A method for classifying agricultural crops includes using satellite data (optical and radar), a step of pre-processing the satellite data, using a classification algorithm(s) to build classification maps, and a step of validating the resulting classification maps. In situ data collected by ground surveys are used to form training, test and validation samples. At the stage of building land cover maps (by applying the classification algorithm to the combined satellite data), all available satellite data channels are used, raster land cover maps are improved by using vector polygon boundaries.
Comments are closed