Spaceborne P-band broadband SAR ionospheric polarization dispersion error compensation method

Abstract

This invention discloses a method for compensating for ionospheric polarization dispersion errors in spaceborne P-band broadband SAR, relating to the field of synthetic aperture radar ionospheric effect error compensation. Specifically, firstly, a range-to-Fourier transform is performed on a two-dimensional SAR image affected by polarization dispersion to obtain a four-channel frequency domain image. Then, the image is transformed from a linear polarization basis to a circular polarization basis and processed to obtain image Z. By summing all azimuth points in each range direction of image Z, a one-dimensional complex array is obtained. The phase is extracted from each point in the array, and phase unwrapping and deblurring are performed to obtain the Faraday rotation angle. Finally, polarization dispersion compensation and a range-to-inverse Fourier transform are performed to obtain a time-domain image. This invention utilizes the SAR image itself to estimate the Faraday rotation angle, enabling accurate compensation for polarization dispersion even when errors exist in the Earth's magnetic field strength.

Description

Technical Field

[0001] This invention relates to the field of ionospheric effect error compensation in synthetic aperture radar, specifically a method for compensating for ionospheric polarization dispersion error in spaceborne P-band broadband SAR. Background Technology

[0002] Synthetic Aperture Radar (SAR) is a high-resolution microwave imaging system with features such as all-weather, all-day, and wide-area Earth observation. Therefore, SAR systems have unique advantages in disaster monitoring, environmental monitoring, marine monitoring, resource exploration, and military applications.

[0003] P-band SAR has strong penetration capabilities through leaf clusters and shallow ground surfaces, and has broad application prospects in forest biomass measurement and detection of concealed targets under vegetation cover. Fully polarimetric SAR transmits and receives electromagnetic waves using different polarization methods. The four sets of HH, HV, VH, and VV (horizontally polarized transmission and horizontally polarized reception, horizontally polarized transmission and vertically polarized reception, vertically polarized transmission and horizontally polarized reception, and vertically polarized transmission and vertically polarized reception) data can comprehensively acquire the scattering characteristics of a target to any electromagnetic wave in the observation direction. Compared with traditional SAR, fully polarimetric SAR greatly improves the ability to acquire target scattering information. However, it suffers from the Faraday rotation effect (the rotation angle is the Faraday rotation angle) when linearly polarized electromagnetic waves pass through the ionosphere, causing the polarization surface to rotate relative to the incident wave under the influence of the electromagnetic field. Therefore, when broadband SAR signals pass through the ionosphere, the variation of the Faraday rotation angle with the signal frequency must be considered, leading to distortion of the amplitude spectrum of the received signal and the occurrence of polarization dispersion.

[0004] In existing technologies, assuming that polarization dispersion occurs in the range direction and one direction of the SAR image is the azimuth direction and the other is the range direction, the contrast is calculated by taking the modulus of the images from the four channels (HH, HV, VH, and VV) and then superimposing them. The total electron content (TEC) of the ionosphere is estimated using the maximum contrast autofocus algorithm, and the Faraday rotation angle is calculated based on the TEC and the Earth's magnetic field strength to compensate for polarization dispersion. However, when there are errors in the Earth's magnetic field strength, existing technologies cannot accurately compensate for polarization dispersion. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a method for compensating for ionospheric polarization dispersion errors in spaceborne P-band broadband SAR. By estimating the Faraday rotation angle using the SAR image itself, it can accurately compensate for polarization dispersion when there are errors in the Earth's magnetic field strength.

[0006] The specific steps of the spaceborne P-band broadband SAR ionospheric polarization dispersion error compensation method are as follows:

[0007] The steps of this invention are as follows:

[0008] Step 1: For a two-dimensional SAR image affected by polarization dispersion, perform a Fourier transform on the signal at each azimuth time point of the image to the frequency domain, obtaining the frequency domain image M. HH M HV M VH and M VV .

