Multi-angle one-dimensional scattering center three-dimensional reconstruction method

Abstract

The invention discloses a multi-angle one-dimensional scattering center three-dimensional reconstruction method, and belongs to the technical field of radar signal processing. According to the method, through a scattering center dimension raising method combining subspace decomposition, RANSAC, spherical projection and visibility screening, a three-dimensional scattering center space structure of a target is recovered on the basis of a multi-angle one-dimensional distance scattering center; and the three-dimensional scattering center space distribution of the target can be deduced only by depending on multi-angle one-dimensional broadband distance domain scattering center estimation. According to the method, the problem that the three-dimensional scattering center cannot be recovered when only small-angle or discrete angular domain measurement is carried out is solved, and the availability of the three-dimensional scattering center in a non-cooperative observation scene is remarkably improved; the method has the advantages of high robustness and strong anti-noise and false scattering center capability. The method does not depend on continuous angle coverage, can only depend on non-missing data to realize three-dimensional scattering center inversion when part of angles are missing, and is especially suitable for non-cooperative targets such as maneuvering aircrafts, ships and satellites.

Description

Technical Field

[0001] This invention relates to radar signal processing technology, specifically to a method for three-dimensional reconstruction of a one-dimensional scattering center from multiple angles. Background Technology

[0002] One important method for radar target identification is utilizing the distribution characteristics of scattering centers (i.e., strong scattering points). Because different targets have different physical structures, the distribution of their scattering centers also differs. Based on the far-field scattering data obtained by radar, parameters such as the position and intensity of the scattering centers can be extracted, serving as key features for target identification. In high-frequency electromagnetic scattering theory, the scattering of complex targets can be obtained by superimposing multiple physically significant scattering centers. Radar systems can acquire equivalent range scattering centers from different azimuths through multi-angle observations. Extracting the position and intensity of these scattering centers is of significant value for target identification, structural characterization, and stealth assessment.

[0003] Traditional scattering center extraction methods, such as Multiple Signal Classification (MUSIC), typically only obtain one-dimensional range-domain scattering centers from a single angle. Reconstructing three-dimensional scattering centers requires wide-angle inverse synthetic aperture radar (ISAR) imaging or high-precision attitude measurement. However, in non-cooperative scenarios, the angular coverage of a target is often discrete, discontinuous, or even unknown, making it difficult to infer its three-dimensional position based solely on a one-dimensional scattering center. Existing methods lack a mechanism to unify multi-angle range information into a single geometric framework, resulting in large errors and poor reliability in the dimensionality estimation of scattering centers. Furthermore, traditional multi-angle matching methods are susceptible to noise and spurious scattering centers, leading to frequent failures in direct matching. Therefore, how to utilize multi-angle low-dimensional scattering information to achieve dimensionality-upgraded reconstruction of three-dimensional scattering centers has become a key technical challenge. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention proposes a three-dimensional reconstruction method for one-dimensional scattering centers from multiple angles. First, the precise one-dimensional scattering center distances from each angle are extracted. Then, the Random Sample Consensus (RANSAC) algorithm is used to fit a sphere to obtain the approximate location of the three-dimensional scattering center. Next, the true scattering centers are screened, and finally, clustering is performed to achieve three-dimensional reconstruction of the scattering center.

