A multi-dimensional synthetic aperture radar registration method and device

Abstract

The present invention discloses a multi-dimensional synthetic aperture radar registration method and device, which performs consistency processing on signals during the imaging process to achieve multi-band consistency imaging. The method includes: selecting a reference band, obtaining the range offset and azimuth offset of each band relative to the reference band based on the phase center position of the antenna of each band, the reference trajectory and the reference drift angle of each band; performing range signal resampling, range spectrum bandpass filtering and range signal interception to obtain processed range pulse pressure data; performing azimuth resampling, spectrum bandpass filtering and azimuth signal interception to obtain processed SAR images; using calibration points to check the registration accuracy between each band image and the reference band image; if the registration accuracy is greater than 1 pixel, using calibration points to correct the range offset and azimuth offset of each non-reference band image relative to the reference band.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-band synthetic aperture radar data processing, and in particular to a multi-dimensional synthetic aperture radar registration method and device. Background Art

[0002] As an active remote sensing technology, Synthetic Aperture Radar (SAR) has the ability to observe the earth all day and all weather, and has important application potential in the fields of natural resource monitoring, terrain mapping, forest biomass estimation, etc. With the continuous development of SAR technology, SAR has expanded from single polarization to multi-polarization, from a single band to multiple bands, and systems with multi-band multi-polarization and other multi-dimensional simultaneous observation capabilities continue to emerge. For example, Germany's F-SAR has the ability to obtain full polarization data in the five bands of P, L, S, C, and X; the domestic multi-dimensional SAR has the ability to observe the six bands of P, L, S, C, X, and Ka at the same time.

[0003] Multi-dimensional SAR can provide data in multiple bands. The scattering characteristics and penetration capabilities of objects in different bands are different, which is conducive to further promoting the development of SAR industry applications. However, due to differences in bandwidth, beam width, antenna phase center installation position, etc. in different bands, there are geometric distortions between SAR images in different bands. How to achieve multi-dimensional SAR multi-band registration is an important prerequisite for the comprehensive application of multi-dimensional SAR multi-band data.

[0004] In recent years, a lot of research has been carried out to achieve SAR registration. Most scholars start with the images after SAR imaging processing, borrow the registration methods of computer vision and optical images, use the features or grayscale information of the image for matching, and then build a geometric relationship model for geometric transformation to achieve image registration, but the registration accuracy is relatively low. There are also a few scholars who process the SAR data at the imaging stage, use high-precision external DEM, and study and analyze the sources of errors in the SAR imaging process, perform error compensation in the imaging process, and achieve image self-registration after SAR imaging. At present, the research on registration is mainly aimed at the same band. Due to the differences in the scattering characteristics and geometric differences of the objects in different bands, the registration technology directly starting from the image often fails to match, and the registration technology processed from the imaging process is limited to a certain band, but the antenna phase center is different. At the same time, the registration accuracy is affected by the accuracy of the DEM and the approximate processing of error compensation. Therefore, facing the joint observation application needs of multi-dimensional SAR in multiple bands, it is of great research significance to improve the registration accuracy of multi-dimensional SAR in multiple bands. Summary of the invention

[0005] In view of this, the present invention provides a multi-dimensional synthetic aperture radar alignment method and device, which can solve the technical problems of coupling errors caused by different band sampling intervals, antenna phase center positions, time synchronization, motion compensation, etc.

[0006] In order to solve the above technical problems, the present invention is implemented as follows.

[0007] A multi-dimensional synthetic aperture radar registration method, the method comprising the following steps:

[0008] Step S1: Obtain a multi-band echo data set I={I 1 ,…,I r ,…,I num}; Calculate the sampling delay correction value of each band using the internal calibration data in the echo data; Generate the reference trajectory of each band using the position and attitude data; Obtain the phase center position of each band antenna based on the installation position of each band antenna; Obtain the reference drift angle of each band based on the reference trajectory of each band; Select the reference band, and obtain the distance offset D of each band relative to the reference band based on the phase center position of each band antenna, the reference trajectory and the reference drift angle of each band. 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}; where num is the number of bands in the system, 1≤i≤num-1, 1≤r≤num, I r is the echo data corresponding to band r, D i is the distance offset of non-reference band i relative to the reference band, T i is the azimuth offset of the non-reference band i relative to the reference band;

[0009] Step S2: for each of the distance offsets D relative to the reference band i , perform the following operations: perform range pulse compression processing on the echo signal, based on D i The corresponding band distance spectrum bandwidth, sampling delay correction value, and D i , perform range signal resampling, range spectrum bandpass filtering and range signal interception; obtain the processed range pulse pressure data, denoted as D i ';

[0010] Step S3: for each of the azimuth offsets T relative to the reference band i, perform the following operations: perform motion compensation and Stolt interpolation on the processed distance pulse pressure data, based on the T i The azimuth spectrum bandwidth, pulse repetition frequency, and T of the corresponding band i , perform azimuth resampling, spectrum bandpass filtering and azimuth signal interception; obtain the processed SAR image, denoted as T i ';

[0011] Step S4: using calibration points to check the registration accuracy between each band image and the reference band image;

[0012] Step S5: If the registration accuracy is greater than 1 pixel, the distance offset and azimuth offset of each non-reference band image relative to the reference band are corrected using the calibration points, and the process proceeds to step S2; if the registration accuracy is less than or equal to 1 pixel, the method ends.