[0009] M HH For horizontally polarized transmission and horizontally polarized reception of images, M HV For horizontally polarized transmission and vertically polarized reception of images, M VH For vertically polarized transmission and horizontally polarized reception of images, M VV To transmit images with vertical polarization and receive them with vertical polarization;

[0010] Step 2: Convert the frequency domain image M HH M HV M VH and M VV Transforming to the circular polarization basis yields the two-dimensional image Z. 12 and Z 21 ;

[0011] Z 12 =j*M HH +j*M VV +M VH -M HV

[0012] Z 21 =j*M HH +j*M VV -M VH +M HV

[0013] Step 3: Transform the 2D image Z... 12 The value of each pixel in the image is conjugate and then compared with the value of the two-dimensional image Z. 21 Multiply them to obtain the image Z;

[0014]

[0015] Step 4: Sum all azimuth points in the Z-axis of the image to obtain a one-dimensional complex array, and extract the phase for each point in the array.

[0016] Step 5: Perform phase unwrapping on the phase of each extracted point, and take one-quarter of the original phase value as the estimated Faraday rotation angle.

[0017] Step 6: Estimate each Faraday rotation angle within the bandwidth. Go separately Blurring, obtaining the Faraday rotation angle

[0018] Specifically:

[0019] First, for each Faraday rotation angle within the bandwidth Calculate the total electron content (TEC) of the ionosphere at each frequency corresponding to each value, and calculate the average value TEC1; at the same time, calculate TEC2 when the Faraday rotation angle is π / 2 and f = 0;

[0020] The calculation formula is as follows:

[0021]

[0022] Where Q is a constant; θ represents the angle between the direction of the Earth's magnetic field and the direction of radar electromagnetic wave propagation; B represents the intensity of the Earth's magnetic field; f c f is the carrier frequency, and f is the frequency corresponding to the current Faraday rotation angle;

[0023] When TEC1-n*TEC2 reaches the given TEC range, obtain the integer n, and then calculate the Faraday rotation angle. Subtract n times Get to Blurred Faraday rotation angle Right now

[0024] Step 7: Utilize the deblurred Faraday rotation angle Polarization dispersion is compensated to obtain the frequency domain images of the compensated HH, HV, VH, and VV channels. and

[0025] The compensation formula is as follows:

[0026]

[0027]

[0028]

[0029]

[0030] Step 8: Convert the frequency domain image and

[0031] Perform a distance-to-inverse Fourier transform to obtain the time-domain image.

[0032] The advantages of this invention are:

[0033] The present invention provides a method for compensating for ionospheric polarization dispersion error in spaceborne P-band broadband SAR. By summing all azimuth points in each distance direction of the image Z and extracting the phase, the Faraday rotation angle that varies within the bandwidth is obtained. This allows the Faraday rotation angle that varies within the bandwidth to be estimated using the SAR image itself without relying on the Earth's magnetic field strength, thereby compensating for polarization dispersion. Attached Figure Description

[0034] Figure 1 This is a flowchart of the spaceborne P-band broadband SAR ionospheric polarization dispersion error compensation method of the present invention;

[0035] Figure 2 This is an HH channel city image that is not affected by polarization dispersion in a specific embodiment of the present invention;

[0036] Figure 3 This is an HH channel mountain image that is not affected by polarization dispersion in a specific embodiment of the present invention. Detailed Implementation

[0037] To facilitate understanding and implementation of the present invention by those skilled in the art, the present invention will be further described in detail and in depth below with reference to the accompanying drawings.

[0038] This invention provides a method for compensating for ionospheric polarization dispersion error. It utilizes the SAR image itself and employs the Bickel & Bates method to estimate the Faraday rotation angle. By taking advantage of the different Faraday rotation angles corresponding to different distances in the image, the range of variation of the Faraday rotation angle within the bandwidth is estimated. After phase unwrapping and deblurring, the Faraday rotation angle is used to compensate for polarization dispersion.

[0039] The aforementioned spaceborne P-band broadband SAR ionospheric polarization dispersion error compensation method, such as... Figure 1 As shown, the specific steps are as follows:

[0040] Step 1: For the 2D SAR image affected by polarization dispersion, perform a Fourier transform (range-to-Fourier transform) on the signal at each azimuth time step to obtain the frequency domain image M affected by polarization dispersion. HH M HV M VH and M VV .

