A Common-Platform Bistatic Forward-Looking SAR Imaging Method and System

Abstract

This invention discloses a shared-platform bistatic forward-looking SAR imaging method and system, comprising: acquiring left forward-looking radar data, right forward-looking radar data, velocity and acceleration data to obtain bistatic synthetic aperture radar echo data; performing forward-looking region imaging on the acquired bistatic synthetic aperture radar echo data; performing false target removal processing on the forward-looking region imaging using a support vector machine algorithm or a clustering algorithm to obtain a masked image after false target removal; extracting the real target points of the original image from the bistatic synthetic aperture radar echo data based on the masked image to obtain the final forward-looking synthetic aperture radar image; this invention solves the problem of false points in the backpropagation algorithm when processing forward-looking imaging by introducing support vector machines and clustering algorithms, ultimately achieving reliable shared-platform forward-looking imaging.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a common-platform bistatic forward-looking SAR imaging method and system. Background Technology

[0002] Synthetic Aperture Radar (SAR) imaging is a method of acquiring target information using radar technology. Unlike traditional optical imaging techniques, SAR actively transmits microwave signals, and the target reflects the echo, thereby generating an image containing target information. A key feature of SAR imaging is its ability to perform imaging at any time and under any weather conditions. It is widely used in military and aerospace fields, such as geological research, ocean observation, and rescue reconnaissance. In the civilian sector, SAR technology is also developing towards miniaturization, lightweight design, high resolution, and bistatic / multistatic capabilities.

[0003] Current synthetic aperture radar (SAR) imaging studies primarily focus on frontal and oblique forward-looking modes. However, in forward-looking scenarios, two point targets symmetrically positioned along the beam centerline exhibit identical Doppler trajectories as the radar advances, leading to left-right blurring in the forward-looking imaging results. Current solutions in forward-looking imaging include monopulse forward-looking imaging, array antenna super-angle resolution technology, and dual-base SAR radar forward-looking imaging. Monopulse forward-looking imaging technology has a relatively simple hardware structure and execution algorithm. However, when applied to multiple strongly scattering targets or continuously distributed targets, the limited number of monopulse angle measurement channels prevents the differentiation of multiple targets within the same resolution cell, resulting in "angle flicker" and reduced imaging quality. Sum-difference amplitude angle measurement based on reconstructed estimates improves the positioning accuracy of targets within the antenna coverage area, validating the proposed method's effectiveness in improving forward-looking image clarity. However, symmetrical targets appearing on the left and right sides of the forward-looking direction exhibit "left-right blurring" due to their identical Doppler frequencies, making them indistinguishable in monopulse imaging because they are located within the same resolution cell. Super-angle resolution (SAR) technology using array antennas relies on large-aperture antennas to achieve azimuth super-resolution. It improves image contrast and signal-to-noise ratio by coherently smoothing multiple images. However, due to the system's complexity and the high requirements for platform size, it is not suitable for small to medium-sized platforms such as those used in automobiles. Dual-base synthetic aperture radar (DAR) forward-looking imaging typically addresses forward-looking ambiguity by placing transmitters and receivers on separate platforms. However, the presence of remote transmission and reception and the resulting dependence on multiple platforms complicate the overall structural model, making it impossible to complete transmission, reception, and processing on a single platform.

[0004] Therefore, existing technologies still need improvement. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a common platform dual-base forward-looking SAR imaging method and system to address the shortcomings of existing forward-looking synthetic aperture radar imaging methods, which suffer from low imaging quality and complex overall structural models.

[0006] The technical solution adopted by this 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] Data from the left forward-looking radar, right forward-looking radar, velocity, and acceleration are collected to obtain bistatic synthetic aperture radar echo data.

[0009] The acquired bistatic synthetic aperture radar echo data is used to perform forward-looking region imaging. The false target removal process is performed on the forward-looking region imaging using a support vector machine algorithm or a clustering algorithm to obtain a mask image after false target removal.

[0010] Based on the masked image, the true target points of the original image in the bistatic synthetic aperture radar echo data are extracted to obtain the final forward-looking synthetic aperture radar image.

[0011] In one implementation, the acquisition of left forward-looking radar data, right forward-looking radar data, velocity, and acceleration data includes, prior to:

[0012] The communication module initiates a connection request to two single-base radar modules, and when the two single-base radar modules successfully connect and synchronize, the bistatic forward-looking synthetic aperture radar imaging system is run.

