CN102778672A - Method for estimating phase errors of multi-polarization SAR (synthetic aperture radar) - Google Patents

Abstract

The invention relates to a method for estimating phase errors of multi-polarization SAR (synthetic aperture radar) on the basis of a phase gradient auto-focus algorithm, and belongs to the technical field of SAR imaging. The method includes estimating phase errors of various polarized images by the phase gradient auto-focus algorithm, and compensating the phase errors for the polarized images by phase error estimation values to obtain compensated polarized SAR images; and computing contrast ratios and improvement factors of the contrast ratios of the polarized images before and after the phase errors are compensated, and using the phase error estimation value corresponding to the best improvement factor of the corresponding contrast ratio as an estimation value of the phase errors of the polarized SAR images. When the phase errors of the polarized images are compensated by the phase gradient auto-focus algorithm, a focusing effect of the images with the compensated errors is improved along with increase of the contrast ratios of the images. Accordingly, the various polarized images have a good consistently focusing effect after the phase errors of the polarized images are compensated by the phase error estimation value obtained by the method.

Description

A kind of phase error estimation method that is applied to multipolarization SAR

Technical field

The invention belongs to SAR (synthetic-aperture radar) technical field of imaging, be specifically related to a kind of phase error estimation method that is applied to multipolarization SAR based on PGA algorithm (phase gradient autofocus algorithm).

Background technology

In the SAR imaging processing, the kinematic error of Texas tower can cause echo data along the orientation to there being phase error, if phase error is greatly to causing the orientation to defocusing, then must to its accurately estimate with the compensation.Generally; Phase error can be calculated by the motion sensor data on the machine of carrying; But because the measuring accuracy of motion sensor is limited; Cause the calculated value and the actual value deviation of phase error very big, so need to adopt the phase error estimation method based on echo data, for example PGA algorithm.

When the contrast of polarization diagrams picture is big, adopt the PGA algorithm to the polarization image compensation after, the focusing effect of image is obvious.Conventional P GA algorithm proposes to single polarization SAR data; Yet because same atural object has different scattering forms for the electromagnetic wave of different polarization modes; The multipolarization SAR image of same observation scene presents different scattering characteristics; To its respectively Using P GA algorithm can obtain the phase error estimation and phase error value of different accuracy, thereby cause the focusing effect of each polarization diagrams picture inconsistent, influence that follow-up polarization is synthetic, processing such as target classification and face of land parametric inversion.

Summary of the invention

Given this, the present invention proposes a kind of phase error estimation method that is applied to multipolarization SAR, can obtain the phase error estimation and phase error value of the same precision of each polarization diagrams picture, realize that the focusing effect of each polarization diagrams picture is consistent based on the PGA algorithm.

The concrete performing step of the inventive method is following:

Step 1, utilize the PGA algorithm to estimate that respectively HH polarization diagrams picture, VH polarization diagrams picture and VV polarization diagrams picture (use symbol I _HH, I _VH, I _VVExpression) phase error, the estimated value that obtains three groups of phase errors (is used symbol φ _HH(u), φ _VH(u), φ _VV(u) expression, wherein u is that the orientation is to the time domain coordinate).

${\mathrm{\φ}}_{\mathrm{HH}}\left(u\right)=\underset{m=1}{\overset{u}{\mathrm{\Σ}}}\arg \left\{\mathrm{FFT}\right[g_{\mathrm{HH}}\left(m\right)]\·\mathrm{conj}\{\mathrm{FFT}\left[g_{\mathrm{HH}}(m-1)\right]\left\}\right\}---\left(1\right)$

${\mathrm{\φ}}_{\mathrm{VH}}\left(u\right)=\underset{m=1}{\overset{u}{\mathrm{\Σ}}}\arg \left\{\mathrm{FFT}\right[g_{\mathrm{VH}}\left(m\right)]\·\mathrm{conj}\{\mathrm{FFT}\left[g_{\mathrm{VH}}(m-1)\right]\left\}\right\}---\left(2\right)$

${\mathrm{\φ}}_{\mathrm{VV}}\left(u\right)=\underset{m=1}{\overset{u}{\mathrm{\Σ}}}\arg \left\{\mathrm{FFT}\right[g_{\mathrm{VV}}\left(m\right)]\·\mathrm{conj}\{\mathrm{FFT}\left[g_{\mathrm{VV}}(m-1)\right]\left\}\right\}---\left(3\right)$

Wherein, arg representes to get phase operation, and FFT representes Fourier transform, g _HH(m), g _VH(m), g _VV(m) the phase history data of expression HH polarization diagrams picture, VH polarization diagrams picture and VV polarization diagrams picture, wherein m be the orientation of phase history data to subscript, conj representes to get conjugation.

