Unlock instant, AI-driven research and patent intelligence for your innovation.

SVM and AHP combined polarimetric SAR image landslide automatic detection method

An automatic detection and image technology, applied in character and pattern recognition, instruments, calculations, etc., can solve the problems of many false alarms and low accuracy rate, and achieve the effect of suppressing false alarms and improving the accuracy rate of landslide detection

Pending Publication Date: 2022-06-03
PLA STRATEGIC SUPPORT FORCE INFORMATION ENG UNIV PLA SSF IEU
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem of low accuracy and many false alarms in the detection of landslide debris flow coverage areas in single-temporal polarization SAR images under complex backgrounds, this invention proposes an automatic detection method for landslide debris flow coverage areas combined with SVM and AHP. Compared with existing methods, the method of the present invention has a high degree of automation and has a better detection success rate

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • SVM and AHP combined polarimetric SAR image landslide automatic detection method
  • SVM and AHP combined polarimetric SAR image landslide automatic detection method
  • SVM and AHP combined polarimetric SAR image landslide automatic detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0044] like figure 1 As shown, an embodiment of the present invention provides a method for automatic detection of landslides in polarimetric SAR images combining SVM (Support Vector Machine) and AHP, including the following steps:

[0045] S101: Preprocess the post-disaster polarimetric SAR image, and calculate a polarimetric coherence matrix of the pre-processed post-disaster polarimetric SAR image;

[0046] Specifically, the preprocessing of polarimetric SAR images includes image cropping, terrain correction, azimuth angle compensation and filtering operations in sequence on post-disaster polarimetric SAR images. In order to improve the classification result of the polarimetric SAR image, this embodiment compensates for the polarimetric azimuth shift caused by the terrain slope. The filtering operation at this stage adopts refined LEE filtering with a window size of 5×5, which can reduce the adverse effect of speckle noise on the landslide detection effect.

[0047] For t...

Embodiment 2

[0056] On the basis of the foregoing embodiment, as an implementable manner, the extraction of polarization features based on the polarization coherence matrix specifically includes:

[0057] Step A1: Perform Yamaguchi decomposition on the polarization coherence matrix to obtain four scattering powers, which are respectively the surface scattering power P s , the dihedral angle scattering power P d , volume scattering power P v and the helical scattering power P h , normalize the four scattering powers;

[0058] Specifically, the Yamaguchi decomposition has been widely used to quantitatively analyze the fundamental scattering mechanisms of different landforms. This decomposition yields four scattering powers: the surface scattering power P s , the dihedral angle scattering power P d , volume scattering power P v and the helical scattering power P h , the decomposition model is as follows:

[0059] T=f s T s +f d T d +f v T v +f h T h =P s +P d +P v +P h (1...

Embodiment 3

[0077] Analytic Hierarchy Process (AHP) is a multi-attribute decision-making method based on operations research. The method decomposes decision-making tasks into different hierarchical structures, and then compares the factors that affect the task results in pairs, evaluates the grades and assigns them according to their importance, and forms a judgment matrix with the assignments of each factor.

[0078] On the basis of the above embodiments, as an implementable manner, in step S102, the AHP analysis method is used to determine the surface scattering mechanism region according to the polarization characteristics, and threshold segmentation and morphological processing are performed to obtain the suspected landslide region Specifically include:

[0079] Step B1: Construct a judgment matrix as shown in the following table according to the polarization characteristics:

[0080] Table 1 Judgment matrix and importance weights of parameters

[0081]

[0082] Specifically, acc...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides an SVM and AHP combined polarized SAR image landslide automatic detection method. The method comprises the following steps: 1, preprocessing a post-disaster polarized SAR image, and calculating a polarized coherence matrix of the post-disaster polarized SAR image; 2, extracting polarization characteristics based on the polarization coherence matrix, determining a surface scattering mechanism region by adopting an AHP analysis method according to the polarization characteristics, and performing threshold segmentation and morphological processing to obtain a suspected landslide region; 3, constructing a feature vector by using all elements of the polarization coherence matrix, carrying out SVM supervised classification on the preprocessed post-disaster polarization SAR image according to the feature vector, and carrying out binarization and morphological processing on a classification result to obtain a landslide region classification result; and step 4, performing logic and operation on the suspected landslide area obtained in the step 2 and the landslide area classification result obtained in the step 3, wherein the operation result is the landslide area. According to the invention, false alarms caused by other background ground objects on landslide detection can be inhibited, and the accuracy of landslide detection is improved.

Description

technical field [0001] The invention relates to the field of full-polarization synthetic aperture radar (PolSAR) image processing and image encryption technology, in particular to a landslide and debris flow damaged area detection method combining Support Vector Machine (SVM) and Analytic Hierarchy Process (AHP). Background technique [0002] Existing technology (e.g. Plank S, D,Eisank C,et al.Comparing object-basedlandslide detection methods based on polarimetric SAR and optical satelliteimagery—A case study in Taiwan[C] / / Proceedings of the 7th InternationalWorkshop on Science and Applications of SAR Polarimetry and PolarimetricInterferometry, POLinSAR.2015 : 27-30; Wang Xingling, Hu Deyong, Tang Hong, Shu Yang. Landslide Information Extraction from Airborne Full Polarization SAR Image Based on Bayes Decision[J]. Remote Sensing of Land and Resources, 2014, 26(02): 121-127) Although both It can detect the damaged area of ​​landslide and debris flow well, but it has poor di...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06V10/764G06K9/62G06V10/20
CPCG06F18/2411
Inventor 牛朝阳张浩波刘伟高欧阳李润生邹玮琦胡涛
Owner PLA STRATEGIC SUPPORT FORCE INFORMATION ENG UNIV PLA SSF IEU