Wetland Information Extraction In The Region Of Suangtaizi Estuary Of Liaodong Bay Based On Object-oriented Technology

Suangtaizi Estuary Wetland located in the north of Liaodong Bay， covering the area of1214.72km~2， is one of wetlands with best perserved ecosystem in the world.The wetland inthe area plays an very important role in the researches on wetlands species diversityprotection. The main wetland vegetation types include suaeda grass(suaeda heteroptera) andreeds(Phragmites communis)， forming the famous red beach landscape and the world’slargest reed field. The studies on wetland mapping in this area have been of importance.In this paper，Landsat TM images in the year of2000and2010are chose as the basicdata source.Firstly， the spectral characteristics of different land cover types are analyzedusing some enhancement methods， which provide the theoretical foundation of wetlandinformation extraction.Secondly， the object-oriented classification technology of wetland indentification fromremote sensing images is explored. The multi-stage， multi-scale segmentation method isused to extract wetland types. Based on the detailed analysis of shape， texture， spectralfeatures of land cover types， vegetation index， the feature space for the characteristics ofthe study area is built. Thirdly， after image segmentation， the construction of the featurespace， accurate wetlands are indentified using object-oriented classification method. Theclassification results are validated through the field data. The analysis and testing results showthat the object-oriented classification method meets the demands on the classificationaccuracy. The analysis of wetland change trend provides an auxiliary reference for the regionin the future work of wetlands protection.Due to natural factors and human activities， the natural wetland area decreased greatlyin the past ten years. The natural ecological environment has been destroyed. A large amountof natural wetland are converted into a artificial wetland.With the aquaculture rising in the1990s， the red Suaeda beach is seriously degraded. The growth of population may cause theexpansion of paddy field and construction land.