The invention discloses a
remote sensing-based
landslide mass volume estimation method. The method is characterized by comprising the steps of obtaining a post-disaster
remote sensing image and a pre-disaster
terrain data DEM of a to-be-estimated
landslide mass region; performing boundary extraction on the
landslide mass region in the post-disaster
remote sensing image, and extracting and obtaining feature points of a landslide mass boundary; selecting a
reference plane; based on the feature points of the landslide mass boundary, performing
surface fitting on a landslide mass to obtain a fitted landslide mass surface, and calculating the volume between the fitted landslide mass surface and the
reference plane; according to the DEM, extracting an original
terrain surface, and calculating the volume between the original
terrain surface and the
reference plane; and fitting a difference between the volume between the landslide mass surface and the reference plane, and the volume between the original terrain surface and the reference plane to serve as a landslide mass volume. According to the method, the
estimation precision of the landslide mass volume can be improved, so that the
estimation time is greatly shortened.