Technique of the second satellite of CBERS-1 image multi-stage information extraction on land use and land cover

Since Oct. 21st 2003, on which the second satellite of CBERS-1 was successfully launched, professionals in the field of remote sensing pay more attention if information about the Earth´s surface can be effectively extracted from its collection data. A method on extracting LULC information is discussed using data imaging on Feb. 2nd 2003 in this paper. Remote sensing has inherent correlations with geo-knowledge. There are only few men interpreting remote sensing image with spectral knowledge. In this paper, a technique of multi-stage information extraction on LULC is present considering geo-knowledge by analyzing single band valve of sample end members, profile spectrum and constructing spectral band index. Different separable information of LULC is extracted at different stages, during which certain geo-knowledge is fused. In the end, this method is applied to an experiment area-Yun Nan DaLi, which obtains a better classified result compared with other supervised classified and unsupervised classified methods at the same area. Hence, it is useful to extract land use and land cover information from CBERS-1 image and extent application fields of CBERS-1