Study On The Distribution And Identification Methods Of Landslides In Qinling-daba Mountains

Landslides are characterized by unpredictability,strong suddenness,wide distribution,and great destructiveness.Although geoscientists have used a series of identification and monitoring methods to investigate and warn landslides for a long time,the problem that landslides cause huge losses to people's lives and property has not been fundamentally solved.At present,Remote Sensing(RS)images is used in the early identification and investigation of geological disasters is one of the important research directions in the field of landslide disasters.In the aspect of landslide identification,geoscientists at home and abroad have gradually shifted from the visual interpretation and human-computer interactive interpretation to the automatic identification of geological hazards.With the development of computer technology,automatic identification of geological hazards has gradually matured,and object-oriented and pixel-based computer identification methods have been developed.The article is based on the project "Investigation of Geological Hazards in Zhenba County in the Middle Reaches of the Hanjiang River",taking Zhenba County and its surrounding area in the Qinling-Daba Mountains as the research area.Through data collection and field disasters investigation in the area,the distribution law of landslides is summarized,and quantifiable interpretation signs are extracted to intelligently interpret landslides.The main research methods and results of this article are as follows:1.The article combines the geological data collected in the research area and the field geological disasters survey to make statistics and analysis of the correlation between the distribution of landslides and landform,slope,aspect,slope type,distance from rivers and roads in the study area.Finally,the distribution law and geological identification signs of landslides in the study area are summarized.2.By collecting high-resolution remote sensing images and terrain data in the study area,construct three-dimensional geological models of the study area.On this basis,colors,image textures,and topographic interpretation signs of landslides in the study area are proposed.This article summarizes a set of visual interpretation methods for landslides in the Qinling-Daba Mountains,and provides theoretical guidance for intelligent landslide RS interpretation.3.This paper compares object-oriented and pixel-based landslide identification methods,showing that object-oriented landslide identification method is more suitable for landslide identification in the study area,and explores the optimal segmentation scale and object classification in the study area.This paper uses quantifiable identification signs of landslides to interpret the areas where landslides are concentrated.The results show that,compared with visual interpretation,object-oriented identification method has advantages in terms of recognition efficiency,integrated analysis of various feature information,output of results.Object-oriented identification method can identify landslides and greatly reduces the workload of remote sensing interpretation.With continuous development of computer technology and improvement of remote sensing images,intelligent interpretation will eventually replace visual interpretation and human-computer interactive interpretation to achieve complete automation.The research and achievements of this paper lay a foundation for the further development of intelligent landslide RS interpretation.