Near-infrared drone
agricultural monitoring.
NIRDAM makes precision farming affordable by combining low-cost microclimate stations, drones and machine learning instead of expensive satellite imagery.
NIRDAM (Near Infra-Red Drone Agricultural Monitoring) aims to lower agricultural risk and improve yields with an affordable precision-farming technique. Parameters that used to be available mainly through expensive satellite imagery are monitored with low-cost microclimate stations and unmanned aerial vehicles (UAVs).
Data gathering
Throughout the project, microclimate stations with advanced sensors are deployed on the soil types typical of Vojvodina — including chernozem, cambisol, albeluvisol, terra rossa and fluvisol — providing a continuous stream of ground-truth data for training and validating the models.
Enhancing agriculture with UAV
Drones scan the fields with near-infrared spectroscopy, and software processes the images to estimate soil properties across the whole flyover area. Machine learning is used to build NIR-image-based estimators for each soil type, with both station-dependent and station-independent approaches.
Decision support
Focusing on leaf wetness, solar radiation and pH, the project develops a machine-learning decision-support system that predicts the need for irrigation and fertilisation and assesses plant health.
About Panonit
Panonit is a software engineering company from Novi Sad, Serbia, building software, hardware and products for legal, healthcare, finance and enterprise clients worldwide — and its own technology through in-house R&D.