[0005] The multi-angle one-dimensional scattering center three-dimensional reconstruction method of the present invention includes the following steps: 1) Extracting one-dimensional scattering centers from multiple angles: The far-field of scattering at various angles is extracted from the far-field of scattering measured by broadband, and the corresponding one-dimensional scattering center in the spherical coordinate system is extracted from the far-field of scattering at each angle, so as to obtain the set of one-dimensional scattering centers at various angles in the spherical coordinate system. 2) Extract the approximate location of the three-dimensional scattering center: The one-dimensional scattering center is transformed from the spherical coordinate system to the rectangular coordinate system; the random sampling consensus algorithm (RANSAC) is used to extract one-dimensional scattering centers from the set of one-dimensional scattering centers; the projection relationship between the three-dimensional scattering center and the one-dimensional scattering center is used to fit the sphere to obtain the candidate points of the three-dimensional scattering center; and the approximate location of the three-dimensional scattering center is obtained by dividing the scattering unit. 3) Screening for true three-dimensional scattering centers: The approximate location of the three-dimensional scattering center is reprojected onto various angles to obtain a one-dimensional projection set of the three-dimensional scattering center; the distance error is further calculated, and the one-dimensional projection of the three-dimensional scattering center is filtered according to the distance error to obtain a filtered set of three-dimensional scattering centers in the whole space. 4) Three-dimensional scattering center clustering in full space: Hierarchical clustering based on distance thresholds is used to cluster the selected full-space three-dimensional scattering centers to obtain their precise locations. The scattering coefficients of the full-space three-dimensional scattering centers are then obtained using the least squares method, thus yielding the final three-dimensional scattering centers.

[0006] In step 1), the radar receives the scattered far-field data as frequency domain echo data. From the broadband measured scattered far-field, the scattered far-field at each angle is extracted based on the elevation and azimuth angles. The corresponding one-dimensional scattering center in spherical coordinates is extracted from the scattered far-field at each angle using the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT). , Given the number of one-dimensional scattering centers, we obtain the set of one-dimensional scattering centers at various angles in spherical coordinates. , For the first in spherical coordinate system l The distance from each one-dimensional scattering center to the origin, and the azimuth angle and pitch angle related, l =1, … , L Angles include pitch and azimuth. The pitch angle needs to be set to be greater than 45°. A pitch angle of 0~360° can be considered as the entire space.

[0007] In step 2), extracting the approximate location of the three-dimensional scattering center includes the following steps: a) According to the projection relationship, transform the one-dimensional scattering center from the spherical coordinate system to the rectangular coordinate system based on its elevation and azimuth angles to obtain the set of one-dimensional scattering centers at various angles in the rectangular coordinate system. , , and For the first in the rectangular coordinate system l The coordinates of a one-dimensional scattering center; b) Using RANSAC, three points are randomly selected from the set of one-dimensional scattering centers at various angles in the rectangular coordinate system each time. The three selected points and the origin are used as four points on the sphere to fit the sphere. The intersection of the line connecting the origin and the center of the sphere with another point on the sphere is used as a candidate point for the three-dimensional scattering center. c) In a Cartesian coordinate system, the target area is divided into three-dimensional scattering units, and a coarse threshold is set; d) Determine whether the three-dimensional scattering center is located in the scattering unit based on the number of candidate points contained in the scattering unit and the coarse threshold. Set the center of all scattering units with a number of candidate points greater than the coarse threshold as the coarse location of the three-dimensional scattering center in the whole space, and establish the correlation between the one-dimensional scattering center and the three-dimensional scattering center.

[0008] In step 2) a), the pitch angle from a single angle and azimuth The estimated one-dimensional scattering center is the true three-dimensional scattering center at the pitch angle. and azimuth Projecting the one-dimensional scattering center from spherical coordinates to Cartesian coordinates based on its elevation and azimuth angles: in, , and For the first in the rectangular coordinate system l The coordinates of a one-dimensional scattering center; the projection relationship between the actual three-dimensional scattering center and the one-dimensional scattering center is as follows: in, , and Let be the coordinates of the three-dimensional scattering center corresponding to the one-dimensional scattering center in a rectangular coordinate system; the above equation shows that the projection points of the three-dimensional scattering center at a fixed position in any direction are distributed in the coordinate system. Center of the sphere, radius is On the ball.

[0009] In step 2)b), RANSAC selects a sufficient number of candidate points, ranging from 100,000 to 500,000.

[0010] In step 2)c), the target range is divided into scattering units in the shape of small cubes in a rectangular coordinate system. The size of the small cubes is determined according to the size of the target range, and is taken as 1 / 1000 to 1 / 50 of the side length. The rough threshold is 100 to 5000, which is 1 / 1000 to 1 / 100 of the number of candidate points.

