A unique event took place in the Kaluga region from May 8th till May 11th — the first automatic system in the country for early diagnostics of apple tree diseases was tested.
Apple trees get sick – unfortunately, this is a widespread phenomenon. Moreover, when the disease can be determined, its consequences are already irreversible — the appearance of apples is significantly damaged, and it is problematic to sell them. And so that diseases do not spoil the fruits, apple trees are treated with special means. And not only sick trees — it is impossible to distinguish them from healthy ones at an early stage. All this, of course, affects the environmental friendliness of apples.
The company “Soyuzsnab”, which owns extensive apple orchards, decided to cope with this problem. At her request, Tsuru Robotics, Skolkovo resident, in collaboration with the Skoltech research group, developed a system that would recognize the disease at an early stage.
Scientists took an apple tree, infected it with a scab (this is the most common disease of apple trees), and studied it on a spectrometer — in what range of the spectrum the first signs of the disease are visible. And this range was found.
Then it was time for the field, or rather garden tests. The company “Microgarden” (a scientific-technical division of “Soyuzsnab” and the resident of agro-industrial Skolkovo cluster) with the financial support of Foundation For Assistance To Small Innovative Enterprises held a test in one of the gardens in Kaluga region. The date was chosen because apple trees diseases begin to appear during the flowering period — in May is such a period in the central region of Russia. The system developed by Tsuru Robotics uses three cameras — a conventional one, a thermal imager, and a special camera that takes pictures in five ranges.
The structure is mounted on a tractor that moves along the trees and shoots.
The first experiment was successful. The system tested about 30 rows of apple trees of various varieties.
Now the collected data is tested by computer algorithms for recognizing the disease. And only sick trees will need to be treated with special tools.
In the future, by the way, the system will be able to recognize whether the tree is cured or needs to be re-processed.