Seed purity is an important feature of seed batches – unwanted seeds contaminating crop seeds lead to inhomogeneous harvests, or even endanger the cultivation success. While seeds from other cultivars or crop species already reduce the quality of the harvest, seeds from weeds, invasive species, or toxic plants can convey serious dangers to the whole cultivation.
Employing advanced imaging technologies and machine-learning algorithms, we at LemnaTec provide tools to discriminate between seed types.
The automated image processing detects seeds in images and assigns them to species or cultivars. Using state-of-the-art machine learning algorithms, the systems can be trained to any seed type, thus allowing to discriminate and classify individual phenotypic features in homogenous seed batches. The tools allow to discriminate between cultivars of one species, differentiate between species of one genus, or identify contaminations through weed or foreign seed.
Our toolbox is ready to cope with your seed-related task – the SeedAIxpert systems provide manual or automatic solutions.
Example image of seed purity analysis and seed phenotyping: detecting non-seed objects (encircled black) among seeds (encircled red), together with measuring length, width, and area of all seeds.