Crop Classification is a highly relevant topic in agriculture for farmers, authorities and companies selling to farmers alike.
Data on the area of grown crops is not available during the grow season. Data is polled from farmers in the beginning of the year and aggregated on a district level.
Computer Vision algorithms can detect crop types based on phenological criteria in an optical sphere at a later stage. However, analyzing radar data and training the algorithm with historical data increases the accuracy of classification significantly.
Based on the information generated with accurate and timely crop classification, harvest forecasts can be improved significantly, fertilization recommendations can be provided to farmers and on-the-spot controls of EU subsidies can be disposed.