Study on the Subsiding Land Extraction from Landsat Tm Image Supported by Gis and Domain Knowledge

Zhou Xing-dong,Du Pei-jun,Guo Da-zhi
DOI: https://doi.org/10.1007/s11769-003-0081-9
IF: 3.101
2003-01-01
Chinese Geographical Science
Abstract:The subsiding land can be extracted from Remote Sensing image based on its spectral and spatial features. The features of subsiding land caused by mining, especially its RS information features and relative knowledge are proposed. Three methods can be used to extract subsiding land from RS image. The first is to categorize the region into different blocks (or layers) according to their features and apply corresponding strategies for each block, the second is to identify the changeable region based on GIS firstly and then to classify those regions, and the third is to post-process the classified image by traditional methods or ANN(Artificial Neural Network) methods based on domain knowledge and GIS. Two direct extraction methods are introduced also. One is the extraction based on the water accumulating property of subsiding land, and the other is based on the dynamic change of land cover in subsiding land.
What problem does this paper attempt to address?