14-6 Digital Analysis System to Evaluate Peanut Maturity: Predicting Yield and Grade.

See more from this Division: Students of Agronomy, Soils and Environmental Sciences (SASES)
See more from this Session: Symposium--National Student Research Symposium Oral Contest: Session 2
Sunday, October 21, 2012: 5:20 PM
Duke Energy Convention Center, Room 208, Level 2
Share |

Blaire Colvin1, Diane Rowland1, Wilson Faircloth2 and Jason Ferrell1, (1)Agronomy, University of Florida, Gainesville, FL
(2)USDA-ARS, NPRL, Dawson, GA
The current accepted method for determining peanut maturity and harvest prediction is the color class method developed by Williams and Drexler in 1981. This has proven to provide a relative measure of maturity but has challenges and limitations. It requires an individual to remove the exocarp of the hull and visually categorize the mesocarp color.  This results in large variability in color classification from person to person and can result in significant error in the final harvest prediction. It also requires a degree of expertise and experience in using the method  and is very time consuming. Finally, the method was based on the color development in the Florunner cultivar which is no longer commercially produced, and there is concern that color development in more recent cultivars may differ.  Due to these issues there is a need for an objective color sorting method that can be tailored to varying peanut cultivars. We have developed a digital method to acquire and analyze pod images for maturity and harvest predictions. Replicated plots were established in Florida in 2010 and 2011 with serial harvests of Georgia Green and GA06G cultivars. At harvest, yield and grade were evaluated and pod samples were collected from each plot for color classification using the traditional visual method. These same samples were also imaged and analyzed with color classification software. Tests were run to determine the optimum color definition and mathematical ratio of the resulting pixel analysis for pod color classification.  Samples from all plots were digitally analyzed and the resulting data was tested to determine if it could accurately categorize pod color.  The ultimate goal of this work is to provide an objective method to predict peanut maturity that can be easily utilized by growers, extension agents, and consultants.
See more from this Division: Students of Agronomy, Soils and Environmental Sciences (SASES)
See more from this Session: Symposium--National Student Research Symposium Oral Contest: Session 2