Using OpenCV to detect product availability on a shelf
Guys,
I’m going to create a product/service that will enable checking of a product’s availability in shelves, via the comparison of pictures taken by a fixed camera and the corresponding planogram. I'm looking for the appropiate tecnologies right now.
My intention is to create a product/service that will provide the necessary information that a given product is not available or doesn’t comply with its programmed position in the store planogram. An in-store, fixed to the ceiling or top of shelf, cell phone/camera that will send pictures every 5 minutes to a central system, which in turn will process them, compare them to the planogram for that shelf, and then take action (email, SMS or call the appropriate people ) to re-stock the product or correct the planogram.
Is this something that OpenCV can be used for? What are your recommendations?
I was looking at a paper "“How to Use SIFT Vectors to Analyze an Image with Database Templates”", and hopefuly find an SDK for what they are doing.
Your prompt recommendations are very much appreciated.
Best regards,
Hans
This can certainly be done with openCV. I think looking into camera's like the industrial variants of the kinect could be quite interesting. However, since the camera's are actually fixed, this can almost certainly be done with simple frame differencing, checking if there is a change on the shelf over time. Would try to focus on simple techniques first before making it complicated.
This is however (like seen in the response beneath) a problem solved by many research groups which could actually help you out there simply by contract. If you interested in that, let me know, I can also provide you with some connections to people doing stuff like this.
The previous answer was deleted as self-advertising, contrary to forum rules