Method and device for identifying earthquake landslide in complex mountainous area, medium and equipment

By identifying and fusion of remote sensing images and DEM data in the complex mountainous seismic landslide area, and using the improved U-Net network model and attention mechanism module to identify, the problem of poor seismic and landslide recognition in existing technologies is solved. Realize higher identification accuracy.

CN119380187AInactive Publication Date: 2025-01-28CHINA EARTHQUAKE DISASTER PREVENTION CENT
View PDF 3 Cites 2 Cited by

Patent Information

Application Number
CN202411416906.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-11
Publication Date
2025-01-28
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the identification of seismic landslides in complex mountainous areas, the accuracy of the identification results is relatively poor due to the heterogeneity and spatial inhomogeneity of seismic landslides.

Method used

Through the remote sensing images and DEM grid data of the obtained area, the characteristic extraction and fusion of the DEM grid data, combined with the improved U-Net network model, use the attention mechanism module to give the features to the characteristics of the characteristics to identify the earthquake and landslide.

Benefits of technology

Improve the accuracy of earthquake landslide recognition, and can more accurately obtain the identification results of the seismic landslide in the test area.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119380187A_ABST
    Figure CN119380187A_ABST
Patent Text Reader

Abstract

The invention discloses a complex mountainous area seismic landslide identification method and device, a medium and equipment, and relates to the technical field of seismic landslide identification, and the method comprises the steps: carrying out the feature extraction and fusion of an obtained remote sensing image of a to-be-detected area and DEM raster data, and obtaining a fused remote sensing image; based on the U-Net network model, an attention mechanism module is added after jump connection of a last layer of a decoder of the U-Net network model, and an improved U-Net network model is constructed; and inputting the fused remote sensing image into the improved U-Net network model, training the improved U-Net network model to obtain an identification model, inputting the remote sensing image of the area to be detected into the identification model to carry out seismic landslide identification, and obtaining a seismic landslide identification result. According to the method, the terrain features and the optical remote sensing features of the earthquake landslide in the complex mountainous area are fused, so that the identification result of the earthquake landslide in the to-be-detected area is accurately obtained.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Landslide and surrounding ground feature identification method and system based on multi-modal feature fusion

    CN115830469A

  • Landslide detection method of double-branch network combining spectrum and digital elevation model

    CN116205122A

  • Landslide detection method and system based on multi-scale feature fusion

    CN116740521A

Cited By

  • Terraced field water and soil conservation measure damage judgment method and device and storage medium

    CN120808182A

  • Landslide detection method based on optical satellite remote sensing and DEM data adaptive fusion

    CN122157008A