A method for improving elevation resolution based on array 3D SAR motion error

Abstract

The present invention provides a method for improving elevation resolution based on array three-dimensional SAR motion error, which can directly use the array three-dimensional SAR non-ideal linear motion error to increase the physical effective baseline length, thereby improving elevation resolution. The present invention uses the non-ideal linear motion error to increase the original physical effective baseline length, and simultaneously performs two-dimensional joint modeling of the observation scene in azimuth and altitude, and realizes scene reconstruction through two-dimensional joint solution, thereby improving the Rayleigh resolution in elevation without losing the Rayleigh resolution in azimuth and the Rayleigh resolution in range.

Description

Technical Field

[0001] The present invention relates to the field of radar processing technology, and in particular to a method for improving elevation resolution based on array three-dimensional SAR motion error. Background Art

[0002] Array 3D SAR can obtain a 3D image of a scene in a single flight, which makes up for the shortcomings of optical sensors. It can observe targets all day and all weather, and can obtain more ground information based on its own microwave characteristics. It has long been studied in depth by experts and scholars in related fields at home and abroad. The higher resolution of array 3D SAR in elevation is achieved by adding multiple antennas in sequence in the elevation direction to form an antenna array. Its elevation resolution is determined by the length of the elevation baseline. The number of baselines determines the number of targets that can be distinguished, and the baseline length and signal-to-noise ratio determine the accuracy of elevation measurement. However, in actual systems, the number of baselines is generally small, and most baselines are non-uniformly distributed, and in some cases are even time-varying baselines, so the elevation resolution is difficult to improve due to system limitations.

[0003] In order to improve the elevation resolution of array 3D SAR, domestic scholars have proposed different methods. In a published paper, “Synthetic Aperture Radar Three-Dimensional Imaging: From Tomography, Array to Microwave Vision [J]. Journal of Radar, 2019, 8(6).”, it is proposed that the elevation resolution capability of array interferometric SAR comes from angle resolution. According to radar resolution theory, the resolution is inversely proportional to the antenna size. The higher the resolution, the longer the antenna size. To achieve a 1m 3D imaging resolution, an airborne radar antenna of about 20 to 50m is required, which is difficult to achieve in engineering and has become a core problem restricting high-resolution 3D imaging. In another published paper, “3D surface reconstruction of layover areas in continuous terrain for multi-baseline SAR interferometry using a curve model [J]. International Journal of Remote Sensing, 2015, 36(7-8): 2093-2112. A super-resolution algorithm based on curve model constraint is proposed in this paper. The continuous natural terrain is expressed as a piecewise curve model, and the traditional target resolution imaging problem based on correlation matched filtering is transformed into the optimal estimation problem of curve model parameters. This method realizes three-dimensional super-resolution imaging under the condition of small-size antenna, but does not use motion error to improve elevation resolution.

[0004] In actual situations, there will be errors in the movement of the carrier aircraft. When the carrier aircraft performs non-ideal linear motion, the baseline also performs non-ideal linear motion, thereby extending the physical effective baseline length in space, which is the motion error in the traditional sense. However, the current research on motion errors is all about compensating for them, and there are few studies on using non-ideal linear motion errors to improve elevation accuracy. Due to the limitations of the carrier platform, the cross-heading baseline length of the array 3D SAR is short and the elevation accuracy is limited. Therefore, the motion error can be used to simultaneously perform two-dimensional joint modeling of the observation scene in azimuth and altitude, and the scene reconstruction can be achieved through two-dimensional joint solution, thereby improving the elevation accuracy. So far, there has been no report on the direct application of this technology to improve the elevation accuracy of array 3D SAR. Summary of the invention

[0005] In view of this, the present invention provides a method for improving elevation resolution based on array three-dimensional SAR motion error, which can directly use the array three-dimensional SAR non-ideal linear motion error to increase the physical effective baseline length, thereby improving elevation resolution.

