Self-focusing imaging method and device based on adaptive segmented Doppler compensation

Abstract

The application relates to a self-focusing imaging method and device based on adaptive segmental Doppler compensation. The method comprises the following steps: acquiring an original echo signal under a radar forward-looking long-aperture condition; constructing an adaptive criterion with a phase error as a constraint condition according to the signal; dynamically dividing the full aperture based on the adaptive criterion to obtain a plurality of adaptive sub-apertures; respectively performing imaging processing on echo signals in each adaptive sub-aperture to obtain a plurality of sub-aperture images; establishing a cross-segment coherent mapping model between different sub-aperture images; converting the cross-segment coherent mapping model into a two-dimensional optimization problem in which a distance direction compensation operator and an azimuth direction compensation operator are separable; iteratively solving the two-dimensional optimization problem by using an alternating direction multiplier method to obtain an optimal compensation operator; and compensating and coherently reconstructing each sub-aperture image by using the optimal compensation operator to output a full-focusing image. The method can realize high-resolution and high-energy focusing self-focusing imaging of an extended target and a complex scene.

Description

Technical Field

[0001] This application relates to the field of radar imaging technology, and in particular to a self-focusing imaging method and apparatus based on adaptive segmented Doppler compensation. Background Technology

[0002] Forward-looking radar imaging, with its all-weather, all-day detection capabilities, holds significant application value in areas such as autonomous navigation of unmanned platforms, low-altitude target detection, and environmental perception in complex scenarios. To overcome the resolution limitations imposed by the finite azimuth aperture under forward-looking observation conditions, it is typically necessary to utilize platform motion to generate long-term coherent accumulation, thereby expanding the equivalent synthetic aperture to improve azimuth resolution. Besides utilizing motion synthetic aperture, introducing multiple-input multiple-output (MIMO) array technology is also an effective way to enhance forward-looking resolution. By arranging multiple transmit and receive units on the platform, MIMO technology uses orthogonal waveforms to form an equivalent virtual array in space, expanding the equivalent spatial aperture without increasing physical size and improving the spatial sampling degrees of freedom. However, limited by the payload size, hardware cost, and system complexity of unmanned platforms, simply increasing the physical array length or number of channels is insufficient to meet the demands of ultra-high resolution. Therefore, with limited array aperture, it is still necessary to combine platform motion with long-term coherent accumulation to achieve spatiotemporal joint aperture expansion. However, under long aperture conditions, the slant range history caused by the relative motion of the target no longer satisfies the traditional linear approximation. The Doppler history exhibits significant nonlinear changes over time, leading to problems such as Doppler spectrum broadening, enhanced spatially varying phase errors, and image defocusing. Especially in close-range, highly maneuverable scenarios, these higher-order nonlinear errors accumulate rapidly over time, making it difficult for traditional compensation methods based on global linear models to achieve stable focusing.

[0003] Existing research on nonlinear motion errors under long aperture conditions mainly includes two categories: global compensation methods and sub-aperture segmentation methods. Global compensation methods typically construct a unified high-order compensation function to correct range travel and phase errors. For example, methods based on nonlinear Chirp Scaling (NCS) can compensate for some high-order motion errors in the frequency domain, but they rely on a globally consistent error model. When the error exhibits significant spatial variation under long aperture conditions, it is difficult to maintain consistent focusing performance across the entire imaging area. In contrast, sub-aperture segmentation decomposes the long-term nonlinear motion process into multiple locally quasi-stationary sub-apertures, which can reduce the complexity of the compensation model to some extent, and is therefore widely used in long aperture imaging processing. However, most existing segmentation methods use fixed-length divisions, lacking the ability to adaptively adjust to changes in the instantaneous motion state and nonlinearity of the target. When the aperture length is not chosen appropriately, excessively long sub-apertures can lead to the failure of local linear models, while excessively short sub-apertures can reduce coherence accumulation gain. In addition, even after independent compensation of each sub-aperture, there is still a phase mismatch problem between segments, which can easily introduce energy dispersion and focusing degradation during the full aperture reconstruction process.

