Landslide detection method based on double-flow feature extraction network, unmanned aerial vehicle and medium

The embodiment of the invention discloses a landslide detection method based on a double-flow feature extraction network, an unmanned aerial vehicle and a medium. The method comprises the following steps: determining an inspection area comprising a plurality of inspection nodes; determining a target inspection path according to the plurality of inspection nodes; routing inspection is performed according to the target routing inspection path and a to-be-detected image is shot; inputting the to-be-detected image into a pre-trained double-flow feature extraction network model for feature extraction to obtain a landslide boundary feature map and a mountain feature map; performing multi-scale fusion processing on the landslide boundary feature map and the mountain feature map to obtain a target feature map; determining the vegetation coverage rate of the inspection area according to the target feature map; rainfall information is obtained, and a landslide detection result is determined according to the rainfall in