A synthetic aperture radar imaging method based on joint statistical inference

Abstract

This invention belongs to the field of synthetic aperture radar (SAR) imaging technology. It proposes a SAR imaging method based on joint statistical inference. By processing the radar echo signal in the range direction, a linear imaging model is established. Based on this, a target feature constraint model is introduced to characterize the sparsity and structural continuity of the imaging target. Combined with modeling methods for the statistical characteristics of interference under complex electromagnetic environments, a joint statistical modeling framework for target features and environmental interference is constructed. Then, by constructing a hierarchical inference structure, joint inference of target features and environmental interference is performed, and a numerical inference method is used to solve for the posterior distribution, thereby achieving stable reconstruction of the imaging target. This invention can effectively suppress background interference under complex electromagnetic interference conditions, maintain the structural integrity of the target, and improve imaging accuracy, structural preservation capability, and adaptability to complex environments.

Description

Technical Field

[0001] This invention belongs to the field of synthetic aperture radar imaging technology, and in particular relates to a synthetic aperture radar imaging method based on joint statistical inference. Background Technology

[0002] Synthetic Aperture Radar (SAR) is a microwave remote sensing tool with advantages such as all-weather, all-time capability, long range, and high resolution. By utilizing the relative motion between the target and the radar, a finite aperture can be equivalently synthesized into a long-aperture antenna, thereby obtaining high-resolution azimuth imaging results. The essence of SAR imaging is... Solve the linear inverse problem, where This represents the observed data or echo signal. The target feature to be recovered. For Fourier dictionaries, This is due to additive environmental interference. In practical applications, due to factors such as complex background interference, incomplete observations, and model uncertainties, synthetic aperture radar imaging often exhibits underdeterministic, high-dimensional, or even ill-conditioned characteristics. Therefore, traditional imaging methods often struggle to obtain stable and reliable imaging results.

[0003] Against this backdrop, the development of Compressive Sensing (CS) technology has provided a theoretical foundation for sparse signal reconstruction, enabling high-quality imaging even under limited observation conditions. Currently, synthetic aperture radar (SAR) imaging methods based on joint statistical inference, utilizing CS theory for interference suppression and sparse feature learning, mainly fall into two categories: convex optimization algorithms and statistical Bayesian algorithms. While convex optimization algorithms offer global solvability and convergence guarantees, they cannot provide confidence assessments for signal estimation, thus limiting their robustness in environments with strong interference or noise. Statistical Bayesian algorithms, on the other hand, can leverage the uncertainty of echo signals to improve signal recovery accuracy, while simultaneously providing interval estimates including confidence levels. Their accuracy and robustness are superior to convex optimization algorithms.

[0004] However, Bayesian algorithms place certain requirements on prior modeling. When it is difficult to form a concise joint inference structure between the prior model and the observation model, the posterior distribution is usually difficult to directly obtain a computable expression, thus increasing the complexity of the imaging inference process and the difficulty of engineering implementation. Secondly, existing variational Bayesian inference frameworks can only handle closed posterior solutions of a few typical sparse priors such as Laplace priors. In practical applications, single or static priors cannot represent refined features and it is difficult to maintain both the sparsity and structural continuity of target features under complex interference conditions.

[0005] Furthermore, existing convex optimization and Bayesian inference methods generally rely on additive white Gaussian noise models when modeling environmental interference. However, the forms of interference in real electromagnetic environments are far more complex than ideal models. In addition to common additive white Gaussian noise, there are also electromagnetic interference, ground clutter, and other typical non-Gaussian interferences. Therefore, traditional single additive interference models are no longer sufficient to accurately characterize complex interference features and cannot meet the requirements for imaging accuracy and robustness in multi-source interference environments.

[0006] In summary, traditional imaging methods struggle to maintain a stable ability to quantitatively characterize the fine features of a target in complex electromagnetic environments. Summary of the Invention

[0007] In view of this, the present invention aims to propose a synthetic aperture radar imaging method based on joint statistical inference to solve the problem that traditional imaging methods are difficult to maintain stable quantitative characterization of the fine features of targets in complex electromagnetic environments.

