A sea surface background target SAR image simulation method

Abstract

This invention discloses a method for simulating SAR images of targets against a sea surface background. The invention relates to the field of synthetic aperture radar (SAR) signal simulation technology and includes the following steps: constructing a triangular mesh model, dividing the sea surface and identifying target regions, obtaining simulated radar parameters, pre-calculating and storing the complex reflection coefficients of the triangular surface elements of the sea surface using the Fresnel reflection formula, emitting radar beams using the bouncing ray method based on a hierarchical bounding box tree and a collaborative parallel computing architecture, tracking the contribution of the scattered field, calculating the frequency domain RCS data, coherently superimposing the frequency domain RCS data using the sinc function, and generating a SAR simulation image by combining grayscale mapping, and integrating the tail and wake of the target to be identified to obtain the final SAR simulation image containing the sea surface background and the target. This method enables the radar beam tracking process to be performed efficiently in complex scenarios, greatly improving simulation performance.

Description

Technical Field

[0001] This invention relates to the field of synthetic aperture radar signal simulation technology, specifically a method for simulating SAR images of sea surface background targets. Background Technology

[0002] Synthetic Aperture Radar (SAR), with its advantages of all-weather, all-day, long-range, and high-resolution observation, has become a key technological support for maritime target reconnaissance and battlefield situational awareness. However, acquiring SAR images is extremely difficult in combat scenarios involving maritime targets. Therefore, SAR imaging simulation technology has become a core bridge connecting equipment performance and combat applications, and its technical level directly determines the realism of combat simulation and the effectiveness of decision-making. While significant progress has been made in SAR imaging simulation technology for sea surfaces, the two main approaches—echo signal-based and image feature-based—still face core bottlenecks. The former requires RCS calculations for multiple angles and frequencies before performing echo simulations, which is time-consuming and involves massive amounts of data, failing to meet real-time simulation requirements and also unable to obtain images of arbitrary frequencies and angles through interpolation. The latter approximates the scattering center using several point scattering sources, resulting in images with poor realism and representation, making it difficult to support accurate target identification and situation assessment, and hindering the demonstration of SAR reconnaissance effectiveness. Furthermore, existing technologies have limited research on wake detection, restricting the development of related identification technologies and operational applications. Therefore, there is an urgent need for a sea surface SAR image simulation method that can accurately simulate the target, the surrounding sea surface and the wake structure, and meet the requirements of real-time calculation and the realism of the simulated image.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a SAR image simulation method for sea surface background targets to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for simulating SAR images of targets against a sea surface background, comprising the following steps: Construct a triangular mesh model of the monitored sea area, determine the sea surface area and the target area to be identified in the triangular mesh model, divide the sea surface area and the target area to be identified into several sea surface triangular elements and target triangular elements, and add material properties to each triangular element. The simulated radar parameters are obtained, and based on the simulated radar parameters and the material properties of the sea surface triangular element, the complex reflection coefficient of the sea surface triangular element at the radar beam incident angle is pre-calculated and stored using the Fresnel reflection formula. A hierarchical bounding box tree is constructed based on all triangular facets in the triangular mesh model, and a collaborative parallel computing architecture is set up. Under the collaborative parallel computing architecture, radar beams are emitted to the triangular mesh model based on the bouncing ray method, and radar beam tracking is performed in combination with the hierarchical bounding box tree. Based on the complex reflection coefficient of the sea surface and the material properties of the target to be identified, the scattering field contribution of each triangular facet is calculated using the physical optics method. The frequency domain RCS data of each triangular facet is obtained by integrating the field strength based on the scattering field contribution. Based on the frequency domain RCS data of each triangular element, the original SAR simulation image is generated by coherent superposition using the sinc function and combined with grayscale mapping processing. Based on the sailing speed of the target to be identified, the tail wake of the target to be identified is determined. By combining the original SAR simulation image and the tail wake of the target to be identified, a SAR simulation image containing the sea surface background and the target to be identified is obtained.

[0006] Furthermore, the material property identifier is used to describe the material properties of the triangular element, including complex permittivity and conductivity; the triangular mesh model is divided into several sea surface triangular elements and the target triangular element is identified, and a material property identifier is added to each triangular element.

[0007] Furthermore, the complex reflection coefficient of the sea surface triangular element includes the horizontally polarized complex reflection coefficient and the vertically polarized complex reflection coefficient, and the simulated radar parameters include the radar beam incident direction and the radar wavelength. The specific formula used to calculate the horizontal polarization complex reflection coefficient of a triangular surface element on the sea surface is as follows: In the formula, Let be the horizontal polarization complex reflection coefficient of the i-th sea surface triangular element. The refractive index of the radar beam in air. The refractive index of the radar beam in seawater. Let be the angle between the incident direction of the radar beam and the direction of the normal vector of the i-th sea surface triangular element. The angle between the radar beam reflection direction and the direction of the normal vector of the i-th sea surface triangular element is denoted as , where i is the index of the sea surface triangular element. The specific formula used to calculate the vertically polarized complex reflection coefficient is as follows: In the formula, Let be the vertical polarization complex reflection coefficient of the i-th sea surface triangular element.

