SVM Lithological Classification of PolSAR Image in Yushigou Area, Qilian Mountain

2013 
Concerning the scarce geological exploration of the Yushigou by human, located in Qilian Mountain, which is an important origin of polymetallic deposits, and by means of radar remote sensing image characterized by large volume information about the geology to identify regional lithology which is helpful to direct the mine; in this paper, the RADARSAT-2 data was selected after preprocessing, to establish the sample of different lithologies by comparison and analysis on the features of different polarization imaging. In addition, the eight kinds of lithologies were obtained, such as, the Ordovician sewer cluster, volcanic and ultramafic rocks, etc. using the Support Vector Machine (SVM) classification method combined with the characteristics of color, shape and texture; and then compared with the practical results from field survey, proving the high accuracy of the classification.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    1
    References
    4
    Citations
    NaN
    KQI
    []