[0013] Preferably, the step S1 comprises:

[0014] Step S11: Calculate the sampling delay correction value of each band according to the internal calibration data in the echo data;

[0015] Step S12: Project the longitude and latitude position coordinates in the position and attitude data into Gaussian plane coordinates, and then perform linear fitting on the Gaussian plane coordinates of the position to generate reference trajectories and reference drift angles of each band; based on the installation positions of the antennas of each band, obtain the phase center positions of the antennas of each band; based on the phase center positions of the antennas of each band, perform coordinate rotation on the actual flight trajectory and reference trajectory of each band to generate the antenna phase center trajectory of each band;

[0016] Step S13: According to the antenna phase center trajectory of each band and the reference drift angle, the position vector of the antenna phase center of each band is used to calculate the distance offset D of each band relative to the reference band = {D 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}.

[0017] Preferably, the step S2 comprises:

[0018] Step S21: Using D i The echo data of the corresponding band and D i The range reference function of the corresponding band completes the range pulse compression processing;

[0019] Step S22: According to D iThe minimum sampling frequency of D i The distance pulse pressure data of the corresponding band is resampled in the distance direction;

[0020] Step S23: Estimate D i The distance of the corresponding band is the range of the spectrum bandwidth, with D i The minimum value of the distance to spectrum bandwidth in the band is D i The distance of the corresponding band is subjected to frequency domain bandpass filtering on the pulse pressure data;

[0021] Step S24: According to D i The sampling delay correction value of the corresponding band, system sampling delay, D i The range offset and range beam width of the corresponding band are i The distance pulse pressure data of the corresponding band is intercepted to obtain the processed distance pulse pressure data.

[0022] Preferably, the step S3 comprises:

[0023] Step S31: Select the oblique angle of the reference band and calculate T i The Doppler center frequency of the corresponding band is i The processed distance pulse pressure data of the corresponding band is subjected to motion compensation and Stolt interpolation processing;

[0024] Step S32: According to T i The minimum PRF value for T i The data of the corresponding band is resampled in azimuth and the azimuth pulse compression process is completed;

[0025] Step S33: Estimation of T i The azimuth spectrum width of the corresponding band ranges in T i The minimum value of the azimuth spectrum width of the corresponding band is T i The data of the corresponding band is subjected to frequency domain bandpass filtering;

[0026] Step S34: According to the azimuth offset, i The data of the corresponding band is used to intercept the azimuth signal.

[0027] Preferably, the step S4 comprises:

[0028] Step S41: using the geographic coordinates of the calibration point and image auxiliary data, the calibration point is roughly positioned according to the range Doppler model;

[0029] Step S42: performing sinc interpolation on the image in a preset range around the calibration point to estimate the position of the sub-pixel calibration point;

[0030] Step S43: Calculate the coordinate difference of the calibration point between each non-reference band image and the reference band image.

[0031] A multi-dimensional synthetic aperture radar registration device, the device comprising:

[0032] Data acquisition module: configured to acquire a multi-band echo data set I={I 1 ,…,I r ,…,I num}; Calculate the sampling delay correction value of each band using the internal calibration data in the echo data; Generate the reference trajectory of each band using the position and attitude data; Obtain the phase center position of each band antenna based on the installation position of each band antenna; Obtain the reference drift angle of each band based on the reference trajectory of each band; Select the reference band, and obtain the distance offset D of each band relative to the reference band based on the phase center position of each band antenna, the reference trajectory and the reference drift angle of each band. 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}; where num is the number of bands in the system, 1≤i≤num-1, 1≤r≤num, I r is the echo data corresponding to band r, D i is the distance offset of non-reference band i relative to the reference band, T i is the azimuth offset of the non-reference band i relative to the reference band;

[0033] The distance pulse pressure data acquisition module is configured to obtain the distance pulse pressure data for each of the distance offsets D relative to the reference band. i , perform the following operations: perform range pulse compression processing on the echo signal, based on D i The corresponding band distance spectrum bandwidth, sampling delay correction value, and D i , perform range signal resampling, range spectrum bandpass filtering and range signal interception; obtain the processed range pulse pressure data, denoted as D i ';