[0008] Furthermore, the specific method for constructing a hierarchical bounding box tree based on all triangular facets in the triangular mesh model is as follows: Based on the BVH tree structure, a hierarchical bounding box tree is constructed. All sea surface triangular facets and target triangular facets in the triangular mesh model are traversed, and a minimum axis-aligned bounding box is generated for each triangular facet. The minimum axis-aligned bounding box completely encloses a single triangular facet with the minimum volume. A bottom-up aggregation strategy is adopted to group spatially adjacent triangular facets according to the order of bounding box overlap from high to low, and a parent bounding box is constructed for each group of triangular facets. The aggregation is recursively performed until a root node that encloses the entire triangular mesh model is formed, and finally a multi-level bounding box tree structure is formed. The logic behind radar ray beam tracking using a hierarchical bounding box tree is as follows: Tracking priorities are prioritized based on the hierarchical structure of the bounding box tree, decreasing sequentially from the upper-level parent bounding box to the lower-level triangular facets. During radar ray beam tracking, starting from the root node of the hierarchical bounding box tree, the tree structure is illuminated and intersected according to the tracking priority. Specifically, the radar ray beam first intersects with the upper-level parent bounding box of the hierarchical bounding box tree. If the radar ray beam has no intersection with a certain level parent bounding box, then all minimum axis-aligned bounding boxes and triangular facets under that parent bounding box do not need further detection. If an intersection exists, the system continues to traverse the minimum axis-aligned bounding boxes included in the parent bounding box containing that intersection until the intersecting single triangular facet is located.

[0009] Furthermore, the specific design logic of the collaborative parallel computing architecture is as follows: A collaborative parallel computing architecture is constructed based on a CPU+GPU collaborative parallel computing mode. Specifically, it includes a parallelization scheme based on OpenMP, which includes the construction and optimization of hierarchical bounding box trees, the initialization of radar beam emission parameters, the loading and allocation of material property data, and the final aggregation and RCS integration calculation of scattered field data. GPUs are used for core parallel computing tasks. Based on the CUDA architecture, several radar beams are allocated to the GPU's computing cores to achieve parallel execution of radar beam tracking. Each GPU core independently completes the tracking of a radar beam's propagation path, hierarchical intersection detection with the bounding box tree, reflection direction calculation, and phase accumulation to calculate the frequency domain RCS data of each triangular element.

[0010] Furthermore, the specific method used to calculate the frequency domain RCS data of each triangular element is as follows: The intersection points of the radar beam and the triangular element surface are determined using the Möller-Trumbore triangle intersection algorithm. Then, through a hierarchical bounding box tree and a collaborative parallel computing architecture, the triangular elements intersecting the radar beam in different reflection directions are identified. Finally, the scattering field contribution of each triangular element is calculated using physical optics methods. The specific method used is as follows: In the formula, The contribution to the scattered field of the p-th triangular element. Radar wave number, Let be the distance from the p-th triangular element to the radar echo observation point. Let be the unit normal vector at the intersection of the radar beam and the p-th triangular element. The vector of the incident magnetic field strength of the radar. The path factor for reflection to the intersection of the p-th triangular element is determined by the complex reflection coefficient and the number of reflections corresponding to the triangular element. The scattering direction is a unit vector. The imaginary unit, For the p-th triangular element, the effective scattering surface is... Let be the area element on the surface of the triangular element. This represents the position vector at the intersection of the radar beam and the p-th triangular element, where p is the index of all triangular elements; The specific formula used to obtain the frequency domain RCS data of each triangular element based on the contribution of the scattered field through field strength integration is as follows: In the formula, For the RCS data of the p-th triangular element, The intensity of the incident electric field for radar.

[0011] Furthermore, the specific method for generating the original SAR simulation image is as follows: Based on the RCS data of each triangular element, the RCS data of each triangular element is mapped in two dimensions along the slant range-azimuth plane of the triangular mesh model, and the mapped two-dimensional plane is divided into several two-dimensional resolution cells. The RCS values ​​of all triangular elements within the coverage area of ​​each two-dimensional resolution cell are superimposed to determine the RCS value of each two-dimensional resolution cell. Based on the RCS values ​​of each two-dimensional resolution cell, the original SAR simulation image is generated by coherent superposition using the sinc function, combined with grayscale mapping processing and standard SAR imaging processing. The standard SAR imaging processing specifically involves: using the standard range-Doppler algorithm, and through range compression, range migration correction, and azimuth compression methods, to generate the original SAR simulation image containing the target to be identified and the sea surface background.

