High quality data on seeds and seedlings are of interest in plant breeding, seed production, seed testing labs, for seed treatments, plant variety offices, plant cultivar protection, in gene banks, and research.
Seed and seedling quality assessments require numerical data on seeds and seedlings. These include seed and seedling counts and dimensions, for instance length of emerging roots and shoots. Beyond counting and measuring, detailed assessments determine quality features, e.g. whether seedling germination is normal, or whether seed batches contain foreign seeds.
Digital imaging and advanced image processing deliver well-documented reliable data that are repeatable and can be standardized. Machine learning facilitates training the algorithms to quality criteria on seedling normality, seed purity, and many more user-defined features. The human factor that introduces bias whenever a rating is done by visual inspection is minimized, because all results are gained independently from personal impressions.