[0004] Under long-aperture conditions, higher-order nonlinear terms in the target slant range history cause significant changes in the Doppler history over time, making traditional compensation methods based on a globally consistent error model unsuitable. Taking the classic Phase Gradient Autofocus (PGA) method as an example, its core assumption is that all scattering points share a unified phase error function, and it uses strong scattering points in the image domain to estimate the global phase error. However, in near-field forward-looking long-aperture scenes, the slant range change rate and Doppler history of targets at different spatial locations differ significantly, resulting in a significantly spatially variable phase error. In this case, globally consistent compensation cannot simultaneously address the focusing requirements of different regions. Furthermore, when there are few stable strong scattering points or extended targets in the scene, the phase estimation accuracy of PGA further decreases, easily introducing local energy diffusion and sidelobe lift. Therefore, relying solely on global phase correction is insufficient to solve the nonlinear defocusing problem under long-aperture conditions.

[0005] To reduce the modeling complexity of nonlinear motion errors in long-aperture imaging, the sub-aperture segmentation method approximates the full-aperture motion process as multiple locally quasi-stationary sub-processes by shortening the local coherence accumulation time. This allows for stable compensation using low-order models within each sub-aperture. This type of method effectively alleviates the accumulation of high-order phase errors caused by long-term accumulation, thus demonstrating good applicability in long-aperture imaging. However, most existing methods employ fixed-length sub-aperture divisions, which essentially assume that the degree of nonlinearity remains consistent across different stages. But in forward-looking high-maneuver scenarios, the relative motion state of the target continuously changes over time, and the degree of Doppler nonlinearity varies significantly across different time periods. Fixed-length divisions struggle to simultaneously balance local model accuracy and coherence accumulation gain: excessively long sub-apertures cause the local linearity assumption to fail, while excessively short sub-apertures reduce azimuth resolution. Furthermore, after independent compensation of each sub-aperture, the difference between the target position mapping relationship and the residual phase error makes it difficult to achieve strict coherent alignment between segments, thus affecting the overall aperture energy reconstruction effect. Summary of the Invention

[0006] Therefore, it is necessary to provide a self-focusing imaging method and device based on adaptive segmented Doppler compensation that can achieve high-resolution, high-energy focusing self-focusing imaging in extended targets and complex scenes, in order to address the above-mentioned technical problems.

[0007] A self-focusing imaging method based on adaptive piecewise Doppler compensation, the method comprising: Acquire the raw echo signal under radar forward-looking long aperture conditions.

[0008] Based on the original echo signal, an adaptive criterion with phase error as the constraint is constructed, and the full aperture is dynamically divided based on the adaptive criterion to obtain multiple adaptive sub-apertures.

[0009] The echo signals within each adaptive sub-aperture are processed separately to obtain multiple sub-aperture images.

[0010] A cross-segment coherence mapping model is established between images with different sub-apertures. The model is used to characterize the spatial position offset and residual phase error between sub-apertures.

[0011] The cross-segment coherent mapping model is transformed into a two-dimensional optimization problem in which the range compensation operator and the azimuth compensation operator can be separated. The alternating direction multiplier method is used to iteratively solve the two-dimensional optimization problem to obtain the optimal compensation operator.

[0012] The optimal compensation operator is used to compensate and coherently reconstruct each sub-aperture image, and the fully focused image is output.

[0013] A self-focusing imaging device based on adaptive piecewise Doppler compensation, the device comprising: The signal acquisition module is used to acquire the raw echo signal under radar forward-looking long aperture conditions.

[0014] The adaptive segmentation module is used to construct an adaptive criterion with phase error as a constraint based on the original echo signal, and to dynamically divide the full aperture based on the adaptive criterion to obtain multiple adaptive sub-apertures.

[0015] The sub-aperture imaging module is used to perform imaging processing on the echo signals in each adaptive sub-aperture to obtain multiple sub-aperture images.

[0016] The model building module is used to establish a cross-segment coherence mapping model between images with different sub-apertures. The model is used to characterize the spatial position offset and residual phase error between sub-apertures.

