Co-platform double-base foresight SAR imaging method and system

Through the common platform dual-base forward-view SAR imaging method, the support vector machine and clustering algorithm are used to process the double-base synthetic aperture radar echo data, which solves the problem of false points in forward-view imaging, and realizes high-quality forward-view imaging. For small and medium-sized platforms, the problems of low imaging quality and complex structure in the prior art are solved.

CN119087436AActive Publication Date: 2024-12-06SHENZHEN UNIV

Patent Information

Application Number
CN202411199717.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-12-06
Estimated Expiration
2044-08-29

AI Technical Summary

Technical Problem

Existing synthetic aperture radar imaging has problems of low imaging quality and complex overall structural model in front-view scenes, especially because the Doppler process of the two point targets with symmetrical beam centerline when the radar is advanced is the same, resulting in blurring left and right. It is a phenomenon, and it is difficult for the prior art to realize effective forward imaging on small and medium-sized platforms.

Method used

Using the common platform dual-base forward-view SAR imaging method, false target elimination of double-base synthetic aperture radar echo data by collecting left front-view radar data, right front-view radar data, velocity and acceleration data, combined with support vector machine algorithm or clustering algorithm 处理,得到虚假目标消除后的掩码图像,从而提取真实目标点,实现高质量的前视合成孔径雷达成像。

Benefits of technology

By introducing support vector machines and clustering algorithms, the problem of false points in forward vision imaging is solved, reliable common platform forward vision imaging is realized, and the imaging quality of forward vision synthetic aperture radar is improved, and it is suitable for small and medium-sized platforms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119087436A_ABST
    Figure CN119087436A_ABST
Patent Text Reader

Abstract

The invention discloses a common-platform double-base foresight SAR imaging method and system, and the method comprises the steps: collecting left foresight radar data, right foresight radar data, and speed and acceleration data, and obtaining double-base synthetic aperture radar echo data; foresight area imaging is carried out on the collected bistatic synthetic aperture radar echo data, false target elimination processing is carried out on foresight area imaging through a support vector machine algorithm or a clustering algorithm, and a mask image after false target elimination is obtained; extracting a real target point of an original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain a final foresight synthetic aperture radar image; according to the method, the support vector machine and the clustering algorithm are introduced to solve the problem of false points existing in the back propagation algorithm in the process of processing foresight imaging, and finally reliable co-platform foresight imaging is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a common platform dual-base forward-looking SAR imaging method and system. Background Art

[0002] Synthetic Aperture Radar (SAR) imaging is a method of obtaining target information through radar technology. Unlike traditional optical imaging technology, SAR generates images with target information by actively emitting microwave signals and reflecting echoes from the target. SAR imaging is characterized by the ability to perform imaging at any time and in any weather conditions. It is widely used in military and aerospace fields, such as geological research, ocean observation, rescue and reconnaissance, etc. In the civilian field, SAR is also developing in the direction of miniaturization, light weight, high resolution, dual-multi-base, etc.

[0003] At present, synthetic aperture radar imaging mainly studies the front side view and oblique forward view modes. In the forward view scene, since the two point targets symmetrical on the beam centerline have the same Doppler history to the radar when the radar moves forward, the forward view imaging result has the problem of left and right blur. At present, there are solutions such as single pulse forward view imaging technology, array antenna super angle resolution technology, and dual-base synthetic aperture radar forward view imaging technology in the field of forward view imaging. The hardware structure and execution algorithm of single pulse forward view imaging technology are relatively simple. However, when applied to multiple strong scattering targets or continuously distributed targets, the number of single pulse angle measurement channels is limited and multiple targets in the same resolution unit cannot be distinguished, resulting in "angle flicker" phenomenon and reducing imaging quality. The angle measurement based on the sum and difference ratio of the reconstructed estimate improves the positioning accuracy of the target within the antenna coverage range, verifying that the proposed method effectively improves the clarity of the forward view image. However, symmetrical targets appearing on the left and right sides of the forward view direction will have the problem of "left and right blur" due to the same Doppler frequency. In single pulse imaging, they cannot be separated because they are located in the same resolution unit. Array antenna super-angle resolution technology relies on large-aperture antennas to achieve azimuth super-resolution, and improves image contrast and signal-to-noise ratio by coherently smoothing multiple images. However, due to the complexity of the system, it has high requirements for the size of the platform structure and cannot be applied to small and medium-sized platforms such as automobiles. Dual-base synthetic aperture radar forward imaging usually separates the transmitter and receiver on different platforms to solve the forward blur problem, but due to the dependence on multiple platforms for transmission and reception in different locations, the overall structural model is complicated and transmission, reception and processing cannot be completed on the same platform.

[0004] Therefore, the prior art needs to be improved. Summary of the invention

[0005] The technical problem to be solved by the present invention is that, in view of the defects of the prior art, the present invention provides a common platform dual-base forward-looking SAR imaging method and system to solve the problems of low imaging quality and complex overall structural model in the existing forward-looking synthetic aperture radar imaging method.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] In a first aspect, the present invention provides a common platform bistatic forward-looking SAR imaging method, comprising:

[0008] Collect left front-looking radar data, right front-looking radar data, velocity and acceleration data to obtain bistatic synthetic aperture radar echo data;

[0009] Performing forward-view area imaging on the collected bistatic synthetic aperture radar echo data, and performing false target elimination processing on the forward-view area imaging by using a support vector machine algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated;

[0010] The real target points of the original image in the bistatic synthetic aperture radar echo data are extracted according to the mask image to obtain the final forward-looking synthetic aperture radar imaging.