[0008] To achieve the above objectives, the technical solution of the present invention is implemented as follows: This invention provides a synthetic aperture radar imaging method based on joint statistical inference, comprising: Synthetic Aperture Radar (SAR) echo signals are preprocessed to obtain range-compressed domain echo data. Based on the range-compressed domain echo data, a linear observation model between the range-compressed domain echo data and the imaging target features is constructed. The signal preprocessing includes range migration correction and range compression processing. By introducing target feature constraint information to describe the spatial distribution characteristics of imaging targets, statistical modeling is performed on the sparse characteristics and structural continuity characteristics of imaging targets in the spatial domain, and a target feature constraint model is constructed to characterize the sparse characteristics and structural continuity of imaging targets. Environmental interference statistics are introduced to describe the statistical characteristics of interference in complex electromagnetic environments. Multimodal and non-Gaussian interference in range-compressed domain echo data are modeled to construct an environmental interference model to characterize complex electromagnetic environments. Under the unified statistical inference framework, the target feature constraint information and the environmental interference statistical information are incorporated into the same joint posterior model to construct a joint posterior distribution model that includes observation data, imaging target features and interference statistical parameters. The joint posterior distribution model is reconstructed by a structured decomposition method, and then iteratively solved by a numerical inference method to obtain the posterior estimation result of the imaging target features. The posterior estimation result is then output as the synthetic aperture radar imaging result.

[0009] Furthermore, the linear observation model is expressed as: in, For range-compressed domain echo data, The target feature to be recovered. For Fourier dictionaries, This is due to additive environmental interference.

[0010] Furthermore, the target feature constraint information includes constraint terms for describing the sparsity characteristics of the target space and constraint terms for describing the structural continuity of the target space.

[0011] Furthermore, the introduction of environmental interference statistics to describe the statistical characteristics of interference in complex electromagnetic environments, modeling multimodal and non-Gaussian interference in range-compressed domain echo data, and constructing an interference model to characterize complex electromagnetic environments includes: The environmental interference statistics are characterized by a mixture distribution model, and a Gaussian mixture model is used to model the interference distribution, constructing an interference model to characterize complex electromagnetic environments. Its probability distribution is expressed as follows: in, This indicates that the mean is 0 and the covariance matrix is... The complex Gaussian distribution, and They represent the first The mixture weighting coefficients and variance of the Gaussian are: , Let be the order of the Gaussian mixture.

[0012] Furthermore, under the unified statistical inference framework, the target feature constraint information and the environmental interference statistical information are incorporated into the same joint posterior model to construct a joint posterior distribution model that includes observation data, imaging target features, and interference statistical parameters, including: The target feature constraint information is introduced into the joint posterior model through a prior probability model, and a joint prior form containing sparse constraint terms and structural continuity constraint terms is adopted.

[0013] Furthermore, the step of incorporating the target feature constraint information and the environmental interference statistical information into the same joint posterior model within a unified statistical inference framework to construct a joint posterior distribution model containing observation data, imaging target features, and interference statistical parameters also includes: Based on the linear observation model and the environmental disturbance model, a conditional probability model for the observation data under given target characteristics and disturbance statistical parameters is introduced to describe the generation mechanism of the observation data.

[0014] Furthermore, the numerical inference method employs a posterior solution method that combines a near-end update mechanism with a random sampling mechanism.

[0015] Compared with existing technologies, the synthetic aperture radar imaging method based on joint statistical inference described in this invention has the following advantages: (1) The synthetic aperture radar imaging method based on joint statistical inference described in this invention introduces target feature constraint information and environmental interference statistical information under a unified inference framework, and performs joint modeling and collaborative inference on imaging target features and complex interference background. This effectively reduces the influence of complex electromagnetic environment and non-Gaussian interference on imaging results, thereby significantly improving the robustness of synthetic aperture radar imaging in complex environments.

[0016] (2) The synthetic aperture radar imaging method based on joint statistical inference described in this invention considers the sparsity of the target in spatial distribution and the continuity of local structure during the target feature modeling process. Through the synergistic effect of multiple target feature constraint information, it can better maintain the contour structure and detail information of the target while improving the imaging contrast and resolution, thus avoiding the structural damage or detail loss problems common in existing methods.

[0017] (3) The synthetic aperture radar imaging method based on joint statistical inference described in this invention can effectively characterize the multi-mode interference characteristics that may exist in the echo data by modeling and introducing the statistical characteristics of interference in complex electromagnetic environment, making the imaging process more adaptable to the form of interference distribution, thereby improving the reliability of imaging results under strong interference and non-ideal observation conditions.

[0018] (4) The synthetic aperture radar imaging method based on joint statistical inference described in this invention constructs a joint inference structure for synthetic aperture radar imaging, integrates target feature constraint information and environmental interference statistical information into the same inference framework, and performs joint inference through decomposition and coordination, effectively avoiding the performance limitations caused by a single model or a single prior constraint, and improving the overall expressive power of the imaging model.

