Abstract: The development of the space remote sensing technology provides a wealth of usable remote sensing data for geologist. In these remote sensing data ASTER has broad spectral bandwidths from VNIR to TIR and can meet the requirements for classifications of most minerals and rocks. In addition to the conventional data processing methods, the hyperspectral techniques for rock and mineral information recognition and extraction may also be used to process the multiband data of ASTER and yield more accurate mineral and rock classification results. The GIS evaluation model may be used to process altered rock information interpreted by ASTER, which may produce accurate remote sensing anomalies and guide mineral exploration.