[0017] The optimization solution module is used to transform the cross-segment coherent mapping model into a two-dimensional optimization problem in which the range compensation operator and the azimuth compensation operator can be separated. The alternating direction multiplier method is used to iteratively solve the two-dimensional optimization problem to obtain the optimal compensation operator.

[0018] The output module is used to compensate and coherently reconstruct each sub-aperture image using the optimal compensation operator, and output a fully focused image.

[0019] The aforementioned self-focusing imaging method and device based on adaptive piecewise Doppler compensation firstly, dynamically divides the entire aperture by constructing an adaptive criterion with phase error as a constraint. This allows for real-time adjustment of the sub-aperture length according to the nonlinear changes in the target's motion state, ensuring that the signal within each sub-aperture satisfies the local quasi-linearity condition. This fundamentally avoids the problems of traditional fixed-length segments leading to local linear model failure due to excessive length or decreased coherence accumulation gain due to excessive length, effectively suppressing Doppler spectrum broadening and spatially varied phase error enhancement caused by higher-order nonlinear terms under long aperture conditions. Secondly, by establishing a cross-segment coherent mapping model between images of different sub-apertures and transforming this model into a two-dimensional optimization problem with separable range and azimuth compensation operators, joint characterization and decoupling processing of spatial position offset and residual phase error are achieved, significantly improving the flexibility of spatially varied error compensation in long aperture scenarios. Finally, the Alternating Directional Multiplier Method (ADMM) is employed to iteratively solve this two-dimensional optimization problem. This method efficiently obtains the optimal compensation operator while avoiding the enormous computational overhead of directly solving high-dimensional, strongly coupled inverse problems, thereby achieving strict phase alignment and coherent reconstruction among multiple sub-apertures. In summary, this method can significantly suppress nonlinear defocusing, improve azimuth resolution and energy focusing capability in long-aperture forward-looking imaging, and exhibit good focusing consistency and reconstruction fidelity for extended targets and complex scenes. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the forward-looking radar imaging geometry model in one embodiment; Figure 2 This is a flowchart illustrating a self-focusing imaging method based on adaptive piecewise Doppler compensation in one embodiment. Figure 3 This is a structural block diagram of a self-focusing imaging device based on adaptive segmented Doppler compensation in one embodiment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0022] To analyze the nonlinear motion error in long-aperture forward-looking imaging, a system is established as follows: Figure 1 The image shows the forward-looking radar imaging geometry model. The radar platform is mounted at a height of... On the motion platform, at a constant speed along The target flies at a constant speed in a straight line along its axis. The line of sight between the target and the radar is defined as the radar's line of sight, and the beam direction is at the elevation angle. α The angle between the radar line of sight and the Z-axis, and the beam pointing azimuth angle. θ The angle between the radar line of sight projection on the ground and the X-axis.

[0023] Select the target point location within the imaging area as The point symmetrical about the Y-axis is The position of the radar platform changes over time, slow time... At that time, the coordinates of the transmitting antenna are The coordinates of the receiving antenna are .Target launch slant range and the i The receiving slant range of each received signal It is expressed as follows: ; ; Among them, the i The azimuth coordinates of each receiving antenna element are x i .

[0024] Two-way slant distance is Because radar electromagnetic waves propagate at near the speed of light, and the platform's movement speed is much less than the speed of light, radar echo acquisition generally adopts a "stop-go-stop" working mode: at the moment of single-azimuth sampling, the radar transmits a signal at the instantaneous position, and the platform remains stationary until the echo reception is completed; after the echo reception is completed, the platform moves to the next azimuth sampling position, and the cycle is repeated to complete the all-round echo acquisition.

[0025] The signal transmitted by the radar's central antenna is a linear frequency modulated pulse signal, which can be represented as follows: ; in Indicates a fast time. For rectangular window functions, T This represents the time-domain width of the rectangular window function. Indicates the operating carrier frequency of the transmitted signal. This indicates the frequency modulation slope.