[0006] To achieve the above object, the technical solution of the present invention is:

[0007] A method for improving elevation resolution based on array three-dimensional SAR motion error of the present invention comprises the following steps:

[0008] Step 1, determining the non-ideal linear motion trajectory of the array 3D SAR according to the POS data, includes the following sub-steps:

[0009] Step 11, determining the POS data range within the synthetic aperture corresponding to the scene according to the scene information;

[0010] Step 12, adjusting the sampling interval of the POS data within this range so that it can describe the changes of non-ideal linear motion while meeting the requirements corresponding to the number of scene points;

[0011] Step 13, scaling the POS data within the sampling range accordingly according to the scene grid point size;

[0012] Step 2, determining the location of the scene grid points, wherein the azimuth grid spacing is calculated according to the azimuth Rayleigh resolution, and the height grid spacing is smaller than the height Rayleigh resolution, thereby achieving height super-resolution processing;

[0013] Step 3, generating observation equations according to the motion trajectory and scene grid position;

[0014] Step 4: Substitute the radar data into the observation equation, solve the observation equation under the constraint of target sparsity, and obtain the reconstruction result.

[0015] Wherein, the step 2 includes the following sub-steps:

[0016] Step 21, determining the Rayleigh resolution in azimuth and the Rayleigh resolution in elevation according to the non-ideal linear motion trajectory;

[0017] Step 22, adjusting the sampling interval of the scattering points of the observation scene according to the Rayleigh resolution of the non-ideal linear motion in the elevation direction, so that the interval of the scattering points of the observation scene in the elevation direction is smaller than the Rayleigh resolution of the non-ideal linear motion in the elevation direction;

[0018] Step 23, according to the azimuth Rayleigh resolution of the non-ideal linear motion, the interval of the scattering points of the observed scene in azimuth is made equal to the azimuth Rayleigh resolution of the non-ideal linear motion.

[0019] Wherein, the step 4 includes the following sub-steps:

[0020] Step 41, determining the elevation unambiguous range according to the bandwidth of the non-ideal linear motion trajectory and the number of channels;

[0021] Step 42, adjusting the sparse constraint of the observed scene target according to the unambiguous range until the requirement is met;

[0022] Step 43, finally solving the observation equation.

[0023] In step 11, the POS data range in the synthetic aperture corresponding to the scene is determined according to the scene information, and the number N of cross-heading channels and the bandwidth B of the array 3D SAR are determined according to the radar parameters. r , beam center viewing angle θ, beam width And the scene's closest slant distance R min , determine the number of sampling points M in azimuth according to the processing range; calculate the aperture length formula L according to the beam width of the transmitted signal of the array 3D SAR system sar =R min *θ*3.14 / 180, where L sar is the subaperture length, according to X = x end -x 1 Calculate the total aperture length, where X is the total aperture length, x 1 and x end It is the coordinate value corresponding to the start and end of the azimuth direction in the radar POS data. The number of sub-apertures N = X / L sar , and then determine the position corresponding to POS by the azimuth coordinates corresponding to the scene.

[0024] Among them, the observation equation A=exp(-2jπ(a*F a +s*F s), where a is the horizontal coordinate of the observation scene grid, s is the vertical coordinate of the observation scene grid, and F a is the horizontal coordinate of the scene observation space spectrum, F s is the ordinate of the scene observation space spectrum.

[0025] Beneficial effects:

[0026] 1. The present invention utilizes the non-ideal linear motion error to increase the original physical effective baseline length, and simultaneously performs two-dimensional joint modeling of the observation scene in azimuth and altitude, and realizes scene reconstruction through two-dimensional joint solution, thereby improving the Rayleigh resolution in elevation without losing the Rayleigh resolution in azimuth and the Rayleigh resolution in range.

[0027] 2. The present invention determines the elevation unambiguous range according to the bandwidth and the number of channels of the non-ideal linear motion trajectory; adjusts the sparse constraints of the observed scene targets according to the unambiguous range until the requirements are met, and finally solves the observation equation. Under the premise of not losing the performance of the target, the target point position can be correctly reconstructed under both the ideal linear trajectory and the non-ideal linear trajectory.