[0011] In step 3), screening for true three-dimensional scattering centers includes the following steps: a) Reproject the approximate location of the three-dimensional scattering center in the entire space onto various angles based on the pitch and azimuth angles to obtain a one-dimensional projection set of the three-dimensional scattering center. , M The number of approximate locations of the three-dimensional scattering centers. Let m be the distance from the one-dimensional projection of the m-th 3D scattering center to the origin, where m=1. … , M ; b) Calculate the distance error based on the distance from the one-dimensional projection of the three-dimensional scattering center to the origin and the corresponding one-dimensional scattering center in spherical coordinates. ; c) Set error threshold Distance error at this angle If the angle is within a certain range, then the view is considered visible; a visibility threshold is set. Only when the three-dimensional scattering center is visible within the visibility threshold is the approximate location of the three-dimensional scattering center considered to be the true three-dimensional scattering center, thus obtaining the filtered set of three-dimensional scattering centers in the entire space. , N The number of three-dimensional scattering centers in the entire space after screening. Let n be the distance from the one-dimensional projection of the nth 3D scattering center to the origin, where n=1. … , N .

[0012] In step 3)b), distance error : Among them, the The distance from each one-dimensional scattering center to the origin Distance from the m-th three-dimensional scattering center projected one-dimensionally to the origin The same angle means the same pitch angle and azimuth .

[0013] In step 3)c), the error threshold The visibility threshold is 1 / 4 to 3 / 4 of the side length of the scattering unit. The value is 0.05~0.2, meaning that the approximate location of the three-dimensional scattering center is considered to be the true three-dimensional scattering center only when it is visible at more than 5% to 20% of all angles. (Original) The approximate locations of the three-dimensional scattering centers, after filtering, became... N A selected full-space three-dimensional scattering center.

[0014] In step 4), the clustering of the selected full-space three-dimensional scattering centers includes the following steps: The selected set of full-space three-dimensional scattering centers Each data point is treated as an independent cluster. The distance between all clusters is calculated (initially the distance between points), and a distance threshold is set. The distance should be 2 to 5 times the side length of the scattering unit; find the two closest clusters, if the distance between the two clusters is less than a distance threshold. The two clusters are merged, and the new coordinates of the merged cluster are the average of the coordinates of the two clusters. This process is repeated to find the two nearest clusters until the distance between the two nearest clusters is greater than a distance threshold. If the current clusters are all present, the loop stops, and the final clustering result is the total number of clusters. The final output clusters are... The precise location of the three-dimensional scattering center in the entire space is obtained by using the least squares method to assign a value to the precise location of the three-dimensional scattering center in the entire space based on the far-field of scattering received by the radar, and then calculating the scattering coefficient of the three-dimensional scattering center in the entire space to obtain the final three-dimensional scattering center.

[0015] Advantages of this invention: (1) This invention solves the problem that the three-dimensional scattering center cannot be recovered when only small angle or discrete angular domain measurements are available: By using spherical projection geometric constraints and RANSAC fitting mechanism, this invention can use the multi-angle distance scattering center to infer the three-dimensional position, which significantly improves the availability of the three-dimensional scattering center in non-cooperative observation scenarios.

[0016] (2) It has high robustness and strong resistance to noise and false scattering centers: RANSAC has a natural ability to suppress multipath, noise and false peaks. Combined with the spherical geometric consistency constraint, it can effectively eliminate scattering points that do not meet the geometric relationship, making the three-dimensional reconstruction results more stable and reliable.

[0017] (3) Applicable to actual measurement scenarios of discontinuous angular domains, missing angles or maneuvering targets: This invention does not rely on continuous angle coverage. When some angles are missing, three-dimensional scattering center inversion can be achieved by relying only on the data that is not missing. It is particularly suitable for non-cooperative targets such as maneuvering aircraft, ships and satellites.

