Early detection of wheat germination using image analysis with a preliminary exploration of its onset prediction based on imbibition development

Research on imbibition and germination is currently a topic of great interest among researchers.One cost-effective yet efficient approach to monitoring imbibition and germination is the use of image analysis.In this study, a total of 404 winter wheat grains (Triticum aestivum L.) of four different varieties were examined.Image data were captured using a high-resolution camera controlled by a computer.The grains were analysed for their size and shape described by the image moment (η 2,0 ).Fourteen parameters were proposed, which were subsequently used to characterise the development of the grains during imbibition and germination.The progression of the image moment indicated that wheat grain imbibition and germination could easily be divided into three distinct phases: in the first, hydration occurs predominantly in the embryo, followed by the endosperm in the second.The third phase consists of the germination itself.The minimum of the image moment course marks the onset of germination, occurring at a stage when radicle development is still imperceptible to human evaluators.Germination identified in this manner can thus be partially predicted from the aforementioned parameters.The linear model yielded a coefficient of determination of 0.5017.