Research on automatic extraction technique of surface water based on remote sensing feature index
基于遥感特征指数的地表水体自动提取技术研究

Aimed at extracting surface water information from the mass remote sensing data effectively, and improving the efficiency of automatic extraction, a new method is presented in this paper by producing a curve based on Normalized Difference Vegetation Index (NDVI), Normalized Difference Build-up Index (NDBI), and Modified Normalized Difference Water Index (MNDWI). According to the analysis, a rule is found, that is, the DN value of surface water increases monotonically in the curve, the DN value of vegetation decreases monotonically, but other typical objects, like soils and vegetation, do not have the character. Based on this rule, an automatic extraction model is built by the ERDAS IMAGINE software. Compared with the existing methods, the precision of the model is better than those of others. At last, the surface water is automatically classified by the decision tree model under the ARCGIS software context.