A polarimetric SAR image feature extraction method

By using a two-stage target decomposition model, the problem of scattering symmetry assumption in polarimetric SAR images is solved. The model decomposes the target into multiple scattering components, enabling detailed description and accurate differentiation of ground features and improving the interpretation effect of polarimetric SAR images.

CN117111063BActive Publication Date: 2026-07-10SHANGHAI RADIO EQUIP RES INST
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SHANGHAI RADIO EQUIP RES INST
Filing Date
2023-08-22
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing polarimetric SAR image target decomposition methods fail to adequately consider the scattering symmetry assumption, resulting in the appearance of negative energy and the loss of physical meaning in the decomposition results. Furthermore, they are difficult to distinguish between symmetric and asymmetric scattering cases and fail to effectively compensate for even-order scattering caused by rotation.

Method used

A two-stage target decomposition model is adopted, including polarization pointing angle compensation, ellipticity angle compensation, power ratio calculation of scattering asymmetric components, and adaptive volume scattering model. The polarimetric SAR image is decomposed into scattering symmetric and asymmetric parts, and further decomposed into surface, even, adaptive volume, spiral, directional dipole and composite asymmetric scattering components.

Benefits of technology

It enables more detailed descriptions of different land features, extracts more representative scattering information, decomposes the asymmetric part of scattering in detail, effectively distinguishes various land feature types, and improves the accuracy of polarimetric SAR image interpretation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117111063B_ABST
    Figure CN117111063B_ABST
Patent Text Reader

Abstract

This invention relates to a method for feature extraction from polarimetric SAR images, comprising the following steps: acquiring raw polarimetric SAR image data to obtain the coherence matrix of the polarimetric SAR image; performing polarization pointing angle compensation and ellipticity angle compensation on the coherence matrix; calculating the proportion of power corresponding to the asymmetric scattering component to the total power; decomposing the coherence matrix after ellipticity angle compensation into a symmetric scattering part and an asymmetric scattering part, obtaining the coherence matrix and scattering power of each part; determining an adaptive volume scattering model; decomposing the obtained symmetric scattering part into surface scattering component, even-order scattering component, and adaptive volume scattering component; and decomposing the asymmetric scattering part into helical scattering component, directional dipole scattering component, and composite asymmetric scattering component. This invention can provide a more detailed description of ground features in different states, effectively extracting scattering information from different ground features, and playing a significant role in analyzing the scattering characteristics of different ground features.
Need to check novelty before this filing date? Find Prior Art