Method for decomposing image four components of polarization synthetic aperture radar

Abstract

The invention discloses a method for decomposing image four components of a polarization synthetic aperture radar and relates to analysis of the scattering characteristic of a polarization SAR image land object, in particular to a computing method of the scattering type of a building which has the non-zero orientation angle and is hard to process in an existing polarization SAR image decomposing method. As it is shown in the graph (1), the method includes the following steps that firstly, in terms of pixels corresponding to the building with the non-zero orientation angle in the image, a screening computing method is provided; secondly, in terms of the building with the non-zero orientation angle, a generalized even-time scattering model which can describe the scattering feature of the building is provided; thirdly, a non-negative feature value decomposing technology is provided for resolving the scattering components. By means of the method for decomposing the image four components of the polarization synthetic aperture radar, the defects of an original scattering model can be effectively avoided, a satisfying image decomposing result on the high-resolution full-polarized SAR image can be obtained and the method has the advantages of being novel, innovative, simple and practical.

Description

technical field [0001] The invention relates to the extraction of scattering information of Polarimetric Synthetic Aperture Radar (POLSAR) images, modeling of scattering types, estimation and acquisition of scattering components, and specifically a four-component decomposition method of Polarimetric Synthetic Aperture Radar images. Background technique [0002] Traditional synthetic aperture radar (Synthetic Aperture Radar, SAR) uses a single polarization channel to transmit and receive electromagnetic waves, and can only obtain the scattering characteristics of ground objects for electromagnetic waves of a certain polarization mode, and the acquired scattering information of ground objects is limited. As a new development direction of SAR, POLSAR transmits and receives electromagnetic waves with different polarization modes through two channels, and obtains four sets of measurement data to form a complete polarization base to detect the scattering characteristics of electrom...

AI Technical Summary

Problems solved by technology

[0004] Yamaguchi decomposition is currently the commonly used four-component decomposition method for polarimetric SAR in the world. Although this method can link the scattering mechanism of the actual ground object with the physical scattering model of the ground object, it is limited by the description ability of the model, and there are some problems in the calculation results. At the same time, this method also has the problem of overestimating the power of the volume scattering component, especially for high-resolution polarimetric SAR images, the above limitations are more obvious

Specifically, for high-resolution full-polarization SAR images, artificial buildings in urban areas usually form an included angle with the azimuth of the radar, causing the orientation angle of these artificial buildings to be different from 0 degrees, which in turn causes such ground objects to be in the When doing Yamaguchi four-component decomposition, the volume scattering component is too high, which does not match the actual ground object scattering type

The method of orientation angle compensation is often used in the world. After the orientation angle of the scattering matrix of the target is rotated to 0 degrees, polarization decomposition is performed; however, as the orientation angle increases, the effect of orientation angle compensation decreases, and the orientation of buildings When the angle reaches 45 degrees, the compensation effect of the orientation angle is the worst. Therefore, in the decomposition results, the power of the scattering component of the ground object is still dominant, and the correct scattering mechanism of this type of ground object cannot be obtained, which brings difficulties to the identification of the ground object type.

Images