Smartphone app might help when picking peanuts

Georgia’s peanut crop generates greater than $600 million yearly, however figuring out the optimum harvest time for the crop may be tough. Present crop evaluation instruments are time consuming and susceptible to human error, resulting in hundreds of thousands of {dollars} in misplaced yield annually.

Researchers within the College of Georgia School of Engineering are working to assist farmers enhance the standard and amount of their yield by offering them with sooner and extra accurate data. In a brand new research printed in Springer Precision Agriculture, the UGA group presents a transportable, smartphone-based system for peanut harvest evaluation.

One of many main strategies now used to find out peanut maturity is a peanut profile board, an eye fixed check primarily based on the shut relationship between the pod’s coloration and its maturity. This technique requires farmers to dig up samples from their area for evaluation by a county agent. The harvest day for the peanuts is estimated primarily based on the share of black peanuts in a pattern.

From days to minutes

“As peanut fields are typically broadly scattered, the gathering, delivery and evaluation can take days to finish,” stated Rui Li, the research’s lead writer and a 2020 graduate of UGA with a Ph.D. in engineering. “In distinction, it takes three to 5 minutes to acquire a printed report from our system.”

To deal with the drawbacks of present evaluation strategies, the UGA researchers built-in smartphone expertise with a transportable picture station. The semi-automated system permits farmers to gather, retailer and entry harvest-related knowledge at their very own farm and wirelessly ship the data to a county agent for additional analysis.

To make use of the system, a farmer locations samples on a board designed to carry peanuts in particular person slots and slides the board into the underside of the picture station. Inserting a smartphone on the prime the picture station, the person snaps an image of the crop pattern. From there, the group’s smartphone software analyzes the colour of the peanuts and feeds the data to a web-based database. The applying permits the person to customise their report by inputting data akin to the sector location, the kind of peanut and the illness situation.

In accordance with the researchers, this faster evaluation instrument will permit farmers to repeatedly monitor the maturity of their crop and account for components akin to soil kind and climate.

“Beneath-matured or over-matured peanuts can have a excessive degree of aflatoxin, a toxin produced by fungi discovered on some agricultural crops,” stated Li. “If a farmer can repeatedly observe the extent of peanut maturity, they’ll be capable of begin the harvesting course of in an acceptable time window and cut back the yield loss attributable to aflatoxin.”

The researchers carried out 52 on-site assessments to validate the system’s accuracy. In contrast with human graders, the system demonstrated error charges of 9 % in calculating the share of black-and-brown peanuts; 18.8 % in calculating the share of black peanuts; and 5.8 % in calculating the entire variety of peanuts in a pattern. Regardless of these discrepancies, the researchers say the proof-of-concept system demonstrated the potential to offer quick and dependable outcomes to peanut growers.

Enhancements to return

“The system may be improved in two methods,” stated Li. “One is to additional optimize the light conditions for constant picture high quality, and the opposite is to enhance the colour detection perform utilizing a data-driven algorithm.”

As soon as a enough quantity of human grading knowledge is recorded within the database, the group believes machine studying algorithms may additional enhance the system’s accuracy. It might be attainable, they stated, to develop a predictive mannequin for peanut harvest primarily based on the information.

The venture, supported by Nationwide Science Basis ICORP and Venturewell grants in addition to the Georgia Peanut Fee, has generated two pending utility patents.

The buyer wins

“The event of this technique shouldn’t be solely a big time saver for peanut farmers, it should enhance our means to find out maturity which is straight associated to peanut taste for the buyer,” stated Don Koehler, government director of the Georgia Peanut Fee.

The venture, supported by Nationwide Science Basis ICORP and Venturewell grants, has generated two pending utility patents.

“The farmers (collaborating within the area assessments) have been strongly eager about the advantages of the smartphone software. They just like the portability and low value of the system,” stated Li.