[0034] The pulse pressure data azimuth processing module is configured to calculate the azimuth offset T relative to the reference band for each T i , perform the following operations: perform motion compensation and Stolt interpolation on the processed distance pulse pressure data, based on the T i The azimuth spectrum bandwidth, pulse repetition frequency, and T of the corresponding bandi , perform azimuth resampling, spectrum bandpass filtering and azimuth signal interception; obtain the processed SAR image, denoted as T i ';

[0035] SAR image registration accuracy measurement module: configured to use calibration points to check the registration accuracy between each band image and the reference band image;

[0036] Judgment module: If the registration accuracy is greater than 1 pixel, the calibration points are used to correct the range and azimuth offsets of each non-reference band image relative to the reference band, triggering the range pulse pressure data acquisition module; if the registration accuracy is less than or equal to 1 pixel, the registration ends.

[0037] Preferably, the data acquisition module includes:

[0038] Correction submodule: configured to calculate the sampling delay correction value of each band according to the internal calibration data in the echo data;

[0039] Center trajectory acquisition submodule: configured to project the longitude and latitude position coordinates in the position and attitude data into Gaussian plane coordinates, and then perform linear fitting on the Gaussian plane coordinates of the position to generate the reference trajectory and reference drift angle of each band; based on the installation position of the antenna of each band, obtain the phase center position of the antenna of each band; based on the phase center position of the antenna of each band, perform coordinate rotation on the actual flight trajectory and reference trajectory of each band to generate the antenna phase center trajectory of each band;

[0040] Offset acquisition submodule: configured to calculate the distance offset D of each band relative to the reference band based on the antenna phase center trajectory and reference drift angle of each band and the position vector of the antenna phase center of each band. 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}.

[0041] Preferably, the distance pulse pressure data acquisition module comprises:

[0042] Range pulse compression: configured to utilize D i The echo data of the corresponding band and D i The range reference function of the corresponding band completes the range pulse compression processing;

[0043] Resampling submodule: Configured according to D i The minimum sampling frequency of D iThe distance pulse pressure data of the corresponding band is resampled in the distance direction;

[0044] Frequency domain bandpass filter submodule: configured to estimate D i The distance of the corresponding band is the range of the spectrum bandwidth, with D i The minimum value of the distance to spectrum bandwidth in the band is D i The distance of the corresponding band is subjected to frequency domain bandpass filtering on the pulse pressure data;

[0045] Distance pulse pressure submodule: configured according to D i The sampling delay correction value of the corresponding band, system sampling delay, D i The range offset and range beam width of the corresponding band are i The distance pulse pressure data of the corresponding band is intercepted to obtain the processed distance pulse pressure data.

[0046] Preferably, the pulse pressure data position processing module comprises:

[0047] The second correction submodule is configured to select the oblique angle of the reference band and calculate T i The Doppler center frequency of the corresponding band is i The processed distance pulse pressure data of the corresponding band is subjected to motion compensation and Stolt interpolation processing;

[0048] Azimuth processing submodule: configured according to T i The minimum PRF value for T i The data of the corresponding band is resampled in azimuth and the azimuth pulse compression process is completed;

[0049] Frequency domain bandpass filter submodule: configured to estimate T i The azimuth spectrum width of the corresponding band ranges in T i The minimum value of the azimuth spectrum width of the corresponding band is T i The data of the corresponding band is subjected to frequency domain bandpass filtering;

[0050] Intercept submodule: configured to adjust the T according to the azimuth offset i The data of the corresponding band is used to intercept the azimuth signal.

[0051] Preferably, the SAR image registration accuracy measurement module includes:

[0052] Coarse positioning submodule: configured to use the geographic coordinates of the calibration point and image auxiliary data to perform coarse positioning of the calibration point according to the range Doppler model;

[0053] Estimation submodule: configured to perform sinc interpolation on the image within a preset range around the calibration point to estimate the sub-pixel level calibration point position;

[0054] Coordinate difference determination submodule: configured to calculate the coordinate difference of the calibration point between each non-reference band and the reference band image.

[0055] Beneficial effects:

[0056] (1) The present invention analyzes the parameter differences of the antenna phase center position vector, sampling frequency, bandwidth and PRF during the imaging process of each band, performs consistency processing on the signal during the imaging process, realizes multi-band consistent imaging, and greatly reduces the geometric distortion between the images of each band;

[0057] (2) After the multi-band consistency imaging, the present invention can recalculate the offset based on a small number of calibration points, and then perform iterative optimization processing to effectively improve the registration accuracy between the multiple bands. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the flow chart of the multi-dimensional synthetic aperture radar method of the present invention;

[0059] Figure 2 It is a schematic flow chart of the multi-dimensional synthetic aperture radar refinement method of the present invention;

[0060] Figure 3 Schematic diagram of different antenna center positions of the present invention;

[0061] Figure 4 It is a schematic diagram of the structure of the multi-dimensional synthetic aperture radar device of the present invention. DETAILED DESCRIPTION

[0062] The present invention is described in detail below in conjunction with the accompanying drawings and embodiments.