[0019] (5) The synthetic aperture radar imaging method based on joint statistical inference described in this invention adopts a hierarchical modeling and step-by-step inference process to decompose the complex imaging problem into multiple feasible sub-problems. While ensuring inference accuracy, it has good engineering feasibility. At the same time, the inference framework can flexibly introduce or adjust target feature constraint information and environmental interference statistical information according to different application scenarios, and has good scalability and applicability.

[0020] (6) The synthetic aperture radar imaging method based on joint statistical inference described in this invention improves the resolution, contrast and structural feature preservation of the imaging results. The synthetic aperture radar imaging results obtained by this invention can provide more stable and reliable basic data support for subsequent target detection, recognition and analysis tasks. Attached Figure Description

[0021] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart of a synthetic aperture radar imaging method based on joint statistical inference as described in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the synthetic aperture radar imaging geometric model in the synthetic aperture radar imaging method based on joint statistical inference described in Embodiment 1 of the present invention; Figure 3 This is a hierarchical Bayesian directed acyclic graph used in the synthetic aperture radar imaging method based on joint statistical inference described in Embodiment 1 of the present invention to jointly model target features and environmental interference. Figure 4 This is a schematic diagram of the joint posterior inference of target features and environmental interference in the synthetic aperture radar imaging method based on joint statistical inference described in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the imaging results obtained by different imaging processing methods in the synthetic aperture radar imaging method based on joint statistical inference described in Embodiment 1 of the present invention; Figure 6 This is a schematic diagram of the synthetic aperture radar imaging device according to Embodiment 2 of the present invention; Figure 7 This is a schematic diagram of the synthetic aperture radar imaging terminal described in Embodiment 3 of the present invention. Detailed Implementation

[0022] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0023] Example 1 To address the problem of difficulty in quantifying and characterizing the fine features of targets in complex electromagnetic environments in the field of synthetic aperture radar imaging, this embodiment provides a synthetic aperture radar imaging method based on joint statistical inference.

[0024] Due to structural bottlenecks in the modeling level of synthetic aperture radar (SAR) imaging systems, improvements to single modules are insufficient to address the challenges of complex environmental interference and multi-feature representation. Therefore, it is necessary to approach the modeling from the perspective of the overall system, jointly modeling target features and complex environmental interference, constructing a unified information fusion framework, and thus solving the key technical problem of refined target feature extraction under complex electromagnetic environments.

[0025] Based on the aforementioned technical problems, this embodiment provides a synthetic aperture radar imaging method based on joint statistical inference that performs collaborative quantization modeling of multiple target features and achieves stable posterior inference under multimodal non-Gaussian interference conditions.

[0026] Figure 1 This is a flowchart of a synthetic aperture radar imaging method based on joint statistical inference as described in Embodiment 1 of the present invention. See also... Figure 1 This synthetic aperture radar imaging method based on joint statistical inference specifically includes the following steps: Step 1: Perform signal preprocessing on the synthetic aperture radar echo signal to obtain range compressed domain echo data, and construct a linear observation model between the range compressed domain echo data and the imaging target features based on the range compressed domain echo data; wherein, the signal preprocessing includes range migration correction and range compression processing.

[0027] Using the above steps, by applying signal processing methods such as range migration correction and range compression, the range migration error in the echo signal is corrected, and the echo data is converted to the range compression domain, providing a reliable data foundation for subsequent imaging processing and target feature modeling.

[0028] Step 2: Introduce target feature constraint information to describe the spatial distribution characteristics of the imaging target, statistically model the sparse characteristics and structural continuity characteristics of the imaging target in the spatial domain, and construct a target feature constraint model to characterize the sparse characteristics and structural continuity of the imaging target, so as to achieve the collaborative characterization of different features.

[0029] By using the above steps, constraint information is introduced to characterize the features of the imaging target, and the sparse characteristics and structural continuity features of the target in the spatial domain are modeled to improve the stability and accuracy of target feature recovery in complex electromagnetic environments.

[0030] Step 3: Introduce environmental interference statistics to describe the statistical characteristics of interference in complex electromagnetic environments, model multimodal and non-Gaussian interference in range-compressed domain echo data, and construct an environmental interference model to characterize complex electromagnetic environments.

[0031] By using the above steps, statistical information for characterizing the interference characteristics of complex electromagnetic environments is introduced to model the multi-source and non-ideal interference features presented in synthetic aperture radar echo signals, thereby enhancing the robustness of the imaging process under strong interference conditions.