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

[0014] The left forward-looking synthetic aperture radar data in the dual-base synthetic aperture radar echo data is imaged based on the backpropagation algorithm, and the left offset is compensated by the offset compensation strategy to obtain the left forward-looking synthetic aperture radar image M.

[0015] The delay factor of the hardware delay compensation strategy is used to compensate for the delay of the right forward-looking synthetic aperture radar data in the dual-base synthetic aperture radar echo data. The right forward-looking synthetic aperture radar data is imaged based on the backpropagation algorithm. The right offset is compensated for by the offset compensation strategy to obtain the right forward-looking synthetic aperture radar image N.

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

[0017] Data domain matching is performed on the left forward-looking synthetic aperture radar data and the right forward-looking synthetic aperture radar data in the dual-base synthetic aperture radar echo data.

[0018] The compensation factors corresponding to the left offset compensation strategy and the right offset compensation strategy are calculated based on the matching results, and the delay factor corresponding to the hardware delay compensation strategy is also calculated.

[0019] In one implementation, the step of performing false target removal processing on the forward-looking region imaging using a clustering algorithm to obtain a mask image after false target removal includes:

[0020] Acquire the left forward-looking synthetic aperture radar image M at time i. i And right forward-looking synthetic aperture radar image N i ;

[0021] Constructing the relation matrix:

[0022]

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

[0024] The relationship matrix is ​​fed into the clustering algorithm to classify the real target points and false target points in the forward-looking scene, and the real target points are extracted based on the binary classification results.

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

[0026] In one implementation, the step of performing false target removal processing on the forward-looking region imaging using a support vector machine algorithm to obtain a mask image after false target removal includes:

[0027] Construct a data source X. For the left forward-looking synthetic aperture radar image M and the right forward-looking synthetic aperture radar image N, take (M(m,n), N(m,n)) as X(m,n) and traverse all pixels to obtain the 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, the step of extracting the real target points of the original image from the bistatic synthetic aperture radar echo data based on the mask image to obtain the final forward-looking synthetic aperture radar image 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, eliminate the false target points of the original image, and obtain the final forward-looking synthetic aperture radar image.

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

[0032] The left front radar is used to collect data from the left forward-looking radar.

[0033] Right front radar, used to collect data from the right front-view radar;

[0034] The inertial navigation module is used to collect the speed at the moment the radar is operating.

[0035] And the main control unit, used to realize bistatic forward-looking synthetic aperture radar imaging;

[0036] The main control terminal includes:

[0037] The data acquisition module is used to acquire data from the left forward-looking radar, the right forward-looking radar, velocity and acceleration data, and obtain bistatic synthetic aperture radar echo data.

[0038] The mask image module is used to perform forward-looking region imaging on the acquired bi-base synthetic aperture radar echo data, and to perform false target elimination processing on the forward-looking region imaging through support vector machine algorithm or clustering algorithm to obtain a mask image after false target elimination.

[0039] The synthetic aperture radar imaging module is used to extract the real target points of the original image from the bi-base synthetic aperture radar echo data based on the mask image, so as to obtain the final forward-looking synthetic aperture radar image.

[0040] Thirdly, 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 the common-platform bistatic forward-looking SAR imaging program, when executed by the processor, is used to implement the operation of the common-platform bistatic forward-looking SAR imaging method as described in the first aspect.

[0041] Fourthly, the present invention also provides a medium, which is a computer-readable storage medium storing 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 as described in the first aspect.

[0042] The present invention, by employing the above technical solution, has the following effects:

[0043] This invention obtains bistatic synthetic aperture radar (SAR) echo data by collecting data from left and right forward-looking radars, as well as velocity and acceleration data. It then performs forward-looking region imaging on the collected SAR echo data and uses a support vector machine (SVM) algorithm or clustering algorithm to remove false targets from the forward-looking region imaging, obtaining a masked image after false target removal. The true target points of the original image in the SAR echo data can be extracted from the masked image to obtain the final SAR image. This invention solves the problem of false points in the backpropagation algorithm when processing forward-looking imaging by introducing SVM and clustering algorithms, ultimately achieving reliable co-platform forward-looking imaging and improving the imaging quality of SAR. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0045] Figure 1 This is a flowchart of the dual-base forward-looking SAR imaging method of the common platform of the present invention.

[0046] Figure 2 This is a schematic diagram of the dual-base forward-looking SAR imaging system of the common platform of the present invention.

