Crop Classification Using Different Color Spaces and RBF Neural Networks

Agricultural activities could represent an important sector for the economy of certain countries. In order to maintain control of this sector, it is necessary to schedule censuses on a regular basis, which represents an enormous cost. In recent years, different techniques have been proposed with the objective of reducing the cost and improving automation, these cover from Personal Digital Assistants usage to satellite image processing. In this paper, we described a methodology to perform a crop classification task over satellite images based on the Gray Level Co-Occurrence Matrix (GLCM) and Radial Basis Function (RBF) neural network. Furthermore, we study how different color spaces could be applied to analyze satellite images. To test the accuracy of the proposal, we apply the methodology over a region and we present a comparison by evaluating the efficiency using three color spaces and different distance classifiers.