[0012] Furthermore, the logic for determining the tail wake of the target to be identified is as follows: based on the target's speed, the height of the tail wake is determined, and the formula for calculating the tail wake height is as follows: In the formula, When navigating to identify the target, the coordinates of the sea surface area are: The tail wake is high at the end. For the half-length of the target to be identified, The width of half the boat of the target to be identified. The draft of the target to be identified. The y-axis is the angle between the direction of the wake propagation and the positive x-axis, where the x-axis is a coordinate axis with the positive direction being due east from sea level, and the y-axis is a coordinate axis perpendicular to the x-axis at sea level. y and y' are the x and y coordinates of the sea level region, respectively; Based on the tail wake height of the target being identified during its navigation, the tail wake height is superimposed onto the corresponding part of the sea surface region in the triangular mesh model, and the tail wake of the target being identified during its navigation is mapped onto the corresponding part of the original SAR simulation image to obtain a SAR simulation image containing the sea surface background and the target being identified.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This scheme constructs a hierarchical bounding box tree and sets up a collaborative parallel computing architecture, enabling the radar ray beam tracking process to be performed efficiently in complex scenarios, greatly improving simulation performance. By applying the bouncing ray method, it can more comprehensively simulate the scattering and reflection process of light waves on the sea surface and target surface, making the final generated SAR image richer in detail and able to effectively reflect the shape, material and performance of the target in the sea environment. By combining the physical optics method to calculate the scattering field contribution of each triangular element, more accurate frequency domain RCS data can be obtained, thereby further improving the image resolution and recognition rate, as well as the success rate of target detection and recognition. By performing coherent superposition of frequency domain RCS data using the sinc function and combining it with grayscale mapping to generate the original SAR simulation image, the visual effect of the image is further enhanced. By combining the sailing speed of the target to be identified to determine the tail wake, the dynamic performance of the target on the sea surface can be realistically reproduced, thereby simulating the behavioral characteristics of the target in the real marine environment to a certain extent. This further enhances the practicality and analytical value of SAR images, provides highly accurate simulation images, and provides data support for radar detection and identification research. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 SAR simulation images of sea state 1, 3 and 5 regions; Figure 3 RCS-mapped SAR simulation images for sea state levels 1, 3, and 5; Figure 4 Simulated image of the tail wake height of the target to be identified during navigation. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0016] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0017] Example: Please see Figures 1-4 The present invention provides a technical solution: A method for simulating SAR images of targets against a sea surface background, comprising the following steps: Step 1: Construct a triangular mesh model of the monitored sea area, determine the sea surface area and the target area to be identified in the triangular mesh model, divide the sea surface area and the target area to be identified into several sea surface triangular elements and target triangular elements, and add material properties to each triangular element.

[0018] The material property identifier is used to describe the material properties of the triangular element, including complex permittivity and conductivity. The method used to determine the specific sea surface region and the target region to be identified in the triangular mesh model is as follows: using image processing techniques, such as threshold segmentation and contour detection, the specific sea surface and the target region to be identified in the triangular mesh model are identified and divided to determine the corresponding regions. The triangular mesh model is divided into several sea surface triangular elements and target identification triangular elements. A material attribute identifier is added to each triangular element, describing its material properties. Specific steps include: establishing a triangular mesh model of the monitored sea area in three-dimensional space; designing the triangular elements based on the actual sea area's topography and target features to ensure sufficient resolution; creating a material attribute identifier for each material attribute, containing information such as the different materials and their corresponding complex permittivity and conductivity; and determining the location and material properties of each triangular element by traversing the triangular mesh model. The material properties of each triangular element are specifically determined by the material within that element. For example, the material of the sea surface triangular element is seawater, and the material of the target identification triangular element is the corresponding construction material. This is determined by the material set in the model, and by consulting existing scientific literature and research reports to find complex permittivity and conductivity data for similar materials, or by using material performance databases such as MatWeb and NIST Materials Data to find the electrical property data of the construction material.

[0019] Step 2: Obtain simulated radar parameters. Based on the simulated radar parameters and the material properties of the sea surface triangular element, pre-calculate and store the complex reflection coefficient of the sea surface triangular element at the radar beam incident angle using the Fresnel reflection formula.

[0020] The complex reflection coefficient of the sea surface triangular element includes the horizontal polarization complex reflection coefficient and the vertical polarization complex reflection coefficient, and the simulated radar parameters include the radar beam incident angle and the radar wavelength. The specific formula used to calculate the horizontal polarization complex reflection coefficient of a triangular surface element on the sea surface is as follows: In the formula, Let be the horizontal polarization complex reflection coefficient of the i-th sea surface triangular element. The refractive index of the radar beam in air. The refractive index of the radar beam in seawater. Let be the angle between the incident direction of the radar beam and the direction of the normal vector of the i-th sea surface triangular element. The angle between the radar beam reflection direction and the direction of the normal vector of the i-th sea surface triangular element is denoted as , where i is the index of the sea surface triangular element. The specific formula used to calculate the vertically polarized complex reflection coefficient is as follows: In the formula, Let be the vertical polarization complex reflection coefficient of the i-th sea surface triangular element.

