GAINESVILLE, FL (UF/IFAS) — Pricing fruit comes down to a bit of art and some science for farmers. But new technology from the University of Florida may remove some of that guesswork.
“Growers often rely on their gut to estimate a price for their crop,” said Daniel Lee, a professor of agricultural and biological engineering at the UF Institute of Food and Agricultural Sciences.
Lee led new research to develop artificial intelligence that will help farmers obtain accurate strawberry flower counts through images of the plants.
Like many farmers, strawberry growers must negotiate prices with buyers, such as grocery store chains, three to five weeks before they harvest the fruit. But it takes three to five weeks for a strawberry plant to go from growing flowers to producing fruit, UF/IFAS researchers say.
The delay from flowering to producing fruit causes farmers to guess on crop prices, said Natalia Peres, a plant pathology professor at the UF/IFAS Gulf Coast Research and Education Center in Balm, Florida. To more precisely estimate how much they will charge for their harvest, growers need to know how many flowers are on each strawberry plant so they can tell how much fruit will grow from those plants.
Soon, UF/IFAS researchers believe, growers will be able to use a new technique to accurately count the strawberries’ flowers.
That method, which scientists call an “artificial intelligent strawberry flower recognition and visual representation system,” is essentially a camera that takes images of the strawberry plant.
For the new research, UF/IFAS scientists built the imaging system at the UF Gainesville campus, then used it to take pictures of strawberry plants at the UF/IFAS Gulf Coast Research and Education Center and the UF/IFAS Plant Science Research and Education Unit in Citra, Florida.
Scientists then created flower distribution maps, using artificial intelligence and compared counts from images from the system to manual counts in the field and manual counts from the images. They discovered that counts from the images made by the system matched the manual counts from images. Growers can use those flower counts to predict their yield so they can give a better cost estimate to retailers, UF/IFAS researchers said.
As the next step, Lee sees UF/IFAS researchers completing development of an automated flower-counting system and modeling yield prediction according to weather conditions, in particular temperature. That is because it can take longer for fruit to ripen when temperatures are below 60 degrees than when temperatures are above that.
“Growers will be able use the results to market their fruit with more precision,” Lee said.