QUANTITATIVE ESTIMATION OF CORN ENDOSPERM VITREOSITY BY VIDEO IMAGE-ANALYSIS

Vitreosity, or hardness, is an important grain quality factor for com. An increasing amount of research is aimed at understanding the genetic and biochemical basis of vitreosity and at improving vitreosity of high-lysine (opaque-2) com through breeding strategies. Previous methods for quantifying vitreosity were destructive, subjective, or required sophisticated equipment and expertise. This study evaluated a simple video image analysis procedure for quantifying vitreosity and determined how various processing steps affected the results. Kernels were surrounded with modeling clay and viewed on a light box with a monochrome video camera. The video signal was captured to a personal computer and analyzed with commercially available hardware and software. A segregating F2 population from a cross of vitreous Pool 29 QPM X nonvitreous B73o2 was classified visually into 10% steps of vitreosity. High correlations were observed between visual classification and average grayscale values of captured video images at all stages of processing. Grayscale value was inversely proportional to kernel thickness. Removing image background and eliminating a segment corresponding to the embryo area increased the average grayscale range and resolution, but adjustment for kernel thickness did not substantially improve the correlation. This method allows researchers and breeders to quantify vitreosity of corn and other cereal grains on a continuous scale with readily available equipment and expertise and overcomes the problems of subjectivity, destructiveness, and complexity associated with other approaches.