Germination analysis is key to maximizing crop productivity.

Seed germination and seedling establishment is a crucial process that influences the quality and yield of a crop. Seeds that germinate effectively and uniformly, and according to an expected germination rate, are likely to yield better crops. Therefore, tests monitor the germination progress to enable an assessment of the seed quality. Currently, crop breeders and agricultural laboratories conduct and record these tests manually – a time consuming and error-prone procedure. SeedGerm is a new, AI-enabled process that tests seed samples in a more efficient and low-cost manner. (Mirage)

Seed testing

Germination data is useful for seed certification, developing guidance on sowing density and insurance models. Seed vigor, describing rates and uniformity of seed emergence, is a vital input for actions on faster canopy closure, weed suppression, and crop yields.

Measurement of seeds’ germination and vigor is a painstaking and manual process requiring inspection by qualified staff or external agencies.

The Earlham Institute, the John Innes Centre, Syngenta (NYSE: SYT), and NIAB collaborated on the development of SeedGerm. It is an easy-to-use, low-cost, and scalable solution to this problem.

How it works

SeedGerm uses software for machine-learning-based image analysis in combination with hardware for automated imaging.

The solution targets the assessment of various traits of seeds, especially germination and vigor.

Cameras take photographs of the germination process inside a special cabinet. Images from these cabinets document the development of the seed right from the stage of the soaking water, the emergence of its root, and changes thereafter.

The machine learning software receives training to assess how the germination is progressing.

Josh Colmer, a Ph.D. student in the Anthony Hall Group at EI and first author on the paper, said: “To apply machine learning so effectively to test seed germination marks an exciting step forward, especially as the learnings from this project can inform a variety of image-based analyses with wide-ranging applications in crop research.”

Based on the current progress, experts from Syngenta are confident that SeedGerm can successfully measure the health of seedlings and their germination rate for major crop species such as tomato and rapeseed.

SeedGerm may ultimately replace the current practice of manually recording seed germination, and provide invaluable insights for seed certification, seed insurance, and sowing practices.

“The developments and learnings from SeedGerm are truly a big step forward in automation and generating high quality and reliable data in scoring seed germination,” said Dr. Rene Benjamins, Senior Scientist at Syngenta Seeds. “This will help seed companies like Syngenta in providing the best quality to their customers.”

Related Story: Ant Financial to Develop Blockchain-Based Agro-Monitoring Product With Bayer

Image Credit: Flickr