Pre-harvest yield estimates are essential for most agricultural businesses, as they help them organize and manage logistics when it comes to transport, supplies, labor, etc. Overestimating yields could lead to higher costs, while the opposite could mean a potential waste of produce. This is only one of the many aspects that makes precision agriculture an essential tool to improve efficiency.
Kiwifruit is a leading product in many countries such as Italy, China and New Zealand. However, there is still not a lot of information regarding IT systems to estimate kiwifruit yields. Researchers from the University of Trento, in collaboration with Metacortex S.r.l. in Torcegno (TN) created a device that enables the estimate of kiwifruit productions.
"We developed an artificial vision system that is completely automatic and non-invasive to estimate kiwifruit yields in a specific orchard. The system is made up mainly of an optical sensor mounted on a mini-tractor that moves within the orchard at low speed. The images acquired are processed and the results obtained help make an estimate," explain the researchers.
The system was tested in two large orchards and results showed an error margin between 6 and 15% of the system, which has already been in commercial use for the past 2 years.
Compared with the traditional manual estimates, it has been proven how this system helps save considerable time and reduce estimate errors, especially when talking about large scale agriculture.
Source: Mekhalfi Mohamed L., Nicolò Carlo, Ianniello Ivan, Calamita Federico, Goller Rino, Barazzuol Maurizio, Melgani Farid, 'Vision System for Automatic On-Tree Kiwifruit Counting and Yield Estimation', 2020, Sensors, Vol. 20 (15), 4214.