[0028] 3. In the present invention, the number of cross-direction channels N and the bandwidth B of the array three-dimensional SAR are determined according to radar parameters. r , beam center viewing angle θ, beam width And the scene's closest slant distance R min The number of azimuth sampling points M is determined according to the processing range. The model parameters are relatively simple and the model can be applied to a variety of situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 is a flow chart of the method of the present invention.

[0030] Figure 2 It is the ideal straight line flight trajectory diagram of the X-band array three-dimensional SAR system.

[0031] Figure 3 It is a non-ideal straight line flight trajectory diagram of the X-band array three-dimensional SAR system.

[0032] Figure 4 It is an information map of the scattering point locations in the X-band observation scene.

[0033] Figure 5 This is a comparison chart of the reconstruction results of ideal straight line and non-ideal straight line flight trajectories under the same signal-to-noise ratio loss.

[0034] Figure 6 This is a graph of the reconstruction accuracy of ideal straight line and non-ideal straight line flight trajectories with different bandwidths under the same signal-to-noise ratio loss.

[0035] Figure 7This is a graph of the reconstruction accuracy of ideal straight line and non-ideal straight line flight trajectories under different signal-to-noise ratio losses.

[0036] Figure 8 This is a graph of the reconstruction accuracy of ideal straight line and non-ideal straight line flight trajectories under different numbers of array elements under the same signal-to-noise ratio loss.

[0037] Fig. 9 It is the result of reconstructing three-dimensional imaging of a building in X-band using the method of the present invention. DETAILED DESCRIPTION

[0038] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0039] The present invention provides a method for improving elevation resolution by utilizing the non-ideal linear motion error of array three-dimensional SAR. Different from the conventional SAR that compensates for the non-ideal linear motion error, the present invention utilizes the non-ideal linear motion error to increase the original physical effective baseline length, and simultaneously performs two-dimensional joint modeling of the observation scene in azimuth and altitude, and realizes scene reconstruction through two-dimensional joint solution. Among them, the motion error of non-ideal linear motion can increase the physical effective baseline length, which can break through the traditional elevation Rayleigh resolution that does not utilize motion error. The implementation steps of the present invention are as follows:

[0040] Step 1, determining the non-ideal linear motion trajectory of the array 3D SAR according to the POS data, includes the following sub-steps:

[0041] Step 11, determining the POS data range within the synthetic aperture corresponding to the scene according to the scene information;

[0042] In this embodiment, the number N of cross-heading channels and the bandwidth B of the array 3D SAR are determined according to the radar parameters. r , beam center viewing angle θ, beam width And the scene's closest slant distance R min , determine the number of azimuth sampling points M according to the processing range, and the specific parameters are as follows:

[0043] Table 1 Parameter settings of this embodiment

[0044]

[0045] The aperture length formula is calculated based on the beam width of the transmitted signal of the array 3D SAR system: L sar =R min *θ*3.14 / 180, where L sar is the subaperture length, according to X = x end -x 1 Calculate the total aperture length, where X is the total aperture length, x 1 and x endIt is the coordinate value corresponding to the start and end of the azimuth in the radar POS data, and finally uses N = X / L sar The number of sub-apertures is calculated, and then the position corresponding to the POS is determined by the azimuth coordinates corresponding to the scene.

[0046] Step 12, adjust the sampling interval of the POS data within this range so that it can describe the changes in non-ideal linear motion while meeting the requirements corresponding to the number of scene points; in this embodiment, each channel of the flight trajectory within a synthetic aperture of the array 3D SAR system is discretized into 32 sampling points in azimuth.

[0047] Step 13, scale the POS data in this range after sampling according to the size of the scene grid point. When the number of cross-heading channels is 12, the ideal straight line and non-ideal straight line motion trajectories obtained by scaling the radar POS data are as follows: Figure 2 and Figure 3 As shown. At this time, the azimuth bandwidth corresponding to the ideal straight line and non-ideal straight line motion is B al and B ac , B al =B ac , the elevation bandwidth is B sl and B sc , the scene observation space spectrum coordinates are marked as (F a ,F s ).

