Sign up for our daily Newsletter and stay up to date with all the latest news!

Subscribe I am already a subscriber

You are using software which is blocking our advertisements (adblocker).

As we provide the news for free, we are relying on revenues from our banners. So please disable your adblocker and reload the page to continue using this site.

Click here for a guide on disabling your adblocker.

Sign up for our daily Newsletter and stay up to date with all the latest news!

Subscribe I am already a subscriber
Ron van Dooren, Beltech:

Smart cameras predict future development of cuttings

Smart cameras see things people don't see, or miss. If this concerns any specific wavelengths of the electromagnetic spectrum not visible to the human eye, we speak of multi-spectral technique. And we looked for a guy to tell you all about it, meet Ron van Dooren from Beltech, a Dutch company specialized in a variety of industrial applications of Vision technology. As it happened, Ron is now rolling out a project in a greenhouse, at a young plant grower eager to improve the quality of the cuttings it produces.

Wide spectrum
"Multi-spectral cameras are able to register light both visible and invisible to the human eye, ranging from ultraviolet on one side to infrared on the other", Ron explains. "This offers many interesting possibilities. Perhaps most famously, one can think of cameras depicting warmth, making it possible to see whether or not a house is isolated well. The same technique is used in the fresh industry, for example to see through a cucumber and detect possible internal damages, which do not yet show on the surface."

Applied to cuttings
A Dutch plant grower, however, believes Vision might also improve the quality of the cutting material he produces. "We now investigate to what extent quality of the plant to be can be predicted", Ron says. "Ideally this techniques offers a possibility to separate strong from lesser quality in an early stage. We are now busy developing the physical set-up."

Once installed testing can start. This is expected to take several weeks, as predictions will only show later, once the cutting actually develops. "Three questions are key for us, to have the camera work properly: it should be able to distinguish which one is good, which one is not good, and where good borders bad. There the software we built making machine self learning possible should be optimized."

"One of the challenges is specifying the exact wavelengths that will tell the future. This is complicated, because the predictive value of specific wavelengths might differ by plant. Once these two requirements are met, I'm sure we'll build a machine that will repay itself in no time."

Ron: "Development of different types of harvest robots (on the photo an artist impression of the Harvester) are made possible only by multi-spectral technology.

Technology will help
In the future of greenhouse production, technology won't outsmart human intelligence, Ron thinks. However, efficiency will grow, waste will be minimized, and technology will definitely help to do so. "As long as these are the "reasons motivating us, we are on the right track. As the world population grows and environmental requirements get ever more strict, improving technology is the way to tackle these issues."

For more information:
Ron van Dooren
Beltech B.V.
The Netherlands
+31 (0)40 257 1929
[email protected]

Publication date: