Supermarket Identify

As a side note, please excuse me if it is not great because I am 12. Thanks a lot.

Inspiration

I was inspired to do this after seeing people struggle to scan each barcode. This led to large queues. It was then I got the idea

What it does

Supermarket Identify takes away the need for any barcodes and barcode scanners. Simply take out your phone, take a picture and send it. We'll do the rest. Using ML it works out what is selected.

How I built it

I started off by using TensorFlow and python combined with the COCO Dataset. To implement taking a picture I used CV2 as well as PIL.

Challenges I ran into

It was hard to code a program that lets chrome use the webcam. This was the hardest bit.

Accomplishments that I'm proud of

This is my first ML project and I have learnt so much.

What I learned

I have learned about how to effectively use TensorFlow and the Model Zoo.

What's next for Supermarket Identify

I think I am going to add some more family-friendly features and make it quicker.

Built With

• cv2
• google
• jupyter-notebook
• pil
• python
• tensorflow