AI-basert tidlig deteksjon av plantesykdommer ved bruk av dronebilder

Offentlig prosjektsammendrag

The project aims to develop a technology for early detection of crop diseases and anomalies using drone imagery and machine learning. In current farming practices, diseases are often detected too late, leading to yield loss and inefficient use of inputs such as crop protection chemicals. The main challenge is to identify early-stage disease signals that are difficult to distinguish from normal variations in crops caused by environmental conditions such as lighting, soil and growth stages. In addition, there is limited availability of high-quality labeled data. The project will address these challenges through systematic data collection from pilot fields, preprocessing and structuring of drone imagery, and development and testing of machine learning models. The solution will be iteratively improved through validation in real field conditions.