The new SeedLab, manufactured by spectral imaging experts Videometer, is part of state-of-the-art equipment available at CHAP’s Digital Phenotyping Laboratory at partner Rothamsted Research.
The robot uses advanced spectral imaging technologies and machine learning algorithms to automatically inspect and sort seed, based on physical and chemical features, working with varying sizes from oilseed rape to maize.
Kristian Ryder Thomsen, Vision Developer of Videometer said: “As the SeedLab is completely automated, it makes seed analysis incredibly easy and efficient. For small seeds such as oilseed rape, this means analysing up to 2500 per minute and for larger such as maize, up to 200 per minute, depending on the complexity of the analysis.
“The automation includes image capture and an end of measurement summary report, making this a powerful tool for those wishing to enhance their understanding of seed or grain quality.”
Analysis begins with a vibration unit that distributes the seeds evenly onto a moving belt. After imaging and grading, a pick and place robotic arm then automatically separates different grains into vessels such as cups or petri-dishes.
The process is customisable for variables such as picking speed, collection vessel, classification model, and illumination.
Research Assistant for CHAP, Faye McDiarmid, said: “The SeedLab is a valuable addition to the imaging technology CHAP has available at the Digital Phenotyping Lab; offering an integrated analysis and sorting capability for both seed and grain.
“We’re excited to be able to offer this as part of collaborative projects, both grant-funded and commercial.”
For information about CHAP’s new SeedLab, e-mail enquiries@chap-solutions.co.uk
