Efficiency assessment of the of Support Vector Machines for land use classification using Landsat ETM+ data (Case study: Ilam Dam Catchment)

Land use classification using remotely sensed images is one of the most common applications in remote sensing, and many algorithms have been developed and applied for this purpose in the literature. This study investigates the efficiency of Support Vector Machines algorithms in image classification. Support Vector Machines (SVMs) are a group of supervised classification algorithms of machine learning that have been used in the remote sensing filed. The classification accuracy produced by SVMs may show variation depending on the choice of the kernel function. In this study, SVMs were used for land use classification of Ilam dam catchment using Land sat ETM+ data. The classification using SVM method was implemented automatically by using four kernel types, linear, polynomial, radial basis, sigmoid and the results were analyzed thoroughly. Results showed that SVMs, especially with use of radial, polynomial and linear function kernels, outperform the maximum likelihood classifier in terms of overall (about 10%) and kappa coefficient(about 15%) accuracies. So, this study verifies the efficiency and capability of SVMs in classification of remote sensed images.