Distributed multi-platform multi-angle airborne sar moving target detection and velocity estimation method

Abstract

The application relates to the technical field of radar signal processing and moving target detection, in particular to a distributed multi-platform multi-angle airborne SAR moving target detection and speed estimation method, which comprises the following steps: establishing a multi-platform observation model; acquiring multi-channel SAR echo data of each radar platform; performing channel registration and differential processing on the multi-channel SAR echo data to obtain differential echo of each radar platform; forming a moving target response image of each radar platform based on the differential echo of each radar platform; fusing the moving target response images of the multiple radar platforms, extracting a candidate position set of a moving target from the fused response image, and extracting distance change information of the moving target under each radar platform; constructing a multi-angle two-dimensional speed estimation model to estimate the two-dimensional speed of the target at the candidate position; and evaluating and screening the candidate position to output the position and two-dimensional speed of the final target. The method can realize multi-platform SAR moving target detection, positioning and two-dimensional speed estimation.

Description

Technical Field

[0001] The embodiments of this application relate to the field of radar signal processing and moving target detection technology, and particularly to a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method and a computer-readable storage medium. Background Technology

[0002] Synthetic Aperture Radar (SAR) possesses all-weather, all-time imaging capabilities and is widely used in ground surveillance and moving target detection. Traditional SAR imaging typically assumes a static scene. When a moving target is present, its echo is affected by factors such as range travel, Doppler shift, and phase changes, resulting in defocusing, offset, or trailing in the imaging, thus reducing the accuracy of target detection and parameter estimation. Ground Moving Target Indication (GMTI) and Displaced Phase Center Antenna (DPCA) technologies suppress stationary clutter and preserve moving target signals by differentially analyzing echoes from adjacent time points. However, single-platform DPCA has limitations: it can only obtain motion information of the target relative to a single observation direction, making it difficult to directly recover the complete velocity vector of the target in the two-dimensional plane of the ground. Furthermore, single-platform or single-view velocity measurement methods are susceptible to factors such as observation geometry, pulse repetition frequency, phase ambiguity, and imaging defocusing. Especially when the target velocity is high or the observation angle is insufficient, a single platform struggles to reliably obtain complete motion parameters. Therefore, this paper proposes a moving target detection and velocity estimation method suitable for distributed multi-platform, multi-angle SAR-GMTI, which has significant engineering application value. Summary of the Invention

[0003] In view of this, embodiments of this application propose a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method and a computer-readable storage medium, aiming to solve the technical problems of insufficient stability of existing single-platform SAR-GMTI methods in complex backgrounds, incomplete single-view velocity information, and limited accuracy of two-dimensional velocity estimation, so as to achieve robust moving target detection and high-precision two-dimensional velocity vector estimation under multi-platform collaborative observation.

[0004] To achieve the above objectives, embodiments of this application propose a distributed, multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method, the method comprising the following steps: A multi-platform observation model is established, which includes multiple radar platforms with different heading angles, and each radar platform conducts collaborative detection of the same observation area. Acquire multi-channel SAR echo data from various radar platforms; Channel registration and differential processing were performed on the multi-channel SAR echo data of each radar platform to suppress static background and obtain differential echoes of each radar platform. Based on the differential echoes of each radar platform, moving target response images of each radar platform are generated. The system fuses moving target response images from multiple radar platforms and extracts a set of candidate locations for moving targets from the fused moving target response images. Extract the distance variation information of each candidate location in the candidate location set under each radar platform; Based on distance change information and multi-platform observation models, a multi-angle two-dimensional velocity estimation model is constructed, and the two-dimensional velocity of the target at each candidate location is estimated using the multi-angle two-dimensional velocity estimation model. The candidate positions are comprehensively evaluated and screened. The target at the candidate position with the highest comprehensive evaluation index is taken as the final target, and the position and two-dimensional velocity of the final target are output.

[0005] To achieve the above objectives, embodiments of this application also propose a computer-readable storage medium storing a computer program that, when executed by a processor, enables a distributed multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method as described above.

[0006] This application proposes a distributed multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method, which establishes a multi-platform observation model. This model includes multiple radar platforms with different heading angles, each collaboratively detecting the same observation area. Multi-channel SAR echo data from each radar platform is acquired. Channel registration and differential processing are performed on the multi-channel SAR echo data of each radar platform to suppress static backgrounds, resulting in differential echoes for each platform. Moving target response images for each radar platform are generated based on these differential echoes. The moving target response images from multiple radar platforms are fused, and a set of candidate positions for moving targets is extracted from the fused images. Range variation information for each candidate position in the set is extracted from each radar platform. Based on the range variation information and the multi-platform observation model, a multi-angle two-dimensional velocity estimation model is constructed, and the velocity is estimated using multiple... A two-dimensional velocity estimation model is proposed to estimate the two-dimensional velocity of a target at each candidate location. The candidate locations are comprehensively evaluated and screened, and the target at the candidate location with the highest comprehensive evaluation index is selected as the final target. The position and two-dimensional velocity of the final target are then output. This scheme utilizes multiple platforms with different heading angles to collaboratively detect the same observation area. Each platform is equipped with a dual-channel or multi-channel receiving structure, and stationary backgrounds are suppressed through channel registration and differential processing. Furthermore, using multi-angle observation information of the target from different platforms, a target motion parameter estimation model is constructed, thereby achieving moving target detection, localization, and two-dimensional velocity estimation. Based on this, this scheme can solve the technical problems of insufficient stability in moving target detection under complex backgrounds, incomplete single-view velocity information, and limited accuracy of two-dimensional velocity estimation in existing single-platform SAR-GMTI methods, thus achieving robust moving target detection and high-precision two-dimensional velocity vector estimation under multi-platform collaborative observation.