[0041] M HH For horizontally polarized transmission and horizontally polarized reception of images, M HV For horizontally polarized transmission and vertically polarized reception of images, M VH For vertically polarized transmission and horizontally polarized reception of images, M VV To transmit images with vertical polarization and receive them with vertical polarization;

[0042] Step 2: Convert the frequency domain image M HH M HV M VH and M VV Transforming to the circular polarization basis yields the two-dimensional image Z. 12 and Z 21 ;

[0043] Z 12 =j*M HH +j*M VV +M VH -M HV

[0044] Z 21 =j*M HH +j*M VV -M VH +M HV

[0045] j is the imaginary unit.

[0046] Step 3: Transform the 2D image Z... 12 The value of each pixel in the image is conjugate and then compared with the value of the two-dimensional image Z. 21 Multiply them to obtain the image Z;

[0047]

[0048] Step 4: Sum all azimuth points in the Z-axis of the image to obtain a one-dimensional complex array, and extract the phase for each point in the array.

[0049] Step 5: Perform phase unwrapping on the phase of each extracted point, and take one-quarter of the original phase value as the estimated Faraday rotation angle.

[0050] Step 6: Estimate each Faraday rotation angle within the bandwidth. Go separately Blurring, obtaining the Faraday rotation angle

[0051] Specifically:

[0052] First, for each Faraday rotation angle within the bandwidth Calculate the total electron content (TEC) of the ionosphere at each corresponding frequency and obtain the average value TEC1; simultaneously, calculate the Faraday rotation angle. TEC2 when f=0;

[0053] The calculation formula is as follows:

[0054]

[0055] Where Q is a constant; θ represents the angle between the direction of the Earth's magnetic field and the direction of radar electromagnetic wave propagation; B represents the intensity of the Earth's magnetic field; f c f is the carrier frequency, and f is the frequency corresponding to the current Faraday rotation angle;

[0056] When TEC1-n*TEC2 reaches the given TEC range, obtain the integer n, and then calculate the Faraday rotation angle. Subtract n times Get to Blurred Faraday rotation angle Right now

[0057] Step 7: Utilize the deblurred Faraday rotation angle Polarization dispersion is compensated to obtain the frequency domain images of the compensated HH, HV, VH, and VV channels. and

[0058] The compensation formula is as follows:

[0059]

[0060]

[0061]

[0062]

[0063] Step 8: Convert the frequency domain image and Perform a distance-to-inverse Fourier transform to obtain the time-domain image.

[0064] Example:

[0065] Step 1: When the SAR signal passes through the ionosphere, it generates a Faraday rotation angle Ω(f) under the influence of electric and magnetic fields:

[0066]

[0067] Where Q is a constant; and Q = 2.365 × 10 4 The unit is A·m 2 / kg; θ represents the angle between the direction of the Earth's magnetic field and the direction of radar electromagnetic wave propagation; B represents the intensity of the Earth's magnetic field, in Wb / m. 2 ;f c Here, f is the carrier frequency, f is the frequency of the baseband signal, and TEC is the total electron content of the ionosphere, measured in TECU, where 1 TECU = 102 16 m -2 .

[0068] Step 2: Since the Faraday rotation angle Ω(f) is related to the operating frequency f+f of the SAR system... c This is related to the fact that different frequency components correspond to different Ω(f), which leads to polarization dispersion;

[0069] as follows:

[0070]

[0071] in This represents the scattering matrix of a point target affected by polarization dispersion. Represents the scattering matrix of a real point target (unaffected by polarization dispersion). It is a Faraday rotation matrix.