[0047] Figure 3 This is a schematic diagram of the front-side cascaded system in this invention.

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

[0049] Figure 5 This is a flowchart of the forward-looking imaging processing algorithm in this invention.

[0050] Figure 6 This is a schematic diagram of the nine random point targets in this invention.

[0051] Figure 7 This is a schematic diagram of the radar BP imaging results on the right side in this invention.

[0052] Figure 8 This is a schematic diagram of the radar BP imaging results on the left side in this invention.

[0053] Figure 9 This is a schematic diagram of the processing results of the support vector machine scheme in this invention.

[0054] Figure 10This is a schematic diagram of the Kmeans clustering algorithm processing results in this invention.

[0055] Figure 11 This is an optical scene diagram based on actual measured data from this invention.

[0056] Figure 12 This is a schematic diagram of the radar BP imaging results on the left side in this invention.

[0057] Figure 13 This is a schematic diagram of the radar BP imaging results on the right side in this invention.

[0058] Figure 14 This is a schematic diagram of the forward-looking imaging result of the method proposed in this invention.

[0059] Figure 15 This is a functional schematic diagram of the terminal in one implementation of the present invention.

[0060] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0062] Exemplary methods

[0063] Current synthetic aperture radar (SAR) imaging studies primarily focus on frontal and oblique forward-looking modes. However, in forward-looking scenarios, two point targets symmetrically positioned along the beam centerline exhibit identical Doppler trajectories as the radar advances, leading to left-right blurring in the forward-looking imaging results. Current solutions in forward-looking imaging include monopulse forward-looking imaging, array antenna super-angle resolution technology, and dual-base SAR radar forward-looking imaging. Monopulse forward-looking imaging technology has a relatively simple hardware structure and execution algorithm. However, when applied to multiple strongly scattering targets or continuously distributed targets, the limited number of monopulse angle measurement channels prevents the differentiation of multiple targets within the same resolution cell, resulting in "angle flicker" and reduced imaging quality. Sum-difference amplitude angle measurement based on reconstructed estimates improves the positioning accuracy of targets within the antenna coverage area, validating the proposed method's effectiveness in improving forward-looking image clarity. However, symmetrical targets appearing on the left and right sides of the forward-looking direction exhibit "left-right blurring" due to their identical Doppler frequencies, making them indistinguishable in monopulse imaging because they are located within the same resolution cell. Super-angle resolution (SAR) technology using array antennas relies on large-aperture antennas to achieve azimuth super-resolution. It improves image contrast and signal-to-noise ratio by coherently smoothing multiple images. However, due to the system's complexity and the high requirements for platform size, it is not suitable for small to medium-sized platforms such as those used in automobiles. Dual-base synthetic aperture radar (DAR) forward-looking imaging typically addresses forward-looking ambiguity by placing transmitters and receivers on separate platforms. However, the presence of remote transmission and reception and the resulting dependence on multiple platforms complicate the overall structural model, making it impossible to complete transmission, reception, and processing on a single platform.

[0064] To address the above technical problems, this invention provides a co-platform bistatic forward-looking SAR imaging method. This method acquires bistatic synthetic aperture radar (SAR) echo data by collecting data from the left and right forward-looking radars, as well as velocity and acceleration data. Forward-looking region imaging is then performed on the acquired SAR echo data. False target removal processing is then performed on the forward-looking region imaging using a support vector machine (SVM) algorithm or clustering algorithm to obtain a masked image after false target removal. The true target points of the original image in the SAR echo data can be extracted from the masked image to obtain the final SAR image. This invention solves the problem of false points in the backpropagation algorithm when processing forward-looking imaging by introducing SVM and clustering algorithms, ultimately achieving reliable co-platform forward-looking imaging and improving the imaging quality of SAR.

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

[0066] Step S100: Collect left forward-looking radar data, right forward-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 based on a system implementation of a common-platform bistatic forward-looking SAR imaging method provided in this embodiment.

[0068] Overall, this embodiment first uses a communication module to build a multi-sensor cascaded working system (including building a cascaded acquisition system for forward-looking dual millimeter-wave radars and a real-time velocity acquisition system); then, it performs forward-looking region imaging on the acquired bistatic SAR (Synthetic Aperture Radar) echo data; finally, it introduces the SVM (Support Vector Machine, a type of generalized linear classifier that performs binary classification of data in a supervised learning manner) algorithm and clustering algorithm to solve the problem of false points in the BP (backpropagation) algorithm when dealing with forward-looking imaging, and finally achieves reliable co-platform forward-looking imaging.

