A MODEL FOR AUTOMATIC EXTRACTION OF WATER
INFORMATION FORM THE CITY ZONE BASED ON SPOT-5 IMAGE
CAO Kai 1,3, JIANG Nan 3, LI Xin-guo 2,3,LV Heng 3
1.Department of Urban and Resource, Nanjing University, Nanjing 210093, China; 2.Graduate School of The Chinese Academy of Sciences, Beijing 100039, China; 3.Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China
With the main city zone as the research area, the authors studied the method for extracting water information in SPOT-5 image. The water and shadow information was obtained from the image by setting the division line in band SWIR. By using spectral, spatial and temporal characteristics, such as the index of shape, the decision tree model for automatic extraction of water information in city zones from SPOT-5 image can be set up , so as to get the water information from the area. For estimating the precision of the model, a comparison was made between the model and the supervised classification method in the whole area and in a special zone with much building shadow. The results show that in the whole area the precision of the model is about 2.5% higher than that of the supervised classification method, and in the special zone the precision improvement reaches 11.6%. Besides, the model has good transplanting function, with only slight differences in the division line in different cases.
曹凯, 江南, 李新国, 吕恒. 基于SPOT-5图像的城市水体自动提取模型研究[J]. 国土资源遥感, 2005, 17(4): 24-27.
CAO Kai, JIANG Nan, LI Xin-Guo, LV Heng. A MODEL FOR AUTOMATIC EXTRACTION OF WATER
INFORMATION FORM THE CITY ZONE BASED ON SPOT-5 IMAGE. REMOTE SENSING FOR LAND & RESOURCES, 2005, 17(4): 24-27.