 This research investigated the use of object-based image analysis and advanced machine learning methods for the mapping of nine major summer crops from Aster satellite images. The best performing classifier was a combination of support vector machines and a second layer of support vector machines, which achieved an overall accuracy of 89%. This result was significantly better than the conventional decision tree classifier, which had an overall accuracy of 79%. This article was authored by Jose M. Pena, Pedro A. Gutierrez, Cesar Jervas Martinez, and others.