Step 2, utilize phase error estimation and phase error value φ _HH(u), φ _VH(u), φ _VV(u) respectively HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams are looked like to carry out phase compensation.With symbol I (φ _HH), I (φ _VH), I (φ _VV) HH polarization diagrams picture, VH polarization diagrams picture and the VV polarization diagrams picture of expression behind the phase error compensation.

The contrast of each polarization diagrams picture before step 3, the calculating phase error compensation is with symbol C [I _HH], C [I _VH], C [I _VV] expression; Simultaneously, the contrast of each polarization diagrams picture behind the calculating phase error compensation is with symbol C [I (φ _HH)], C] I (φ _VH)], C [I (φ _VV)] expression;

$C\left[I_{\mathrm{ij}}\right]=\sqrt{\frac{E\left\{{\left|I_{\mathrm{ij}}\right|}^2\right\}}{E{\left\{\right|I_{\mathrm{ij}}\left|\right\}}^2}-1}---\left(4\right)$

$C\left[I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right]=\sqrt{\frac{E\left\{{\left|I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right|}^2\right\}}{E{\left\{\right|I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\left|\right\}}^2}-1}---\left(5\right)$

Wherein, ij is the polarization of electromagnetic wave mode, and ij=HH, VH, VV; E{} is for getting equal Value Operations.

Step 4, calculate the contrast improvement factor of HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams picture respectively, use symbol γ _HH(φ _HH), γ _VH(φ _VH), γ _VV(φ _VV) expression;

${\mathrm{\γ}}_{\mathrm{ij}}\left({\mathrm{\φ}}_{\mathrm{ij}}\right)=\frac{C\left[I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right]-C\left[I_{\mathrm{ij}}\right]}{C\left[I_{\mathrm{ij}}\right]},$ ij＝HH,VH,VV (6)

Step 5, get the phase error estimation and phase error value of the maximum pairing phase error of contrast improvement factor as each polarization diagrams picture, φ representes with symbol.

$\mathrm{\φ}=\underset{\mathrm{ij}=\mathrm{HH},\mathrm{VH},\mathrm{VV}}{\arg \max }\left\{{\mathrm{\γ}}_{\mathrm{ij}}\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right\}---\left(7\right)$

Beneficial effect

The invention provides a kind of phase error estimation method that is applied to multipolarization SAR, choose the phase error estimation and phase error value of each polarization diagrams picture, and utilize this value that each polarization diagrams is looked like to carry out phase compensation according to the contrast improvement factor based on the PGA algorithm.When the polarization diagrams image contrast is big; Utilize the PGA algorithm that polarization diagrams is looked like to compensate and to obtain apparent in view focusing effect; Therefore; The present invention utilizes the corresponding phase error estimation and phase error value of maximum-contrast improvement factor that polarization diagrams is looked like to compensate, and the phase error estimation and phase error precision difference that can solve because of each polarization diagrams picture causes the inconsistent problem of focusing effect, can obtain focusing effect preferably.

Description of drawings

Fig. 1 is the implementing procedure figure of specific embodiment of the present invention;

Fig. 2 is each the polarization SAR image behind the phase error compensation that two kinds of distinct methods of use obtain in the specific embodiment of the present invention; Wherein, the HH polarization SAR image of Fig. 2 (a) after for the phase error compensation that obtains through conventional P GA algorithm; The HH polarization SAR image of Fig. 2 (b) after for the phase error compensation that obtains through the inventive method; The VV polarization SAR image of Fig. 2 (c) after for the phase error compensation that obtains through conventional P GA algorithm; The VV polarization SAR image of Fig. 2 (d) after for the phase error compensation that obtains through the inventive method; The VH polarization SAR image of Fig. 2 (e) after for the phase error compensation that obtains through conventional P GA algorithm; The VH polarization SAR image of Fig. 2 (f) after for the phase error compensation that obtains through the inventive method.