[0018] This invention utilizes a scattering center dimensionality enhancement method combining subspace decomposition, RANSAC, spherical projection, and visibility filtering to recover the three-dimensional scattering center spatial structure of a target based on multi-angle one-dimensional range scattering centers. The three-dimensional scattering center spatial distribution of the target can be inferred solely based on multi-angle one-dimensional broadband range domain scattering center estimation. Attached Figure Description

[0019] Figure 1 This is a flowchart of the multi-angle one-dimensional scattering center three-dimensional reconstruction method of the present invention; Figure 2 This is a schematic diagram illustrating the transformation of the one-dimensional scattering centers at various angles from the spherical coordinate system to the rectangular coordinate system using the multi-angle one-dimensional scattering center three-dimensional reconstruction method of the present invention. Figure 3 This is a schematic diagram of the three-dimensional scattering center and the one-dimensional projection point of the three-dimensional scattering center in this invention; Figure 4 This is a schematic diagram illustrating the approximate location of the three-dimensional scattering center obtained by the multi-angle one-dimensional scattering center three-dimensional reconstruction method according to the present invention; Figure 5 This is a schematic diagram illustrating the precise location of the three-dimensional scattering center in the entire space obtained by the multi-angle one-dimensional scattering center three-dimensional reconstruction method according to the present invention. Detailed Implementation

[0020] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0021] This embodiment uses a specific aircraft target simulation result as an example to illustrate the specific implementation of the present invention, restoring the three-dimensional scattering center spatial structure of the target based on the multi-angle one-dimensional range scattering center. The frequency of the aircraft target is set to 9.75~10.25GHz, with an interval of 10MHz; the pitch angle is set to 45~135°, with an interval of 5°; and the elevation angle is set to 0~360°, with an interval of 2°.

