Muslims living abroad in non-muslim countries , or maybe just spending their vacations , stuggle to find Halal food .. that's why we thought of Halal Food app.
What it does
It's a mobile app that tells you wether the product you are holding between your hands is halal or not . All you need to do is take a picture of the back of the product , where the list of ingredients is , our program will turn the ingredients in The picture into text (translate it if needed) and then compare it to a list of haram ingredients already in set in a database .
How we built it
We used Deep Learning Based OCR for Text to extract text from the picture . NLP to translate the text if needed . A database of haram ingredients .
Challenges we ran into
The dataset of halal or haram ingredients doesn't exist , this is why we have to create it .
Accomplishments that we're proud of
Make a clear roadmap towards the implementation of the model
What we learned
We learned more about OCR , NLP , classifications , RAM management
What's next for Halal Food
Hahal food can be developy into a web application , thst is linked to a server holding a lzrge data bzse of the official ingredients used in every product , the data will be collected directly from producers ,through audits or through government's institutions . In addition , every picture ever taken of the same produxt will be used to train our model so that in recognise the product directly without repeating the text extraction process over znd over each time . Furthermore , it will use the bar code to identify the product directly taking less time , each time!