Embodiment

The invention provides a kind of phase error estimation method that is applied to multipolarization SAR (synthetic-aperture radar) based on the PGA algorithm; Its basic thought is: the phase error estimation and phase error value that adopts each polarization diagrams picture of conventional P GA algorithm computation; Utilize this value that every width of cloth polarization diagrams is looked like to carry out phase error compensation; Calculate the contrast of each polarization diagrams picture of phase error compensation front and back; And try to achieve the contrast improvement factor, select the phase error estimation and phase error value of the corresponding phase error estimation and phase error value of maximum contrast improvement factor for use as each polarization diagrams picture.

Below in conjunction with Figure of description and embodiment the inventive method is described further.

To observation scene image applications the inventive method that respectively polarizes for certain cities and towns, implementing procedure is as shown in Figure 1, and concrete steps are:

Step 1, utilize conventional P GA algorithm respectively HH polarization diagrams picture, VH polarization diagrams picture and VV polarization diagrams picture (to be used symbol I _HH, I _VH, I _VVExpression) carries out phase error estimation and phase error, obtain three groups of phase error estimation and phase error values and (use symbol φ _HH(u), φ _VH(u), φ _VV(u) expression, wherein u is that the orientation is to the time domain coordinate).

${\mathrm{\φ}}_{\mathrm{HH}}\left(u\right)=\underset{m=1}{\overset{u}{\mathrm{\Σ}}}\arg \left\{\mathrm{FFT}\right[g_{\mathrm{HH}}\left(m\right)]\·\mathrm{conj}\{\mathrm{FFT}\left[g_{\mathrm{HH}}(m-1)\right]\left\}\right\}---\left(8\right)$

${\mathrm{\φ}}_{\mathrm{VH}}\left(u\right)=\underset{m=1}{\overset{u}{\mathrm{\Σ}}}\arg \left\{\mathrm{FFT}\right[g_{\mathrm{VH}}\left(m\right)]\·\mathrm{conj}\{\mathrm{FFT}\left[g_{\mathrm{VH}}(m-1)\right]\left\}\right\}---\left(9\right)$

${\mathrm{\φ}}_{\mathrm{VV}}\left(u\right)=\underset{m=1}{\overset{u}{\mathrm{\Σ}}}\arg \left\{\mathrm{FFT}\right[g_{\mathrm{VV}}\left(m\right)]\·\mathrm{conj}\{\mathrm{FFT}\left[g_{\mathrm{VV}}(m-1)\right]\left\}\right\}---\left(10\right)$

Wherein, arg representes to get phase operation, and FFT representes Fourier transform, g _HH(m), g _VH(m), g _VV(m) the phase history data of expression HH polarization diagrams picture, VH polarization diagrams picture and VV polarization diagrams picture, wherein m be the orientation of phase history data to subscript, conj representes to get conjugation.

Step 2, utilize phase error estimation and phase error value φ respectively _HH(u), φ _VH(u), φ _VV(u) HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams are looked like to carry out phase compensation, obtain HH polarization diagrams picture, VH polarization diagrams picture and VV polarization diagrams picture behind the phase error compensation, use symbol I (φ respectively _HH), I (φ _VH), I (φ _VV) expression.

The contrast of the HH polarization diagrams picture before step 3, the calculating phase error compensation, VH polarization diagrams picture, VV polarization diagrams picture is with symbol C [I _HH], C [I _VH], C [I _VV] expression; Simultaneously, the contrast of the HH polarization diagrams picture behind the calculating phase error compensation, VH polarization diagrams picture, VV polarization diagrams picture is with symbol C [I (φ _HH)], C [I (φ _VH)], C [I (φ _VV)] expression;

$C\left[I_{\mathrm{ij}}\right]=\sqrt{\frac{E\left\{{\left|I_{\mathrm{ij}}\right|}^2\right\}}{E{\left\{\right|I_{\mathrm{ij}}\left|\right\}}^2}-1}---\left(11\right)$

$C\left[I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right]=\sqrt{\frac{E\left\{{\left|I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right|}^2\right\}}{E{\left\{\right|I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\left|\right\}}^2}-1}---\left(12\right)$

Wherein, ij is the polarization of electromagnetic wave mode, and ij=HH, VH, VV; E{} is for getting equal Value Operations.