[0007] Optionally, multi-platform observation models include: One radar platform; among them, the first The heading angle of each radar platform is Its heading unit vector is represented as follows: ; No. The projected positions of the radar platforms on the ground plane are shown below: ; in, For the platform The initial ground projection position, For the platform Flight speed, For slow-time variables; The position of the target to be detected on the ground plane is shown as follows: ; in, The initial ground position of the target. The target is two-dimensional velocity.

[0008] Optionally, acquire multi-channel SAR echo data from various platforms, including: Imaging observations of the observation area were conducted using various radar platforms to obtain corresponding multi-channel SAR echo data; among them, for the first... The echo data received by each radar platform is represented as follows: ; in, Indicates the first The target amplitude coefficient received by each radar platform For distance-oriented sampling variables, For radar wavelength, Let be the envelope function after distance compression. If the bandwidth of the transmitted linear frequency modulated signal is B, then ; Indicates the first A radar platform in slow-time variables Distance-oriented sampling variables SAR echo data received at the location; Target and the The instantaneous slant range between the radar platforms is expressed as follows: ; The target is located on the ground. For the first The height of each radar platform For the goal in slow time The three-dimensional spatial coordinates of the location For the first Each radar platform in slow time The three-dimensional spatial coordinates of the location.

[0009] Optionally, channel registration and differential processing are performed on the multi-channel SAR echo data of each radar platform to suppress static background, resulting in differential echoes for each radar platform, including: The multi-channel SAR echo data of each radar platform are registered between channels according to the principle of equivalent phase center, so that the echoes of static scenes are aligned in different channels. The registered channel signals are differentially processed to obtain differential echoes; the expression for the differential echo is as follows: ; in, Indicates the first The first receiving channel echo of the radar platform Indicates the first The echo from the second receiving channel of the radar platform The slow time interval is used to register the equivalent phase center.

[0010] Optionally, based on the differential echoes of each radar platform, a moving target response image for each radar platform is generated, including: A two-dimensional ground imaging grid was established for the observation area; Based on the instantaneous slant range from grid points in a two-dimensional ground imaging grid to each radar platform, the complex echo of the corresponding range cell is extracted from the echo data, and phase compensation and slow-time accumulation are performed to obtain the first... The original SAR images of each radar platform are expressed as follows: ; in, , representing imaging grid points To the Instantaneous slant range of each radar platform; The differential echoes from each radar platform are subjected to the same spatial focusing process to obtain the first... The moving target response image of a radar platform is expressed as follows: ; in, Indicates the first Each radar platform at the imaging grid point The moving target response image at the location, its subscript This indicates differential processing.

[0011] Optionally, moving target response images from multiple radar platforms are fused, and a candidate set of moving target locations is extracted from the fused moving target response images, including: The moving target response images of each radar platform are normalized in amplitude, and the expression is as follows: ; Based on the moving target response images after amplitude normalization processing of each radar platform, a multi-platform fused response map is constructed, the expression of which is as follows: ; in, For the first The integration weight of each platform; Based on local peaks or salient regions in the multi-platform fusion response map, a set of candidate locations for moving targets is extracted; the expression for the set of candidate locations for moving targets is as follows: ; in, Indicates the candidate position.

[0012] Optionally, the distance variation information of each candidate location in the candidate location set under each radar platform is extracted, including: For each candidate position in the candidate position set The reference slant distance at each slow time is determined by the following expression: ; in, ; In the differential echo of the corresponding radar platform, a range search window centered on the reference slant range is set to extract local peak values, thus obtaining the range response trajectory of the candidate position under that radar platform. Its expression is as follows: ; Among them, distance response trajectory This represents the value extracted from the local peak response of the differential echo, varying with slow time. Changing distance observation sequence; By linearly fitting the distance response trajectory, the distance change information is obtained, and its expression is as follows: ; Among them, the results obtained from different radar platforms It reflects the projection information of the target velocity in different observation directions.