[0072] Where M HH M HV M VH M VV These represent the measured HH, HV, VH, and VV channel images, respectively; S HH S HV S VH S VV These represent the actual HH, HV, VH, and VV channel images, respectively; expanding the above equation yields:

[0073] M HH =cos 2 Ω(f)·S HH +cosΩ(f)sinΩ(f)·(S HV -S VH )-sin 2 Ω(f)·S VV

[0074] M HV =cos 2 Ω(f)·S HV -cosΩ(f)sinΩ(f)·(S HH +S VV )+sin 2 Ω(f)·S VH

[0075] M VH =cos 2 Ω(f)·S VH +cosΩ(f)sinΩ(f)·(S HH +S VV )+sin 2 Ω(f)·S HV

[0076] M VV =cos 2Ω(f)·S VV +cosΩ(f)sinΩ(f)·(S HV -S VH )-sin 2 Ω(f)·S HH

[0077] M HH For a single point target, it is represented by the measured horizontally polarized emission and horizontally polarized reception scattering components. For a two-dimensional image, it is represented by the measured horizontally polarized emission and horizontally polarized reception image, which is composed of countless single point targets.

[0078] Polarization dispersion can cause distortion of the amplitude spectrum of the received signal, resulting in defocusing of the SAR image at range.

[0079] Step 3: Perform a range-to-Fourier transform on the time-domain SAR image affected by polarization dispersion to obtain the frequency-domain image M. HH M HV M VH M VV ;

[0080] Step 4: Transform the frequency domain image affected by polarization dispersion to a circularly polarized basis to obtain the two-dimensional image Z. 12 and Z 21 ;

[0081] Step 5: Estimate the Faraday rotation angle using data from the circularly polarized basis;

[0082] The two-dimensional image Z 12 The value of each pixel in the image is conjugate and then compared with the value of the two-dimensional image Z. 21 Multiply to obtain image Z. Sum all azimuth points in each distance direction of image Z and extract the phase. Through phase unwrapping and deblurring, estimate the Faraday rotation angle that varies within the bandwidth.

[0083] Step 6: Use the deblurred Faraday rotation angle to compensate for polarization dispersion; and perform a range-direction inverse Fourier transform on the frequency domain image after polarization dispersion compensation to obtain the time domain image.

[0084] This implementation case uses urban and mountain scenes from ALOS PALSAR2's fully polarimetric data for simulation verification, such as... Figure 2 and Figure 3 The images shown are city and mountain images in the HH channel, which are unaffected by polarization dispersion. The size of each image is 400 in the azimuth direction and 400 in the range direction. The horizontal direction is the range direction and the vertical direction is the azimuth direction.

[0085] First, the simulation introduces noise and polarization dispersion. The HH, HV, VH, and VV channel images, which are unaffected by polarization dispersion, are superimposed with random noise of 20dB and then subjected to a range-to-Fourier transform. Next, the HH, HV, VH, and VV channel images after the range-to-Fourier transform are subjected to amplitude-weighted polarization dispersion. Here, the parameter f... c =435MHz, |f|≤42.5MHz, TEC=50TECU, θ=0°, B=3.5×10 -5 Wb / m 2 Then, a distance-to-inverse Fourier transform is performed to obtain a time-domain image affected by polarization dispersion.

[0086] Next, polarization dispersion is compensated, where the given TEC range is 40-60TECU. When TEC1-n*TEC2 belongs to 40-60TECU, n = -2 is obtained.

[0087] The results show that the correlation coefficients of the images after polarization dispersion compensation using this invention and the images unaffected by polarization dispersion were calculated, and the correlation coefficients of the HH, HV, VH, and VV channels in urban and mountain scenes are shown in Tables 1 and 2.

[0088] Table 1

[0089] HH HV VH VV There is currently a 10% magnetic field error. 0.999 0.7264 0.7845 0.9979 There is currently a 20% magnetic field error. 0.9806 0.5107 0.6169 0.9612 This method 1 0.9999 0.9999 1

[0090] Table 2

[0091] HH HV VH VV There is currently a 10% magnetic field error. 0.9988 0.8217 0.8304 0.9985 There is currently a 20% magnetic field error. 0.9756 0.6426 0.655 0.9684 This method 0.9997 0.9997 0.9997 0.9997

[0092] In existing methods for estimating TEC using maximum contrast, for urban scenarios, when there is a 10% error in the Earth's magnetic field strength (i.e., the magnetic field strength is 0.9 times the actual Earth's magnetic field strength), the correlation coefficients of the HH and VV channels are close to 1, with the correlation coefficients of the HV and VH channels being 0.7264 and 0.7845, respectively. When there is a 20% error in the Earth's magnetic field strength (i.e., the magnetic field strength is 0.8 times the actual Earth's magnetic field strength), the correlation coefficients of the HH, HV, VH, and VV channels are 0.9806, 0.5107, 0.6169, and 0.9612, respectively.