[0011] In one implementation, the collecting of left front-looking radar data, right front-looking radar data, speed and acceleration data includes:

[0012] A connection request is initiated to two monostatic radar modules through the communication module, and when the two monostatic radar modules are successfully connected and synchronized, a dual-base forward-looking synthetic aperture radar imaging system is operated.

[0013] In one implementation, the performing forward-looking region imaging on the collected bistatic synthetic aperture radar echo data includes:

[0014] Imaging the left front-looking synthetic aperture radar data in the bistatic synthetic aperture radar echo data based on a back-propagation algorithm, and performing left offset compensation through an offset compensation strategy to obtain a left front-looking synthetic aperture radar image M;

[0015] The right front-looking synthetic aperture radar data in the bistatic synthetic aperture radar echo data is delayed compensated by the delay factor of the hardware delay compensation strategy, the right front-looking synthetic aperture radar data is imaged based on the back propagation algorithm, and right offset compensation is performed by the offset compensation strategy to obtain a right front-looking synthetic aperture radar image N.

[0016] In one implementation, the forward-looking region imaging of the collected bistatic synthetic aperture radar echo data includes:

[0017] Performing data domain matching on left front-looking synthetic aperture radar data and right front-looking synthetic aperture radar data in the bistatic synthetic aperture radar echo data;

[0018] The compensation factors corresponding to the left offset compensation strategy and the right offset compensation strategy are calculated according to the matching result, and the delay factor corresponding to the hardware delay compensation strategy is calculated.

[0019] In one implementation, performing false target elimination processing on the front view area imaging by using a clustering algorithm to obtain a mask image after the false targets are eliminated includes:

[0020] Get the left front-view synthetic aperture radar image M at time i i and the right forward synthetic aperture radar image N i ;

[0021] Construct the relationship matrix:

[0022]

[0023] Among them, M i (m,n) and N i (m,n) represent M i and N i The value at the pixel (m,n);

[0024] The relationship matrix is ​​passed into the clustering algorithm to classify the true target points and false target points in the forward-looking scene, and the true target points are extracted according to the binary classification results;

[0025] Based on the extracted real target points, a mask image of the final prediction result is constructed.

[0026] In one implementation, performing false target elimination processing on the front view area imaging by a support vector machine algorithm to obtain a mask image after the false targets are eliminated includes:

[0027] Construct a data source X, take (M(m,n), N(m,n)) as X(m,n) for the left front synthetic aperture radar image M and the right front synthetic aperture radar image N, and traverse all pixel points to obtain a data source matrix X;

[0028] Input X into the trained support vector machine model to obtain the mask image of the final prediction result.

[0029] In one implementation, extracting the real target point of the original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain the final forward-looking synthetic aperture radar imaging includes:

[0030] The mask image is used to multiply the left forward-looking synthetic aperture radar image M or the right forward-looking synthetic aperture radar image N in the original image, to extract the real target points of the original image, to eliminate the false target points of the original image, and to obtain the final forward-looking synthetic aperture radar imaging.

[0031] In a second aspect, the present invention provides a common platform dual-base forward-looking SAR imaging system, comprising:

[0032] Left front radar, used to collect left front-looking radar data;

[0033] Right front radar, used to collect right front-looking radar data;

[0034] Inertial navigation module, used to collect the speed of the radar at the time of operation;

[0035] and the main control end, which is used to realize bistatic forward-looking synthetic aperture radar imaging;

[0036] Wherein, the main control end includes:

[0037] A data acquisition module is used to collect left front-looking radar data, right front-looking radar data, speed and acceleration data to obtain bistatic synthetic aperture radar echo data;

[0038] The mask image module is used to perform forward-viewing area imaging on the collected bistatic synthetic aperture radar echo data, and perform false target elimination processing on the forward-viewing area imaging by a support vector machine algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated;

[0039] The synthetic aperture radar imaging module is used to extract the real target points of the original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain the final forward-looking synthetic aperture radar imaging.

[0040] In a third aspect, the present invention provides a terminal, comprising: a processor and a memory, wherein the memory stores a common platform bistatic forward-looking SAR imaging program, and when the common platform bistatic forward-looking SAR imaging program is executed by the processor, it is used to implement the operation of the common platform bistatic forward-looking SAR imaging method as described in the first aspect.

[0041] In a fourth aspect, the present invention further provides a medium, which is a computer-readable storage medium, and which stores a common platform bistatic forward-looking SAR imaging program. When the common platform bistatic forward-looking SAR imaging program is executed by a processor, it is used to implement the operation of the common platform bistatic forward-looking SAR imaging method as described in the first aspect.

[0042] The present invention adopts the above technical solution to achieve the following effects:

[0043] The present invention acquires bistatic synthetic aperture radar echo data by collecting left forward-looking radar data, right forward-looking radar data, speed and acceleration data; performs forward-looking region imaging on the acquired bistatic synthetic aperture radar echo data, performs false target elimination processing on the forward-looking region imaging by a support vector machine algorithm or a clustering algorithm, and acquires a mask image after the false targets are eliminated; the real target points of the original image in the bistatic synthetic aperture radar echo data can be extracted according to the mask image, and the final forward-looking synthetic aperture radar imaging can be acquired; the present invention solves the problem of false points existing in the back-propagation algorithm in processing forward-looking imaging by introducing a support vector machine and a clustering algorithm, and finally realizes reliable common-platform forward-looking imaging, thereby improving the imaging quality of the forward-looking synthetic aperture radar. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0045] Figure 1 It is a flow chart of the CCP platform bistatic forward-looking SAR imaging method of the present invention.