[0069] The hardware block diagram of a common-platform dual-base forward-looking SAR imaging system provided in this embodiment is as follows: Figure 2 As shown, the system includes: a left front-side radar for acquiring left-side forward-looking radar data; a right front-side radar for acquiring right-side forward-looking radar data; an inertial navigation module for acquiring the radar's velocity at the time of operation; and a main control unit for implementing bistatic forward-looking SAR imaging. The overall control commands for the system are issued by the main control unit. A set of components consisting of the control unit and radar is referred to as a single-base radar module. The designed forward-looking SAR imaging hardware system requires two single-base radar modules, respectively mounted on the left and right front sides of the vehicle platform. The introduction of the inertial navigation module helps the system acquire the radar's velocity data at the time of operation.

[0070] In this embodiment, before collecting data from the left forward-looking radar, the right forward-looking radar, and velocity and acceleration data, it is also necessary to establish a connection between the two single-base radar modules and the main control terminal.

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

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

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

[0074] First, the master control unit initiates connection requests through the left and right monopolar radar modules of the communication module. If the connection is successful, the monopolar radar module enters the ready stage. If the connection fails, the control unit / industrial control computer rejects the connection; the master control unit then re-initiates the connection request. This avoids asynchronous radar data acquisition caused by command issuance.

[0075] If the two single-base radar modules successfully connect and synchronize, the system starts running and collects data from the left front-looking radar, the right front-looking radar, and data such as velocity and acceleration. 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 velocity is the velocity data at the radar's operating moment obtained through the inertial navigation module. The acceleration is calculated based on the acquired velocity data.

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

[0077] Step S200: Forward-looking region imaging is performed on the acquired bistatic SAR echo data. False targets are eliminated by using an SVM algorithm or clustering algorithm to obtain a masked image after false target elimination.

[0078] In this embodiment, before performing forward-looking region imaging on the acquired bistatic SAR echo data, it is necessary to perform data matching on 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 one implementation of this embodiment, the step of performing forward-looking region imaging on the acquired bistatic SAR echo data includes: performing data domain matching on the left and 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 based on the matching results; and calculating the delay factor corresponding to the hardware delay compensation strategy.

[0080] In this embodiment, a schematic diagram of the forward-looking cascaded system (suitable for vehicle-mounted forward-looking SAR imaging systems) is shown below. Figure 3 As shown in the image, the blue shaded area represents the overlapping region of the left and right front radars in terms of coverage. As the car moves forward, the left and right radars accumulate data in the overlapping region for the same amount of time. Therefore, from a SAR image perspective, the imaging results of the left and right radars in the overlapping region are approximately the same.

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

[0082]

[0083] Where R is the SAR detection range, and L gapθ represents the placement interval between the two radars on the left and right sides, and θ is the angle between the edge sensed by the radar and the plane on which the radar is placed.

[0084] Let LF be the result of extracting the overlapping region from the left front radar SAR image, and RF be the result of extracting the overlapping region from the right front radar SAR image. The correlation between LF and RF is calculated using the Pearson coefficient for matching, as shown in the following formula:

[0085]

[0086] Where, x i and y i These are the observations for LF and RF, respectively. and These are the mean values ​​of LF and RF, respectively. The closer r is to 1, the higher the matching degree of LF and RF. Through multiple matching, the nth frame of right front radar SAR data that best matches the mth frame of left front radar SAR data 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, from the perspective of echo data model, with Figure 4 Taking a 5-second echo data model as an example, the horizontal axis represents the number of pulse transmissions, and the vertical axis represents the number of distance sampling points. With a pulse repetition frequency (PRF) of 10,000, the radar acquires 50,000 pulses within a 5-second working cycle. To ensure imaging quality, 500 pulses are considered as one imaging data segment. To ensure continuous and stable SAR imaging, the step length (STEP) between frames is set to 100 pulses. Echoes 1 to 500 correspond to the first frame, echoes 101 to 600 correspond to the second frame, and so on until the last frame is processed. The delay factor corresponding to the hardware delay compensation strategy can be calculated from the frame interval, using the following formula:

[0088]

[0089] Because Δt is caused by hardware differences between different radar modules, Δt between radar modules is fixed without any changes to the radar modules. Therefore, after the cascaded system is deployed, hardware latency matching is only required during the initial startup; thereafter, the latency compensation is a fixed amount.