[0063] The present invention discloses a multi-dimensional synthetic aperture radar registration method and device, which obtains system parameters of different bands, performs spatiotemporal consistency processing in the imaging process based on the spatial geometric relationship of antenna phase centers, and realizes multi-dimensional SAR multi-band image registration.

[0064] like Figure 1-2 As shown, the multi-dimensional synthetic aperture radar registration method includes:

[0065] Step S1: Obtain a multi-band echo data set I={I 1 ,…,I r ,…,I num}; Calculate the sampling delay correction value of each band using the internal calibration data in the echo data; Generate the reference trajectory of each band using the position and attitude data; Obtain the phase center position of each band antenna based on the installation position of each band antenna; Obtain the reference drift angle of each band based on the reference trajectory of each band; Select the reference band, and obtain the distance offset D of each band relative to the reference band based on the phase center position of each band antenna, the reference trajectory and the reference drift angle of each band. 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}; where num is the number of bands in the system, 1≤i≤num-1, 1≤r≤num, I r is the echo data corresponding to band r, D i is the distance offset of non-reference band i relative to the reference band, T i is the azimuth offset of the non-reference band i relative to the reference band;

[0066] Step S2: for each of the distance offsets D relative to the reference band i , perform the following operations: perform range pulse compression processing on the echo signal, based on D i The corresponding band distance spectrum bandwidth, sampling delay correction value, and D i The parameters are then used to perform range signal resampling, range spectrum bandpass filtering and range signal interception. The processed range pulse pressure data is obtained, which is recorded as D i ';

[0067] Step S3: for each of the azimuth offsets T relative to the reference band i , perform the following operations: perform motion compensation and Stolt interpolation on the processed distance pulse pressure data, based on the T i The azimuth spectrum bandwidth, pulse repetition frequency, and T of the corresponding band i Parameters such as azimuth resampling, spectrum bandpass filtering and azimuth signal interception are performed; the processed SAR image is obtained, denoted as T i ';

[0068] Step S4: using calibration points to check the registration accuracy between each band image and the reference band image;

[0069] Step S5: If the registration accuracy is greater than 1 pixel, the distance offset and azimuth offset of each non-reference band image relative to the reference band are corrected using the calibration points.i Updated to D i1 , T i Update to T i1 ; Enter step S2; If the registration accuracy is less than or equal to 1 pixel, the method ends.

[0070] The present invention uses the multi-band images acquired by the multi-dimensional SAR system in one flight, performs signal processing such as bandpass filtering based on the spatial geometric relationship of the antenna phase center and the system parameters of each band, and realizes multi-band SAR image registration. The basic idea of ​​the present invention is to achieve the purpose of multi-dimensional SAR consistent imaging processing through time consistency processing and spatial consistency processing, so as to achieve image registration accuracy better than 1 pixel after multi-band imaging.

[0071] The step S1 comprises:

[0072] Step S11: Calculate the sampling delay correction value of each band according to the internal calibration data in the echo data;

[0073] Step S12: Project the longitude and latitude position coordinates in the position and attitude data into Gaussian plane coordinates, and then perform linear fitting on the Gaussian plane coordinates of the position to generate reference trajectories and reference drift angles of each band; based on the installation positions of the antennas of each band, obtain the phase center positions of the antennas of each band; based on the phase center positions of the antennas of each band, perform coordinate rotation on the actual flight trajectory and reference trajectory of each band to generate the antenna phase center trajectory of each band;

[0074] Step S13: According to the antenna phase center trajectory of each band, the reference bias angle, etc., using the position vector of the antenna phase center of each band, calculate the distance offset D of each band relative to the reference band = {D 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}.

[0075] The step S2 comprises:

[0076] Step S21: Using D i The echo data of the corresponding band and D i The range reference function of the corresponding band completes the range pulse compression processing;

[0077] Step S22: According to D i The minimum sampling frequency of D i The distance pulse pressure data of the corresponding band is resampled in the distance direction;

[0078] Step S23: Estimate D i The distance of the corresponding band is the range of the spectrum bandwidth, with D i The minimum value of the distance to spectrum bandwidth in the band is D i The distance of the corresponding band is subjected to frequency domain bandpass filtering on the pulse pressure data;

[0079] Step S24: According to D i The sampling delay correction value of the corresponding band, system sampling delay, D i The range offset and range beam width of the corresponding band are i The distance pulse pressure data of the corresponding band is intercepted to obtain the processed distance pulse pressure data.