[0048] Step 2, determining the location of the scene grid points, wherein the azimuth grid spacing is calculated according to the azimuth Rayleigh resolution, and the height grid spacing is smaller than the height Rayleigh resolution, thereby achieving height super-resolution processing, including the following sub-steps:

[0049] Step 21, determining the azimuth Rayleigh resolution and the elevation Rayleigh resolution according to the non-ideal linear motion trajectory; specifically, according to the ideal straight line and the non-ideal straight line motion trajectory of the array 3D SAR system, the azimuth Rayleigh resolution and the elevation Rayleigh resolution can be determined to be:

[0050] (B a Represents B al and B ac )

[0051] (B s Represents B sl and B sc )

[0052] Step 22, according to the elevation of non-ideal linear motion, the Rayleigh resolution ρ s, continuously adjust the sampling interval of the scattering points of the observation scene until the interval of the scattering points of the observation scene in the elevation direction is less than the elevation Rayleigh resolution ρ of the non-ideal linear motion s , to meet the super-resolution requirements;

[0053] Step 23, according to the azimuth Rayleigh resolution ρ of the non-ideal linear motion a , adjust the sampling interval of the scattering points of the observation scene, and make the interval of the scattering points of the observation scene in azimuth equal to the azimuth Rayleigh resolution ρ of the non-ideal linear motion a , the observation scene grid coordinates are marked as (a , s).

[0054] Step 3, generating an observation equation according to the motion trajectory and the scene grid position; in this embodiment, the observation equation A=exp(-2jπ(a*F a +s*F s ));

[0055] Step 4, substitute the radar data into the observation equation, solve the observation equation under the constraint of target sparsity, and obtain the reconstruction result, including the following sub-steps:

[0056] Step 41, determine the elevation unambiguous range according to the bandwidth of the non-ideal linear motion trajectory and the number of channels; in this embodiment, the array 3D SAR system observes the scene scattering point position information as follows Figure 4 As shown in the figure, there are four targets, with their azimuths evenly distributed at 2m and 4m, and their heights evenly distributed at 3m and 5m. Based on the bandwidth B scaled by non-ideal linear motion and the number of cross-heading channels N, the elevation unambiguous range is determined as

[0057] Step 42, adjusting the sparse constraint of the observed scene target according to the unambiguous range until the requirement is met;

[0058] Step 43, finally solving the observation equation.

[0059] By comparison Figure 5 It can be seen from the reconstruction results that when the signal has azimuth side lobes, the target point position can be correctly reconstructed under both the ideal straight line trajectory and the non-ideal straight line trajectory without losing the performance of the target through the method of the present invention.

[0060] By comparing the reconstruction results, it can be seen that in a flight under the array 3D SAR system, the processing process fits the non-ideal straight line motion trajectory into the ideal straight line motion trajectory, which will sacrifice the physical effective baseline length and lose the elevation resolution. As a result, in the case of high precision (when the two targets are close), the target cannot be correctly distinguished in the elevation direction, and the scene information cannot be reconstructed without distortion, such as Figure 5 As shown. Figure 6 As shown in Figure 1, a hundred Monte Carlo simulation experiments were conducted on target reconstruction in two situations under different bandwidth ratios. Under the same flight conditions, the reconstruction probability using non-ideal linear motion errors is significantly higher than the reconstruction probability without using non-ideal linear motion errors. Figure 7 As shown in Figure 1, a hundred Monte Carlo simulation experiments were conducted on target reconstruction in two situations under different signal-to-noise ratios. Under the same flight conditions, the reconstruction probability using non-ideal linear motion errors is significantly higher than the reconstruction probability without using non-ideal linear motion errors. Figure 8 As shown in Figure 1, a hundred Monte Carlo simulation experiments were conducted on target reconstruction in two situations with different numbers of array elements. Under the same flight conditions, the reconstruction probability using non-ideal linear motion errors is significantly higher than the reconstruction probability without using non-ideal linear motion errors. Fig. 9 As shown, the three-dimensional imaging result of a building in the X-band is reconstructed by the method of the present invention. The above results show that under the same processing conditions, the present invention can use the non-ideal linear motion error to increase the physical effective baseline length and significantly improve the Rayleigh resolution in the elevation direction.