[0032] Step 4: Under the unified statistical inference framework, the target feature constraint information and the environmental interference statistics information are incorporated into the same joint posterior model to construct a joint posterior distribution model that includes observation data, imaging target features and interference statistical parameters.

[0033] Using the above steps, target feature constraint information and environmental interference statistical information are jointly modeled and collaboratively inferred under a unified inference framework. A joint posterior model of target features and complex environment is constructed, providing an inference basis for subsequent posterior distribution solution and imaging result generation.

[0034] Step 5: Reconstruct the joint posterior distribution model through structured decomposition, and iteratively solve the joint posterior distribution model through numerical inference to obtain the posterior estimation result of the imaging target features. Then, output the posterior estimation result as the synthetic aperture radar imaging result.

[0035] Using the above steps, the posterior distribution of the auxiliary variable can be solved by using a random sampling numerical inference method to statistically sample the posterior distribution of the split variable and the dual variable introduced in the inference process described in step 4. In one embodiment, the Markov chain Monte Carlo method can be used for statistical sampling. A numerical update mechanism suitable for non-smooth constraint terms is applied to process the non-smooth prior structure, thereby obtaining high-resolution synthetic aperture radar imaging results and providing confidence intervals.

[0036] This method reconstructs the complex inference structure to achieve joint inference of target features and environmental interference under a unified constraint framework, thereby obtaining stable and reliable imaging results in complex electromagnetic environments. Simultaneously, this method establishes a dual-hybrid prior model for target features and the complex electromagnetic environment. By introducing auxiliary variables, the complex posterior inference problem is reconstructed into more manageable subproblems, and alternating updates are performed under consistency constraints to obtain an approximate estimate of the posterior distribution and corresponding statistics, thus achieving high-resolution synthetic aperture radar imaging.

[0037] Figure 2 This is a schematic diagram of the synthetic aperture radar (SAR) imaging geometric model in the SAR imaging method based on joint statistical inference described in Embodiment 1 of the present invention. See also... Figure 2 The specific preprocessing procedure for synthetic aperture radar signals in step 1 is as follows: Airborne synthetic aperture radar platform at speed The radar moves at a constant speed along a predetermined track. Through the relative motion between the radar platform and the imaging area, an equivalent synthetic aperture is formed in the heading direction, thereby enabling observation and imaging of the imaging area. The radar illuminates the imaging area multiple times during the platform's movement, acquiring the target's scattered echo signals at different observation angles.

[0038] During the imaging process, the synthetic aperture radar continuously transmits periodic, wide-bandwidth linear frequency modulated (LFM) signals and simultaneously receives the scattered echoes from the target. The LFM transmitted signal can be expressed as: in The time width is T The rectangle function, For the center frequency, To adjust the frequency, To save time.

[0039] After demodulation and imaging preprocessing, the radar echo signal is subjected to range migration correction using Image Formation Processing (IFP). The corrected data is then further interpolated using the Polar Format Algorithm (PFA) to obtain the synthetic aperture radar echo data in the wavenumber domain. in, This indicates additive interference during the transmission and reception of echo signals. Represents range beam. The distance between the airborne radar platform and the center of the target scene. For radar platform operating speed, For azimuth beam, This is a slow time.

[0040] Perform a Fourier transform on the range data to obtain the range compressed domain data: in, This is expressed as the response function in the range direction. It is represented as a linear phase in the azimuth direction.

[0041] In summary, after synthetic aperture radar signal transmission, echo reception, and imaging domain signal processing, the relationship between radar echo data and imaging target features can be modeled as the following linear observation model: in, For range-compressed domain echo data, The target feature to be recovered. For Fourier dictionaries, This is due to additive environmental interference.

[0042] In one embodiment of this example, the target feature constraint information includes constraint terms for describing the sparsity characteristics of the target space and constraint terms for describing the structural continuity of the target space.

[0043] For example, the specific steps of step 2 are as follows: Based on the linear observation model established in step 1, in order to achieve stable recovery of the refined features of the imaging target under complex electromagnetic environments, target feature constraint information is introduced to describe the statistical properties of the imaging target features. The statistical characteristics are modeled.

[0044] Target feature constraint information is used to characterize the typical features and attributes of the imaging target in the spatial domain, including but not limited to the sparsity of the spatial distribution of the target's reflectance features and the structural continuity between adjacent resolution units. This target feature constraint information serves as an information source in the joint inference process, guiding the imaging results to maintain spatial structural features consistent with the target's physical characteristics while satisfying observational consistency.