[0021] It should be noted that the polarization complex reflection coefficient is an important parameter describing the reflection characteristics of electromagnetic waves at different medium interfaces. Depending on the polarization of the incident electromagnetic wave, it can be divided into horizontal polarization and vertical polarization. In the case of a triangular surface element on the sea surface, the calculation formula for the complex reflection coefficient is specifically based on the Fresnel reflection formula, expressing the reflection coefficients for horizontal and vertical polarization respectively. In the specific simulation process, the corresponding polarization complex reflection coefficient is selected for calculation based on the polarization of the incident electromagnetic wave.

[0022] The refractive index of the radar beam in air can be obtained by using data from experiments or literature under specific conditions such as temperature, air pressure, and altitude, and is expressed as a complex number. The refractive index of the radar beam in seawater can be obtained by directly measuring the refractive index of the seawater sample using a refractometer; or by consulting relevant literature in marine science and hydrology to find refractive index data under different temperature and salinity conditions, which is also expressed as a complex number.

[0023] The specific method for obtaining the angle between the radar beam incident angle direction and the normal vector direction of the i-th sea surface triangular element is as follows: it is calculated by the radar emission direction and the sea surface normal vector. The sea surface triangular element normal vector can be obtained by the geometric properties of the triangular element on the sea surface, specifically by the position of the three vertices of the triangular element. Using measuring equipment such as a radar system, the emission angle of the radar wave can be obtained by the sensor. The specific method for obtaining the angle between the radar beam reflection direction and the normal vector direction of the i-th sea surface triangular element is as follows: The reflection angle is the angle between the radar wave reflection direction and the sea surface normal vector. According to the law of signal reflection, the reflection angle is directly calculated from the incident angle.

[0024] This method can effectively solve the problem of material homogenization with a small increase in computational load, achieving a balance between real-time computation and the realism of simulation images.

[0025] Step 3: Construct a hierarchical bounding box tree based on all triangular elements in the triangular mesh model, and set up a collaborative parallel computing architecture. Under the collaborative parallel computing architecture, a radar beam is emitted to the triangular mesh model based on the bouncing ray method. The radar beam is tracked in combination with the hierarchical bounding box tree. Based on the complex reflection coefficient of the sea surface and the material properties of the target to be identified, the scattering field contribution of each triangular element is calculated using the physical optics method. The frequency domain RCS data of each triangular element is obtained by integrating the field strength based on the scattering field contribution.

[0026] The specific method for constructing a hierarchical bounding box tree based on all triangular facets in the triangular mesh model is as follows: Based on the BVH tree structure, a hierarchical bounding box tree is constructed. All sea surface triangular facets and target triangular facets in the triangular mesh model are traversed, and a minimum axis-aligned bounding box is generated for each triangular facet. The minimum axis-aligned bounding box completely encloses a single triangular facet with the smallest volume. In this embodiment, the AABB bounding box is used as the minimum axis-aligned bounding box. A bottom-up aggregation strategy is adopted to group spatially adjacent triangular facets according to the order of bounding box overlap from high to low, and a parent bounding box is constructed for each group of triangular facets. The aggregation is recursively performed until a root node that encloses the entire triangular mesh model is formed, and finally a multi-level bounding box tree structure is formed. The logic behind radar ray beam tracking using a hierarchical bounding box tree is as follows: Tracking priorities are prioritized based on the hierarchical structure of the bounding box tree, decreasing sequentially from the upper-level parent bounding box to the lower-level triangular facets. During radar ray beam tracking, starting from the root node of the hierarchical bounding box tree, the tree structure is illuminated and intersected according to the tracking priority. Specifically, the radar ray beam first intersects with the upper-level parent bounding box of the hierarchical bounding box tree. If the radar ray beam has no intersection with a certain level parent bounding box, then all minimum axis-aligned bounding boxes and triangular facets under that parent bounding box do not need further detection. If an intersection exists, the system continues to traverse the minimum axis-aligned bounding boxes included in the parent bounding box containing that intersection until the intersecting single triangular facet is located.

[0027] BVH trees are a hierarchical data structure based on spatial partitioning. Their core objective is to reduce invalid intersection detection of rays and surface elements through a logic that progresses from coarse screening to fine inspection.

[0028] The specific design logic of the collaborative parallel computing architecture is as follows: A collaborative parallel computing architecture is constructed based on a CPU+GPU collaborative parallel computing mode. Specifically, it includes a parallelization scheme based on OpenMP, which includes the construction and optimization of hierarchical bounding box trees, initialization of radar beam emission parameters, loading and allocation of material property data, and final aggregation and RCS integration calculation of scattered field data. GPUs are used for core parallel computing tasks. Based on the CUDA architecture, several radar beams are allocated to the GPU's computing cores to achieve parallel execution of radar beam tracking. Each GPU core independently completes the tracking of a radar beam's propagation path, hierarchical intersection detection with the bounding box tree, reflection direction calculation, and phase accumulation to calculate the frequency domain RCS data of each triangular element.