[0013] Optionally, based on distance variation information and a multi-platform observation model, a multi-angle two-dimensional velocity estimation model is constructed, and the two-dimensional velocity of the target at each candidate location is estimated using this model, including: For the i-th radar platform, define the vector: ; Construct the relation, whose expression is as follows: ; in, , for the first The horizontal speed of each radar platform; The equations of multiple radar platforms are combined into matrix form, and their expression is as follows: ; in, ; The two-dimensional velocity of the target at the candidate location is estimated using the weighted least squares method, and its expression is as follows: ; in, This is a weight matrix used to represent the reliability of observation results from different radar platforms.

[0014] Optionally, a comprehensive evaluation and screening process is performed on each candidate position, and the target at the candidate position corresponding to the highest comprehensive evaluation index is selected as the final target. The position and two-dimensional velocity of the final target are then output, including: For each candidate location, its comprehensive evaluation index is calculated, and its expression is as follows: ; in, Indicates the intensity of the fusion response. Indicates the degree of multi-platform support. This represents the residuals of the motion parameter estimation; The weighting coefficients for the fusion response intensity, The weighting coefficient for the degree of multi-platform support. Weighting coefficients for estimating residuals of motion parameters, and ; All candidate positions are sorted according to comprehensive evaluation indicators, and candidate positions with low indicators are eliminated. The target at the candidate location with the largest comprehensive evaluation index is selected as the final target, and its position estimate and two-dimensional velocity estimate are used as the final target's position and two-dimensional velocity. Output the final target's position and two-dimensional velocity. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies of this application will be briefly introduced below. Obviously, the following drawings are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. The drawings described herein are only used to explain this application and are not intended to limit this application.

[0016] Figure 1 This is a flowchart of a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method provided in one embodiment of this application; Figure 2 This is a flowchart of a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation process provided in one embodiment of this application; Figure 3 This is a differential image of the original SAR image and the phase center offset antenna provided in one embodiment of this application; Figure 4 This is a schematic diagram of the candidate positions after the fusion of the three-platform differential responses provided in one embodiment of this application; Figure 5This is a schematic diagram illustrating the estimation of the position and velocity of a moving target in a complex static scene provided in one embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. Those skilled in the art will understand that many technical details have been presented in the embodiments of this application to facilitate better understanding. However, the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments. The division of the following embodiments is for ease of description and should not constitute any limitation on the specific implementation of this application. The following embodiments can be combined with and referenced by each other without contradiction.

[0018] Synthetic Aperture Radar (SAR) possesses all-weather, all-time imaging capabilities and is widely used in ground surveillance and moving target detection. Traditional SAR imaging typically assumes a static scene. When a moving target is present, its echo is affected by factors such as range travel, Doppler shift, and phase changes, resulting in defocusing, offset, or trailing in the imaging, thus reducing the accuracy of target detection and parameter estimation. Ground Moving Target Indication (GMTI) and Displaced Phase Center Antenna (DPCA) technologies suppress stationary clutter and preserve moving target signals by differentially analyzing echoes from adjacent time points. However, single-platform DPCA has limitations: it can only obtain motion information of the target relative to a single observation direction, making it difficult to directly recover the complete velocity vector of the target in the two-dimensional plane of the ground. Furthermore, single-platform or single-view velocity measurement methods are susceptible to factors such as observation geometry, pulse repetition frequency, phase ambiguity, and imaging defocusing. Especially when the target velocity is high or the observation angle is insufficient, a single platform struggles to reliably obtain complete motion parameters. Therefore, this paper proposes a moving target detection and velocity estimation method suitable for distributed multi-platform, multi-angle SAR-GMTI, which has significant engineering application value.

[0019] In view of this, embodiments of this application propose a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method (also referred to as a distributed multi-platform multi-angle SAR-GMTI moving target detection and two-dimensional velocity estimation method) and a computer-readable storage medium, aiming to solve the problems of insufficient stability of existing single-platform SAR-GMTI methods in complex backgrounds, incomplete single-view velocity information, and limited accuracy of two-dimensional velocity estimation, so as to achieve robust moving target detection and high-precision two-dimensional velocity vector estimation under multi-platform collaborative observation.

[0020] Specifically, addressing the technical problems of existing single-platform SAR-GMTI methods, such as insufficient stability in moving target detection under complex backgrounds, incomplete single-view velocity information, and limited accuracy in two-dimensional velocity estimation, this application proposes a distributed multi-platform multi-angle SAR-GMTI moving target detection and two-dimensional velocity estimation method. This method utilizes multiple platforms with different heading angles to collaboratively detect the same observation area. Each platform is equipped with a dual-channel or multi-channel receiving structure, and stationary backgrounds are suppressed through channel registration and differential processing. Then, using the multi-angle observation information of the target from different platforms, a target motion parameter estimation model is constructed, thereby realizing moving target detection, localization, and two-dimensional velocity estimation.

[0021] One embodiment of this application proposes a distributed, multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method, applied to an electronic device. The electronic device can be a terminal or a server; this embodiment and subsequent embodiments will use a server as an example. The implementation details of the distributed, multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method proposed in this embodiment are described below. The following implementation details are provided for ease of understanding and are not essential for implementing this solution.

