Elizabeth Kurcinka, University of Minnesota Agronomy & Plant Genetics, Agronomy and Plant Genetics, 1991 Buford Circle, St Paul, MN 55108
Color has proven to be a useful measurement of aesthetic quality in turfgrass. Recently, digital image analysis has been utilized to examine turf color by measuring hue, saturation and brightness (HSB) of images. Grass seed production systems may also be able to utilize digital color analysis to predict spring nutrient requirements or seed yield. A study was conducted to determine whether plant nitrogen content or seed yield could be predicted using color image analysis. Digital images were collected on Kentucky bluegrass (Poa pratensis) and perennial ryegrass (Lolium perenne) research plots April through June 2006 in Roseau and Lake of the Woods counties in Northwestern Minnesota. Plots received 67,67+45 (split) 112, 135,112+45 (split) and 157 N kg ha-1 of four nitrogen fertilizer formulations (urea, ESN® polymer coated urea, ammonium nitrate and ammonium sulfate) applied in the fall and split applications in the early spring. Images and plant samples were collected throughout the growing season. Images were analyzed using Adobe® Photoshop® to determine average levels of HSB per image. Nitrogen rates effected plant color and tissue nitrogen content. Tissue color was primarily correlated with tissue nitrogen content and also seed yield. These preliminary results suggest digital image analysis may be a diagnostic tool used to determine spring nutrient levels of grass seed production crops in Northwestern Minnesota.