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

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

[0092] Step S201: Based on the BP algorithm, the left forward-looking SAR data in the bistatic SAR echo data is imaged, and the left offset is compensated by the offset compensation strategy to obtain the left forward-looking SAR image M.

[0093] Step S202: Delay compensation is performed on the right forward-looking SAR data in the dual-base SAR echo data using the delay factor of the hardware delay compensation strategy, and the right forward-looking SAR data is imaged based on the BP algorithm. Right offset compensation is then performed using the offset compensation strategy to obtain the right forward-looking SAR image N.

[0094] In this embodiment, after the hardware system acquires left and right forward-looking bistatic radar data, it first compensates for the hardware delay between radar modules by performing compensation on a specific radar module. The compensation factor is obtained through an algorithm in the data domain matching section. This algorithm eliminates the SAR image mismatch problem caused by hardware differences. Specifically, when performing hardware delay compensation on a specific radar module in the system, it can be either left or right forward-looking radar data.

[0095] like Figure 5 As shown, taking hardware delay compensation for right forward-looking radar data as an example, the process of imaging the forward-looking region from the acquired bistatic SAR echo data includes:

[0096] First, based on the delay factor of the hardware delay compensation strategy calculated during data domain matching, delay compensation is performed on the right forward-looking SAR data in the bistatic SAR echo data to eliminate the SAR image mismatch problem caused by hardware differences. Then, the BP imaging algorithm is used to image the left forward-looking SAR data and the hardware-delay-compensated right forward-looking SAR data in the bistatic SAR echo data. Finally, with the center position of the bistatic radar as the imaging center, the amount of offset of the radar beam centers on the left and right sides from the imaging center is compensated to obtain the left forward-looking SAR image M and the right forward-looking SAR image N.

[0097] Specifically, when performing offset compensation, the centerline of the bistatic radar is first determined, and then the centerlines of the left and right radar beams are determined respectively. By calculating the offset between the centerline of the left radar beam and the centerline of the bistatic radar, left offset compensation is performed based on the calculated offset to obtain the left forward-looking SAR image M. Similarly, by calculating the offset between the centerline of the right radar beam and the centerline of the bistatic radar, right offset compensation is performed based on the calculated offset to obtain the right forward-looking SAR image N.

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

[0099] Step S203: Perform false target removal processing on the forward-looking region imaging using an SVM algorithm or a clustering algorithm to obtain a mask image after false target removal.

[0100] In this embodiment, after obtaining the left forward-looking SAR image M and the right forward-looking SAR image N, false target removal processing is performed on the left forward-looking SAR image M and the right forward-looking SAR image N using either of the two schemes (i.e., using the SVM algorithm or the clustering algorithm) to obtain the mask image after false target removal.

[0101] Specifically, in one implementation of this embodiment, the first scheme employs a clustering algorithm. The step of performing false target removal processing on the forward-looking region imaging using the clustering algorithm to obtain a masked image after false target removal includes: acquiring the left forward-looking synthetic aperture radar image M at time i. i And right forward-looking synthetic aperture radar image N i Construct a relation matrix; input the relation matrix into the clustering algorithm to classify the real target points and false target points in the forward-looking scene, and extract the real target points based on the binary classification results; construct the mask image of the final prediction result based on the extracted real target points.

[0102] In this embodiment, in the first scheme above, a relation matrix is ​​constructed using left forward-looking SAR data and right forward-looking SAR data. This relation matrix reflects the intensity change of the radar on the left and right sides of the forward-looking scene for the same coordinate point, that is, there is a large difference between the real target point and the false target point in the relation matrix. Then, a clustering algorithm is used to classify these real points and false points to obtain the mask of the prediction result.

[0103] Specifically, clustering algorithms rely on the construction of a relation matrix. For two radar images acquired at the same time but from different viewpoints, if the relation matrix can accurately depict the features of real and false targets, the clustering algorithm can separate real and false target points based on these features. As an example, the clustering algorithm process is as follows:

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

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

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

[0107] 4) Pass the relation matrix into the kmeans clustering algorithm, extract the true target points based on the binary classification results, and construct a mask.