[0022] For example, the method provided in this embodiment is applicable to the detection, localization and motion state estimation of ground moving targets in a complex static background under the condition of collaborative observation by multiple airborne SAR platforms.

[0023] The specific process of the distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method proposed in this embodiment can be described as follows: Figure 1 As shown, it includes: Step 101: Establish a multi-platform observation model.

[0024] The multi-platform observation model includes multiple radar platforms with different heading angles, and each radar platform conducts collaborative detection of the same observation area.

[0025] For example, a multi-platform observation model may include: One radar platform; among them, the first The heading angle of each radar platform is Its heading unit vector is represented as follows: ; No. The projected positions of the radar platforms on the ground plane are shown below: ; in, For the platform The initial ground projection position, For the platform Flight speed, For slow-time variables; The position of the target to be detected on the ground plane is shown as follows: ; in, The initial ground position of the target. The target is two-dimensional velocity.

[0026] Step 102: Obtain multi-channel SAR echo data from each radar platform.

[0027] For example, imaging observations of the observation area are performed by various radar platforms to obtain corresponding multi-channel SAR echo data; among which, for the first... The echo data received by each radar platform is represented as follows: ; in, Indicates the first The target amplitude coefficient received by each radar platform For distance-oriented sampling variables, For radar wavelength, Let be the envelope function after distance compression. If the bandwidth of the transmitted linear frequency modulated signal is B, then ; Indicates the first A radar platform in slow-time variables Distance-oriented sampling variables SAR echo data received at the location; Target and the The instantaneous slant range between the radar platforms is expressed as follows: ; The target is located on the ground. For the first The height of each radar platform For the goal in slow time The three-dimensional spatial coordinates of the location For the first Each radar platform in slow time The three-dimensional spatial coordinates of the location.

[0028] It is understandable that the above echo signal and slant range formula can characterize the impact of target motion on the phase and range history of multi-platform SAR echoes.

[0029] Step 103: Perform channel registration and differential processing on the multi-channel SAR echo data of each radar platform to suppress static background and obtain the differential echo of each radar platform.

[0030] For example, for the multi-channel SAR echo data (also known as multi-channel echo) of each radar platform, inter-channel registration is performed based on the principle of equivalent phase center to align the echoes of a static scene in different channels. Subsequently, the registered channel signals are differentially processed to obtain differential echoes; the expression for the differential echoes is as follows: ; in, and They represent the first The echoes from the two receiving channels of the radar platform The slow time interval corresponding to the registration of the equivalent phase center. Specifically, where, Indicates the first The first receiving channel echo of the radar platform Indicates the first The echo from the second receiving channel of the radar platform.

[0031] It is understandable that the static background has high consistency after registration and is effectively suppressed after differencing; while the moving target has positional changes at adjacent observation times, so the residual response is retained in the differencing results.

[0032] Step 104: Based on the differential echoes of each radar platform, generate moving target response images for each radar platform.

[0033] A two-dimensional ground imaging grid is established for the observation area, specifically as follows: ; in, This represents a grid point within the ground plane to be imaged. This two-dimensional ground point is then expanded into three-dimensional coordinates: Then the imaging grid points To the The instantaneous slant distance of each platform is: .

[0034] Based on the instantaneous slant range, the complex echo of the corresponding range cell is extracted from the echo data, and phase compensation and slow-time accumulation are performed to obtain the first... Original SAR images from each platform: ; Similarly, the differential echoes from each platform are subjected to the same spatial focusing process to obtain the... Dynamic target response images of each platform: ; in, Indicates the first Each radar platform at the imaging grid point The moving target response image at the location, its subscript This indicates differential processing. The image primarily reflects the response intensity of a moving target at the corresponding observation angle.

[0035] It should be noted that the original SAR image is used to characterize the overall scattering distribution of the observation area, while the moving target response image is used to highlight the moving target and suppress the static background. It mainly reflects the response intensity of the moving target at the corresponding observation angle, providing input for subsequent multi-platform fusion detection and velocity estimation.

[0036] Step 105: Fuse moving target response images from multiple radar platforms and extract a set of candidate locations for moving targets from the fused moving target response images.

[0037] The moving target response images of each radar platform are normalized in amplitude, and the expression is as follows: ; Based on the moving target response images after amplitude normalization processing of each radar platform, a multi-platform fused response map is constructed, the expression of which is as follows: ; in, For the first The integration weight of each platform; Based on local peaks or salient regions in the multi-platform fusion response map, a set of candidate locations for moving targets is extracted; the expression for the set of candidate locations for moving targets is as follows: ; in, Indicates the candidate position.