[0022] The multi-angle one-dimensional scattering center three-dimensional reconstruction method in this embodiment, such as Figure 1 As shown, it includes the following steps: 1) Extracting one-dimensional scattering centers from multiple angles: The radar receives scattered far-field data in the frequency domain. From the broadband measured scattered far-field, the scattered far-field at each angle is extracted based on the elevation and azimuth angles. The corresponding one-dimensional scattering center in spherical coordinates is extracted from the scattered far-field at each angle using the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT). A maximum of 20 scattering centers can be extracted from each angle. Given the number of one-dimensional scattering centers, we obtain the set of one-dimensional scattering centers at various angles in spherical coordinates. , For the first in spherical coordinate system l The distance from each one-dimensional scattering center to the origin, and the azimuth angle and pitch angle related, l =1, … , L ; 2) Extract the approximate location of the three-dimensional scattering center: a) Angles include pitch and azimuth, from a single angle azimuth... and pitch angle The estimated one-dimensional scattering center is the true three-dimensional scattering center at the azimuth angle. and pitch angle The projection below; will the first l A one-dimensional scattering center is transformed from a spherical coordinate system to a rectangular coordinate system based on its elevation and azimuth angles: in, , and For the first in the rectangular coordinate system l By considering the coordinates of each one-dimensional scattering center, we obtain the set of one-dimensional scattering centers at various angles in a Cartesian coordinate system. ; Transforming the one-dimensional scattering centers at various angles in the spherical coordinate system to the rectangular coordinate system O-XYZ, the results are as follows: Figure 2 As shown; The projection relationship between the true three-dimensional scattering center and the one-dimensional scattering center is as follows: in, , and Let be the coordinates of the three-dimensional scattering center corresponding to the one-dimensional scattering center in a rectangular coordinate system; the above equation shows that the projection points of the three-dimensional scattering center at a fixed position in any direction are distributed in the coordinate system. Center of the sphere, radius is According to the projection relationship on the sphere; This conclusion can be explained more intuitively from a geometric perspective: the line connecting the three-dimensional scattering center and its one-dimensional projection point is always perpendicular to the angular direction. Therefore, the three-dimensional scattering center A, the origin O, and the three one-dimensional projections D1, D2, and D3 of the three-dimensional scattering centers lie on the same sphere. AOD1, AOD2, and AOD3 each form a right-angled triangle, with the one-dimensional scattering center located at the right-angled vertex and changing position with the angular direction, as shown below. Figure 3 As shown; according to geometric principles: the right-angled vertices of all right triangles with the same line segment as their hypotenuse lie on a sphere with that line segment as its diameter; therefore, the projection point of a fixed three-dimensional scattering center in any direction must lie on a sphere with the line connecting the three-dimensional scattering center and the origin as its diameter; the spherical distribution law of the projection point of a fixed-position three-dimensional scattering center is the basis of this invention; it is the main basis for obtaining the spatial position of the scattering center (the intersection of the diameter passing through the origin and the sphere); b) Using RANSAC, three points are randomly selected from the set of one-dimensional scattering centers at various angles in the rectangular coordinate system each time. The three selected points and the origin are used as four points on the sphere to fit the sphere. The intersection of the line connecting the origin and the center of the sphere with the other point on the sphere is used as the candidate point of the three-dimensional scattering center. 200,000 samples are taken, and the number of candidate points is 200,000. c) If the three selected points are not projections of the same 3D scattering center, the resulting 3D positions will deviate from the target position and be very sparse; if the three selected points are projections of the same 3D scattering center, the resulting 3D positions will be concentrated around the target position, which may be caused by resolution, electromagnetic simulation calculation errors, etc. However, it should be noted that the true candidate points of the same 3D scattering center are very close to each other, while the false candidate points are randomly distributed in a larger space. Inspired by the difference in the distribution of true and false candidates, the target range is divided into scattering units of small cubes in a rectangular coordinate system. The size of the small cube is determined according to the size of the target range, and the length of the volume diagonal of the small cube is equal to the distance resolution; the rough threshold is 300. d) Determine whether the three-dimensional scattering center is located within a scattering unit based on the number of candidate points and a coarse threshold. Set the center of all scattering units with a number of candidate points greater than the coarse threshold as the coarse location of the three-dimensional scattering center in full space, such as... Figure 4 As shown, the correlation between one-dimensional scattering centers and three-dimensional scattering centers was established; 3) Screening for true three-dimensional scattering centers: a) Reproject the approximate location of the three-dimensional scattering center in the entire space onto various angles based on the pitch and azimuth angles to obtain a one-dimensional projection set of the three-dimensional scattering center. , M The number of approximate locations of the three-dimensional scattering centers. Let m be the distance from the one-dimensional projection of the m-th 3D scattering center to the origin, where m=1. … , M ; b) Calculate the distance error based on the one-dimensional distance and the corresponding one-dimensional scattering center in spherical coordinates. : in, For the first The distance from each one-dimensional scattering center to the origin The distance from the m-th three-dimensional scattering center to the origin via a one-dimensional projection is given; the angles of the two projections are the same, meaning they correspond to the same azimuth angle. and pitch angle ; c) Set the error threshold to 3 / 4 of the scattering unit's side length, and the distance error at this angle... If the distance is less than the error threshold, the angle is considered visible. A visibility threshold of 0.2 is set; only when the three-dimensional scattering center is visible within this threshold range is its approximate location considered the true three-dimensional scattering center, thus obtaining the filtered set of all-space three-dimensional scattering centers. , N The number of three-dimensional scattering centers in the entire space. Let n be the distance from the one-dimensional projection of the nth 3D scattering center to the origin, where n=1. … , N ; 4) Three-dimensional scattering center clustering in full space: The selected set of full-space three-dimensional scattering centers Each data point is treated as an independent cluster. The distance between all clusters is calculated (initially the distance between points), and a distance threshold is set. The distance is three times the side length of the scattering unit; find the two closest clusters, if the distance between the two clusters is less than a distance threshold. The two clusters are merged, and the new coordinates of the merged cluster are the average of the coordinates of the two clusters. This process is repeated to find the two nearest clusters until the distance between the two nearest clusters is greater than a distance threshold. If the current clusters are all present, the loop stops, and the final clustering result is the total number of clusters. The final output clusters are... The precise location, that is, the precise location of the three-dimensional scattering center in full space, such as Figure 5As shown, the amplitude of the three-dimensional scattering center, i.e., the scattering coefficient, varies with the angle; the amplitude of the three-dimensional scattering center differs at different angles. Using the least squares method, the far-field scattering received by the radar is used to assign precise positions to the three-dimensional scattering centers in the entire space, and the scattering coefficients of the three-dimensional scattering centers in the entire space are calculated to obtain the final three-dimensional scattering centers. The final three-dimensional scattering centers include the precise positions and scattering coefficients of the three-dimensional scattering centers in the entire space. Through the multiple final three-dimensional scattering centers obtained, the distribution characteristics of the target's scattering centers are determined, and these characteristics can provide effective information for downstream tasks such as radar target identification.