[0045] Within the statistical modeling framework, the target feature constraint information can be introduced into the joint inference process in a priori form, the specific mathematical expression of which can be selected according to application requirements. In one implementation, a prior model that simultaneously includes sparsity constraint terms and structural continuity constraint terms can be used to jointly constrain the imaging target features. For example, a priori form based on a generalized Gaussian distribution and total variational constraints can be expressed as follows: in It is a positive coefficient, used to balance the influence of different constraint information in the inference process; Used to characterize the sparsity of the spatial distribution of features of the imaging target, where This is an adjustable parameter used to adjust the strength of sparsity; Secondly , This is a two-dimensional discrete gradient operator used to describe the local continuity and structural consistency of imaging target features in the spatial domain. The aforementioned prior form serves as a feasible implementation of target feature constraint information, illustrating the specific methods for introducing this constraint information.

[0046] Using the above-described configuration, the introduced target feature constraint information can guide the sparsity of the imaging results while ensuring the continuity of the imaging target structure, preventing the loss of target structure due to the introduction of sparsity. It works in conjunction with the environmental interference statistics introduced in subsequent steps to constrain and update the imaging target features within a unified inference framework.

[0047] In one embodiment of this example, step 3 specifically includes the following steps: Based on the linear observation model established in step 1, to characterize the statistical characteristics of environmental interference and noise in radar echoes under complex electromagnetic environments, interference statistics are introduced to describe the distribution characteristics of environmental interference. This involves analyzing the interference term... Modeling the statistical behavior.

[0048] Environmental interference statistics are used to describe the multimodal, non-Gaussian, and statistically varying interference components that may exist in actual radar echoes, reflecting the impact of complex electromagnetic environments on the imaging process. By incorporating these interference statistics into the inference process, the model's adaptability to anomalous interference and background clutter changes can be effectively improved, thus providing a more reasonable statistical description for subsequent joint inference.

[0049] In subsequent step 4, within a unified statistical modeling framework, environmental disturbance statistics can be characterized using a mixture distribution model. The specific mathematical form of this model can be selected based on the actual disturbance characteristics. Specifically, a Gaussian mixture model can be used to model the statistical distribution of the disturbance term, and its probability distribution can be expressed as: in, This indicates that the mean is 0 and the covariance matrix is... The complex Gaussian distribution, and They represent the first The mixture weighting coefficients and variance of the Gaussian are: , Let be the order of the Gaussian mixture.

[0050] In one embodiment of this example, step 3 further includes the following steps: The target feature constraint information is introduced into the joint posterior model through a prior probability model, and a joint prior form containing sparse constraint terms and structural continuity constraint terms is adopted.

[0051] For details, see Figure 3 and Figure 4To facilitate modeling the uncertainty of interference statistical parameters within a unified inference framework and improve the model's adaptability to complex environmental changes, a hierarchical Bayesian modeling approach is introduced to hierarchically model the mixture weights and covariance parameters of the interference distribution. Through this method, complex interference statistical characteristics can be incorporated into the joint inference process, working in conjunction with target feature constraints to robustly update the imaging target features.

[0052] Using the above setup, the introduced environmental interference statistical model and its hierarchical modeling method serve as a feasible way to realize environmental interference statistical information. Its specific distribution form, mixing order, and parameter hierarchy can be flexibly selected and adjusted according to the actual electromagnetic environment characteristics and interference statistical features to adapt to the interference modeling needs under different imaging scenarios.

[0053] In one embodiment of this example, step 4 further includes the following steps: Based on the linear observation model and the environmental disturbance model, a conditional probability model for the observation data under given target characteristics and disturbance statistical parameters is introduced to describe the generation mechanism of the observation data.

[0054] Based on the introduction of target feature constraint information and environmental interference statistical information in steps 2 and 3 respectively, a joint statistical inference model is constructed that includes observation data, imaging target features and environmental interference statistical parameters to achieve robust estimation of imaging target features under complex electromagnetic environments.

[0055] Based on the linear observation model established in step 1 and combined with the modeling results of the environmental disturbance statistical characteristics in step 3, a conditional probability model (i.e., likelihood function) of the observation data under given target characteristics and disturbance statistical parameters is introduced to describe the generation mechanism of the observation data. In one embodiment, according to the environmental disturbance statistical model adopted in step 3, the likelihood function can be expressed in Gaussian mixture form: After introducing the likelihood function and the target feature constraint information given in step 2, a joint posterior probability model containing observation data, target features, and interference statistical parameters can be constructed. The joint posterior probability distribution can be expressed as: The first term describes the observation consistency constraint, while the latter two terms correspond to the target feature sparsity constraint and structural continuity constraint introduced in step 2, respectively.