[0038] For example, local peaks or significant regions can represent local maxima or highlighted areas in the fused image where the response value is significantly higher than the surrounding background. Its core function is to reduce the complexity of subsequent multi-platform distance variation extraction and two-dimensional velocity estimation from a full-image search to a few high-probability locations, thereby significantly reducing computational load, improving detection efficiency, and enhancing the accuracy of real targets and suppressing false responses by overlaying information from multiple angles.

[0039] For example, an adaptive threshold detection method based on a constant false alarm rate can be used to determine local peaks or significant regions in the multi-platform fusion response map, and then extract a set of candidate locations for moving targets. For instance, firstly, a two-dimensional sliding window containing a target unit, a protection unit, and a reference unit is set for each pixel in the multi-platform fusion response map; then, based on the statistical characteristics of the pixel response values ​​within the reference unit, an adaptive detection threshold corresponding to a preset false alarm probability is calculated. Finally, the response value of the target unit is compared with the adaptive detection threshold; if the response value of the target unit is greater than or equal to the adaptive detection threshold, the pixel is marked as a candidate location; otherwise, it is determined to be background and can be removed. Optionally, non-maximum suppression or spatial clustering can be performed on the marked candidate locations to eliminate redundant candidate locations.

[0040] Specifically, the preset false alarm probability is taken as follows: In application scenarios with high tolerance for false alarms (i.e., allowing for more false alarms), the preset false alarm probability can be... The magnitude of the value; in application scenarios where the system's false alarm tolerance requirement is low (i.e., strict suppression of false alarms is required), the preset false alarm probability can be... The magnitude of the value.

[0041] It should be noted that in the field of radar target detection, the false alarm probability is usually set to... arrive In terms of magnitude, in the SAR-GMTI system, constant false alarm rate (CFAR) processing not only obtains reliable moving target tracks but also avoids saturation of the GMTI processing system due to an excessive number of false alarm targets. Therefore, in scenarios involving distributed multi-platform, multi-angle collaborative detection and fusion response maps with multiple candidate points, The magnitude of the false alarm probability can balance detection sensitivity and system processing burden, ensuring that the number of candidate points entering the subsequent comprehensive evaluation and screening stages is controllable.

[0042] Step 106: Extract the distance change information of each candidate location in the candidate location set under each radar platform.

[0043] For example, for each candidate location in the candidate location set The reference slant distance at each slow time is determined by the following expression: ; in, ; In the differential echo of the corresponding radar platform, a range search window centered on the reference slant range is set to extract local peak values, thus obtaining the range response trajectory of the candidate position under that radar platform. Its expression is as follows: ; Among them, distance response trajectory This represents the value extracted from the local peak response of the differential echo, varying with slow time. Changing distance observation sequence; Near the center slow time, the change in target slant range can be approximated as: ; in, For the first The rate of change of distance observed by each radar platform; By linearly fitting the distance response trajectory, the distance change information is obtained, and its expression is as follows: ; Among them, the results obtained from different radar platforms It reflects the projection information of the target velocity in different observation directions.

[0044] Step 107: Based on distance change information and multi-platform observation model, construct a multi-angle two-dimensional velocity estimation model, and estimate the two-dimensional velocity of the target at each candidate location using the multi-angle two-dimensional velocity estimation model.

[0045] For example, for the i-th radar platform, define the vector: ; The rate of change of the target's distance relative to the radar platform satisfies: ; in, , for the first The horizontal speed of each radar platform; The resulting relational expression is as follows: ; The equations of multiple radar platforms are combined into matrix form, and their expression is as follows: ; in, ; The two-dimensional velocity of the target at the candidate location is estimated using the weighted least squares method, and its expression is as follows: ; in, This is a weight matrix used to represent the reliability of observation results from different radar platforms.

[0046] Step 108: Perform a comprehensive evaluation and screening of each candidate position, and take the target at the candidate position with the largest comprehensive evaluation index as the final target, and output the position and two-dimensional velocity of the final target.

[0047] After obtaining the initial set of candidate locations, this embodiment further evaluates the comprehensive credibility of each candidate location. For the k-th candidate location, its response intensity in the multi-platform fused moving target response map, its support level in the observation results of multiple platforms, and the residual of motion parameter estimation obtained from the multi-platform observation equation are calculated. Based on the above factors, a comprehensive evaluation index for candidate locations is constructed.

[0048] For each candidate location, its comprehensive evaluation index is calculated, and its expression is as follows: ; in, Indicates the intensity of the fusion response. Indicates the degree of multi-platform support. This represents the residuals of the motion parameter estimation; The weighting coefficients for the fusion response intensity, The weighting coefficient for the degree of multi-platform support. Weighting coefficients for estimating residuals of motion parameters, and ; In one embodiment, it is preferable to , , .

[0049] All candidate locations were ranked according to a comprehensive evaluation index, and those with weak response, insufficient multi-platform support, or large geometric residuals were eliminated. Finally, the target at the candidate location with the highest comprehensive evaluation index was selected as the final target. Specifically: ; The target at the candidate location with the highest comprehensive evaluation index is selected as the final target, and its position and velocity estimates are used as the final target position and two-dimensional velocity estimates, as expressed below: ; Among them, the position estimation result is used to characterize the spatial position of the target in the observation area, and the velocity estimation result is used to characterize the two-dimensional motion state of the target in the ground plane.

