CropID Prototype

The CropID system will classify horticultural crops using a machine learning approach integrating; multi-spectral satellite imagery, synthetic aperture radar data, soil properties, physical field characteristics. Image processing algorithms will be used to segment crops based on spectral reflectance, colour and texture. Satellite images acquired through the year will allow the system to build knowledge about individual fields. Soil properties will be used to focus the analysis on areas suitable for specific crops. The Launchpad grant will support the development of a working prototype which will demonstrate the capabilities of the system based on the requirements of a potential customer. Future developments could provide crop identification services for land use monitoring at a national scale. Other services could also be developed to provide crop health and yield forecasts to farmers.