[0026] Assume the backscattering coefficient of the target point is The azimuth window function is , No. i The point target echo received by each receiving array element at a corresponding time can be represented in the following form: ; After down-conversion and pulse compression, the baseband received signals for each channel of the echo analog signal are as follows: ; Where B is the transmission signal bandwidth. It is the carrier wavelength.

[0027] echo After imaging processing, the obtained imaging result vector is At this point, the forward-looking imaging result exhibits left and right Doppler blur, which can be addressed by using a matrix composed of multi-channel guide vectors. This can be solved using beamforming technology.

[0028] In one embodiment, such as Figure 2 As shown, a self-focusing imaging method based on adaptive piecewise Doppler compensation is provided, which is then applied to... Figure 1 Taking the imaging geometry model in the image as an example, the following steps are included: Step 202: Obtain the raw echo signal under radar forward-looking long aperture conditions.

[0029] Specifically, Figure 1The model does not consider the nonlinear variation characteristics of the target slant range history under long-term coherent accumulation conditions. Under short aperture conditions, the relative motion of the target can usually be approximated by a linear model, in which case the echo phase can maintain good coherence throughout the aperture; however, as the synthetic aperture time increases, the higher-order terms in the target slant range history will gradually strengthen, causing the echo phase to deviate from the ideal linear model, thus leading to nonlinear changes in the Doppler history, and further causing spectral broadening and image defocusing. Therefore, it is necessary to further analyze the higher-order nonlinear errors in the slant range history under long aperture conditions. The baseband received signals of each channel are optimized to: ignore... ,exist Phase term Taylor expansion yields: ; Based on geometric relationships, The linear portion is compensated, and the compensated slope distance is... Over time That is, the number of pulses The changes are as follows: ; Phase Term With pulse number The changes are as follows: ; Based on the relationship between the instantaneous Doppler frequency and the phase derivative, the instantaneous Doppler frequency of the target can be obtained as follows: .

[0030] Step 204: Based on the original echo signal, construct an adaptive criterion with phase error as the constraint, and dynamically divide the full aperture based on the adaptive criterion to obtain multiple adaptive sub-apertures.

[0031] Specifically, to suppress the higher-order phase distortion caused by the long aperture, a sub-aperture partitioning strategy is adopted to approximate the nonlinear observation process as a series of locally quasi-stationary processes. Assume the time of the original long aperture is... Divide it into The non-overlapping sub-aperture, the first The time for the aperture segment is ,have: ; Ignoring the influence of array elements, let In the Within a single aperture, the instantaneous two-way slant range of the radar platform relative to the target can be expressed as: ; in, For the first The relative position of the segment target.

[0032] Initial slant distance from the target in the first segment The relationship between them is ; and There is a relationship between them: ; exist to and Taylor expansion yields: ; Therefore: ; The distortions are manifested in the slant range and azimuth angle, respectively. .

[0033] The first can be obtained The first receiving antenna element The relationship between the first echo segment and the second echo segment is approximately as follows: ; Can obtain and The relationship between them: ; and They represent the first The correspondence between the imaging results of the first aperture segment and the first aperture segment on the target is as follows: ; Define the target position from the perspective of the first segment. Equal to the target reference position Substituting, we can get Initial reference positioning of the target The relationship between them: ; Therefore, the first The relationship between the imaging results of segment 1 and the target location in segment 2 is not only related to... It is related to the target's coordinates as well. Considering the entire imaging scene, the relationships between segments cannot be directly expressed using the first... Two-dimensional matrix of imaging results of segment 1 and segment 2. and Now it is vectorized into a column vector and It is possible to obtain Mapped to The equation is:

[0034] in, It is a resampling matrix used to correct spatial position offsets caused by slant range travel; It is a diagonal phase compensation matrix, responsible for eliminating residual higher-order phase errors and space-varying Doppler modulation. It's an error interference.

[0035] Step 206: Perform imaging processing on the echo signals within each adaptive sub-aperture to obtain multiple sub-aperture images.