[0050] Output the final target position and two-dimensional velocity estimation results.

[0051] This application proposes a distributed multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method, which establishes a multi-platform observation model. This model includes multiple radar platforms with different heading angles, each collaboratively detecting the same observation area. Multi-channel SAR echo data from each radar platform is acquired. Channel registration and differential processing are performed on the multi-channel SAR echo data of each radar platform to suppress static backgrounds, resulting in differential echoes for each platform. Moving target response images for each radar platform are generated based on these differential echoes. The moving target response images from multiple radar platforms are fused, and a set of candidate positions for moving targets is extracted from the fused images. Range variation information for each candidate position in the set is extracted from each radar platform. Based on the range variation information and the multi-platform observation model, a multi-angle two-dimensional velocity estimation model is constructed. This scheme employs a multi-angle two-dimensional velocity estimation model to estimate the two-dimensional velocity at each candidate location. The candidate locations are then comprehensively evaluated and selected, with the location corresponding to the highest comprehensive evaluation index being chosen as the final target. The position and two-dimensional velocity of this final target are then output. This scheme utilizes multiple platforms with different heading angles to collaboratively detect the same observation area. Each platform is equipped with a dual-channel or multi-channel receiving structure, and stationary backgrounds are suppressed through channel registration and differential processing. Furthermore, using multi-angle observation information from different platforms, a target motion parameter estimation model is constructed, thereby achieving moving target detection, localization, and two-dimensional velocity estimation. Based on this, the scheme addresses the technical problems of insufficient stability in moving target detection under complex backgrounds, incomplete single-view velocity information, and limited accuracy in two-dimensional velocity estimation in existing single-platform SAR-GMTI methods. This enables robust moving target detection and high-precision two-dimensional velocity vector estimation under multi-platform collaborative observation.

[0052] like Figure 2 As shown, this embodiment proposes a distributed multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method. This embodiment is based on a collaborative observation network consisting of M distributed sensor platforms (such as radar, denoted as platform 1 to platform M), aiming to solve the problem of accurate real-time tracking of moving targets in complex backgrounds. The observation data from each platform are processed through the following process: S1: Establishment of the observation model.

[0053] S2: Clutter suppression and imaging.

[0054] S3: Differential image fusion for moving target detection.

[0055] For example, moving target detection can be performed by differential image fusion to extract candidate points (i.e., candidate locations mentioned above).

[0056] S4: Multi-platform joint two-dimensional velocity inversion.

[0057] S5: Determine the target point.

[0058] For example, the target point can be the target at the candidate position in the above embodiment, specifically the target at the candidate position corresponding to the maximum comprehensive evaluation index.

[0059] S6: Determine the target position estimate and two-dimensional velocity estimate of the target point, and use them as the final target position and two-dimensional velocity.

[0060] For a detailed description of S1 to S6, please refer to the above embodiments, which will not be repeated here.

[0061] To verify the effectiveness of the method of the present invention, the simulation experimental parameters in Table 1 are given here.

[0062] The flight platform consists of three units at an altitude of 4000m. The antenna spacing between each platform is 0.6m, and the flight speed is 150m / s for all platforms. The heading angles for platforms A, B, and C are as follows: The initial carrier frequency was 10 GHz, the frequency bandwidth was 150 MHz, and the pulse repetition frequency was 500 Hz. The experimentally tested moving target velocity was 36.1 m / s. The processing flow is as follows: Figure 1 As shown.

[0063] Table 1 Simulation Experiment Parameters

[0064] Figure 3 The original SAR imaging results and corresponding Displaced Phase Center Antenna (DPCA) differential results obtained from platforms A, B, and C under different observation directions are presented. The left column shows the original synthetic aperture radar (SAR) images for each platform; the static background scattering is strong, and due to the different observation geometries of different platforms, the background texture and sidelobe structure show significant differences. The right column shows the DPCA images for each platform; it can be seen that the static background is significantly suppressed after differential processing, while moving targets exhibit prominent local responses in the DPCA images of all three platforms. Star-shaped markers represent the true target positions, and square markers represent the final estimated positions; the two are basically consistent in the DPCA images of the three platforms, indicating that multi-angle DPCA differential processing can enhance the observability of moving targets relative to the static background.

[0065] Figure 4This is a candidate target image obtained by fusing the differential responses of three platforms, used to generate candidate initial positions for moving targets. In the image, brightness levels represent the intensity of the normalized fused response; higher brightness indicates a stronger moving target response at that grid point. Circular markers represent candidate moving target positions extracted from the fused moving target response image, star markers represent the actual moving target positions, and square markers represent the final estimated positions. The method proposed in this application does not rely solely on local peaks in the single-platform echo-generated image for decision-making, but further combines multi-platform, multi-angle observation consistency and candidate position quality evaluation for screening. Therefore, although multiple candidate positions exist in the fused image, the final estimated position still stably converges to the vicinity of the actual target, demonstrating the role of multi-platform information fusion in suppressing false peaks and improving detection reliability.

