Russia to Train Neural Network to Monitor Wheat and Cabbage Health
Researchers are building an image dataset that will teach AI to recognize crop diseases and threats using drone and field photos.
Scientists at the Buryat State Agricultural Academy have begun developing a comprehensive dataset on the growth stages of spring wheat and white cabbage to enable automated crop monitoring powered by artificial intelligence.
Using drones and on-the-ground field surveys, specialists—together with experts from the Russian Agricultural Center—are capturing thousands of high-resolution images from active farmland. Students then label each photo by category, and experienced agronomists verify the annotations to ensure up to 95% accuracy. The resulting datasets will train a neural network to identify crop health issues and potential threats directly from images.
Field Monitoring and Crop Protection
Once trained, the algorithms will be integrated into Agrika, a national digital agriculture platform that lets farmers monitor fields in real time, automatically detect early signs of disease, and receive AI-generated recommendations for plant protection and yield preservation.
The team aims to complete a dataset of at least 5,000 images by 2030, covering all major growth phases of the plants. The project could become a cornerstone for precision agriculture in Russia, helping farmers boost efficiency, reduce losses, and make data-driven decisions about crop care.