[0029] Because the triangular mesh model formed by the target to be identified and the sea surface background is an ultra-large target, the number of facets in its 3D model often reaches hundreds of thousands to millions. If the bouncing ray method is directly used to detect the intersection of all rays with facets, the computational load will increase exponentially with the number of facets, resulting in a severely time-consuming ray tracing process. This cannot meet the real-time requirements of frequent switching of observation angles and dynamic adjustment of imaging parameters in combat simulations. To address this, this paper constructs a collaborative parallel computing architecture combining BVH tree space acceleration and CPU+GPU, breaking through the efficiency bottleneck from two dimensions: reducing unnecessary computation and improving hardware computing power utilization. Multi-core synchronous operation significantly improves the processing speed of ray reflection iteration. Meanwhile, the high bandwidth memory of the GPU supports the rapid reading of massive triangular face data and BVH tree node data, avoiding data transmission bottlenecks. The spatial acceleration of the BVH tree and the parallel computing of CPU+GPU complement each other. The BVH tree reduces the number of face elements detected by a single ray, reducing the workload of each round of calculation. The parallel architecture of CPU+GPU performs multi-core parallel processing to complete the calculation task of massive rays at the same time. The combination of the two breaks through the efficiency limitations of the traditional computing mode from the two aspects of single-ray efficiency improvement and multi-ray parallel processing, providing core technical support for the real-time performance of simulation.

[0030] The bouncing ray method is a highly efficient numerical algorithm in the field of high-frequency electromagnetic computation. This method integrates geometric optics to handle the ray propagation and reflection process, and combines physical optics to calculate the final field strength. It can efficiently solve the multiple scattering problem of complex targets such as aircraft and ships. This method has the significant advantages of high computational efficiency and low memory usage, so it is widely used in scenarios where the wavelength is much smaller than the target size to solve the RCS of electrically large targets.

[0031] The specific method used to calculate the frequency domain RCS data of each triangular element is as follows: The intersection points of the radar beam and the triangular element surface are determined using the Möller-Trumbore triangle intersection algorithm. Then, through a hierarchical bounding box tree and a collaborative parallel computing architecture, the triangular elements intersecting the radar beam in different reflection directions are identified. Finally, the scattering field contribution of each triangular element is calculated using physical optics methods. The specific method used is as follows: In the formula, The contribution to the scattered field of the p-th triangular element. This is the radar wavenumber, specifically calculated based on the radar wavelength. Let be the distance from the p-th triangular element to the radar echo observation point. Let be the unit normal vector at the intersection of the radar beam and the p-th triangular element. The vector of the incident magnetic field strength of the radar. The path factor for reflection to the intersection of the p-th triangular element is determined by the complex reflection coefficient and the number of reflections corresponding to the triangular element. The scattering direction is a unit vector. The imaginary unit, For the p-th triangular element, the effective scattering surface is... Let be an area element on the surface of a triangular facet. During integration, the area element can be considered as a tiny region. Generally, uniform sampling is performed on the facet to determine the area element. This represents the position vector at the intersection of the radar beam and the p-th triangular element, where p is the index of all triangular elements; It should be noted that this formula is based on Physical Optics (PO) theory, assuming that the size of the scattering object is much larger than the wavelength. It uses the principle of coherent superposition of electromagnetic waves to calculate the scattered field. The Physical Optics method is suitable for larger scattering objects such as triangular facets because, in this case, the incident wave can be considered a plane wave, neglecting higher-order scattering effects; a unit normal vector is used. The cross-product operation demonstrates the relationship between the surface scattering characteristics and the direction of the incident wave. The introduction of this feature allows the formula to reflect different reflection times and characteristics, enhancing the description of scattering from complex targets. This is achieved through the effective scattering surface... Integrating the data on the surface takes into account the contribution of each point on the scattering surface to the overall scattered field. This approach enhances the accuracy of the model and adapts to surface elements of different sizes and shapes. and The introduction of this concept takes into account the phase change of the wave, ensuring an accurate description of the phase during propagation and scattering.

[0032] Where the incident magnetic field strength vector This value can be obtained from the design parameters of the radar system, which are usually provided in the radar system specifications. In simulations, it can be calculated based on information such as the incident direction and intensity of the radar wave; the scattering direction unit vector. It can be calculated from the radar emission direction and the reflection characteristics of the target, and determined by the position of the observation point and the position of the triangular element; The path factor at the intersection of the p-th triangular element is determined. The specific method is as follows: based on the path factor of the previous reflection, combined with the complex reflection coefficient at the current reflection point and the path difference between the reflection points, the path factor is updated according to wave theory. When electromagnetic waves propagate in a medium, they will experience phenomena such as reflection, refraction and interference. In the process of calculating the path factor of multiple reflections, the interaction between the previous and subsequent waves must be taken into account for updating. The path difference between reflection points specifically refers to the path difference between the previous reflection point and the current reflection point. The specific method to obtain it is as follows: determine the three-dimensional coordinates of the previous reflection point, the current reflection point, and the observation point; calculate the distances from the previous reflection point and the current reflection point to the observation point respectively; and subtract the distance from the previous reflection point to the observation point from the distance from the current reflection point to the observation point to obtain the path difference.

