Based on image-based functionalities of the OAK-D along with standard Deep Learning techniques such as object detection, semantic segmentation, etc. We will perform food quality classification (i.e. classifying vegetables based on their quality state). Keeping track of supermarkets (or stores) stocked food’s state can help reduce the amount of wasted food and keep the freshness of the existing goods.

The end-user solution will be an embeeded system (OAK-D Camera + Raspberry Pi) Connected to the internet to deliver a realtime data of the system (Store baskets) to servers that will treat and store date to a database cluster and expose an APIs endpoint to supply dashboards for the end-users interfaces