[0036] Step 208: Establish a cross-segment coherence mapping model between images of different sub-apertures. The model is used to characterize the spatial position offset and residual phase error between sub-apertures.

[0037] Specifically, the complex vector mapping operator is approximately decomposed into a separable matrix operator acting on the range and azimuth directions using the matrix Kronecker product: ; in, It is a distance compensation operator, mainly responsible for correcting slant range aliasing in piecewise linear mappings; This is an azimuth compensation operator designed to eliminate azimuth distortion caused by multi-channel phase deviation and nonlinear motion.

[0038] Due to the nonlinear characteristics of the slope distance evolution under long aperture, simple linear approximation is difficult to completely eliminate the spatial variation error. In order to accurately estimate the above spatial variation mapping parameters, this paper adopts the ADMM optimization method to solve the problem.

[0039] Due to the matrix and pass Since multiplicative coupling exists, directly solving it becomes a non-convex coupled optimization problem. Therefore, it is transformed into a constrained least squares optimization problem, using the Frobenius norm of the matrix residuals to measure the fitting error. The goal is to find the solution that minimizes the fitting error. and Formal expression: ; in, This represents the Frobenius norm of the matrix. The dimension is .

[0040] Step 210: The cross-segment coherent mapping model is transformed into a two-dimensional optimization problem in which the range compensation operator and the azimuth compensation operator can be separated. The alternating direction multiplier method is used to iteratively solve the two-dimensional optimization problem to obtain the optimal compensation operator.

[0041] Specifically, introduce auxiliary variables Make the auxiliary variables satisfy the equality constraints The original problem is transformed into an optimization problem with equality constraints, and the solution is achieved. and The decoupling, after transformation, results in the following problem: ; Introducing Lagrange multipliers and penalty parameters Construct the Lagrange multipliers and the quadratic penalty term, and the corresponding augmented Lagrange function: ; The Alternating Directional Multiplier Method (ADMM) is employed to achieve [optimization] by alternately minimizing the partial derivatives of the augmented Lagrangian function with respect to each optimization variable. , , Successive updates, and through dual variables The update corrects the original constraints and eventually converges to the optimal solution.

[0042] Step 212: Use the optimal compensation operator to compensate and coherently reconstruct each sub-aperture image, and output a fully focused image.

[0043] In the aforementioned self-focusing imaging method based on adaptive piecewise Doppler compensation, firstly, the entire aperture is dynamically divided by constructing an adaptive criterion with phase error as a constraint. This allows for real-time adjustment of the sub-aperture length according to the degree of nonlinear change in the target's motion state, ensuring that the signal within each sub-aperture satisfies the local quasi-linearity condition. This fundamentally avoids the problems of traditional fixed-length segments leading to the failure of the local linear model due to excessive length or the decrease in coherent accumulation gain due to excessive short length. It effectively suppresses Doppler spectrum broadening and spatially varied phase error enhancement caused by higher-order nonlinear terms under long-aperture conditions. Secondly, by establishing a cross-segment coherent mapping model between images of different sub-apertures and transforming this model into a two-dimensional optimization problem with separable range and azimuth compensation operators, joint characterization and decoupling processing of spatial position offset and residual phase error are achieved, significantly improving the flexibility of spatially varied error compensation in long-aperture scenarios. Finally, the Alternating Directional Multiplier Method (ADMM) is employed to iteratively solve this two-dimensional optimization problem. This method efficiently obtains the optimal compensation operator while avoiding the enormous computational overhead of directly solving high-dimensional, strongly coupled inverse problems, thereby achieving strict phase alignment and coherent reconstruction among multiple sub-apertures. In summary, this method can significantly suppress nonlinear defocusing, improve azimuth resolution and energy focusing capability in long-aperture forward-looking imaging, and exhibit good focusing consistency and reconstruction fidelity for extended targets and complex scenes.

