Developing a hardware and software solution for more accurate potato sorting
We leveraged our unique capabilities in computer vision to create a compact hardware unit and user friendly web interface that accurately identifies weight, count, and grade of spuds being loaded into storage.
Question: Can potato growers leverage computer vision to know precisely what count and grades of potatoes are being stored in their cellars?
Answer: Yes, a single camera attached to a portable hardware box can make accurate counts and weights by grade, and identify spud health for each potato loaded into a processor’s cellar.
Over 45 billion tons of potatoes are stored in climate controlled cellars every year in the US, little knowledge of their quality or grades. Not knowing the exactly supply of each cellar leads to inefficient order fulfillment and waste
Our system leverages a custom potato ai model to accurately identify weights, counts, and grades of spuds and display results in a user friendly web interface. Knowing what’s in the cellar enables efficient order placement and fulfillment, increasing satisfaction of purchasers, profit for growers, and yield for processors.