[0080] Furthermore, in this embodiment, according to the system sampling delay t of each band, i , sampling delay correction value Δt i anddy i , calculate the initial slant range R of each band i The calculation formula is as follows. Then the range pulse data is intercepted according to the minimum value of the range spectrum bandwidth.

[0081]

[0082] Where c is the speed of light, dy i is the distance offset.

[0083] The step S3 comprises:

[0084] Step S31: Select the oblique angle of the reference band and calculate T i The Doppler center frequency of the corresponding band is i The processed distance pulse pressure data of the corresponding band is subjected to motion compensation and Stolt interpolation processing;

[0085] Step S32: According to T i The minimum PRF value for T i The data of the corresponding band is resampled in azimuth and the azimuth pulse compression process is completed;

[0086] Step S33: Estimate T i The azimuth spectrum width of the corresponding band ranges in T i The minimum value of the azimuth spectrum width of the corresponding band is T i The data of the corresponding band is subjected to frequency domain bandpass filtering;

[0087] Step S34: According to the azimuth offset, i The data of the corresponding band is used to intercept the azimuth signal.

[0088] The step S4 comprises:

[0089] Step S41: using the geographic coordinates of the calibration point and image auxiliary data, the calibration point is roughly positioned according to the range Doppler model;

[0090] Step S42: performing sinc interpolation on the image in a preset range around the calibration point to estimate the position of the sub-pixel calibration point;

[0091] Step S43: Calculate the coordinate difference of the calibration point between each non-reference band image and the reference band image.

[0092] The step S5 comprises:

[0093] Calculate the coordinate difference of the calibration point of each non-reference band image relative to the reference band image, and convert it into the distance offset D of each non-reference band image relative to the reference band. i1 and the azimuth offset T i1 , D i Updated to D i1 , T i Update to T i1 .

[0094] The following embodiment of the present invention provides a multi-dimensional synthetic aperture radar registration method, the method comprising the following steps:

[0095] Step S101: For each band, extract the internal calibration signal from the radar echo, perform pulse compression and other processing on the internal calibration signal, and calculate the sampling delay correction value Δt i (i=1,2,3…,num).

[0096] Step S102: According to the start and end times of the baseband echo data, the longitude and latitude coordinates of the IMU in the POS are Gaussian projected, and the Gaussian coordinates of the actual flight trajectory are fitted with the least squares linear method to calculate the reference trajectory and the reference drift angle.

[0097] Step S103: Use the reference drift angle to rotate the actual flight trajectory and the reference trajectory, and calculate the attitude angle (yaw angle θ) at each sampling moment according to the IMU trajectory coordinates in the imaging coordinate system, the antenna phase center coordinates in the carrier coordinate system, and the attitude angle (yaw angle θ y , pitch angle θ p and roll angle θ r ), calculate the antenna phase center trajectory coordinates in the imaging coordinate system.

[0098] Step S104: Calculate the offset components of the initial trajectory coordinates along the azimuth and range directions according to the antenna phase center trajectory coordinates and attitude angles in the imaging coordinate system of each band, and record them as dx i,dy i .like Figure 3 As shown, A and B represent the positions of the phase centers of the two antennas, and their coordinates are divided into (x 1 ,y 1 ) and (x 2 ,y 2 ), the calculation is shown below.

[0099]

[0100] Step S105: For each band, perform Fourier transform on the baseband echo data and transform it into the range frequency domain. According to the system parameters such as sampling frequency, sampling time width and modulation frequency, calculate the frequency domain matched filter, multiply it with the transformed echo data, obtain the range pulse compression data, and estimate the spectrum width of the range frequency domain.

[0101] Step S106: performing distance resampling on the distance pulse pressure data of each band according to the minimum value of the sampling frequency of different bands; performing distance spectrum bandpass filtering on each band with the minimum spectrum width of each band as the threshold.

[0102] Step S107: Delay the sampling of each band according to t i , sampling delay correction value Δt i anddy i , calculate the initial slant range R of each band i The calculation formula is as follows. Then the range pulse data is intercepted according to the minimum range beam width.

[0103]

[0104] Step S108: For each band, calculate its Doppler center f according to the following formula: i .

[0105]

[0106] Where v represents the flight speed, θ y and θ r They represent the reference squint angle of the reference band, λ i Indicates the radar wavelength of each band.

[0107] Step S109: for each band, the first-order motion compensation and the second-order motion compensation are calculated respectively using parameters such as the Doppler center and the antenna phase center trajectory, and the second-order motion compensation is performed on the processed range pulse compression data.

[0108] Step S110: performing reference function multiplication and Stolt interpolation in the two-dimensional frequency domain.

[0109] Step S111: for each band, generate a complex conjugate function of the azimuth matched filter and multiply it with the data to complete the azimuth pulse compression processing.