[0044] In one embodiment, a second-order expansion model of the sub-aperture slant range history under long aperture conditions is established. Phase of the segment echo: ; in, The target slant distance, The beam pointing azimuth angle, For the first Phase function of the aperture echo. The carrier wavelength of the transmitted signal. For the first Instantaneous two-way slant distance of the aperture. Let X be the velocity of the radar platform along the X-axis. For the first The duration of the joke's aperture. Let azimuth be the time variable. Derive the formula for spectral broadening: ; in, For spectrum broadening, The instantaneous Doppler frequency, For the first The duration of the joke's aperture.

[0045] According to the Rayleigh criterion, the maximum residual higher-order phase error within each sub-aperture must not exceed [a certain value]. Radius. This physical boundary directly determines the upper limit of degradation in azimuth measurement resolution, corresponding to the following constraint: ; Organize the sub-aperture duration into an explicit constraint: ; in, For platform speed, The target slant distance, The beam pointing pitch angle, The beam pointing azimuth angle, The height of the radar platform relative to the ground. The carrier wavelength of the transmitted signal. For the first The duration of each sub-aperture is determined based on the explicit constraints, resulting in multiple adaptive sub-apertures.

[0046] In one embodiment, a slant range history model between the first sub-aperture and the k-th sub-aperture is established based on the geometric relationship between the radar and the target: ; ; ; in, For the first In the aperture segment, the instantaneous two-way slant distance, expressed in terms of the target position parameter of the first segment, The target slant distance, The beam pointing pitch angle, The beam pointing azimuth angle, This represents the horizontal displacement caused by the platform's motion. A Taylor expansion of the slant distance yields first- and second-order approximate expressions: ; ; We can obtain: ; in, This represents the slant distance deviation. Establish the first... The approximate relationship between the first echo segment and the first echo segment, and the mapping relationship between angle and distance: ; ; in, For the first The receiving antenna, the first The echo signal of the aperture segment, The imaginary unit, It is a complex exponential function. Vectorization of the first... The correspondence between the imaging results of segment 1 and segment 2: ; ; in, For the first The imaging results of the segment, The imaging results for segment 1. For reference slope distance, For matrix vectorization operations, This is a resampling matrix used to correct spatial position offsets caused by slant range travel. This is a diagonal phase compensation matrix used to eliminate residual higher-order phase errors and space-varying Doppler modulation. To mitigate error interference, the mapping relationship is decomposed into a distance compensation operator. and azimuth compensation operator The separable form is used to construct a cross-segment coherent mapping model between images of different sub-apertures: ; ; in, , These are image matrices for different sub-apertures. It is the Frobenius norm. and The dimension is .

[0047] In one embodiment, auxiliary variables and Lagrange multipliers are introduced to transform the coupled optimization problem of the range compensation operator and the azimuth compensation operator into an optimization problem with equality constraints: ; in, For auxiliary variable matrix, Let be the Lagrange multipliers. Construct the augmented Lagrange function for the optimization problem: ; in, To augment the Lagrange function, For penalty parameters, N For azimuth compensation operator, M For the distance compensation operator, Z is an auxiliary variable. , This is a quadratic penalty term. Within the iterative framework of the alternating direction multiplier method, by fixing other variables, the azimuth compensation operator, the range compensation operator, and the auxiliary variables are alternately minimized and updated. After each iteration, the Lagrange multipliers are updated until the original residual and the dual residual satisfy the convergence condition, outputting the final decoupled optimal range compensation operator and optimal azimuth compensation operator.

[0048] In one embodiment, the current range compensation operator and auxiliary variables are fixed, and the azimuth compensation operator is updated by solving a least-squares problem. Using the updated azimuth compensation operator as a known quantity, and with the auxiliary variables fixed, the range compensation operator is updated by solving another least-squares problem. The updated azimuth compensation operator and the range compensation operator are then used to update the auxiliary variables by solving a Sylvester equation.

[0049] In one embodiment, the solution process is shown in Algorithm 1 below:

[0050] It should be understood that, although Figure 2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0051] In one embodiment, such as Figure 3 As shown, a self-focusing imaging device based on adaptive piecewise Doppler compensation is provided, comprising: a signal acquisition module 302, an adaptive piecewise module 304, a sub-aperture imaging module 306, a model building module 308, an optimization solution module 310, and an output module 312, wherein: The signal acquisition module 302 is used to acquire the raw echo signal under radar forward-looking long aperture conditions.