Step 4, the result who utilizes step 3 to obtain calculate the contrast improvement factor of HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams picture respectively, use symbol γ _HH(φ _HH), γ _VH(φ _VH), γ _VV(φ _VV) expression;

${\mathrm{\γ}}_{\mathrm{ij}}\left({\mathrm{\φ}}_{\mathrm{ij}}\right)=\frac{C\left[I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right]-C\left[I_{\mathrm{ij}}\right]}{C\left[I_{\mathrm{ij}}\right]},\mathrm{ij}=\mathrm{HH},\mathrm{VH},\mathrm{VV}---\left(13\right)$

Step 5, get the phase error estimation and phase error value of the maximum pairing phase error estimation and phase error value of contrast improvement factor as each polarization diagrams picture, φ representes with symbol.

$\mathrm{\φ}=\underset{\mathrm{ij}=\mathrm{HH},\mathrm{VH},\mathrm{VV}}{\arg \max }\left\{{\mathrm{\γ}}_{\mathrm{ij}}\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right\}---\left(14\right)$

Can obtain being applied to the high precision phase error estimation and phase error value of multipolarization SAR based on the PGA algorithm through above-mentioned steps.

Fig. 2 (a), Fig. 2 (c), Fig. 2 (e) are respectively and utilize conventional P GA algorithm HH polarization diagrams picture, VV polarization diagrams picture, VH polarization diagrams to be looked like to carry out the result who obtains behind the phase error compensation; Significantly defocusing appears in the as can be seen from the figure soft edge of buildings, the strong scattering point on the roof; Fig. 2 (b), Fig. 2 (d), Fig. 2 (f) are respectively and utilize the inventive method that multipolarization SAR image is carried out the result behind the phase error compensation.Can find out that Fig. 2 (b), Fig. 2 (d) and Fig. 2 (f) compare with Fig. 2 (a), Fig. 2 (c) and Fig. 2 (e) respectively, the focusing effect of buildings is greatly improved.Wherein, 6 figure of Fig. 2 laterally be the orientation to, vertically be the distance to.

In sum, more than being merely a specific embodiment of the present invention, is not to be used to limit protection scope of the present invention.All within spirit of the present invention and principle, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (2)

1. a phase error estimation method that is applied to multipolarization SAR is characterized in that, comprising:

Step 1, utilize the phase gradient autofocus algorithm to estimate the phase error of each polarization diagrams picture of SAR respectively;

Step 2, the result who utilizes step 1 to obtain look like to carry out phase error compensation to each polarization diagrams, the SAR polarization diagrams picture after being compensated;

Step 3, calculate the contrast of each polarization diagrams picture before and after the phase error compensation respectively;

Step 4, calculate the contrast improvement factor of each polarization diagrams picture before and after the phase error compensation respectively;

Step 5, phase error that the contrast improvement factor of maximum is corresponding are as the estimated value of the phase error of each polarization SAR image.

2. a kind of phase error estimation method that is applied to multipolarization SAR according to claim 1 is characterized in that, described contrast improvement factor is meant the contrast rate of change of each polarization diagrams picture of phase error compensation front and back, and computing formula is:

${\mathrm{\γ}}_{\mathrm{ij}}\left({\mathrm{\φ}}_{\mathrm{ij}}\right)=\frac{C\left[I\left({\mathrm{\φ}}_{\mathrm{ij}}\right)\right]-C\left[I_{\mathrm{ij}}\right]}{C\left[I_{\mathrm{ij}}\right]},\mathrm{ij}=\mathrm{HH},\mathrm{VH},\mathrm{VV}$

Wherein, γ _HH(φ _HH), γ _VH(φ _VH), γ _VV(φ _VV) represent the contrast improvement factor of HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams picture respectively; C [I _HH], C [I _VH], C [I _VV] represent the contrast of HH polarization diagrams picture before the phase error compensation, VH polarization diagrams picture, VV polarization diagrams picture respectively; C [I (φ _HH)], C [I (φ _VH)], C [I (φ _VV)] represent the contrast of HH polarization diagrams picture behind the phase error compensation, VH polarization diagrams picture, VV polarization diagrams picture respectively; I _HH, I _VH, I _VVRepresent HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams picture that phase error compensation is preceding respectively; I (φ _HH), I (φ _VH), I (φ _VV) represent HH polarization diagrams picture, VH polarization diagrams picture, VV polarization diagrams picture behind the phase error compensation respectively.

Images