Assessing the Performance of Independent Component Analysis in Remote Sensing Data Processing

Abstract

Independent component analysis (ICA) is a novel method to be considered as a powerful type of analysis in the process of source signal separation. Based on the capabilities of this particular analysis, there will be a hypothesis of applying ICA in the image processing of remote sensing data. This paper aims to introduce the ability of ICA in contrasting and highlighting some area with potential of mineralization. Considering and applying ICA transformation on the ETM+ image of southern Masule, Iran has resulted in finding some favorable points for further investigation. Moreover, sampling program on the indicated area has led to identify some huge, unexpected lithology and dikes. ICA analysis is a robust method even in remote sensing data processing with the high speed and capabilities in separating source signals from noise. © 2012 Indian Society of Remote Sensing.