[0093] For mountain scenes, when there is a 10% error in the Earth's magnetic field strength, the correlation coefficients for the HV and VH channels are 0.8217 and 0.8304, respectively. When there is a 20% error in the Earth's magnetic field strength, the correlation coefficients for the HH, HV, VH, and VV channels are 0.9756, 0.6426, 0.6550, and 0.9684, respectively. In this invention, the correlation coefficients for all four polarization channels are close to 1. Therefore, it can be seen that with existing methods, when there is an error in the Earth's magnetic field strength, the HV and VH channel images after polarization dispersion compensation differ significantly from the HV and VH channel images unaffected by polarization dispersion. However, the images after polarization dispersion compensation using this invention are essentially consistent with the images unaffected by polarization dispersion.

Claims

1. A method for compensating for ionospheric polarization dispersion error in spaceborne P-band broadband SAR, characterized in that, The specific steps are as follows: First, for the two-dimensional SAR image affected by polarization dispersion, the signal at each azimuth time point of the image is Fourier transformed to the frequency domain, resulting in the frequency domain image M corresponding to the four channels HH, HV, VH, and VV. HH M HV M VH and M VV ; Then, the frequency domain image M HH M HV M VH and M VV Transforming to the circular polarization basis yields the two-dimensional image Z. 12 and Z 21 By using the two-dimensional image Z 12 The value of each pixel in the image is conjugate and then compared with the value of the two-dimensional image Z. 21 Multiply them to obtain the image Z; Next, sum the values ​​of all azimuth points along each distance in the Z-axis of the image to obtain a one-dimensional complex array. Extract the phase from each point in the array, unwrap the phase, and take one-quarter of the original phase value as the estimated Faraday rotation angle. Meanwhile, the estimated Faraday rotation angles within the bandwidth Go separately Blurring, obtaining the Faraday rotation angle The phase unwrapping and deblurring are then performed to obtain the Faraday rotation angle. The process is as follows: for each Faraday rotation angle within the bandwidth Calculate the total electron content (TEC) of the ionosphere at each frequency for each value, and obtain the average value TEC1; simultaneously, calculate the Faraday rotation angle. TEC2 when f=0; The calculation formula is as follows: Where Q is a constant; θ represents the angle between the direction of the Earth's magnetic field and the direction of radar electromagnetic wave propagation; B represents the intensity of the Earth's magnetic field; f c f is the carrier frequency, and f is the frequency of the baseband signal corresponding to the current Faraday rotation angle; When TEC1-n*TEC2 reaches the given range of total ionospheric electron content TEC, obtain the integer n, and then calculate the Faraday rotation angle. Subtract n times Get to Blurred Faraday rotation angle Right now Finally, using the deblurred Faraday rotation angle Polarization dispersion is compensated to obtain the frequency domain images of the compensated HH, HV, VH, and VV channels. and Then, a distance-to-inverse Fourier transform is performed to obtain the time-domain image.

2. The spaceborne P-band broadband SAR ionospheric polarization dispersion error compensation method as described in claim 1, characterized in that, The frequency domain image M HH M HV M VH and M VV Transforming to the circular polarization basis yields the two-dimensional image Z. 12 and Z 21 The formula is: Z 12 =j*M HH +j*M VV +M VH -M HV Z 21 =j*M HH +j*M VV -M VH +M HV j is the imaginary unit.

3. The spaceborne P-band broadband SAR ionospheric polarization dispersion error compensation method as described in claim 1, characterized in that, The compensation formula is as follows: M HH For horizontally polarized transmission and horizontally polarized reception of images, M HV For horizontally polarized transmission and vertically polarized reception of images, M VH For vertically polarized transmission and horizontally polarized reception of images, M VV To transmit images with vertical polarization and receive them with vertical polarization.

Images