Application of High Resolution Satellite Imagery for Wetlands Cover Classification Using Object-oriented Method

Wetlands are considered an integral part of the global ecosystem, the development of the remote sensing technology makes us obtain very abundant information of wetlands, especially with the appearance of high resolution satellite imagery which extends the visual field of the wetlands. The aim of this study is to explore the viability of applying high resolution satellite imagery for inland limnetic wetlands cover classification using object-oriented method. Thesis discussed the selection of best segmentation scale during segmentation procedure using statistical, trial and error method, determined the separability of each wetlands cover structural class using each texture band through z-test method, and a high accuracy of classification result was acquired at last through test.