[0046] Figure 2 It is a schematic diagram of the CCP platform bistatic forward-looking SAR imaging system of the present invention.

[0047] Figure 3 It is a schematic diagram of the front-side cascade system in the present invention.

[0048] Figure 4 It is a schematic diagram of a 5-second echo data model in the present invention.

[0049] Figure 5 It is a flow chart of the forward-looking imaging processing algorithm in the present invention.

[0050] Figure 6 It is a schematic diagram of 9 random point targets in the present invention.

[0051] Figure 7 It is a schematic diagram of the radar BP imaging result on the right side of the present invention.

[0052] Figure 8 It is a schematic diagram of the left radar BP imaging result in the present invention.

[0053] Fig. 9 It is a schematic diagram of the processing result of the support vector machine solution in the present invention.

[0054] Fig.10It is a schematic diagram of the processing result of the Kmeans clustering algorithm in the present invention.

[0055] Fig.11 It is the optical scene diagram of the measured data in the present invention.

[0056] Fig.12 It is a schematic diagram of the left radar BP imaging result in the present invention.

[0057] Fig.13 It is a schematic diagram of the radar BP imaging result on the right side of the present invention.

[0058] Fig.14 It is a schematic diagram of the forward-looking imaging result of the method proposed in the present invention.

[0059] Fig.15 It is a functional principle diagram of a terminal in one implementation of the present invention.

[0060] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0062] Exemplary Methods

[0063] At present, synthetic aperture radar imaging mainly studies the front side view and oblique forward view modes. In the forward view scene, since the two point targets symmetrical on the beam centerline have the same Doppler history to the radar when the radar moves forward, the forward view imaging result has the problem of left and right blur. At present, there are solutions such as single pulse forward view imaging technology, array antenna super angle resolution technology, and dual-base synthetic aperture radar forward view imaging technology in the field of forward view imaging. The hardware structure and execution algorithm of single pulse forward view imaging technology are relatively simple. However, when applied to multiple strong scattering targets or continuously distributed targets, the number of single pulse angle measurement channels is limited and multiple targets in the same resolution unit cannot be distinguished, resulting in "angle flicker" phenomenon and reducing imaging quality. The angle measurement based on the sum and difference ratio of the reconstructed estimate improves the positioning accuracy of the target within the antenna coverage range, verifying that the proposed method effectively improves the clarity of the forward view image. However, symmetrical targets appearing on the left and right sides of the forward view direction will have the problem of "left and right blur" due to the same Doppler frequency. In single pulse imaging, they cannot be separated because they are located in the same resolution unit. Array antenna super-angle resolution technology relies on large-aperture antennas to achieve azimuth super-resolution, and improves image contrast and signal-to-noise ratio by coherently smoothing multiple images. However, due to the complexity of the system, it has high requirements for the size of the platform structure and cannot be applied to small and medium-sized platforms such as automobiles. Dual-base synthetic aperture radar forward imaging usually separates the transmitter and receiver on different platforms to solve the forward blur problem, but due to the dependence on multiple platforms for transmission and reception in different locations, the overall structural model is complicated and transmission, reception and processing cannot be completed on the same platform.

[0064] In view of the above technical problems, an embodiment of the present invention provides a common platform bistatic forward-looking SAR imaging method, which obtains bistatic synthetic aperture radar echo data by collecting left forward-looking radar data, right forward-looking radar data, speed and acceleration data; and performs forward-looking area imaging on the collected bistatic synthetic aperture radar echo data, performs false target elimination processing on the forward-looking area imaging by a support vector machine algorithm or a clustering algorithm, and obtains a mask image after the false targets are eliminated; the real target points of the original image in the bistatic synthetic aperture radar echo data can be extracted according to the mask image to obtain the final forward-looking synthetic aperture radar imaging; the present invention solves the problem of false points existing in the back-propagation algorithm when processing forward-looking imaging by introducing a support vector machine and a clustering algorithm, and finally realizes reliable common platform forward-looking imaging, thereby improving the imaging quality of the forward-looking synthetic aperture radar.

[0065] like Figure 1 As shown, an embodiment of the present invention provides a common platform bistatic forward-looking SAR imaging method, comprising the following steps:

[0066] Step S100, collecting left front-looking radar data, right front-looking radar data, velocity and acceleration data to obtain bistatic SAR echo data.

[0067] In this embodiment, the common platform bistatic forward-looking SAR imaging method is implemented based on a common platform bistatic forward-looking SAR imaging system provided in this embodiment.

[0068] In general, in this embodiment, a multi-sensor cascade working system is first constructed using a communication module (including building a forward-looking dual millimeter-wave radar cascade acquisition and a real-time speed acquisition system); then, the collected dual-base SAR (Synthetic Aperture Radar) echo data is imaged in the forward-looking area; finally, the SVM algorithm (Support Vector Machine, a type of generalized linear classifier that performs binary classification of data in a supervised learning manner) and clustering algorithm are introduced to solve the problem of false points in the BP algorithm (back propagation algorithm) when processing forward-looking imaging, and finally reliable common-platform forward-looking imaging is achieved.