[0033] The specific formula used to obtain the frequency domain RCS data of each triangular element based on the contribution of the scattered field through field strength integration is as follows: In the formula, For the RCS data of the p-th triangular element, The intensity of the incident electric field for radar.

[0034] It should be noted that RCS is an effective indicator for measuring a target's reflectivity, commonly used in radar target identification; the relationship between electric field strength and power is expressed using a square relationship. The ratio of scattered intensity to incident intensity reflects the ability of a surface element to scatter radar signals. This represents the power scattered from the surface element. In free space, the intensity of electromagnetic waves decreases with the square of the distance, therefore, it is necessary to introduce this factor when calculating the RCS. This factor takes into account the propagation loss from the surface element to the receiver. Specifically, the intensity of the scattered signal is inversely proportional to the square of the distance, ensuring that the physical properties of the RCS are consistent with actual observations.

[0035] Step 4: Based on the frequency domain RCS data of each triangular element, coherent superposition is performed using the sinc function, combined with grayscale mapping processing, to generate the original SAR simulation image. Based on the sailing speed of the target to be identified, the tail wake of the target to be identified is determined. The original SAR simulation image and the tail wake of the target to be identified are combined to obtain a SAR simulation image containing the sea background and the target to be identified.

[0036] The specific method for generating the original SAR simulation image is as follows: Based on the RCS data of each triangular element, the RCS data of each triangular element is mapped in two dimensions along the slant range-azimuth plane of the triangular mesh model, and the mapped two-dimensional plane is divided into several two-dimensional resolution cells. The specific steps include: determining the operating frequency, wavelength and observation geometry of the radar system, constructing the slant range and azimuth coordinate system, mapping the RCS value of each triangular element to the corresponding position in the slant range-azimuth coordinate system, calculating the position of its center point in the slant range-azimuth coordinate system for each triangular element, assigning the RCS value of the triangular element to its corresponding slant range-azimuth coordinate system position, setting the size of the two-dimensional resolution cell according to the resolution of the SAR system, and dividing the mapped two-dimensional plane into several two-dimensional resolution cells. The RCS values ​​of all triangular facets within the coverage area of ​​each two-dimensional resolution cell are superimposed to determine the RCS value of each two-dimensional resolution cell. Based on the RCS values ​​of each two-dimensional resolution cell, coherent superposition is performed using the sinc function. The sinc function has wide applications in signal processing, especially in the reconstruction of discrete signals. For SAR image generation, the sinc function is used for coherent superposition to combine the RCS value of each two-dimensional resolution cell with the influence of adjacent cells. In the coherent superposition process, the core role of the sinc function is to smooth the RCS values, so that the mutual influence between different resolution cells is fully considered. For the RCS value of each two-dimensional resolution cell, a weighted superposition is performed using the sinc function based on the RCS values ​​of its adjacent cells. The sinc function is a conventional existing technology, so its specific form will not be elaborated here. Combined with grayscale mapping processing and standard SAR imaging processing, the original SAR simulation image is generated. The grayscale mapping processing specifically includes: converting each RCS value processed by the sinc function into a grayscale value, and using a normalization method to map the RCS value to a grayscale range of 0 to 255.

[0037] The standard SAR imaging processing specifically involves: employing the standard range-Doppler algorithm, and using range compression, range migration correction, and azimuth compression methods to generate a raw SAR simulation image containing the target to be identified and the sea surface background. Range compression primarily focuses the received signal in the range dimension, making the target's echo signal clearer in terms of range. Specific steps include echo signal acquisition, matched filtering, and signal focusing. Range migration correction addresses signal delay variations caused by platform motion, ensuring that signals emitted from different locations are processed under the same conditions. Azimuth compression processes the signal in the azimuth dimension to improve the azimuth resolution of the image. The signals processed by range and azimuth compression are combined to generate the final SAR simulation image, where each pixel value corresponds to the scattering intensity at a specific geographical location. Standard SAR imaging processing is a conventional existing technology and will not be elaborated upon here; please refer to [link to relevant documentation] for details. Figure 2 It is used to display SAR simulation images of the sea surface area to be detected under different sea state levels. Specifically, it displays SAR simulation images of the sea surface area to be detected under sea state levels 1, 3 and 5. Figure 3 The image displayed is a SAR simulation image showing the RCS of the corresponding triangular facets at the corresponding locations under sea state levels 1, 3, and 5. The sea state level is used to reflect the sea surface wave conditions. As the sea state level increases, the sea surface waves are larger and the wave height is higher.