[0056] Since the aforementioned joint posterior distribution contains multiple constraints and statistical parameters, its analytical form is often difficult to solve directly. To improve the computability of posterior inference, the inference concepts of variable splitting and data augmentation are introduced to structurally reconstruct the joint posterior inference problem. Specifically, by introducing splitting variables, the target feature constraints and observation consistency constraints are mapped to different subproblems, and dual variables are used to maintain consistency among the solutions of each subproblem.

[0057] In an optional implementation of this embodiment, a splitting variable is introduced. , and the corresponding dual variables , The joint posterior distribution described above can be rewritten in the following form: Among them, the splitting variable is used to carry different constraint information respectively, and the dual variable is used to guide the solutions of each subproblem to be coordinated and updated under the consistency constraint.

[0058] Based on the aforementioned structured posterior model, a decomposition-coordination variational inference method can be used to infer the joint posterior distribution. First, the split sub-problems are solved iteratively. Then, consistency constraints are used to fuse the solutions to each sub-problem, ultimately obtaining the posterior distribution and statistics of the imaging target features. Based on this posterior distribution, point estimation and uncertainty characterization of the imaging results can be further obtained, enabling robust synthetic aperture radar imaging in complex environments.

[0059] In one embodiment of this implementation, the numerical inference method in step 5 adopts a posterior solution method that combines a proximal update mechanism and a random sampling mechanism.

[0060] Specifically, the steps in step 5 are as follows: Based on the joint posterior inference model constructed in step 4, the joint posterior distribution is solved to obtain specific estimation results of the imaging target features. Since the joint posterior distribution obtained in step 4 simultaneously contains target feature constraint information and environmental interference statistics, and introduces split variables and consistency constraints, its analytical form is usually difficult to solve directly. Therefore, a numerical inference method is used to approximate the solution of the posterior distribution.

[0061] In one implementation, a splitting variable is introduced based on step 4. , and the corresponding dual variables , A posterior solution method combining proximal operators and random sampling is employed to iteratively solve the conditional posterior distribution of each sub-variable. Specifically, for the split variable... Its conditional posterior distribution can be expressed as: The first term corresponds to the target feature sparsity constraint introduced in step 2, and the second term is used to constrain the consistency relationship between the splitting variable and the imaging target feature.

[0062] For split variables Its conditional posterior distribution can be expressed as: The first term describes the structural continuity constraint of the target feature in the spatial domain, and the second term guides the consistency between the split variable and the target feature estimation results.

[0063] For dual variables , Its conditional posterior distribution can be uniformly expressed as: The dual variable is used to adjust the deviation between the split variable and the target feature estimation result during the iteration process, so as to achieve coordinated updating of the solutions of each subproblem under the consistency constraint.

[0064] Based on the aforementioned conditional posterior model, the joint posterior distribution is gradually approximated by alternately updating the splitting variable, dual variable, and imaging target features within a unified inference framework. As the iteration process proceeds, the estimation results of each sub-variable gradually converge under the consistency constraint, thereby obtaining the posterior distribution and statistical properties of the imaging target features.

[0065] Using the above method, based on the posterior distribution, point estimation results of the imaging target features can be further obtained and output as the final synthetic aperture radar imaging result. Through the synergistic effect of the target feature constraint information introduced in step 2, the environmental interference statistics introduced in step 3, and the joint inference structure constructed in step 4, the resolution and structural feature preservation capability of the imaging results can be effectively improved under complex electromagnetic environments and strong interference conditions, achieving high-quality synthetic aperture radar imaging.

[0066] The implementation process and technical effects of the method of the present invention will be further explained below with specific examples: In this example, publicly available ground static scene radar data is used to verify the method of the present invention. The radar system and experimental parameters are shown in Table 1, including parameters such as the radar system operating band, radar detection range, transmitted signal bandwidth, and pulse repetition frequency.

[0067] Table 1 Radar System and Experimental Parameters In the data processing process, the acquired radar echo data is first processed according to step 1. The echo signal is then subjected to signal processing operations such as range migration correction and range compression to obtain range compressed domain echo data.

[0068] Subsequently, following step 2, target feature constraint information is introduced to describe the sparse features and structural continuity of the imaging target, and statistical modeling of the imaging target features is performed. At the same time, following step 3, environmental interference statistical information is introduced to characterize the statistical properties of complex electromagnetic interference, and multimodal and non-Gaussian interference that may exist in the echo data is modeled.

