Mountain torrent disaster hyperspectral remote sensing image identification method and identification system

Abstract

The invention discloses a mountain torrent disaster hyperspectral remote sensing image identification method. The method comprises the following steps: step 1, constructing an unmanned aerial vehicleremote sensing system; step 2, compiling a pre-flight preparation work manual and an unmanned aerial vehicle operation manual of an unmanned aerial vehicle; step 3, testing the working state of an unmanned aerial vehicle-mounted micro hyperspectral imager; step 4, determining image recognition performance and parameters; step 5, collecting data on site by the unmanned aerial vehicle remote sensingimage system; step 6,performing ground object spectral characteristic judgment and mountain torrent characteristic factor evolution analysis and recognition by an image recognition system; step 7, performing calculation and judgment by a recognition method subsystem; and step 8, recognizing mountain torrent characteristic factors, and outputting a mountain torrent ditch mountain torrent characteristic factor spatial distribution diagram of a monitoring area. The method has the advantages of realizing the accuracy, reliability and timeliness of the basic data information of landslide and debris flow areas in the mountain torrent disaster-prone areas. The invention further discloses a mountain torrent disaster hyperspectral remote sensing image identification system.

Description

Technical field [0001] The invention relates to the technical field of mountain torrent disaster monitoring and early warning, and more specifically it is a method for identifying mountain torrent disaster hyperspectral remote sensing images. The invention also relates to an identification system adopted by the method for identifying the hyperspectral remote sensing image of the mountain torrent disaster. Background technique [0002] Against the background of global warming and frequent extreme weather, my country is facing increasing risks of natural disasters in recent years, and is entering a period of frequent natural disasters. Among the many disasters, mountain torrents and their secondary disasters are the most severe flood disasters faced every year, especially in the hilly areas, many towns or residential areas are located at the mouths of mud-rock flows, river valleys and even landslides. Mountainous areas in my country account for about two-thirds of the country's la...

AI Technical Summary

Problems solved by technology

However, the application of this technology in mountain torrent disaster monitoring in my country is still in its infancy and exploration stage

Images