[0038] The logic for determining the tail wake of a target to be identified is as follows: based on the target's speed, the height of the tail wake is determined, and the formula for calculating the tail wake height is as follows: In the formula, When navigating to identify the target, the coordinates of the sea surface area are: The tail wake is high at the end. For the half-length of the target to be identified, The width of half the boat of the target to be identified. The draft of the target to be identified. The y-axis is the angle between the direction of the wake propagation and the positive x-axis, where the x-axis is a coordinate axis with the positive direction being due east from sea level, and the y-axis is a coordinate axis perpendicular to the x-axis at sea level. y and y' are the x and y coordinates of the sea level region, respectively; Based on this, and further considering the effects of water viscosity and ship length, the integral is decomposed into two parts: the bow and the stern. The height of the superimposed wake wave is represented as follows: In the formula, Let be the viscosity coefficient of water. For the ship in position The first basis function value of the wave height at that location. For the ship in position The second basis function value of the wave height at that location. For the ship in position The first basis function value of the wave height at that location; For the ship in position The wave height is defined by the second basis function, where the second basis function characterizes the wave height at the bow and the first basis function characterizes the wave height at the stern. Adjust the x-axis variable for the sea level region; in For and The specific expression is: Based on the tail wake height of the target being identified during its navigation, the tail wake height is superimposed onto the corresponding part of the sea surface region in the triangular mesh model, and the tail wake of the target being identified during its navigation is mapped onto the corresponding part of the original SAR simulation image to obtain a SAR simulation image containing the sea surface background and the target being identified.

[0039] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0040] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0041] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0042] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for simulating SAR images of sea surface background targets, characterized in that, The specific steps include: Construct a triangular mesh model of the monitored sea area, determine the sea surface area and the target area to be identified in the triangular mesh model, divide the sea surface area and the target area to be identified into several sea surface triangular elements and target triangular elements, and add material properties to each triangular element. Obtain simulated radar parameters, and based on the simulated radar parameters and the material properties of the sea surface triangular element, pre-calculate and store the complex reflection coefficient of the sea surface triangular element under the radar beam using the Fresnel reflection formula; A hierarchical bounding box tree is constructed based on all triangular elements in the triangular mesh model, and a collaborative parallel computing architecture is set up. Under the collaborative parallel computing architecture, radar beams are emitted to the triangular mesh model based on the bouncing ray method. The radar beams are tracked in combination with the hierarchical bounding box tree. Based on the complex reflection coefficient of the sea surface and the material properties of the target to be identified, the physical optics method is used to calculate the scattering field contribution of each triangular element. The frequency domain RCS data of each triangular element is obtained by integrating the field strength based on the scattering field contribution. Based on the frequency domain RCS data of each triangular element, the original SAR simulation image is generated through coherent superposition and grayscale mapping using the sinc function. Based on the sailing speed of the target to be identified, the tail wake of the target to be identified is determined. By combining the original SAR simulation image and the tail wake of the target to be identified, a SAR simulation image containing the sea background and the target to be identified is obtained.

2. The SAR image simulation method for sea surface background targets according to claim 1, characterized in that: The material property identifier is used to describe the material properties of the triangular elements, including complex permittivity and conductivity. The triangular mesh model is divided into several sea surface triangular elements and the target triangular elements, and a material property identifier is added to each triangular element.

3. The SAR image simulation method for sea surface background targets according to claim 2, characterized in that: The complex reflection coefficient of the sea surface triangular element includes the horizontal polarization complex reflection coefficient and the vertical polarization complex reflection coefficient, and the simulated radar parameters include the radar beam incident direction and radar wavelength. The specific formula used to calculate the horizontal polarization complex reflection coefficient of a triangular surface element on the sea surface is as follows: In the formula, Let be the horizontal polarization complex reflection coefficient of the i-th sea surface triangular element. The refractive index of the radar beam in air. The refractive index of the radar beam in seawater. Let be the angle between the incident direction of the radar beam and the direction of the normal vector of the i-th sea surface triangular element. The angle between the radar beam reflection direction and the direction of the normal vector of the i-th sea surface triangular element is denoted as , where i is the index of the sea surface triangular element. The specific formula used to calculate the vertically polarized complex reflection coefficient is as follows: In the formula, Let be the vertical polarization complex reflection coefficient of the i-th sea surface triangular element.

4. The SAR image simulation method for sea surface background targets according to claim 2, characterized in that: The specific method for constructing a hierarchical bounding box tree based on all triangular elements in the triangular mesh model is as follows: Based on the BVH tree structure, construct a hierarchical bounding box tree, traverse all sea surface triangular elements and identify target triangular elements in the triangular mesh model, generate a minimum axis-aligned bounding box for each triangular element, and the minimum axis-aligned bounding box completely wraps a single triangular element with the minimum volume. A bottom-up aggregation strategy is adopted to group all spatially adjacent triangular facets according to the order of bounding box overlap from high to low, and to build a parent bounding box for each group of triangular facets. The aggregation is recursively performed until a root node that encloses the entire triangular mesh model is formed, and finally a multi-level bounding box tree structure is formed. The logic behind radar ray beam tracking using a hierarchical bounding box tree is as follows: Tracking priorities are prioritized based on the hierarchical structure of the bounding box tree, decreasing sequentially from the upper-level parent bounding box to the lower-level triangular facets. During radar ray beam tracking, starting from the root node of the hierarchical bounding box tree, the tree structure is illuminated and intersected according to the tracking priority. Specifically, the radar ray beam first intersects with the upper-level parent bounding box of the hierarchical bounding box tree. If the radar ray beam has no intersection with a certain level parent bounding box, then all minimum axis-aligned bounding boxes and triangular facets under that parent bounding box do not need further detection. If an intersection exists, the system continues to traverse the minimum axis-aligned bounding boxes included in the parent bounding box containing that intersection until the intersecting single triangular facet is located.

