The application of selective principal components analysis (SPCA) to a thematic mapper (TM) image for the recognition of geomorphologic features configuration

Abstract

Selective principal components analysis (SPCA) has been applied to highly- and/or little-correlated subgroups of bands. Its usefulness was demonstrated in two ways. First, the final result is a false colour composition based on the first order principal component of each highly correlated subgroup of bands, the resulting image containing more than 95 per cent of the total variance of the six TM bands used. Secondly, the second order principal component of pairs of little-correlated bands will show the information that is unique for each band. Both types of analysis have been applied to characterize the geomorphological units at a site in SW Spain. Both methodologies have demonstrated to be very useful in a difficult to access area, with high vegetation diversity covering quite different geomorphic features.