A high-precision airborne SAR image plane positioning method

Abstract

The application provides a high-precision airborne SAR image plane positioning method. The technical scheme is as follows: two SAR images of different angles of the same target region acquired by the airborne radar with the same imaging mode in one flight are used, at least one homonymic point in the two SAR images is selected, the homonymic point is used to construct a position equivalent equation, the system positioning error of the SAR image is obtained by solving the position equivalent equation, and the system positioning error is used to correct each pixel point in all SAR images in this flight to complete the positioning of the SAR image. The application does not need the ground reference point with the known accurate geographic position information, and does not need to use the accurate orbit information of the carrier, and solves the problem of harsh solving conditions of the existing airborne SAR image positioning method.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of microwave remote sensing, and aims to provide a high-precision airborne SAR (Synthetic Aperture Radar) image plane positioning method. BACKGROUND

[0002] Airborne SAR is flexible and variable, and can image the target area at all times, all weather, and high resolution, while also being able to accurately position the target in the scene, playing an important role in effectively monitoring key targets and areas. With the increasing maturity of airborne high-resolution SAR technology, higher requirements are put forward for target positioning accuracy, so the research on high-precision positioning technology of airborne SAR has important significance and practical value.

[0003] Airborne SAR image positioning can be divided into relative positioning method and absolute positioning method according to whether ground reference points are needed in the positioning process. The relative positioning method, see the literature: Jiang, W.; Yu, A.; Dong, Z. Refined polynomial geometric correction methods for spaceborne SAR image. In Proceedings of the 2016 CIE International Conference on Radar, Guangzhou, China, 10-13 October 2016; pp. 1-4. http: / / dx.doi.org / 10.1109 / RADAR.2016.8059216, this method is based on ground reference points with accurate geographic position information, selects appropriate models such as polynomial model and collinear equation model, and uses the prior position information of ground reference points to back-calculate the position information of image area. However, in some target areas such as ocean, desert, military area, etc., it is often difficult to easily obtain the accurate geographic position information (i.e. latitude, longitude, height) of ground reference points; and this kind of positioning method is easily affected by terrain undulation, number and distribution of ground reference points, making the relative positioning technology based on ground reference points have great limitations.

[0004] Absolute positioning technology, i.e., no ground reference point positioning technology, see the literature: Luo, Y.; Qiu, X.; Dong, Q.; Fu, K. A robust stereo positioning solution for multiview spaceborne SAR images based on Range-Doppler model. IEEE Geosci. Remote Sens. Lett. 2022, 19, 1-5. http: / / dx.doi.org / 10.1109 / LGRS.2020.3048731, this kind of method obtains the position information of the target area according to the imaging mechanism of the airborne radar. This kind of method is easily affected by the motion parameters of the aircraft navigation system provided by the aircraft. When the motion error of the aircraft caused by the high-altitude airflow and the deviation of the motion parameters recorded by the navigation system are large, the positioning accuracy of the airborne SAR image will be seriously affected. SUMMARY

[0005] In order to improve the plane positioning accuracy of the airborne SAR image, the inventors have conducted a lot of research and work. Through many flight tests, the inventors found that the system positioning error of the airborne SAR images of different angles of the same target area obtained by the airborne radar with the same imaging mode in one flight has relatively consistent characteristics. Based on this finding, the present application proposes a high-precision airborne SAR image plane positioning method. Compared with the existing method, the present application can more accurately and conveniently calculate the longitude and latitude position of each pixel in the SAR image, and improve the plane positioning accuracy of the SAR image.

[0006] The technical solution of the present application is: a high-precision airborne SAR image plane positioning method, characterized in that two SAR images of different angles of the same target area obtained by the airborne radar with the same imaging mode in one flight are used, at least one homonymic point is selected in the two SAR images, the homonymic point is used to construct a position equivalent equation, the system positioning error of the SAR image is obtained by solving the position equivalent equation, and the system positioning error is used to correct each pixel point in all SAR images in this flight, and the positioning of the SAR image is completed.

[0007] Taking one homonymic point as an example, the constructed position equivalent equation is:

[0008]

[0009] Wherein, [i1j1] and [i2j2] are pixel positions of a same named point in two airborne SAR images respectively; C1 and C2 are first order term coefficient conversion matrices between pixel positions and geographical positions of the target in the two airborne SAR images respectively, D1 and D2 are constant term coefficient conversion matrices between pixel positions and geographical positions of the target in the two airborne SAR images respectively; ρ r1 and ρ r2 are sampling intervals of the two airborne SAR images in the range direction respectively, ρ a1 and ρ a2 are sampling intervals of the two airborne SAR images in the azimuth direction respectively; Δr and Δa are system positioning errors of the airborne SAR images in the range direction and the azimuth direction respectively in this flight.