[0061] The above actual data experiments show that compared with the existing radar imaging processing process in which the non-ideal linear motion trajectory is fitted into an ideal linear motion for direct imaging processing, the method of the present invention will lose the Rayleigh resolution in the elevation direction. This motion error can be used to increase the physical effective baseline length, thereby improving the Rayleigh resolution in the elevation direction under the same conditions.

[0062] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for improving elevation resolution based on array 3D SAR motion error, characterized in that: The steps include: Step 1, determining the non-ideal linear motion trajectory of the array 3D SAR based on POS data, includes the following sub-steps: Step 11, determining the POS data range within the synthetic aperture corresponding to the scene according to the scene information; Step 12: Adjust the sampling interval of the POS data within this range so that it can describe the changes of non-ideal linear motion while meeting the requirements corresponding to the number of scene points; Step 13: scale the POS data within the sampled range accordingly according to the scene grid point size; Step 2: Determine the scene grid point positions, where the azimuth grid spacing is calculated based on the azimuth Rayleigh resolution, and the altitude grid spacing is smaller than the altitude Rayleigh resolution, thereby achieving altitude super-resolution processing. Step 3: Generate observation equations based on the motion trajectory and scene grid position; Step 4: Substitute the radar data into the observation equation, solve the observation equation under the constraint of target sparsity, and obtain a reconstruction result; Step 4 includes the following sub-steps: Step 41, determining the elevation unambiguous range based on the bandwidth of the non-ideal linear motion trajectory and the number of channels; Step 42: adjusting the sparse constraint of the observed scene objects according to the unambiguous range until the requirements are met; Step 43, finally solving the observation equation; In step 11, the POS data range within the synthetic aperture corresponding to the scene is determined according to the scene information, and the number N and bandwidth B of the array 3D SAR cross-heading channels are determined according to the radar parameters. r , beam center angle θ, beam width And the scene's closest slant distance R min , determine the number of azimuth sampling points M according to the processing range; calculate the aperture length formula L according to the beam width of the transmitted signal of the array 3D SAR system sar =R min *θ*3.14 / 180, where L sar is the subaperture length, according to X=x end -x1 calculates the total aperture length, where X is the total aperture length, x1 and x end It is the coordinate value corresponding to the start and end of the azimuth in the radar POS data, and the number of sub-apertures N = X / L sar , and then determine the position corresponding to POS by the azimuth coordinates corresponding to the scene.

2. The method according to claim 1, wherein The step 2 includes the following sub-steps: Step 21, determining the Rayleigh resolution in azimuth and the Rayleigh resolution in elevation according to the non-ideal linear motion trajectory; Step 22, adjusting the sampling interval of the scattering points of the observation scene according to the Rayleigh resolution of the non-ideal linear motion in the elevation direction, so that the interval of the scattering points of the observation scene in the elevation direction is smaller than the Rayleigh resolution of the non-ideal linear motion in the elevation direction; Step 23 : According to the azimuth Rayleigh resolution of the non-ideal linear motion, the azimuth interval of the scattering points of the observed scene is made equal to the azimuth Rayleigh resolution of the non-ideal linear motion.

3. The method according to claim 1 or 2, wherein: According to the non-ideal linear motion trajectory of the array 3D SAR system and the grid position of the observation scene, the observation equation A=exp(-2jπ(a*F a +s*F s ), where a is the horizontal coordinate of the observation scene grid, s is the vertical coordinate of the observation scene grid, and F a is the horizontal coordinate of the scene observation space spectrum, F s is the vertical coordinate of the scene observation space spectrum.

Images