Abstract: Based on multi-temporal CBERS remotely sensed images of the Zhijiang area of the Yangtze River Basin, this paper obtains terrestrial environmental elements such as water taking, vegetation and structures separately by using Maximum likelihood classifier (MLC), Mahalanobis distance classifier and Support Vector Machine(SVM) ， and makes an analysis of the change relationship between the area's terrestrial environmental elements before and After the impoundment of the Three Gorges Reservoir. The results prove that the CBERS images can be used to gain terrestrial environmental elements in a middle size river basin and their change relations effectively, which would be an useful technical support on the basin 's environment protection and city planning.

Key words: CBERS ,   terrestrial environment  ,   classification of remote sense