[0110] Step S112: Select the minimum value of the pulse repetition frequency of different bands, resample the data in azimuth for each band, and estimate the azimuth spectrum width. Step S113: Select the minimum value of the azimuth spectrum width of different bands as the threshold, and perform bandpass filtering based on the Doppler center of each band. i Perform azimuth data capture to generate focused SLC, amplitude images and auxiliary data.

[0111] Step S114: select a calibration point on the image of each band, perform a rough positioning of the calibration point using a range Doppler model according to image auxiliary data such as the Doppler center and radar wavelength, and obtain the initial image coordinates of the calibration point in each band.

[0112] Step S115: Perform peak search within 128 pixels around the initial image coordinates to obtain the sub-pixel image coordinates of the calibration point in each band. Calculate the coordinate difference between the calibration point image coordinates and the reference band image;

[0113] Step S116: If it is less than 1 pixel, then end; if it is greater than 1 pixel, then convert it to the offset components in the azimuth and distance directions relative to the reference band according to the image pixel size, and record them as dx i ,dy i , then go to step S105 and re-image.

[0114] In this embodiment, based on the above-mentioned multi-dimensional synthetic aperture radar data consistency imaging processing method, multiple band images with the same resolution can be obtained, and the inter-band registration accuracy is better than 1 pixel. This embodiment provides a multi-dimensional synthetic aperture radar data consistency imaging processing method, which achieves the purpose of multi-band consistency imaging by performing sampling delay correction, APC position error elimination, spectrum filtering and other processing during the imaging process, and the multi-band image registration accuracy after imaging is better than 1 pixel.

[0115] like Figure 4 As shown, the present invention provides a multi-dimensional synthetic aperture radar registration device, the device comprising:

[0116] Data acquisition module: configured to acquire a multi-band echo data set I={I 1 ,…,I r ,…,I num}; Calculate the sampling delay correction value of each band using the internal calibration data in the echo data; Generate the reference trajectory of each band using the position and attitude data; Obtain the phase center position of each band antenna based on the installation position of each band antenna; Obtain the reference drift angle of each band based on the reference trajectory of each band; Select the reference band, and obtain the distance offset D of each band relative to the reference band based on the phase center position of each band antenna, the reference trajectory and the reference drift angle of each band. 1 ,…,D i , …, D num-1}, and the azimuth offset of each band relative to the reference band T = {T 1 ,…,T i ,…,T num-1}; where num is the number of bands in the system, 1≤i≤num-1, 1≤r≤num, I r is the echo data corresponding to band r, D i is the distance offset of non-reference band i relative to the reference band, T i is the azimuth offset of the non-reference band i relative to the reference band;

[0117] The distance pulse pressure data acquisition module is configured to obtain the distance pulse pressure data for each of the distance offsets D relative to the reference band. i , perform the following operations: perform range pulse compression processing on the echo signal, based on D i The corresponding band distance spectrum bandwidth, sampling delay correction value, and D i , perform range signal resampling, range spectrum bandpass filtering and range signal interception; obtain the processed range pulse pressure data, denoted as D i ';

[0118] The pulse pressure data azimuth processing module is configured to calculate the azimuth offset T relative to the reference band for each T i , perform the following operations: perform motion compensation and Stolt interpolation on the processed distance pulse pressure data, based on the T i The azimuth spectrum bandwidth, pulse repetition frequency, and T of the corresponding band i , perform azimuth resampling, spectrum bandpass filtering and azimuth signal interception; obtain the processed SAR image, denoted as T i ';

[0119] SAR image registration accuracy measurement module: configured to use calibration points to check the registration accuracy between each band image and the reference band image;

[0120] Judgment module: If the registration accuracy is greater than 1 pixel, the calibration points are used to correct the range and azimuth offsets of each non-reference band image relative to the reference band, triggering the range pulse pressure data acquisition module; if the registration accuracy is less than or equal to 1 pixel, the registration ends.

[0121] The above specific embodiments only describe the design principle of the present invention. The shapes and names of the components in the description may be different and are not limited. Therefore, those skilled in the art in the field of the present invention may modify or replace the technical solutions recorded in the above embodiments; and these modifications and replacements do not deviate from the creative purpose and technical solutions of the present invention and should all fall within the protection scope of the present invention.