[0052] The adaptive segmentation module 304 is used to construct an adaptive criterion with phase error as a constraint based on the original echo signal, and to dynamically divide the full aperture based on the adaptive criterion to obtain multiple adaptive sub-apertures.

[0053] The sub-aperture imaging module 306 is used to perform imaging processing on the echo signals in each adaptive sub-aperture to obtain multiple sub-aperture images.

[0054] Model building module 308 is used to establish a cross-segment coherent mapping model between images of different sub-apertures. The model is used to characterize the spatial position offset and residual phase error between sub-apertures.

[0055] The optimization solution module 310 is used to transform the cross-segment coherent mapping model into a two-dimensional optimization problem in which the range compensation operator and the azimuth compensation operator can be separated, and to iteratively solve the two-dimensional optimization problem using the alternating direction multiplier method to obtain the optimal compensation operator.

[0056] Output module 312 is used to compensate and coherently reconstruct each sub-aperture image using the optimal compensation operator, and output a fully focused image.

[0057] Specific limitations regarding the autofocusing imaging device based on adaptive segmented Doppler compensation can be found in the limitations of the autofocusing imaging method based on adaptive segmented Doppler compensation described above, and will not be repeated here. Each module in the aforementioned autofocusing imaging device based on adaptive segmented Doppler compensation can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0058] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0059] 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 computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink, DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0060] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0061] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A self-focusing imaging method based on adaptive piecewise Doppler compensation, characterized in that, The method includes: Acquire the raw echo signal under radar forward-looking long aperture conditions; Based on the original echo signal, an adaptive criterion with phase error as a constraint is constructed, and the full aperture is dynamically divided based on the adaptive criterion to obtain multiple adaptive sub-apertures. The echo signals within each adaptive sub-aperture are processed separately to obtain multiple sub-aperture images; A cross-segment coherence mapping model is established between images with different sub-apertures. The model is used to characterize the spatial position offset and residual phase error between sub-apertures. The cross-segment coherent mapping model is transformed into a two-dimensional optimization problem in which the range compensation operator and the azimuth compensation operator can be separated. The alternating direction multiplier method is used to iteratively solve the two-dimensional optimization problem to obtain the optimal compensation operator. The optimal compensation operator is used to compensate and coherently reconstruct each of the sub-aperture images to output a fully focused image.

2. The method according to claim 1, characterized in that, Based on the original echo signal, an adaptive criterion with phase error as a constraint is constructed, and the full aperture is dynamically divided based on the adaptive criterion to obtain multiple adaptive sub-apertures, including: A second-order expansion model of the sub-aperture slant distance history under long aperture conditions is established, where the k-th sub-aperture segment corresponds to a duration of... Phase: in, Let k be the target slant distance. Let the k-th beam be the pitch angle. The beam pointing azimuth angle, For the first Phase function of the aperture echo. The carrier wavelength of the transmitted signal. For the first Instantaneous two-way slant distance of the aperture. Let X be the velocity of the radar platform along the X-axis. For azimuth time variables, For the first The duration of the joke's aperture; Derivation of the formula for spectral broadening: in, For spectrum broadening, The instantaneous Doppler frequency; According to the Rayleigh criterion, the maximum residual higher-order phase error within each sub-aperture must not exceed [a certain value]. Radius, set the following constraints: Organize the sub-aperture duration into an explicit constraint: in, For platform speed, The target slant distance, The beam pointing pitch angle, The beam pointing azimuth angle, The height of the radar platform relative to the ground. The carrier wavelength of the transmitted signal; Based on the explicit constraints, the division length of each sub-aperture is determined, resulting in multiple adaptive sub-apertures.

