide
MacBriC
@drdou Yuka does not have an OCR of ingredient list. You have to scan the barcode and it only works if they have the barcode in their database, so it is limited.
Alexander Kozhevin
@drdou @macbric We don't use barcode. We make OCR of additives CODES 202, 1451, etc
Nicolas Roulis
@macbric Yes, I see. The ingredients terms doesn't change like a barcode. So, once you get them in every language, you get technically every product in the world. The point is, the list needs to be really clean compared to a simple barcode, this is the only "con"
Anil Murty
Great work. what do you use for the OCR?