Automated Modelling Of Multimodal Data Processes In Remote Sensing

Automated monitoring of bio-geophysical phenomena, especially those occurring in large areas, requires the use of models obtained from remote sensing data. The interaction of multiple components in the optical data flow and the non-ergodicity of the acquisition process can seriously affect the precision of the models. In order to effectively deal with this situation, we are proposing an iterative semi-supervised learning framework that combines regression analysis leading to the final set of models with an iterative classification process. based on support vector machines (SVM) that generates data sets associated with each statistical modality. This paper presents an application of the proposed method in modeling the concentration of Water pollutants, particularly chlorophyll-a, in inland waters using multimodal satellite data sets. (C) 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.