[0010] The planar positioning result [B'L'] of any pixel [i j] in any acquired SAR image in this flight is calculated by using the following formula:

[0011]

[0012] Wherein, ρ r is the calculated sampling interval of the airborne SAR image in the range direction, ρ a is the calculated sampling interval of the airborne SAR image in the azimuth direction; C is the first order term coefficient conversion matrix between pixel positions and geographical positions of the target in the calculated airborne SAR image, D is the constant term coefficient conversion matrix between pixel positions and geographical positions of the target in the calculated airborne SAR image.

[0013] The following technical effects can be achieved by using the present application:

[0014] The present application is based on the mechanism of influencing the positioning accuracy of the airborne SAR image, the system positioning errors of different view airborne SAR images of the same target region acquired by the airborne radar with the same imaging mode in one flight have relatively consistent characteristics, the position equivalent equation is constructed by using at least one same named point, and the accurate planar position of each pixel in the SAR image is calculated by correctly estimating the system positioning error of the airborne SAR image. The present application does not need the ground reference point with the accurate geographical position information, and does not need to use the accurate orbit information of the carrier, thus solving the problem of harsh calculation conditions of the existing airborne SAR image positioning method. The experimental results of the actual measurement data show that the present application can greatly improve the planar positioning accuracy of the airborne SAR image. The present application has wide application prospects in the fields of SAR image positioning and geographical surveying. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a principle flow schematic diagram of the method of the present application;

[0016] Figure 2 is a schematic diagram of the selection of homonym points and test points in multi-view airborne SAR images;

[0017] Figure 3 is the estimation result of the positioning error of the airborne SAR image system based on homonym points;

[0018] Figure 4 is the positioning error evaluation result of the test points by the existing method and the method herein. DETAILED DESCRIPTION

[0019] The embodiments of the present application are further described below in conjunction with the accompanying drawings.

[0020] The airborne SAR image described herein refers to the initial airborne SAR image, i.e., the SAR image directly transmitted to the ground station after airborne radar imaging; the one flight refers to the continuous flight of the carrier aircraft without landing within the time interval between two imaging of the airborne radar; and the same imaging mode refers to the consistent scanning mode and resolution of the two imaging of the airborne radar. The two airborne SAR images involved in the technical solution of the present application are obtained along two different flight routes in one flight.

[0021] Figure 1 is a schematic diagram of the principle flow of the method proposed herein. As shown in Figure 1 After the airborne radar along two different flight routes obtains different view airborne SAR images of the same target region in the same flight using the same imaging mode, the following processing is performed:

[0022] First step: calculating the affine transformation relationship between the target pixel position and the initial geographic position in the two airborne SAR images; second step: constructing the position equivalence equation according to the homonym points; third step: solving the system positioning error according to the position equivalence equation; fourth step, completing the high-precision planar positioning of each image pixel. The detailed description is as follows:

[0023] First step: calculating the affine transformation relationship between the target pixel position and the initial geographic position in the two airborne SAR images

[0024] This step calculates the affine transformation relationship between the target pixel position and the initial geographic position in each image according to the initial positioning auxiliary parameters of each airborne SAR image.

[0025] The affine transformation relationship between the target pixel position and the initial geographic position in any one of the airborne SAR images is solved by the following process:

[0026] The conversion relationship between the pixel position (i, j) and the geographic position (B, L) in the airborne SAR image can be described by the following formula:

[0027]

[0028] where a, b, c, d, e, f are affine transformation coefficients; C is a first-order term coefficient conversion matrix, i.e. D is a constant term coefficient conversion matrix, i.e. D = [e f]. The pixel positions (1, 1), (1, Height), (Width, Height), (Width, 1) and the geographic positions (B1, L1), (B2, L2), (B3, L3), (B4, L4) of the four corner points of the airborne SAR image can be directly obtained from the positioning auxiliary parameters of the airborne SAR image, where Height and Width are the height and width of the airborne SAR image, respectively. Using the pixel positions and the geographic positions of the four corner points, according to (3), (4) can be obtained:

[0029]

[0030] Let H = [H x H y ], then the first-order term coefficient conversion matrix C = H(1:2, 1:2), and the constant term coefficient conversion matrix D = H(3, :).

[0031] Using (4), the first-order term coefficient conversion matrices C1 and C2 of the pixel positions and the geographic positions of the targets in the two airborne SAR images are obtained, and the constant term coefficient conversion matrices D1 and D2 of the pixel positions and the geographic positions of the targets in the two airborne SAR images are obtained.

[0032] Second step: constructing position equivalence equations according to homonymic points

[0033] In this step, homonymic points are selected in the two airborne SAR images. The homonymic points refer to points with the same true geographic position in the two SAR images and having significant characteristics, such as road intersections, building corner points or significant natural landmarks that are easy to identify in the SAR images. Let the pixel positions of the homonymic points in the two airborne SAR images be [i1 j1] and [i2 j2] respectively, and the position equivalence equations are constructed according to the homonymic points, as shown in (1).

