Comparison of two classification methods for vegetation mapping in Palau.

摘要

The goal of this thesis is to test and develop a classification method using IKONOS imagery to produce a vegetation map for the tropical islands of Palau. Two methods of classification will be tested and compared---the use of per-pixel based classifications and segmentation-based classification. Existing research on this subject reveals that there is no current consensus on the best method of classification for tropical Pacific vegetation. The two different classification methods were analyzed to determine the suitability of IKONOS imagery to map diverse tropical vegetation on the islands of Palau.