[0023] Finally, it should be noted that the purpose of disclosing the embodiments is to help further understand the present invention. However, those skilled in the art will understand that various substitutions and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the present invention should not be limited to the content disclosed in the embodiments, and the scope of protection of the present invention is defined by the claims.

Claims

1. A method for three-dimensional reconstruction of a one-dimensional scattering center from multiple angles, characterized in that, The method includes the following steps: 1) Extracting one-dimensional scattering centers from multiple angles: The far-field of scattering at various angles is extracted from the far-field of scattering measured by broadband, and the corresponding one-dimensional scattering center in the spherical coordinate system is extracted from the far-field of scattering at each angle, so as to obtain the set of one-dimensional scattering centers at various angles in the spherical coordinate system. 2) Extract the approximate location of the three-dimensional scattering center: The one-dimensional scattering center is transformed from the spherical coordinate system to the rectangular coordinate system; the random sampling consensus algorithm RANSAC is used to extract one-dimensional scattering centers from the set of one-dimensional scattering centers; the projection relationship between the three-dimensional scattering center and the one-dimensional scattering center is used to fit the sphere to obtain the candidate points of the three-dimensional scattering center; and the approximate location of the three-dimensional scattering center is obtained by dividing the scattering unit. 3) Screening for true three-dimensional scattering centers: The approximate location of the three-dimensional scattering center is reprojected onto various angles to obtain a one-dimensional projection set of the three-dimensional scattering center; the distance error is further calculated, and the one-dimensional projection of the three-dimensional scattering center is filtered according to the distance error to obtain a filtered set of three-dimensional scattering centers in the whole space. 4) Three-dimensional scattering center clustering in full space: Hierarchical clustering based on distance thresholds is used to cluster the selected full-space three-dimensional scattering centers to obtain the final three-dimensional scattering centers.

2. The method according to claim 1, characterized in that, In step 1), the far-field of scattering at various angles is extracted from the broadband measured far-field based on the elevation and azimuth angles. For each angle, the corresponding one-dimensional scattering center in the spherical coordinate system is extracted using rotation-invariant subspace parameter estimation techniques, thus obtaining the set of one-dimensional scattering centers at each angle in the spherical coordinate system. , The number of one-dimensional scattering centers. For the first in spherical coordinate system l The distance from each one-dimensional scattering center to the origin, and the azimuth angle and pitch angle related, l =1, … , L .

3. The method according to claim 2, characterized in that, In step 2), extracting the approximate location of the three-dimensional scattering center includes the following steps: a) According to the projection relationship, transform the one-dimensional scattering center from the spherical coordinate system to the rectangular coordinate system based on its elevation and azimuth angles to obtain the set of one-dimensional scattering centers at various angles in the rectangular coordinate system. ,in, , and For the first in the rectangular coordinate system l The coordinates of a one-dimensional scattering center; b) Using RANSAC, three points are randomly selected from the set of one-dimensional scattering centers at various angles in the rectangular coordinate system each time. The three selected points and the origin are used as four points on the sphere to fit the sphere. The intersection of the line connecting the origin and the center of the sphere with another point on the sphere is used as a candidate point for the three-dimensional scattering center. c) In a Cartesian coordinate system, the target area is divided into three-dimensional scattering units, and a coarse threshold is set; d) Determine whether the three-dimensional scattering center is located in the scattering unit based on the number of candidate points contained in the scattering unit and the coarse threshold. Set the center of all scattering units with a number of candidate points greater than the coarse threshold as the coarse location of the three-dimensional scattering center in the whole space.

