Enhanced land use/cover classification of heterogeneous tropical landscapes using support vector machines and textural homogeneity

► We used Landsat imagery to classify land use/cover over a large, highly heterogeneous tropical area. ► SVM classifiers outperformed other parametric, non-parametric, and hybrid classifiers. ► Textural homogeneity led to the greatest improvements in classification. ► SVM classifiers maximized the usefulness of textural homogeneity and attained overall, producer's, and user's accuracies of ∼90%. ► Our classification approach seems very well suited to accurately map land use/cover of heterogeneous landscapes.