SEMI-SUPERVISED REMOTE SENSING IMAGE CLASSIFICATION METHODS ASSESSMENT

Supervised and unsupervised learning are two well disseminated and discussed paradigms which define how image classification techniques extract knowledge about the data. A recent learning paradigm, called semi-supervised, comes to solve some limitations of supervised learning, as the amount of information needed to conduce an appropriated learning process. Different models of semi-supervised learning have been proposed in literature, which ones basically explore statistical or clustering data proprieties. This work presents a simulation study on the performance of some semi-supervised learning models, applied in image classification methods.