[0069] The hardware block diagram of a common platform bistatic forward-looking SAR imaging system provided in this embodiment is as follows: Figure 2 As shown, it includes: a left front radar for collecting left front-looking radar data; a right front radar for collecting right front-looking radar data; an inertial navigation module for collecting the speed of the radar when it is working; and a main control end for realizing dual-base forward-looking SAR imaging. Among them, the overall control instructions of the system are issued by the main control end. A set of parts consisting of a control unit and a radar is called a single-base radar module. In the designed forward-looking SAR imaging hardware system, two sets of single-base radar modules are required, which are respectively installed on the left front side and the right front side of the vehicle-mounted platform. At the same time, the introduction of the inertial navigation module can help the system obtain speed data when the radar is working.

[0070] In this embodiment, before collecting the left front-looking radar data, the right front-looking radar data, the speed and the acceleration data, it is also necessary to realize the connection between the two monostatic radar modules and the main control end.

[0071] Specifically, in an implementation of this embodiment, the following steps are included before step S100:

[0072] Step S100a, initiating a connection request to two monostatic radar modules through the communication module, and operating the bistatic forward-looking SAR imaging system when the two monostatic radar modules are successfully connected and synchronized.

[0073] In this embodiment, the workflow of the hardware system is as follows:

[0074] First, the main control end initiates a connection request to the two single-base radar modules on the left and right through the communication module. If the access is successful, the single-base radar module enters the ready stage. If the access fails, the control unit industrial computer rejects it; the main control end re-initiates the connection request to avoid asynchronous radar data collection caused by the issuance of instructions.

[0075] If the two single-base radar modules are successfully connected and synchronized, the system starts running to collect left front-looking radar data, right front-looking radar data, speed, acceleration and other data; wherein the left front-looking radar data and the right front-looking radar data are collected by the left front side radar and the right front side radar respectively; the speed is the speed data obtained by the inertial navigation module at the time of radar working; the acceleration is calculated based on the obtained speed data.

[0076] like Figure 1 As shown, an embodiment of the present invention provides a common platform bistatic forward-looking SAR imaging method, comprising the following steps:

[0077] Step S200, performing forward-viewing area imaging on the collected bistatic SAR echo data, performing false target elimination processing on the forward-viewing area imaging by using a SVM algorithm or a clustering algorithm, and obtaining a mask image after the false targets are eliminated.

[0078] In this embodiment, before imaging the forward-looking area of ​​the collected bistatic SAR echo data, data matching is also required for the bistatic SAR echo data domain to determine the compensation factor corresponding to the offset compensation strategy and the delay factor corresponding to the hardware delay compensation strategy.

[0079] Specifically, in an implementation of the present embodiment, the forward-looking area imaging of the collected bistatic SAR echo data includes: performing data domain matching on the left forward-looking SAR data and the right forward-looking SAR data in the bistatic SAR echo data; calculating the compensation factors corresponding to the left offset compensation strategy and the right offset compensation strategy according to the matching results, and calculating the delay factor corresponding to the hardware delay compensation strategy.

[0080] In this embodiment, the schematic diagram of the forward-looking cascade system (applicable to the vehicle-mounted forward-looking SAR imaging system) is as follows: Figure 3 As shown in the figure, the blue shaded area is the overlapping area of ​​the left front radar and the right front radar in terms of coverage. When the car moves forward, the left radar and the right radar have the same accumulation time for the overlapping area in the same period of time. Therefore, from the perspective of SAR images, the imaging results of the left and right radars in the overlapping area are approximately the same.

[0081] When performing data matching, the overlapping area is first extracted. The calculation formula for the overlapping area in the imaging grid is as follows:

[0082]

[0083] Among them, R is the detection distance of SAR, L is gapis the placement interval between the left and right radars, and θ is the angle between the edge of radar perception and the radar placement plane.

[0084] The result of extracting the overlapping area from the left front radar SAR image is recorded as LF, and the result of extracting the overlapping area from the right front radar SAR image is recorded as RF. The Pearson coefficient is used to calculate the correlation between LF and RF for matching. The formula is as follows:

[0085]

[0086] Among them, x i and i are the observed values ​​of LF and RF, respectively. and are the means of LF and RF, respectively. The closer r is to 1, the higher the matching degree of LF and RF is. Through multiple matching, the right front radar SAR data of the nth frame that best matches the left front radar SAR data of the mth frame is obtained. From the perspective of data acquisition, the acquisition time of the mth frame image in the left front radar echo data corresponds to the acquisition time of the nth frame image in the right front radar echo data.

[0087] like Figure 4 As shown in the figure, from the perspective of echo data model, Figure 4 The 5-second echo data model is taken as an example for analysis, where the horizontal axis represents the number of pulse transmissions, and the vertical axis represents the number of distance sampling points. When the pulse repetition frequency PRF is 10,000, the radar obtains a total of 50,000 pulses within a 5-second working cycle. In order to ensure the imaging quality, 500 pulses are used as an imaging data segment. In order to ensure the continuity and stability of SAR imaging, the step length STEP between frames is set to 100 pulses. The 1st to 500th echoes correspond to the first frame of the image, and the 101st to 600th echoes correspond to the second frame of the image, until the last frame of the image is processed. The delay factor corresponding to the hardware delay compensation strategy can be calculated from the frame interval, and the formula is as follows:

[0088]

[0089] Since Δt is caused by the hardware differences between different radar modules, the Δt between radar modules is fixed without any changes to the radar modules. Therefore, after the cascade system is deployed, it is only necessary to match and obtain the hardware delay at the initial startup. In subsequent operations, the delay that needs to be compensated is a fixed size.

