Multi-source remote sensing data multi-feature fusion lithology automatic classification method and system

The invention discloses a multi-source remote sensing data multi-feature fusion lithology automatic classification method and system. The method comprises the following steps: acquiring multi-source remote sensing data; multi-scale optical remote sensing features are extracted, a spectrum and texture feature space is constructed, and spectrum and texture feature optimization is carried out; radar remote sensing features are extracted, and radar remote sensing feature optimization is carried out for lithology identification; multiple types of terrain factors are extracted, and terrain factor optimization is carried out for lithology identification; and constructing a lithology classification feature space in combination with the optimized spectrum and texture features, radar remote sensing features and topographic factor features, and establishing an automatic lithology classification model based on the lithology classification feature space in combination with a supervised classification algorithm and each li