[0066] Figure 5 This paper presents the position and velocity estimation results of a moving target in a complex static scene. In the figure, the light and dark textures in the background represent a finely scattered background; star-shaped markers represent the true initial position of the moving target, with associated solid arrows representing the true velocity; circular markers represent candidate positions extracted from the multi-platform differential fusion response map; and square markers represent the final estimated position, with associated dashed arrows representing the estimated velocity. As can be seen from the figure, there is a static scattering background and multiple candidate positions, with the target's true position... The existence of both the estimated and actual positions in the vicinity, and the high consistency between the estimated and actual velocity directions, demonstrates that the proposed method can effectively filter out the real moving target from the candidate position set and reliably recover the two-dimensional velocity vector even in the presence of complex static backgrounds and multiple pseudo-candidate responses.

[0067] In summary, the distributed multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method proposed in this embodiment, also known as a distributed multi-platform, multi-angle SAR-GMTI moving target detection and two-dimensional velocity estimation method, establishes a unified model for distributed multi-platform, multi-angle collaborative observation scenarios, no longer limited to a single platform or a fixed number of platforms, and has stronger versatility and scalability. It suppresses static backgrounds through intra-platform channel registration and differential processing, and utilizes distance variation information obtained from different observation directions of multiple platforms to construct a multi-angle velocity constraint model, improving the overall observability of moving targets. It achieves the recovery of two-dimensional motion parameters of the target from local observed velocity, improving the dimensionality and accuracy of parameter estimation, and completing the estimation of the moving target's position and two-dimensional velocity. This invention is applicable to engineering applications of multi-airborne platform, multi-heading collaborative SAR-GMTI systems for moving target detection, localization, and motion parameter estimation in complex ground backgrounds, and has good promotional value.

[0068] The steps described above are for clarity only. In implementation, they can be combined into one step, or some steps can be broken down into multiple steps, as long as they involve the same logical relationship, they are all within the scope of protection of this application. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, without changing the core design of the algorithm and process, are also within the scope of protection of this application.

[0069] Another embodiment of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, can implement a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method as described in the above method embodiments.

[0070] That is, those skilled in the art will understand that all or part of the steps in the above method embodiments can be implemented by a program instructing related hardware. The program is stored in a storage medium and includes several instructions to cause a device (such as a microcontroller, chip, etc.) or processor to execute all or part of the steps of the method described in the method embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0071] Those skilled in the art will understand that the above embodiments are specific implementations of this application, and in practical applications, various changes can be made in form and detail without departing from the spirit and scope of this application. For those skilled in the art, several improvements and modifications can be made without departing from the principles of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.

Claims

1. A distributed, multi-platform, multi-angle airborne SAR moving target detection and velocity estimation method, characterized in that, The method includes: A multi-platform observation model is established, which includes multiple radar platforms with different heading angles, and each radar platform conducts collaborative detection of the same observation area. Acquire multi-channel SAR echo data from various radar platforms; Channel registration and differential processing were performed on the multi-channel SAR echo data of each radar platform to suppress static background and obtain differential echoes of each radar platform. Based on the differential echoes of each radar platform, moving target response images of each radar platform are generated. The system fuses moving target response images from multiple radar platforms and extracts a set of candidate locations for moving targets from the fused moving target response images. Extract the distance variation information of each candidate location in the candidate location set under each radar platform; Based on distance change information and multi-platform observation models, a multi-angle two-dimensional velocity estimation model is constructed, and the two-dimensional velocity of the target at each candidate location is estimated using the multi-angle two-dimensional velocity estimation model. The candidate positions are comprehensively evaluated and screened. The target at the candidate position with the highest comprehensive evaluation index is taken as the final target, and the position and two-dimensional velocity of the final target are output.

2. The method according to claim 1, characterized in that, The multi-platform observation model includes: One radar platform; among them, the first The heading angle of each radar platform is Its heading unit vector It is expressed as follows: ； No. The projected positions of the radar platforms on the ground plane are shown below: ； in, For the first The initial ground projection position of each radar platform, For the first The flight speed of each radar platform For slow-time variables; The position of the target to be detected on the ground plane is shown as follows: ； in, The initial ground position of the target. The target is two-dimensional velocity.