[0034] Third step: solving the system positioning error according to the position equivalence equations

[0035] Let:

[0036]

[0037] In the above formula, a1, b1, c1, d1 are affine transformation coefficients in the first-order term coefficient conversion matrix C1, i.e. a2, b2, c2, d2 are affine transformation coefficients in the first-order term coefficient conversion matrix C2, i.e. According to equations (1) and (5), the system positioning errors Δr and Δa of the airborne SAR image in the range and azimuth directions can be obtained as follows:

[0038]

[0039] Where, [ΔB ΔL]=[i2 j2]·C2+D2-[i1 j1]·C1-D1.

[0040] Figure 2 (a) and Figure 2 (b) Two airborne SAR images acquired during a single flight were used to estimate the system positioning error of the airborne SAR images in the range and azimuth directions using 10 corresponding points Hpi (i = 1, ..., 10) selected from the two images. The obtained system positioning error values ​​are as follows: Figure 3 As shown. Figure 3 The dark-colored rectangles represent the range-oriented system positioning error, and the light-colored rectangles represent the azimuth-oriented system positioning error. The calculated average range-oriented system positioning error is 14.19 m with a variance of 0.02; the average azimuth-oriented system positioning error is 5.33 m with a variance of 0.067. This shows that the system positioning errors of multi-view airborne SAR images estimated using different corresponding points are relatively consistent. Therefore, the results of calculating the system positioning error using one or more corresponding points are similar.

[0041] Step 4: Complete high-precision planar positioning of each pixel in the airborne SAR image.

[0042] The system positioning error of an airborne SAR image estimated from an arbitrary point is taken as the actual system positioning error of the entire airborne SAR image. The actual system positioning error is then used to complete the high-precision planar positioning of each pixel in the SAR image.

[0043] To verify the effectiveness of the method proposed in this invention, we used... Figure 3 The system positioning error of the airborne SAR image calculated from the corresponding point Hp1 is used as the actual system positioning error of any airborne SAR image during this flight, and is further used for... Figure 2 In (a), 10 test points Tpi (i = 1, ..., 10) were selected for localization. The final localization results of the two existing methods and the method proposed in this invention were statistically analyzed and compared. The planar localization errors of the three methods for the 10 test points are as follows: Figure 4 As shown, the solid line marked with "□" represents the planar positioning error curve obtained using this invention; the dashed line marked with "◇" represents the planar positioning error curve obtained using the polynomial model in relative positioning technology; and the dotted line marked with "○" represents the planar positioning error curve obtained using the range-Doppler RD model in absolute positioning technology.Figure 4 It can be seen from the table that the positioning error of each pixel in the airborne SAR image and the overall fluctuation of the positioning error of the method of the present application are significantly smaller than those of the other two methods, which shows that the method of the present application is more robust and has higher planar positioning accuracy than the existing positioning methods of airborne SAR images.

[0044] It should be noted that when solving the system positioning error by using the position equivalent equation, one common point in two or more SAR images can be used, or multiple common points in two SAR images can be used, or multiple common points in two or more SAR images can be used. Using the above situations and the position equivalent equation to solve the system positioning error are all within the protection scope of the present application.

Claims

1. A high-precision airborne SAR image plane positioning method, SAR refers to synthetic aperture radar, characterized in that, Two SAR images of different perspectives of the same target region acquired by airborne radar with the same imaging mode in one flight are used, at least one homonymy point in the two SAR images is selected, a position equivalence equation is constructed using the homonymy point, the system positioning error of the SAR image is obtained by solving the position equivalence equation, and the system positioning error is used to correct each pixel point in all SAR images in this flight to complete the positioning of the SAR image. Wherein, using one homonymy point, the position equivalence equation is constructed as: （1） wherein, and are the pixel positions of a same point in two airborne SAR images, respectively; C 1 and C 2 are the first-order term coefficient conversion matrices between the pixel positions and the geographical positions of the targets in the two airborne SAR images, respectively, D 1 and D 2 are the constant term coefficient conversion matrices between the pixel positions and the geographical positions of the targets in the two airborne SAR images, respectively; ρ r1 and ρ r2 are the sampling intervals of the two airborne SAR images in the range direction, respectively, ρ a1 and ρ a2 are the sampling intervals of the two airborne SAR images in the azimuth direction, respectively; Δ r and Δ a are the system positioning errors of the airborne SAR images in the range direction and the azimuth direction in this flight, respectively.

2. The high-precision airborne SAR image plane positioning method according to claim 1, characterized in that, The planar positioning result of any pixel in any acquired SAR image in this flight is calculated using the following formula :​ （2） wherein, ρ r is a sampling interval in the range direction of the computed airborne SAR image, ρ a is a sampling interval in the azimuth direction of the computed airborne SAR image; C is a first order term coefficient conversion matrix between pixel positions of targets in the computed airborne SAR image and geographical positions, D is a constant term coefficient conversion matrix between pixel positions of targets in the computed airborne SAR image and geographical positions.

3. A high-precision airborne SAR image plane positioning system, characterized in that The airborne SAR image is planarly positioned by using any one of the methods in claims 1 to 2.

Images