Claims

1. A multi-dimensional synthetic aperture radar registration method, characterized in that: include: Step S1: Obtain a multi-band echo data set I = {I1, ..., I r ,…,I num }; Calculate the sampling delay correction value of each band using the internal calibration data in the echo data; Generate the reference trajectory of each band using the position and attitude data; Obtain the phase center position of each band antenna based on the installation position of each band antenna; Obtain the reference drift angle of each band based on the reference trajectory of each band; Select a reference band, and based on the phase center position of each band antenna, the reference trajectory and the reference drift angle of each band, obtain the range offset D of each band relative to the reference band = {D1, ..., D i , …, D num-1 }, and the azimuth offset of each band relative to the reference band T = {T1, ..., T i ,…,T num-1 }; Where num is the number of bands in the system, 1≤i≤num-1, 1≤r≤num, I r is the echo data corresponding to band r, D i is the distance offset of non-reference band i relative to the reference band, T i is the azimuth offset of the non-reference band i relative to the reference band; Step S2: for each of the distance offsets D relative to the reference band i , perform the following operations: perform range pulse compression processing on the echo signal, based on D i The corresponding band distance spectrum bandwidth, sampling delay correction value, and D i , perform range signal resampling, range spectrum bandpass filtering and range signal interception; obtain the processed range pulse pressure data, denoted as D i '; Step S3: for each of the azimuth offsets T relative to the reference band i , perform the following operations: perform motion compensation and Stolt interpolation on the processed distance pulse pressure data, based on the T i The azimuth spectrum bandwidth, pulse repetition frequency, and T of the corresponding band i , perform azimuth resampling, spectrum bandpass filtering and azimuth signal interception; obtain the processed SAR image, denoted as T i '; Step S4: using calibration points to check the registration accuracy between each band image and the reference band image; Step S5: If the registration accuracy is greater than 1 pixel, the distance offset and azimuth offset of each non-reference band image relative to the reference band are corrected using the calibration points, and the process proceeds to step S2; if the registration accuracy is less than or equal to 1 pixel, the method ends; The step S1 comprises: Step S11: Calculate the sampling delay correction value of each band according to the internal calibration data in the echo data; Step S12: Project the longitude and latitude position coordinates in the position and attitude data into Gaussian plane coordinates, and then perform linear fitting on the Gaussian plane coordinates of the position to generate reference trajectories and reference drift angles of each band; based on the installation positions of the antennas of each band, obtain the phase center positions of the antennas of each band; based on the phase center positions of the antennas of each band, perform coordinate rotation on the actual flight trajectory and reference trajectory of each band to generate the antenna phase center trajectory of each band; Step S13: According to the antenna phase center trajectory of each band and the reference bias angle, using the position vector of the antenna phase center of each band, calculate the distance offset D of each band relative to the reference band = {D1, ..., D i , …, D num-1 }, and the azimuth offset of each band relative to the reference band T = {T1, ..., T i ,…,T num-1 }; The step S2 comprises: Step S21: Using D i The echo data of the corresponding band and D i The range reference function of the corresponding band completes the range pulse compression processing; Step S22: According to D i The minimum sampling frequency of D i The distance pulse pressure data of the corresponding band is resampled in the distance direction; Step S23: Estimate D i The distance of the corresponding band is the range of the spectrum bandwidth, with D i The minimum value of the distance to spectrum bandwidth in the band is D i The distance of the corresponding band is subjected to frequency domain bandpass filtering on the pulse pressure data; Step S24: According to D i The sampling delay correction value of the corresponding band, system sampling delay, D i The range offset and range beam width of the corresponding band are i The distance pulse pressure data of the corresponding band is intercepted to obtain the processed distance pulse pressure data.

2. The method according to claim 1, characterized in that The step S3 comprises: Step S31: Select the oblique angle of the reference band and calculate T i The Doppler center frequency of the corresponding band is i The processed distance pulse pressure data of the corresponding band is subjected to motion compensation and Stolt interpolation processing; Step S32: According to T i The minimum PRF value for T i The data of the corresponding band is resampled in azimuth and the azimuth pulse compression process is completed; Step S33: Estimate T i The azimuth spectrum width of the corresponding band ranges in T i The minimum value of the azimuth spectrum width of the corresponding band is T i The data of the corresponding band is subjected to frequency domain bandpass filtering; Step S34: According to the azimuth offset, i The data of the corresponding band is used to intercept the azimuth signal.

3. The method according to claim 2, characterized in that The step S4 comprises: Step S41: using the geographic coordinates of the calibration point and image auxiliary data, the calibration point is roughly positioned according to the range Doppler model; Step S42: performing sinc interpolation on the image in a preset range around the calibration point to estimate the position of the sub-pixel calibration point; Step S43: Calculate the coordinate difference of the calibration point between each non-reference band image and the reference band image.