[0069] Based on this, following step 4, under a unified statistical inference framework, the target feature constraint information and environmental interference statistical information are jointly modeled to construct a joint posterior inference model of multi-source statistical information; and through a decomposition-coordination inference structure, the target feature model and environmental interference model are jointly inferred.

[0070] Finally, following step 5, the joint posterior distribution is solved, and the estimated result of the imaging target features is obtained through the approximate solution of the posterior distribution, and it is output as the final synthetic aperture radar imaging result.

[0071] The imaging results and technical effects are analyzed as follows: Figure 5 This is a schematic diagram of the imaging results obtained by different imaging processing methods in the synthetic aperture radar imaging method based on joint statistical inference described in Embodiment 1 of the present invention. Figure 5 The image shows a comparison of imaging results obtained using different imaging processing methods under the same measured radar data conditions. Figure 5 (a) The imaging results without introducing target feature constraint information and environmental interference statistics; Figure 5 (b) is the imaging result that only incorporates the target's sparse feature constraint information; Figure 5 (c) is the imaging result that only incorporates structural continuity constraint information; Figure 5 (d) is the imaging result obtained by simultaneously introducing target feature constraint information and environmental interference statistical information under a unified joint inference framework according to the method of the present invention.

[0072] Depend on Figure 5 It is evident that without the introduction of statistical modeling constraints, the background interference in the imaging results is quite obvious, and the contrast of the target area is low. When only single target feature constraint information is introduced, although the target sparsity or local structural characteristics are enhanced to some extent, there are still problems such as incomplete structure or insufficient interference suppression.

[0073] In contrast, the method of this invention, through joint modeling and collaborative inference of target feature constraint information and environmental interference statistics, effectively maintains the continuity of the target structure and local detailed features while suppressing background interference from complex electromagnetic environments. As can be seen from the imaging entropy and target to clutter ratio (TCR) shown in the figure, the method of this invention significantly improves both the reduction of imaging uncertainty and the enhancement of target region contrast.

[0074] The above examples illustrate that this invention effectively improves the resolution and structure preservation capability of synthetic aperture radar imaging results under complex electromagnetic environment conditions through collaborative modeling and joint inference of multi-source statistical information. This improves the problem that existing methods struggle to balance imaging accuracy and structural feature preservation under strong interference conditions, providing a more reliable imaging foundation for subsequent target detection and recognition tasks.

[0075] Example 2 Figure 6 This is a schematic diagram of the synthetic aperture radar imaging device according to Embodiment 2 of the present invention. Figure 6 A block diagram of an exemplary apparatus suitable for implementing embodiments of the present invention is shown. Figure 6 The device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention. Figure 6 As shown, this imaging device includes: The preprocessing module 201 is used to perform signal preprocessing on the synthetic aperture radar echo signal to obtain range compressed domain echo data, and to construct a linear observation model between the range compressed domain echo data and the imaging target features based on the range compressed domain echo data; wherein, the signal preprocessing includes range migration correction and range compression processing; The first introduction module 202 is used to introduce target feature constraint information to describe the spatial distribution characteristics of the imaging target, statistically model the sparse characteristics and structural continuity characteristics of the imaging target in the spatial domain, and construct a target feature constraint model to characterize the sparse characteristics and structural continuity of the imaging target. The second introduction module 203 is used to introduce environmental interference statistical information to describe the statistical characteristics of interference in complex electromagnetic environments, to model multi-mode and non-Gaussian interference in range-compressed domain echo data, and to construct an environmental interference model to characterize complex electromagnetic environments. The joint construction module 204 is used to incorporate the target feature constraint information and the environmental interference statistical information into the same joint posterior model under a unified statistical inference framework, and construct a joint posterior distribution model that includes observation data, imaging target features and interference statistical parameters. The solver module 205 is used to reconstruct the joint posterior distribution model through structured decomposition, and to iteratively solve the joint posterior distribution model through numerical inference to obtain the posterior estimation result of the imaging target features. Then, the posterior estimation result is output as the synthetic aperture radar imaging result.

[0076] The synthetic aperture radar imaging device provided in the embodiments of the present invention can execute the synthetic aperture radar imaging method based on joint statistical inference provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method.

