An Ensemble-Based Stacked Sequential Learning Algorithm for Remote Sensing Imagery Classification

摘要: Contextual-based image classification attempts at considering spatial/temporal information during the learning process in order to make smarter. Sequential techniques are one of most used ones perform contextual classification, being based on a two-step process, which traditional noncontextual is followed by more step an extended feature vector. In this paper, we propose two ensemble-based approaches sequential less prone errors, since their effectiveness strongly dependent extension ends up adding wrong predicted label neighborhood samples as new features. The proposed validated context land-cover results considerably better than some state-of-the-art literature.