Studies On The Dynamic Changes Of Landscape Patterns Of Wetland Based On GIS And RS Techniques

Wetland is a particular kind of ecosystem between water and land. It's one of the most diversified ecological landscape and dependent living environments to human being. Wetland is regarded as "kidney of earth" for its strong function of water and element circulation. As constant improvement of cognitive level, wetland issues gained great attentions for its special importance. The research of wetland landscape pattern has become hot point in the world.In this paper, we chose 1986 and 2000 remote sensing images as resource. Based on RS and GIS data processing, it analyzes landscape pattern and dynamic changes of wetland in these areas by fundamental theories and methods of landscape ecology.As improvements of neural network algorithm, this paper classifies wetland landscape in drainage area of Taoer River. In addition, on the demand of electronic vector maps in the geography information system, the paper succeeded in vectorizing raster images by means of math morphology and theories.By landscape ecology and spatial analysis, the paper ends with research combining with RS and GIS. It optimizes algorithms of polygon's overlay and makes overlap analysis of 1986 and 2000 remote sensing images. Then with information of landscape, it completes wetland mapping. Adopting markov model, it summarizes the laws and features on dynamic changes of landscape patterns of wetland.