[0090] In this embodiment, after matching the bistatic SAR echo data, the delay factor corresponding to the hardware delay compensation strategy is calculated, and after corresponding compensation is performed according to the delay factor and the offset compensation strategy, the left front-looking SAR image M and the right front-looking SAR image N can be obtained.

[0091] Specifically, in an implementation of this embodiment, step S200 includes the following steps:

[0092] Step S201, imaging the left front-looking SAR data in the bistatic SAR echo data based on the BP algorithm, and performing left offset compensation through an offset compensation strategy to obtain a left front-looking SAR image M;

[0093] Step S202, delay compensation is performed on the right front-looking SAR data in the bistatic SAR echo data by using the delay factor of the hardware delay compensation strategy, the right front-looking SAR data is imaged based on the BP algorithm, and right offset compensation is performed by using the offset compensation strategy to obtain a right front-looking SAR image N.

[0094] In this embodiment, after the hardware system collects left-front-looking and right-front-looking bistatic radar data, it first compensates for the hardware delay between radar modules, and the compensation factor is obtained in the algorithm of the data domain matching part. The SAR image mismatch problem caused by hardware differences is eliminated at the algorithm level. Among them, when the hardware delay compensation is performed on a certain radar data in the system, it can be the left-front-looking radar data or the right-front-looking radar data.

[0095] like Figure 5 As shown in FIG. 1 , taking the hardware delay compensation of the right forward-looking radar data as an example, the process of imaging the forward-looking area of ​​the collected bistatic SAR echo data includes:

[0096] Firstly, according to the delay factor of the hardware delay compensation strategy calculated during data domain matching, the right front-looking SAR data in the bistatic SAR echo data is delayed compensated to eliminate the SAR image mismatch problem caused by hardware differences; then, the left front-looking SAR data in the bistatic SAR echo data and the right front-looking SAR data after hardware delay compensation are imaged using the BP imaging algorithm; finally, with the center position of the bistatic radar as the imaging center, the amount by which the radar beam centers on the left and right sides offset the imaging center is compensated to obtain the left front-looking SAR image M and the right front-looking SAR image N.

[0097] Specifically, when performing offset compensation, the center line of the bistatic radar is first determined, and then the center lines of the left and right radar beams are respectively determined; by calculating the offset between the center line of the left radar beam and the center line of the bistatic radar, a left offset compensation is performed according to the calculated offset to obtain a left forward-looking SAR image M; similarly, by calculating the offset between the center line of the right radar beam and the center line of the bistatic radar, a right offset compensation is performed according to the calculated offset to obtain a right forward-looking SAR image N.

[0098] Specifically, in an implementation of this embodiment, step S200 further includes the following steps:

[0099] Step S203, performing false target elimination processing on the foresight region imaging by using a SVM algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated.

[0100] In this embodiment, after obtaining the left front-looking SAR image M and the right front-looking SAR image N, false target elimination processing is performed on the left front-looking SAR image M and the right front-looking SAR image N by using any one of two schemes (i.e., using the SVM algorithm or the clustering algorithm) to obtain a mask image after the false targets are eliminated.

[0101] Specifically, in an implementation of this embodiment, the first set of solutions adopts a clustering algorithm, and the clustering algorithm is used to perform false target elimination processing on the front-view area imaging to obtain a mask image after the false target is eliminated, including: obtaining a left front-view synthetic aperture radar image M at time i i and the right forward synthetic aperture radar image N i ; Construct a relationship matrix; pass the relationship matrix into the clustering algorithm, classify the true target points and false target points in the front view scene, and extract the real target points according to the binary classification results; based on the extracted real target points, construct a mask image of the final prediction result.

[0102] In this embodiment, in the first solution above, a relationship matrix is ​​constructed using the left front-looking SAR data and the right front-looking SAR data. The relationship matrix reflects the intensity changes of the left and right radars in the front-looking scene for the same coordinate point, that is, there is a large difference between the true target point and the false target point in the relationship matrix; and then a clustering algorithm is used to classify these true points and false points to obtain a mask (mask image) of the prediction result.

[0103] Specifically, the clustering algorithm relies on the construction of a relationship matrix. For two radar images acquired at the same time but from different perspectives, if the relationship matrix can accurately describe the characteristics of their real targets and false targets, the clustering algorithm can separate the real target points from the false target points based on the characteristics. As an example, the process of the clustering algorithm is as follows:

[0104] 1) The bistatic radar acquisition system acquires the forward-looking data of the left radar and the forward-looking data of the right radar at the same time;

[0105] 2) Imaging the acquisition results to obtain the left front SAR image M n and the right front SAR image N n . Take the left front image M at time i i and the right front image N i :

[0106] 3) Build a relationship matrix Among them, M i (m,n) and N i (m,n) refers to M i and N i The value of the pixel at (m,n).

[0107] 4) The relationship matrix is ​​passed into the kmeans clustering algorithm, and the real target points are extracted based on the binary classification results to construct a mask.

[0108] 5) Obtain the final forward-looking imaging result according to mask·M.

[0109] Specifically, in another implementation of the present embodiment, the second set of solutions adopts an SVM algorithm, and the false target elimination processing is performed on the forward-looking area imaging by the support vector machine algorithm to obtain a mask image after the false target is eliminated, including: constructing a data source X, taking (M(m,n), N(m,n)) as X(m,n) for the left forward-looking synthetic aperture radar image M and the right forward-looking synthetic aperture radar image N, traversing all pixel points to obtain a data source matrix X; and inputting X into a trained support vector machine model to obtain a mask image of a final prediction result.