3. The method according to claim 1, characterized in that, Establish a cross-segment coherence mapping model between images of different sub-apertures, including: Based on the geometric relationship between the radar and the target, a slant range history model is established between the first sub-aperture and the k-th sub-aperture: in, For the first In the aperture segment, the instantaneous two-way slant distance, expressed in terms of the target position parameter of the first segment, The target slant distance, The beam pointing pitch angle, The beam pointing azimuth angle, The horizontal displacement generated by the platform's motion; Performing a Taylor expansion on the slant range yields first-order and second-order approximate expressions: We can obtain: in, This refers to the slope distance deviation; Establish the first The approximate relationship between the first echo segment and the first echo segment, and the mapping relationship between angle and distance: in, For the first The receiving antenna, the first The echo signal of the aperture segment, The imaginary unit, It is a complex exponential function; Vectorization The correspondence between the imaging results of segment 1 and segment 2: in, For the first The imaging results of the segment, The imaging results for segment 1 are as follows. For matrix vectorization operations, This is a resampling matrix used to correct spatial position offsets caused by slant range travel. This is a diagonal phase compensation matrix used to eliminate residual higher-order phase errors and space-varying Doppler modulation. For error interference; The mapping relationship is decomposed into a distance compensation operator. and azimuth compensation operator The separable form is used to construct a cross-segment coherent mapping model between images of different sub-apertures: in, , These are image matrices for different sub-apertures. It is the Frobenius norm. and The dimension is .

4. The method according to claim 3, characterized in that, The two-dimensional optimization problem is solved iteratively using the alternating direction multiplier method to obtain the optimal compensation operator, including: By introducing auxiliary variables and Lagrange multipliers, the coupled optimization problem of the range compensation operator and the azimuth compensation operator is transformed into an optimization problem with equality constraints: in, For auxiliary variable matrix, For Lagrange multipliers; Construct the augmented Lagrangian function for the optimization problem: in, To augment the Lagrange function, For penalty parameters, N For azimuth compensation operator, M For the distance compensation operator, Z is an auxiliary variable. , This is a secondary penalty item; Within the iterative framework of the alternating direction multiplier method, by fixing other variables, the azimuth compensation operator, the range compensation operator, and the auxiliary variables are alternately minimized and updated in sequence. After each iteration, the Lagrange multipliers are updated until the original residual and the dual residual satisfy the convergence condition, and the optimal distance compensation operator and the optimal azimuth compensation operator are finally decoupled.

5. The method according to claim 4, characterized in that, Within the iterative framework of the alternating direction multiplier method, by fixing other variables, the azimuth compensation operator, the range compensation operator, and the auxiliary variables are alternately minimized and updated, including: With the current distance compensation operator and auxiliary variables fixed, the azimuth compensation operator is updated by solving a least squares problem; The updated azimuth compensation operator is used as a known quantity, and the auxiliary variable is fixed. The distance compensation operator is updated by solving another least squares problem. The auxiliary variables are updated by solving a Sylvester equation using the updated azimuth compensation operator and the range compensation operator.

6. A self-focusing imaging device based on adaptive piecewise Doppler compensation, characterized in that, The device includes: The signal acquisition module is used to acquire the raw echo signal under radar forward-looking long aperture conditions; An adaptive segmentation module is used to construct an adaptive criterion with phase error as a constraint based on the original echo signal, and to dynamically divide the full aperture based on the adaptive criterion to obtain multiple adaptive sub-apertures. The sub-aperture imaging module is used to perform imaging processing on the echo signals in each adaptive sub-aperture to obtain multiple sub-aperture images; The model building module is used to establish a cross-segment coherence mapping model between images with different sub-apertures. The model is used to characterize the spatial position offset and residual phase error between sub-apertures. The optimization solution module is used to transform the cross-segment coherent mapping model into a two-dimensional optimization problem in which the range compensation operator and the azimuth compensation operator can be separated, and to iteratively solve the two-dimensional optimization problem using the alternating direction multiplier method to obtain the optimal compensation operator. The output module is used to compensate and coherently reconstruct each of the sub-aperture images using the optimal compensation operator, and output a fully focused image.

Images