Principal component analysis applied to remote sensing

Abstract

The main objective of this article was to show an application of principal component analysis (PCA) which is used in two science degrees. Particularly, PCA analysis was used to obtain information of the land cover from satellite images. Three Landsat images were selected from two areas which were located in the municipalities of Gandia and Vallat, both in the Valencia province (Spain). In the first study area, just one Landsat image of the 2005 year was used. In the second study area, two Landsat images were used taken in the 1994 and 2000 years to analyse the most significant changes in the land cover. According to the results, the second principal component of the Gandia area image allowed detecting the presence of vegetation. The same component in the Vallat area allowed detecting a forestry area affected by a forest fire. Consequently in this study we confirmed the feasibility of using PCA in remote sensing to extract land use information.

Keywords

• PCA
• Land use
• Remote sensing
• Landsat