[0110] In this embodiment, in the second solution above, a point in the real scene has different intensity reflections in the left front view and right front view SAR perspectives, thereby constructing a data set of real target points and false target points, and using this data set to train the SVM model. The trained SVM model classifies the corresponding points in the left front view and right front view SAR data to obtain the final prediction result mask.

[0111] Specifically, SVM is divided into regression model and classification model. Since the common platform dual-base radar belongs to homogeneous multi-source data, the feature vector is constructed according to the different intensity reflections of the same point target by radars with different viewing angles. Therefore, it belongs to the classification model prediction training. As an example, the training process using simulation data is as follows:

[0112] 1) Generate n random point targets, and record these random points as O i; and generate echo signals of the left radar and the right radar for these random point targets respectively.

[0113] 2) Imaging the echo signal to obtain the left front SAR image M n and the right front SAR image N n .

[0114] 3) Take the left front image M at time i i and the right front image N i , according to the random point target O i The specific process of making a data set is as follows:

[0115] a. Construct the data source matrix X and label matrix Y. The matrix dimensions are the same as the left and right front image dimensions.

[0116] b. The elements of the data source matrix X are the same pixel values ​​on the left and right front sides. That is, for M i and N i The same pixel point (m,n), take (M i (m,n),N i (m,n)) as X(m,n). Traverse all the pixels and finally get the data source matrix.

[0117] c. For M i and N i The same pixel (m,n) if it is in O i If it indicates that the point is the target point, then the corresponding position (m,n) in the Y matrix is ​​set to 1; otherwise, it is set to 0. Traverse all pixel points and finally get the label matrix.

[0118] 4) Data division: Divide the data source and the corresponding label matrix into training set and test set.

[0119] 5) Train the model and evaluate the model.

[0120] In this embodiment, the trained SVM can be used for real point extraction of forward-looking imaging, and the use process is as follows:

[0121] First, construct the data source X, treat the left front image M and the right front image N used for front-view imaging, take (M(m,n), N(m,n)) as X(m,n), and traverse all pixels to obtain the data source matrix.

[0122] Then, X is input into the trained SVM to obtain the prediction result mask.

[0123] Finally, the final forward-looking imaging result is obtained according to mask·M.

[0124] like Figure 1As shown, an embodiment of the present invention provides a common platform bistatic forward-looking SAR imaging method, comprising the following steps:

[0125] Step S300: extracting the real target points of the original image in the bistatic SAR echo data according to the mask image to obtain the final forward-looking SAR imaging.

[0126] In this embodiment, after the forward-looking area imaging is subjected to false target elimination processing by using an SVM algorithm or a clustering algorithm, the real target points are extracted from the original image in the bistatic SAR echo data according to the mask image after the false targets are eliminated, so that a forward-looking SAR image with high imaging quality can be obtained.

[0127] Specifically, in an implementation of this embodiment, step S300 includes the following steps:

[0128] Step S301, multiplying the left forward-looking SAR image M or the right forward-looking SAR image N in the original image by the mask image, extracting the real target points of the original image, eliminating the false target points of the original image, and obtaining the final forward-looking SAR imaging.

[0129] In this embodiment, the mask is used to multiply the original image M or N so as to extract the real points, thereby achieving the effect of eliminating false target points and obtaining a forward-looking SAR image with high imaging quality.

[0130] In order to verify the feasibility of the proposed solution, a simulation of the solution is performed in this embodiment.

[0131] The experimental scenario is vehicle-mounted SAR imaging. The system works in SAR forward-looking mode, and the interval between radars is 0.5m.

[0132] Firstly, a certain number of random points are generated to simulate point targets in actual optical scenes;

[0133] Next, the echo signals of the left radar and the right radar about the target point are generated respectively, and random factors are added to the echo signals to describe the scattering characteristics of different target points;

[0134] Finally, the proposed forward-looking imaging algorithm is used for processing, including the clustering algorithm scheme and the SVM scheme.

[0135] like Figure 6 As shown in Figure 1, 9 random points were generated in the experimental scene; the results of BP forward imaging based on the 9 random points are shown in Figure 1. Figure 7 and Figure 8 As shown, Figure 7 The middle is the result of BP forward imaging of the echo data of 9 target points by the radar on the right. Figure 8The middle is the result of BP imaging of the echo data of 9 target points on the left; it can be seen that when the BP algorithm processes forward-looking imaging, false point targets appear because of the same Doppler process, and the false points and real points are symmetrical about the center of the left and right radar beams.

[0136] like Fig. 9 and Fig.10 As shown, Fig. 9 The figure below shows the result of using the dual-base SVM scheme for forward imaging. Fig.10 The figure shows the result of forward-looking imaging using the Kmeans clustering algorithm. The proposed method solves the problem of false point targets when the BP imaging algorithm processes single-base forward-looking imaging.

[0137] like Figures 11 to 13 As shown, Fig.11 The optical scene diagram of the measured data is shown in Fig.12 The image shown in the figure is the result of BP imaging by the radar on the left. Fig.13 The result of the right radar BP imaging is shown in FIG. It can be seen that the left and right BP imaging results have obvious symmetrical false point target problems, which shows that a high-quality forward-looking SAR image can be obtained in this embodiment.