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

[0109] Specifically, in another implementation of this embodiment, the second scheme adopts the SVM algorithm. The step of performing false target removal processing on the forward-looking region imaging using the support vector machine algorithm to obtain the mask image after false target removal includes: 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 pixels to obtain the data source matrix X; and inputting X into the trained support vector machine model to obtain the mask image of the final prediction result.

[0110] In this embodiment, in the second scheme described above, a point in the real scene has different intensities reflected in the left and right forward-looking SAR views. This is used to construct a dataset of real and false target points, and an SVM model is trained using this dataset. The trained SVM model classifies corresponding points in the left and right forward-looking SAR data to obtain the final prediction result mask.

[0111] Specifically, SVM is divided into regression models and classification models. Since the data from a shared-platform bistatic radar is homogeneous and multi-source, feature vectors are constructed based on the different intensity responses of radars from different perspectives to the same target. Therefore, this falls under the category of classification model prediction training. As an example, the training process using simulation data is as follows:

[0112] 1) Generate n random target points, denoted as O(n). iThey also generate echo signals from the left and right radars for these random point targets, respectively.

[0113] 2) Obtain the left front SAR image M by imaging the echo signal. n and right anterior SAR image N n .

[0114] 3) Take the left front image M at time i. i and right anterior image N i According to the random point target O i The specific process for creating the dataset is as follows:

[0115] a. Construct a data source matrix X and a label matrix Y, with the matrix dimensions being the same as the dimensions of the left and right front images.

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

[0117] c. For M i and N i For the same pixel (m,n), if it is in O i If a point is identified as a target point, the corresponding position (m, n) in the Y matrix is ​​set to 1; otherwise, it is set to 0. This process is repeated for all pixels to obtain the final label matrix.

[0118] 4) Data partitioning: 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 ground truth point extraction in forward-looking imaging, and the process is as follows:

[0121] First, construct the data source X. Take the left frontal image M and the right frontal image N to be used for forward vision imaging, and take (M(m,n), N(m,n)) as X(m,n). Iterate through 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 based on mask·M.

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

[0125] Step S300: Extract the real target points of the original image from the bistatic SAR echo data based on the mask image to obtain the final forward-looking SAR image.

[0126] In this embodiment, after performing false target removal processing on the forward-looking region imaging using the SVM algorithm or clustering algorithm, the real target points are extracted from the original image in the bistatic SAR echo data based on the mask image after false target removal, thus obtaining a high-quality forward-looking SAR image.

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

[0128] Step S301: Multiply the left forward-looking SAR image M or the right forward-looking SAR image N in the original image using the mask image to extract the true target points of the original image, eliminate the false target points of the original image, and obtain the final forward-looking SAR image.

[0129] In this embodiment, the real points can be extracted by multiplying the original image M or N with a mask, thereby eliminating false target points and obtaining a high-quality forward-looking SAR image.

[0130] To verify the feasibility of the proposed solution, a simulation was performed on the solution in this embodiment.

[0131] The experimental scenario was vehicle-mounted SAR imaging, with the system operating in SAR forward-looking mode and the radar spacing being 0.5m.

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

[0133] Next, echo signals from the left radar and the right radar about the target point are generated respectively. A random factor is 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 clustering algorithm scheme and SVM scheme.

[0135] like Figure 6 As shown, nine random points were generated in the experimental scene; the results of BP forward-looking imaging based on these nine random points are as follows. Figure 7 and Figure 8 As shown, where, Figure 7 The image in the middle shows the result of forward-looking (BP) imaging of the echo data from nine target points by the radar on the right. Figure 8The middle image shows the result of BP imaging of the echo data of 9 target points on the left. It can be seen that the BP algorithm produces false point targets when processing forward-looking imaging because the Doppler trajectories are the same, and the false points and the real points are symmetrical about the left and right radar beam centers.

[0136] like Figure 9 and Figure 10 As shown, Figure 9 The image in the middle is the result of forward-looking imaging using a dual-base SVM scheme. Figure 10 The results shown are from forward-looking imaging using the K-means clustering algorithm. The results demonstrate that the proposed method solves the problem of spurious point targets in the BP imaging algorithm when handling single-base forward-looking imaging.

[0137] like Figures 11-13 As shown, Figure 11 The image shown is an optical scene diagram based on actual measurement data. Figure 12 The image shown is the result of the left-side radar BP imaging. Figure 13 The image shown is the result of radar BP imaging on the right side. It is evident that the BP imaging results on both the left and right sides exhibit a significant problem of symmetrical false point targets, demonstrating that this embodiment can obtain high-quality forward-looking SAR images.

