Multidimensional extraction of UAV tilt photography-based information of loess landslide and its application

Aiming at the needs of geological disaster investigation and assessment in loess area, a multi-dimensional extraction method of landslide single information based on UAV tilt photographic image and flight control data was explored. Rich information of the Jiajiacun landslide was quantitatively extracted from DEM, DOM and real scene 3D model obtained by image processing, including its slope shape, slope, relative height difference, vegetation coverage, longitudinal slope ratio, surface area and source volume. The stratigraphic lithology and regional structure are analyzed by using relevant geological data. Combined with the regional meteorological data, the spatial and temporal distribution of annual average rainfall of Jiajicun landslide was analyzed, which provides an efficient and accurate qualitative and quantitative reference for the accurate assessment of its stability and risk.