5. The SAR image simulation method for sea surface background targets according to claim 4, characterized in that: The specific design logic of the collaborative parallel computing architecture is as follows: A collaborative parallel computing architecture is constructed based on a CPU+GPU collaborative parallel computing mode. Specifically, it includes a parallelization scheme based on OpenMP, which includes the construction and optimization of hierarchical bounding box trees, initialization of radar beam emission parameters, loading and allocation of material property data, and final aggregation and RCS integration calculation of scattered field data. GPUs are used for core parallel computing tasks. Based on the CUDA architecture, several radar beams are allocated to the GPU's computing cores to achieve parallel execution of radar beam tracking. Each GPU core independently completes the tracking of a radar beam's propagation path, hierarchical intersection detection with the bounding box tree, reflection direction calculation, and phase accumulation to calculate the frequency domain RCS data of each triangular element.

6. The SAR image simulation method for sea surface background targets according to claim 5, characterized in that: The specific method used to calculate the frequency domain RCS data of each triangular element is as follows: The intersection points of the radar beam and the triangular element surface are determined using the Möller-Trumbore triangle intersection algorithm. Then, through a hierarchical bounding box tree and a collaborative parallel computing architecture, the triangular elements intersecting the radar beam in different reflection directions are identified. Finally, the scattering field contribution of each triangular element is calculated using physical optics methods. The specific method used is as follows: In the formula, The contribution to the scattered field of the p-th triangular element. Radar wave number, Let be the distance from the p-th triangular element to the radar echo observation point. Let be the unit normal vector at the intersection of the radar beam and the p-th triangular element. The vector of the incident magnetic field strength of the radar. The path factor for reflection to the intersection of the p-th triangular element is determined by the complex reflection coefficient and the number of reflections corresponding to the triangular element. The scattering direction is a unit vector. The imaginary unit, For the p-th triangular element, the effective scattering surface is Let be the area element on the surface of the triangular face element. This represents the position vector at the intersection of the radar beam and the p-th triangular element, where p is the index of all triangular elements; The specific formula used to obtain the frequency domain RCS data of each triangular element based on the contribution of the scattered field through field strength integration is as follows: In the formula, Here is the RCS data for the p-th triangular element. The incident electric field strength of the radar is given.

7. The SAR image simulation method for sea surface background targets according to claim 6, characterized in that: The specific method for generating the original SAR simulation image is as follows: Based on the RCS data of each triangular element, the RCS data of each triangular element is mapped in two dimensions along the slant range-azimuth plane of the triangular mesh model. The mapped two-dimensional plane is divided into several two-dimensional resolution cells. The RCS values ​​of all triangular elements within the coverage area of ​​each two-dimensional resolution cell are superimposed to determine the RCS value of each two-dimensional resolution cell. Based on the RCS values ​​of each two-dimensional resolution cell, the original SAR simulation image is generated by coherent superposition using the sinc function, combined with grayscale mapping processing and standard SAR imaging processing. The standard SAR imaging processing specifically involves using the standard range-Doppler algorithm, and generating the original SAR simulation image containing the target to be identified and the sea surface background through range compression, range migration correction, and azimuth compression methods.

8. The SAR image simulation method for sea surface background targets according to claim 7, characterized in that: The logic for determining the tail wake of a target to be identified is as follows: Based on the target's speed, the height of the tail wake is determined, and the formula for calculating the tail wake height is as follows: In the formula, When navigating to identify the target, the coordinates of the sea surface area are: The tail wake is high at the end. For the half-length of the target to be identified, The width of half the boat of the target to be identified. The draft of the target to be identified. The y-axis is the angle between the direction of the wake propagation and the positive x-axis, where the x-axis is a coordinate axis with the positive direction being due east from sea level, and the y-axis is a coordinate axis perpendicular to the x-axis at sea level. y and y' are the x and y coordinates of the sea level region, respectively; Based on the tail wake height of the target being identified during its navigation, the tail wake height is superimposed onto the corresponding part of the sea surface region in the triangular mesh model, and the tail wake of the target being identified during its navigation is mapped onto the corresponding part of the original SAR simulation image to obtain a SAR simulation image containing the sea surface background and the target being identified.

Images