[0138] like Figure 14 As shown in Figure 14, the results of the proposed forward-looking imaging algorithm are presented. The results show that the stainless steel fence in the optical scene has a strong reflectivity, resulting in a strong reflection point in the imaging result, corresponding to the area within the red box in the imaging result. The imaging results demonstrate that the proposed method can achieve reliable vehicle-mounted forward-looking imaging.

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

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

[0141] This embodiment acquires dual-base synthetic aperture radar (SAR) echo data by collecting data from the left and right forward-looking radars, as well as velocity and acceleration data. Forward-looking region imaging is then performed on the acquired SAR echo data. False target removal processing is then performed on the forward-looking region imaging using a support vector machine (SVM) algorithm or clustering algorithm to obtain a masked image after false target removal. The true target points of the original image in the SAR echo data can be extracted from the masked image to obtain the final SAR image. This embodiment addresses the problem of false points in the backpropagation algorithm when processing forward-looking imaging by introducing SVM and clustering algorithms, ultimately achieving reliable co-platform forward-looking imaging and improving the SAR imaging quality.

[0142] Exemplary device

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

[0144] The left front radar is used to collect data from the left forward-looking radar.

[0145] Right front radar, used to collect data from the right front-view radar;

[0146] The inertial navigation module is used to collect the speed at the moment the radar is operating.

[0147] And the main control unit, used to realize bistatic forward-looking synthetic aperture radar imaging;

[0148] The main control terminal includes:

[0149] The data acquisition module is used to acquire data from the left forward-looking radar, the right forward-looking radar, velocity and acceleration data, and obtain bistatic synthetic aperture radar echo data.

[0150] The mask image module is used to perform forward-looking region imaging on the acquired bi-base synthetic aperture radar echo data, and to perform false target elimination processing on the forward-looking region imaging through support vector machine algorithm or clustering algorithm to obtain a mask image after false target elimination.

[0151] The synthetic aperture radar imaging module is used to extract the real target points of the original image from the bi-base synthetic aperture radar echo data based on the mask image, so as to obtain the final forward-looking synthetic aperture radar image.

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

[0153] This embodiment acquires dual-base synthetic aperture radar (SAR) echo data by collecting data from the left and right forward-looking radars, as well as velocity and acceleration data. Forward-looking region imaging is then performed on the acquired SAR echo data. False target removal processing is then performed on the forward-looking region imaging using a support vector machine (SVM) algorithm or clustering algorithm to obtain a masked image after false target removal. The true target points of the original image in the SAR echo data can be extracted from the masked image to obtain the final SAR image. This embodiment addresses the problem of false points in the backpropagation algorithm when processing forward-looking imaging by introducing SVM and clustering algorithms, ultimately achieving reliable co-platform forward-looking imaging and improving the SAR imaging quality.

[0154] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 15 As shown.

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

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

[0157] It will be understood by those skilled in the art that Figure 15 The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

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

[0159] In one embodiment, a storage medium is provided, wherein the 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 described above.

[0160] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, database, or other media used in the embodiments provided by this invention can include both non-volatile and volatile memory.

[0161] In summary, this invention provides a co-platform bistatic forward-looking SAR imaging method and system, comprising: acquiring left forward-looking radar data, right forward-looking radar data, velocity and acceleration data to obtain bistatic synthetic aperture radar echo data; performing forward-looking region imaging on the acquired bistatic synthetic aperture radar echo data; performing false target removal processing on the forward-looking region imaging using a support vector machine algorithm or a clustering algorithm to obtain a masked image after false target removal; extracting the real target points of the original image from the bistatic synthetic aperture radar echo data based on the masked image to obtain the final forward-looking synthetic aperture radar image; this invention solves the problem of false points in the backpropagation algorithm when processing forward-looking imaging by introducing support vector machines and clustering algorithms, ultimately achieving reliable co-platform forward-looking imaging.