[0138] like Fig.14 As shown in Figure 14, the result of the proposed forward-looking imaging algorithm is shown. The result shows that the reflection coefficient of the stainless steel fence in the optical scene is strong, and a strong point will appear in the imaging result, corresponding to the red frame part in the imaging result. The imaging result shows that the proposed method can achieve reliable vehicle-mounted forward-looking imaging.

[0139] In this embodiment, different SAR viewing angles provided by two radars are utilized, and a clustering algorithm or SVM solution is introduced to solve the problem of false points caused by the same Doppler history of the target during forward-looking radar BP imaging.

[0140] This embodiment achieves the following technical effects through the above technical solution:

[0141] This embodiment acquires bistatic synthetic aperture radar echo data by collecting left forward-looking radar data, right forward-looking radar data, speed and acceleration data; performs forward-looking region imaging on the acquired bistatic synthetic aperture radar echo data, performs false target elimination processing on the forward-looking region imaging by a support vector machine algorithm or a clustering algorithm, and acquires a mask image after the false targets are eliminated; the real target points of the original image in the bistatic synthetic aperture radar echo data can be extracted according to the mask image, and the final forward-looking synthetic aperture radar imaging can be acquired; this embodiment solves the problem of false points existing in the back-propagation algorithm in processing forward-looking imaging by introducing a support vector machine and a clustering algorithm, and finally realizes reliable common-platform forward-looking imaging, thereby improving the imaging quality of the forward-looking synthetic aperture radar.

[0142] Exemplary Devices

[0143] Based on the above embodiments, the present invention further provides a common platform dual-base forward-looking SAR imaging system, comprising:

[0144] Left front radar, used to collect left front-looking radar data;

[0145] Right front radar, used to collect right front-looking radar data;

[0146] Inertial navigation module, used to collect the speed of the radar at the time of operation;

[0147] and the main control end, which is used to realize bistatic forward-looking synthetic aperture radar imaging;

[0148] Wherein, the main control end includes:

[0149] A data acquisition module is used to collect left front-looking radar data, right front-looking radar data, speed and acceleration data to obtain bistatic synthetic aperture radar echo data;

[0150] The mask image module is used to perform forward-viewing area imaging on the collected bistatic synthetic aperture radar echo data, and perform false target elimination processing on the forward-viewing area imaging by a support vector machine algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated;

[0151] The synthetic aperture radar imaging module is used to extract the real target points of the original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain the final forward-looking synthetic aperture radar imaging.

[0152] This embodiment achieves the following technical effects through the above technical solution:

[0153] This embodiment acquires bistatic synthetic aperture radar echo data by collecting left forward-looking radar data, right forward-looking radar data, speed and acceleration data; performs forward-looking region imaging on the acquired bistatic synthetic aperture radar echo data, performs false target elimination processing on the forward-looking region imaging by a support vector machine algorithm or a clustering algorithm, and acquires a mask image after the false targets are eliminated; the real target points of the original image in the bistatic synthetic aperture radar echo data can be extracted according to the mask image, and the final forward-looking synthetic aperture radar imaging can be acquired; this embodiment solves the problem of false points existing in the back-propagation algorithm in processing forward-looking imaging by introducing a support vector machine and a clustering algorithm, and finally realizes reliable common-platform forward-looking imaging, thereby improving the imaging quality of the forward-looking synthetic aperture radar.

[0154] Based on the above embodiment, the present invention further provides a terminal, whose principle block diagram can be as follows: Fig.15 shown.

[0155] The terminal includes: a processor, a memory, an interface, a display screen and a communication module connected through a system bus; wherein the processor of the terminal is used to provide computing and control capabilities; the memory of the terminal includes a storage medium and an internal memory; the storage medium stores an operating system and a computer program; the internal memory provides an environment for the operation of the operating system and the computer program in the storage medium; the interface is used to connect to external devices; the display screen is used to display corresponding information; and the communication module is used to communicate with a cloud server or other devices.

[0156] When the computer program is executed by a processor, it is used to implement the operation of a common platform bistatic forward-looking SAR imaging method.

[0157] It can be understood by those skilled in the art that Fig.15 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the scheme of the present invention, and does not constitute a limitation on the terminal to which the scheme of the present invention is applied. The specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0158] In one embodiment, a terminal is provided, comprising: a processor and a memory, wherein the memory stores a common platform bistatic forward-looking SAR imaging program, and when the common platform bistatic forward-looking SAR imaging program is executed by the processor, it is used to implement the operation of the above common platform bistatic forward-looking SAR imaging method.

[0159] In one embodiment, a storage medium is provided, wherein the storage medium stores a common platform bistatic forward-looking SAR imaging program, and when the common platform bistatic forward-looking SAR imaging program is executed by a processor, it is used to implement the operation of the above common platform bistatic forward-looking SAR imaging method.

[0160] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and volatile memory.

[0161] In summary, the present invention provides a common platform bistatic forward-looking SAR imaging method and system, comprising: collecting left forward-looking radar data, right forward-looking radar data, speed and acceleration data to obtain bistatic synthetic aperture radar echo data; performing forward-looking area imaging on the collected bistatic synthetic aperture radar echo data, performing false target elimination processing on the forward-looking area imaging through a support vector machine algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated; extracting real target points of the original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain the final forward-looking synthetic aperture radar imaging; the present invention solves the problem of false points existing in the back-propagation algorithm when processing forward-looking imaging by introducing a support vector machine and a clustering algorithm, and finally realizes reliable common platform forward-looking imaging.

