A method and device for establishing and identifying landslide classification models in complex background areas

Abstract

The invention provides a landslide classification model establishment and identification method and device in a complex background area, and relates to the establishment of a class imbalance model and landslide identification. The method for establishing a landslide classification model in a complex background area of ​​the present invention includes: obtaining laser radar data in the research area; constructing a terrain object according to the laser radar data, and determining a terrain object feature vector according to the terrain object to determine a data set ; According to the data set, the classification model parameters and the balance coefficient are jointly optimized to determine the collaborative optimal balance coefficient and the collaborative optimal classification model parameter; according to the collaborative optimal balance coefficient and the collaborative optimal classification model parameter determination A class-balanced robust sensitive feature subset; a landslide classification model is established based on the class-balanced robust sensitive feature subset. The technical proposal of the invention improves the recognition accuracy of landslide remote sensing.

Description

technical field [0001] The invention relates to the technical field of class imbalance model establishment and landslide identification, in particular to a landslide classification model establishment and identification method and device in complex background areas. Background technique [0002] In the prior art, in the field of landslide remote sensing recognition, the problem of class imbalance has not attracted enough attention. There are two problems in the learning method of remote sensing class imbalance: (1) lack of an effective method for optimizing the balance coefficient, resulting in the inability to obtain the optimal (2) The lack of a robust and sensitive feature subset after class balancing cannot provide an optimized data set for subsequent classification. These problems further restrict the improvement of landslide remote sensing identification accuracy. Contents of the invention [0003] The problem solved by the invention is how to improve the recognitio...

AI Technical Summary

Problems solved by technology

[0002] In the prior art, in the field of landslide remote sensing recognition, the problem of class imbalance has not attracted enough attention. There are two problems in the learning method of remote sensing class imbalance: (1) lack of an effective method for optimizing the balance coefficient, resulting in the inability to obtain the optimal (2) Lack of robust sensitive feature subsets after class balance, unable to provide optimized data sets for subsequent classification

These problems further restrict the improvement of landslide remote sensing recognition accuracy.

Images