4. The method according to claim 3, characterized in that, In step 2) a), the pitch angle from a single angle and azimuth The estimated one-dimensional scattering center is the true three-dimensional scattering center at the pitch angle. and azimuth Projecting the one-dimensional scattering center from spherical coordinates to Cartesian coordinates based on its elevation and azimuth angles: The projection relationship between the true three-dimensional scattering center and the one-dimensional scattering center is as follows: in, , and Let be the coordinates of the three-dimensional scattering center corresponding to the one-dimensional scattering center in a rectangular coordinate system; the projection points of the three-dimensional scattering center at a fixed position in any direction are distributed in a coordinate system. Center of the sphere, radius is On the ball.

5. The method according to claim 3, characterized in that, In step 2)c), the target area is divided into scattering units in the shape of small cubes in a rectangular coordinate system. The size of the small cubes is determined according to the size of the target area, and is taken as 1 / 1000 to 1 / 50 of the side length; the rough threshold is 100 to 5000.

6. The method according to claim 3, characterized in that, In step 3), screening for true three-dimensional scattering centers includes the following steps: a) Reproject the approximate location of the three-dimensional scattering center in the entire space onto various angles based on the pitch and azimuth angles to obtain a one-dimensional projection set of the three-dimensional scattering center. , M The number of approximate locations of the three-dimensional scattering centers. Let m be the distance from the one-dimensional projection of the m-th 3D scattering center to the origin, where m=1. … , M ; b) Calculate the distance error based on the distance from the one-dimensional projection of the three-dimensional scattering center to the origin and the corresponding one-dimensional scattering center in spherical coordinates. ; c) Set error threshold Distance error at this angle If the angle is within a certain range, then the view is considered visible; a visibility threshold is set. Only when the three-dimensional scattering center is visible within the visibility threshold is the approximate location of the three-dimensional scattering center considered to be the true three-dimensional scattering center, thus obtaining the filtered set of three-dimensional scattering centers in the entire space. , N The number of three-dimensional scattering centers in the entire space after screening. Let n be the distance from the one-dimensional projection of the nth 3D scattering center to the origin, where n=1. … , N .

7. The method according to claim 6, characterized in that, In step 3)b), distance error : Among them, the The distance from each one-dimensional scattering center to the origin Distance from the m-th three-dimensional scattering center projected one-dimensionally to the origin The same angle means the same pitch angle and azimuth .

8. The method according to claim 6, characterized in that, In step 3)c), the error threshold The visibility threshold is 1 / 4 to 3 / 4 of the side length of the scattering unit. The value is 0.05~0.

2.

9. The method according to claim 6, characterized in that, In step 4), the clustering of the selected full-space three-dimensional scattering centers includes the following steps: Each data point in the filtered set of full-space 3D scattering centers is treated as an independent cluster. The distances between all clusters are calculated, and a distance threshold is set. Find the two closest clusters; if the distance between the two clusters is less than a distance threshold... The two clusters are merged, and the new coordinates of the merged cluster are the average of the coordinates of the two clusters. This process is repeated to find the two nearest clusters until the distance between the two nearest clusters is greater than a distance threshold. If the loop stops, all existing clusters are the precise locations of the three-dimensional scattering centers in the entire space. The precise locations of the three-dimensional scattering centers in the entire space are assigned using the least squares method with the far-field scattering received by the radar. The scattering coefficients of the three-dimensional scattering centers in the entire space are calculated to obtain the final three-dimensional scattering centers.

Images