Author(s)

C. Thompson, R. Lindsay, F. Hochleitner, B. Teixeira & F.C. Popok

Abstract

Pattern recognition and classification has so far remained a largely unexplored area of peat-land science. The objective of this work is to classify the surface patterns displayed on an aerial photo of Kentra Moss, a bog (mire) in Western Scotland, UK. The computer experiments reported in this research present the attributes of image textures calculated using statistical texture measures, e.g., mean, variance, skewness, kurtosis, and entropy. Insignificant attributes are eliminated by using statistical techniques. On the other hand the clusters are obtained by calculating the Euclidean distances among all exemplars. The importance of this work is the provis

Keywords