[0162] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A common platform bistatic forward-looking SAR imaging method, characterized in that: include: Collect left front-looking radar data, right front-looking radar data, velocity and acceleration data to obtain bistatic synthetic aperture radar echo data; Performing forward-view area imaging on the collected bistatic synthetic aperture radar echo data, and performing false target elimination processing on the forward-view area imaging by using a support vector machine algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated; The real target points of the original image in the bistatic synthetic aperture radar echo data are extracted according to the mask image to obtain the final forward-looking synthetic aperture radar imaging.

2. The common platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that: The collecting of left front-view radar data, right front-view radar data, speed and acceleration data previously includes: A connection request is initiated to two monostatic radar modules through the communication module, and when the two monostatic radar modules are successfully connected and synchronized, a dual-base forward-looking synthetic aperture radar imaging system is operated.

3. The common platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that: The forward-looking area imaging of the collected bistatic synthetic aperture radar echo data includes: Imaging the left front-looking synthetic aperture radar data in the bistatic synthetic aperture radar echo data based on a back-propagation algorithm, and performing left offset compensation through an offset compensation strategy to obtain a left front-looking synthetic aperture radar image M; The right front-looking synthetic aperture radar data in the bistatic synthetic aperture radar echo data is delayed compensated by the delay factor of the hardware delay compensation strategy, the right front-looking synthetic aperture radar data is imaged based on the back propagation algorithm, and right offset compensation is performed by the offset compensation strategy to obtain a right front-looking synthetic aperture radar image N.

4. The common platform bistatic forward-looking SAR imaging method according to claim 3, characterized in that: The forward-looking region imaging of the collected bistatic synthetic aperture radar echo data may include: Performing data domain matching on left front-looking synthetic aperture radar data and right front-looking synthetic aperture radar data in the bistatic synthetic aperture radar echo data; The compensation factors corresponding to the left offset compensation strategy and the right offset compensation strategy are calculated according to the matching result, and the delay factor corresponding to the hardware delay compensation strategy is calculated.

5. The common platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that: The method of performing false target elimination processing on the front view area imaging by using a clustering algorithm to obtain a mask image after the false targets are eliminated includes: Get the left front-view synthetic aperture radar image M at time i i and the right forward synthetic aperture radar image N i ; Construct the relationship matrix: Among them, M i (m,n) and N i (m,n) represent M respectively i and N i The value at the pixel (m,n); The relationship matrix is ​​passed into the clustering algorithm to classify the true target points and false target points in the forward-looking scene, and the true target points are extracted according to the binary classification results; Based on the extracted real target points, a mask image of the final prediction result is constructed.

6. The common platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that: The method of performing false target elimination processing on the front view area imaging by using a support vector machine algorithm to obtain a mask image after the false targets are eliminated includes: Construct a data source X, take (M(m,n),N(m,n)) as X(m,n) for the left front synthetic aperture radar image M and the right front synthetic aperture radar image N, and traverse all pixel points to obtain a data source matrix X; Input X into the trained support vector machine model to obtain the mask image of the final prediction result.

7. The common platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that: The step of extracting the real target point of the original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain the final forward-looking synthetic aperture radar imaging comprises: The mask image is used to multiply the left forward-looking synthetic aperture radar image M or the right forward-looking synthetic aperture radar image N in the original image, to extract the real target points of the original image, to eliminate the false target points of the original image, and to obtain the final forward-looking synthetic aperture radar imaging.

8. A common platform dual-static forward-looking SAR imaging system, characterized in that: include: Left front radar, used to collect left front-looking radar data; Right front radar, used to collect right front-looking radar data; Inertial navigation module, used to collect the speed of the radar at the time of operation; and the main control end, which is used to realize dual-base forward-looking synthetic aperture radar imaging; Wherein, the main control end includes: A data acquisition module is used to collect left front-looking radar data, right front-looking radar data, speed and acceleration data to obtain bistatic synthetic aperture radar echo data; The mask image module is used to perform forward-viewing area imaging on the collected bistatic synthetic aperture radar echo data, and perform false target elimination processing on the forward-viewing area imaging by a support vector machine algorithm or a clustering algorithm to obtain a mask image after the false targets are eliminated; The synthetic aperture radar imaging module is used to extract the real target points of the original image in the bistatic synthetic aperture radar echo data according to the mask image to obtain the final forward-looking synthetic aperture radar imaging.

9. A terminal, characterized in that: include: A processor and a memory, wherein the memory stores a common platform bistatic forward-looking SAR imaging program, and when the common platform bistatic forward-looking SAR imaging program is executed by the processor, it is used to implement the operation of the common platform bistatic forward-looking SAR imaging method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a common platform bistatic forward-looking SAR imaging program, which, when executed by a processor, is used to implement the operation of the common platform bistatic forward-looking SAR imaging method according to any one of claims 1 to 7.

Citation Information

Patent Citations

  • Foresight double-base mode-based interference SAR imaging technology

    CN107728144A

  • Equivalent-slant-distance-based double-base time-varying acceleration foresight SAR imaging method

    CN108469612A

  • Fast time domain imaging method based on accelerated trajectory bistatic forward-looking synthetic aperture radar

    CN116359921A

  • Bistatic inverse synthetic aperture radar imaging

    US20140285371A1

Cited By

  • Fast hierarchical Kirchhoff imaging method for MIMO array radar

    CN122194152A