4. A multi-dimensional synthetic aperture radar registration device, characterized in that: The device comprises: Data acquisition module: configured to acquire a multi-band echo data set I = {I1, ..., I r ,…,I num }; Calculate the sampling delay correction value of each band using the internal calibration data in the echo data; Generate the reference trajectory of each band using the position and attitude data; Obtain the phase center position of each band antenna based on the installation position of each band antenna; Obtain the reference drift angle of each band based on the reference trajectory of each band; Select the reference band, and obtain the distance offset D of each band relative to the reference band based on the phase center position of each band antenna, the reference trajectory and the reference drift angle of each band. i , …, D num-1 }, and the azimuth offset of each band relative to the reference band T = {T1, ..., T i ,…,T num-1 }; where num is the number of bands in the system, 1≤i≤num-1, 1≤r≤num, I r is the echo data corresponding to band r, D i is the distance offset of non-reference band i relative to the reference band, T i is the azimuth offset of the non-reference band i relative to the reference band; The distance pulse pressure data acquisition module is configured to obtain the distance pulse pressure data for each of the distance offsets D relative to the reference band. i , perform the following operations: perform range pulse compression processing on the echo signal, based on D i The corresponding band distance spectrum bandwidth, sampling delay correction value, and D i , perform range signal resampling, range spectrum bandpass filtering and range signal interception; obtain the processed range pulse pressure data, denoted as D i '; The pulse pressure data azimuth processing module is configured to calculate the azimuth offset T relative to the reference band for each T i , perform the following operations: perform motion compensation and Stolt interpolation on the processed distance pulse pressure data, based on T i The azimuth spectrum bandwidth, pulse repetition frequency, and T of the corresponding band i , perform azimuth resampling, spectrum bandpass filtering and azimuth signal interception; obtain the processed SAR image, denoted as T i '; SAR image registration accuracy measurement module: configured to use calibration points to check the registration accuracy between each band image and the reference band image; Judgment module: If the registration accuracy is greater than 1 pixel, the calibration points are used to correct the range and azimuth offsets of each non-reference band image relative to the reference band, triggering the range pulse pressure data acquisition module; if the registration accuracy is less than or equal to 1 pixel, the registration ends.

5. The device according to claim 4, characterized in that The data acquisition module comprises: Correction submodule: configured to calculate the sampling delay correction value of each band according to the internal calibration data in the echo data; Center trajectory acquisition submodule: configured to project the longitude and latitude position coordinates in the position and attitude data into Gaussian plane coordinates, and then perform linear fitting on the Gaussian plane coordinates of the position to generate the reference trajectory and reference drift angle of each band; based on the installation position of the antenna of each band, obtain the phase center position of the antenna of each band; based on the phase center position of the antenna of each band, perform coordinate rotation on the actual flight trajectory and reference trajectory of each band to generate the antenna phase center trajectory of each band; Offset acquisition submodule: configured to calculate the distance offset D = {D1, ..., D i , …, D num-1 }, and the azimuth offset of each band relative to the reference band T = {T1, ..., T i ,…,T num-1 }.

6. The device according to claim 5, characterized in that The distance pulse pressure data acquisition module comprises: Range pulse compression: configured to utilize D i The echo data of the corresponding band and D i The range reference function of the corresponding band completes the range pulse compression processing; Resampling submodule: Configured according to D i The minimum sampling frequency of D i The distance pulse pressure data of the corresponding band is resampled in the distance direction; Frequency domain bandpass filter submodule: configured to estimate D i The distance of the corresponding band is the range of the spectrum bandwidth, with D i The minimum value of the distance to spectrum bandwidth in the band is D i The distance of the corresponding band is subjected to frequency domain bandpass filtering on the pulse pressure data; Distance pulse pressure submodule: configured according to D i The sampling delay correction value of the corresponding band, system sampling delay, D i The range offset and range beam width of the corresponding band are i The distance pulse pressure data of the corresponding band is intercepted to obtain the processed distance pulse pressure data.

7. The device according to claim 6, characterized in that The pulse pressure data position processing module comprises: The second correction submodule is configured to select the oblique angle of the reference band and calculate T i The Doppler center frequency of the corresponding band is i The processed distance pulse pressure data of the corresponding band is subjected to motion compensation and Stolt interpolation processing; Azimuth processing submodule: configured according to T i The minimum PRF value for T i The data of the corresponding band is resampled in azimuth and the azimuth pulse compression process is completed; Frequency domain bandpass filter submodule: configured to estimate T i The azimuth spectrum width of the corresponding band ranges in T i The minimum value of the azimuth spectrum width of the corresponding band is T i The data of the corresponding band is subjected to frequency domain bandpass filtering; Intercept submodule: configured to adjust the T according to the azimuth offset i The data of the corresponding band is used to intercept the azimuth signal.

8. The device according to claim 7, characterized in that The SAR image registration accuracy measurement module comprises: Coarse positioning submodule: configured to use the geographic coordinates of the calibration point and image auxiliary data to perform coarse positioning of the calibration point according to the range Doppler model; Estimation submodule: configured to perform sinc interpolation on the image within a preset range around the calibration point to estimate the sub-pixel level calibration point position; Coordinate difference determination submodule: configured to calculate the coordinate difference of the calibration point between each non-reference band and the reference band image.

Images