b'PARTNER CONTENTOriental mustard seed seen in two different ways. At left, we see it as the human eye sees it. Then, we see it in near infrared (NIR) which allows us to see different features of the seed. In fact, there is a single canola seed mixed in with the mustard. Can you spot the difference?Image Analysis is About to Change the Game in Seed TestingI magine being able to take a picture of a seed,of seed and possibly suggest if there is a disease or upload the photo to a computer and collect aother quality issue. Seed colour can be influenced mountain of data pertaining to colour, size andby environmental conditions, yielding insights into shape. what sort of environmental stress might be affect-Imagine looking at that seed using differenting seed. Then there is identifying living from dead wavelength bands, like Ultraviolet or Near-infrared,seed; a recent Danish study from Aarhus University that allow you to see what the human eye normallyshowed viable seeds could be distinguished from cannot. It is all possible with seed image analysis,dead seeds with 92 per cent accuracy using multi-specifically multispectral imagingwhich we arespectral imaging.formally introducing in 2023. It is going to changeWe can count seeds extremely accurately. the game in seed testing. Here is why: Using imaging supported by an algorithm, it is easy It makes seed testing even more informative.to segment the image and visually separate the By Janek Bartel, SGSThe basics of traditional seed testing around theseeds. You can see how many seeds you have, but Canada Laboratoryworld have not changed significantly in decades.also what kind of relationship the population has Manager Purity testing has essentially been done the sameto each other. This can be important for precision way for a over half a century. Microscopes haveplanting and figuring out seeding rate.changed, but not too much else. Image analysisWe can identify seeds better. To the human represents a complete process change in seedeye, two wheat seeds of different classes do not testing. Why?look all that different. Using imaging, we can see Multispectral analysis is a new technology thatsignificant differencesfor example, they are a can make an analysts work even more informa- completely different colour, something you cannot tive in conjunction with classical testing methods,see using traditional seed analysis methods.which benefits the customer. We have scoured theIt makes testing far more efficient. Where it marketplace for technology that we can implementcurrently requires a separate test to do a physical at SGS Canada. At the same time, our in-housepurity and a different test to do 1,000 kernel weight, team is building our own boutique customer toolsit is conceivable that down the road, those could to take image analysis to new heights.come from the same test using optical weighing Multispectral imaging allows us to see in differ- and image analysis.ent wavelengths. Each wavelength provides specificBeginning in 2023, SGS image analysis will information about the seed surface structure. Whyensure consistent, reliable and traceable results, does this matter? allowing clients to rest assured that the final test We can glean insights into seed quality. Imageresults are some of the most accurate results they analysis offers deeper insights into the seeds mor- could hope to get.phology. This can help an analyst identify speciesFor more info, visit cropscience.sgs.caJANUARY 2023GERMINATION.CA 15'