Comparing Digital Image Analysis and Visual Rating of Gamma Ray Induced Perennial Rye Grass (Lolium perenne) Mutants

Abstract

To generate variability in perennial rye grass and to select genotypes responsive to low management, gamma-ray irradiation was used for induction of dwarfness and other quality attributes. The main objective of this study was to identify changes and correlations among turf visual rating and digital image analysis in evaluating turf grass quality under different treatments. Differences were significant among irradiated population with respect to hue angle, brightness and saturation. The correlations of hue and DGCI were significantly positive with all the parameters of visual rating. There were non-significant correlation of brightness with quality and texture, and saturation and texture. The DGCI values were in tune with each of these parameters when the slope of regression line was significantly different from zero (p<0.05). These relationships were better in DGCI and hue (r2 = 0.3531) DGCI and saturation (r2 = 0.3017); DGCI and brightness (r2 = 0.1196) and DGCI and colour (r2 = 0.1725). Non-linear relationship was noticed between DGCI and quality (r2 = 0.0004).