3. The method according to claim 2, characterized in that, The acquisition of multi-channel SAR echo data from each platform includes: Imaging observations of the observation area were conducted using various radar platforms to obtain corresponding multi-channel SAR echo data; among them, for the first... The echo data received by each radar platform is represented as follows: ； in, Indicates the first The target amplitude coefficient received by each radar platform For distance-oriented sampling variables, For radar wavelength, Let be the envelope function after distance compression. If the bandwidth of the transmitted linear frequency modulated signal is ,but ; Indicates the first A radar platform in slow-time variables Distance-oriented sampling variables SAR echo data received at the location; Target and the The instantaneous slant range between the radar platforms is expressed as follows: ； The target is located on the ground. For the first The height of each radar platform For the goal in slow time The three-dimensional spatial coordinates of the location For the first Each radar platform in slow time The three-dimensional spatial coordinates of the location.

4. The method according to claim 3, characterized in that, The process of channel registration and differential processing of multi-channel SAR echo data from each radar platform to suppress static background and obtain differential echoes from each radar platform includes: The multi-channel SAR echo data of each radar platform are registered between channels according to the principle of equivalent phase center, so that the echoes of static scenes are aligned in different channels. The registered channel signals are differentially processed to obtain differential echoes; the expression for the differential echo is as follows: ； in, Indicates the first The first receiving channel echo of the radar platform Indicates the first The echo from the second receiving channel of the radar platform The slow time interval is used to register the equivalent phase center.

5. The method according to claim 4, characterized in that, The process of generating moving target response images for each radar platform based on differential echoes from each platform includes: A two-dimensional ground imaging grid was established for the observation area; Based on the instantaneous slant range from grid points in a two-dimensional ground imaging grid to each radar platform, the complex echo of the corresponding range cell is extracted from the echo data, and phase compensation and slow-time accumulation are performed to obtain the first... The original SAR images of each radar platform are expressed as follows: ； in, , representing imaging grid points To the Instantaneous slant range of each radar platform; The differential echoes from each radar platform are subjected to the same spatial focusing process to obtain the first... The moving target response image of a radar platform is expressed as follows: ； in, Indicates the first Each radar platform at the imaging grid point The moving target response image at the location, its subscript This indicates differential processing.

6. The method according to claim 5, characterized in that, The process involves fusing moving target response images from multiple radar platforms and extracting a candidate set of moving target locations from the fused moving target response images, including: The moving target response images of each radar platform are normalized in amplitude, and the expression is as follows: ； Based on the moving target response images after amplitude normalization processing of each radar platform, a multi-platform fused response map is constructed, the expression of which is as follows: ； in, For the first The integration weight of each platform; Based on local peaks or salient regions in the multi-platform fusion response map, a set of candidate locations for moving targets is extracted; the expression for the set of candidate locations for moving targets is as follows: ； in, Indicates the candidate position.

7. The method according to claim 6, characterized in that, The extraction of distance variation information for each candidate location in the candidate location set under each radar platform includes: For each candidate position in the candidate position set The reference slant distance at each slow time is determined by the following expression: ； in, ; In the differential echo of the corresponding radar platform, a range search window centered on the reference slant range is set to extract local peak values, thus obtaining the range response trajectory of the candidate position under that radar platform. Its expression is as follows: ； Among them, distance response trajectory This represents the value extracted from the local peak response of the differential echo, varying with slow time. Changing distance observation sequence; By linearly fitting the distance response trajectory, the distance change information is obtained, and its expression is as follows: ； Among them, the results obtained from different radar platforms It reflects the projection information of the target velocity in different observation directions.

8. The method according to claim 7, characterized in that, Based on distance change information and a multi-platform observation model, a multi-angle two-dimensional velocity estimation model is constructed. This model is then used to estimate the two-dimensional velocity of the target at each candidate location, including: For the For each radar platform, define a vector: ； Construct the relation, whose expression is as follows: ； in, , for the first The horizontal speed of each radar platform; The equations of multiple radar platforms are combined into matrix form, and their expression is as follows: ； in, ; The two-dimensional velocity of the target at the candidate location is estimated using the weighted least squares method, and its expression is as follows: ； in, This is a weight matrix used to represent the reliability of observation results from different radar platforms.

9. The method according to claim 8, characterized in that, The process of comprehensively evaluating and filtering candidate positions, selecting the target at the candidate position with the highest comprehensive evaluation index as the final target, and outputting the position and two-dimensional velocity of the final target includes: For each candidate location, its comprehensive evaluation index is calculated, and its expression is as follows: ； in, Indicates the intensity of the fusion response. Indicates the degree of multi-platform support. This represents the residuals of the motion parameter estimation; The weighting coefficients for the fusion response intensity, The weighting coefficient for the degree of multi-platform support. Weighting coefficients for estimating residuals of motion parameters, and ; All candidate positions are sorted according to comprehensive evaluation indicators, and candidate positions with low indicators are eliminated. The target at the candidate location with the largest comprehensive evaluation index is selected as the final target, and its position estimate and two-dimensional velocity estimate are used as the final target's position and two-dimensional velocity. Output the final target's position and two-dimensional velocity.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it can implement a distributed multi-platform multi-angle airborne SAR moving target detection and velocity estimation method as described in any one of claims 1 to 9.

Images