Researchers at NC State University are developing a tool based on artificial intelligence and computer vision to help blueberry growers estimate yields and monitor fruit ripeness directly in the field.
The project, led by researcher Jing Zhang from the Translational Plant Phenomics Lab, uses a mobile application that allows users to photograph blueberry plants and automatically obtain estimates of fruit quantity and ripeness percentage.
The technology works through a computer vision model trained with thousands of labeled images, capable of distinguishing ripe and unripe fruit while also identifying individual berries on each plant.
To validate the system, researchers and extension agents collected images from 10 commercial blueberry farms in North Carolina using smartphones and portable cameras. They later compared the automated results with manual field counts.
According to the research team, the tool aims to support harvest planning and labor management, considering that blueberries require multiple picking rounds due to uneven ripening.
The system is currently still under development and being expanded to include new varieties, so it is not yet commercially available to growers.
In addition to blueberries, the NC State team is also applying similar technologies to other crops and agricultural challenges, including strawberry diseases.