[0162] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A common-platform bistatic forward-looking SAR imaging method, characterized in that, include: Data from the left forward-looking radar, right forward-looking radar, velocity, and acceleration are collected to obtain bistatic synthetic aperture radar echo data. The acquired bistatic synthetic aperture radar echo data is used to perform forward-looking region imaging. The false target removal process is performed on the forward-looking region imaging using a support vector machine algorithm or a clustering algorithm to obtain a mask image after false target removal. Based on the mask image, the real target points of the original image in the bistatic synthetic aperture radar echo data are extracted to obtain the final forward-looking synthetic aperture radar image. The forward-looking region imaging of the acquired bistatic synthetic aperture radar echo data includes: The left-looking synthetic aperture radar (SAR) data in the dual-base SAR echo data is imaged using the backpropagation algorithm, and left offset compensation is performed using an offset compensation strategy to obtain the left-looking SAR image. ; The delay factor of the hardware delay compensation strategy is used to compensate for the delay of the right forward-looking synthetic aperture radar (SAR) data in the dual-base SAR echo data. The right forward-looking SAR data is then imaged based on the backpropagation algorithm, and right offset compensation is performed using the offset compensation strategy to obtain the right forward-looking SAR image. ; The preceding steps of performing forward-looking region imaging on the acquired bistatic synthetic aperture radar echo data include: Data domain matching is performed on the left forward-looking synthetic aperture radar data and the right forward-looking synthetic aperture radar data in the dual-base synthetic aperture radar echo data. The compensation factors corresponding to the left offset compensation strategy and the right offset compensation strategy are calculated based on the matching results, and the delay factor corresponding to the hardware delay compensation strategy is also calculated. The step of performing false target removal processing on the forward-looking region imaging using a clustering algorithm to obtain a mask image after false target removal includes: Get Left forward-looking synthetic aperture radar image at time of moment and right forward-looking synthetic aperture radar image ; Constructing the relation matrix: ， in, and Represent and exist The value of a pixel; The relationship matrix is ​​fed into the clustering algorithm to classify the real target points and false target points in the forward-looking scene, and the real target points are extracted based on the binary classification results. Based on the extracted real target points, a mask image is constructed to obtain the final prediction result; The step of performing false target removal processing on the forward-looking region imaging using a support vector machine algorithm to obtain a mask image after false target removal includes: Building a data source For the left forward-looking synthetic aperture radar image and the right forward-looking synthetic aperture radar image ,Pick As The data source matrix is ​​obtained by traversing all pixels. ; Will The input is fed into the trained support vector machine model to obtain the mask image of the final prediction result.

2. The co-platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that, The acquisition of left forward-looking radar data, right forward-looking radar data, velocity, and acceleration data includes, prior to: The communication module initiates a connection request to two single-base radar modules, and when the two single-base radar modules successfully connect and synchronize, the bistatic forward-looking synthetic aperture radar imaging system is run.

3. The co-platform bistatic forward-looking SAR imaging method according to claim 1, characterized in that, The step of extracting the real target points of the original image from the dual-base synthetic aperture radar echo data based on the mask image to obtain the final forward-looking synthetic aperture radar image includes: The left forward-looking synthetic aperture radar image in the original image is obtained using the masked image. Or right forward-looking synthetic aperture radar image Multiply the images to extract the true target points from the original image, eliminate the false target points from the original image, and obtain the final forward-looking synthetic aperture radar image.

4. A shared-platform bistatic forward-looking SAR imaging system, used to implement the shared-platform bistatic forward-looking SAR imaging method as described in any one of claims 1-3, characterized in that, include: The left front radar is used to collect data from the left forward-looking radar. Right front radar, used to collect data from the right front-view radar; The inertial navigation module is used to collect the speed at the moment the radar is operating. And the main control unit, used to realize bistatic forward-looking synthetic aperture radar imaging; The main control terminal includes: The data acquisition module is used to acquire data from the left forward-looking radar, the right forward-looking radar, velocity and acceleration data, and obtain bistatic synthetic aperture radar echo data. The mask image module is used to perform forward-looking region imaging on the acquired bi-base synthetic aperture radar echo data, and to perform false target elimination processing on the forward-looking region imaging through support vector machine algorithm or clustering algorithm to obtain a mask image after false target elimination. The synthetic aperture radar imaging module is used to extract the real target points of the original image from the bi-base synthetic aperture radar echo data based on the mask image, so as to obtain the final forward-looking synthetic aperture radar image.

5. A terminal, characterized in that, include: The processor and memory, wherein the memory stores a common-platform bistatic forward-looking SAR imaging program, which, when executed by the processor, is used to implement the operation of the common-platform bistatic forward-looking SAR imaging method as described in any one of claims 1-3.

6. 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 as described in any one of claims 1-3.

Images