[0077] Example 3 Figure 7 This is a schematic diagram of the synthetic aperture radar imaging terminal provided in Embodiment 3 of the present invention; Figure 7 A block diagram is shown that is suitable for implementing embodiments of the present invention. Figure 7 The terminal shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0078] like Figure 7 As shown, terminal 12 is presented in the form of a general-purpose computing device. The components of terminal 12 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and bus 18 connecting different system components (including system memory 28 and processing unit 16).

[0079] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0080] Terminal 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by terminal 12, including volatile and non-volatile media, removable and non-removable media.

[0081] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Terminal 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (… Figure 7 Not shown; usually referred to as a "hard drive"). Although Figure 7Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present invention.

[0082] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of the present invention.

[0083] Terminal 12 can also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), and with one or more devices that enable a user to interact with terminal 12, and / or with any device that enables terminal 12 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 22. Furthermore, terminal 12 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of terminal 12 via bus 18. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with terminal 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0084] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the synthetic aperture radar imaging method based on joint statistical inference provided in the embodiments of the present invention.

[0085] Example 4 Embodiment 4 of the present invention also provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform the synthetic aperture radar imaging method based on joint statistical inference as described in any of the above embodiments.

[0086] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0087] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0088] The program code contained on a computer-readable medium may be transmitted using any suitable medium, including—but not limited to—wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0089] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0090] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A synthetic aperture radar imaging method based on joint statistical inference, characterized in that, include: Synthetic Aperture Radar (SAR) echo signals are preprocessed to obtain range-compressed domain echo data. Based on the range-compressed domain echo data, a linear observation model between the range-compressed domain echo data and the imaging target features is constructed. The signal preprocessing includes range migration correction and range compression processing. By introducing target feature constraint information to describe the spatial distribution characteristics of imaging targets, statistical modeling is performed on the sparse characteristics and structural continuity characteristics of imaging targets in the spatial domain, and a target feature constraint model is constructed to characterize the sparse characteristics and structural continuity of imaging targets. Environmental interference statistics are introduced to describe the statistical characteristics of interference in complex electromagnetic environments. Multimodal and non-Gaussian interference in range-compressed domain echo data are modeled to construct an environmental interference model to characterize complex electromagnetic environments. Under the unified statistical inference framework, the target feature constraint information and the environmental interference statistical information are incorporated into the same joint posterior model to construct a joint posterior distribution model that includes observation data, imaging target features and interference statistical parameters. The joint posterior distribution model is reconstructed by a structured decomposition method, and then iteratively solved by a numerical inference method to obtain the posterior estimation result of the imaging target features. The posterior estimation result is then output as the synthetic aperture radar imaging result.

2. The method according to claim 1, characterized in that: The linear observation model is expressed as follows: in, For range-compressed domain echo data, The target feature to be recovered. For Fourier dictionaries, This is due to additive environmental interference.

3. The method according to claim 1, characterized in that: The target feature constraint information includes constraint terms used to describe the sparsity characteristics of the target space and constraint terms used to describe the structural continuity of the target space.

4. The method according to claim 1, characterized in that, The introduction of environmental interference statistics to describe the statistical characteristics of interference in complex electromagnetic environments, the modeling of multimodal and non-Gaussian interference in range-compressed domain echo data, and the construction of an interference model to characterize complex electromagnetic environments include: The environmental interference statistics are characterized by a mixture distribution model, and a Gaussian mixture model is used to model the interference distribution, constructing an interference model to characterize complex electromagnetic environments. Its probability distribution is expressed as follows: in, This indicates that the mean is 0 and the covariance matrix is... The complex Gaussian distribution, and They represent the first The mixture weighting coefficients and variance of the Gaussian are: , Let be the order of the Gaussian mixture.

5. The method according to claim 1, characterized in that: Under the unified statistical inference framework, the target feature constraint information and the environmental interference statistical information are incorporated into the same joint posterior model to construct a joint posterior distribution model that includes observation data, imaging target features, and interference statistical parameters, including: The target feature constraint information is introduced into the joint posterior model through a prior probability model, and a joint prior form containing sparse constraint terms and structural continuity constraint terms is adopted.

6. The method according to claim 1, characterized in that: The method of incorporating the target feature constraint information and the environmental interference statistical information into the same joint posterior model within a unified statistical inference framework to construct a joint posterior distribution model that includes observation data, imaging target features, and interference statistical parameters further includes: Based on the linear observation model and the environmental disturbance model, a conditional probability model for the observation data under given target characteristics and disturbance statistical parameters is introduced to describe the generation mechanism of the observation data.

7. The method according to claim 1, characterized in that: The numerical inference method adopts a posterior solution method that combines a near-end update mechanism and a random sampling mechanism.

Images