MIMO millimeter wave radar near field sar imaging method based on scene block division

Abstract

The application discloses a MIMO millimeter wave radar near-field SAR imaging method based on scene block division, comprising the following steps: collecting original echo data; constructing a MIMO radar virtual array equivalent model according to the original echo data; determining the spatial coordinates of the virtual antenna phase center according to the MIMO radar virtual array equivalent model; establishing a multi-static to single-static data conversion model according to the spatial coordinates, and equivalently converting the real array element echo intermediate frequency signal into a virtual array intermediate frequency signal through distance difference and phase difference compensation; and according to the virtual array intermediate frequency signal, dividing the imaging area into multiple sub-blocks according to the focusing constraint condition, performing local imaging on each sub-block, and then obtaining a complete image through spatial coherent superposition. The application significantly improves the focusing effect of scene edge targets, and realizes high-precision, low-distortion and high-efficiency near-field three-dimensional imaging.

Description

Technical Field

[0001] This invention belongs to the field of millimeter-wave radar imaging technology, and particularly relates to a MIMO millimeter-wave radar near-field SAR imaging method based on scene block division. Background Technology

[0002] With the continuous development of semiconductor and radio frequency technologies, the performance of millimeter-wave radio frequency chips has become increasingly mature, leading to the widespread application of millimeter-wave devices. Millimeter-wave radar, due to its high penetration, non-destructive nature to biological tissues, and non-ionizing radiation characteristics, plays a crucial role in non-destructive testing, security scanning, and medical examinations. Near-field imaging, as one of the core applications of millimeter-wave radar, enables precise reconstruction of target details at close range and is widely used in detecting internal defects in materials and capturing the outlines of concealed objects.

[0003] In millimeter-wave imaging systems, large-aperture structures are typically required to achieve high-resolution image reconstruction. Traditional near-field imaging systems usually use single-input single-output (SISO) radar to perform two-dimensional mechanical scanning of the target to acquire two-dimensional or three-dimensional images. The problem with this design is its low data acquisition efficiency, making it difficult to meet the real-time requirements of imaging scenarios. Therefore, multi-input multiple-output (MIMO) multi-station array configurations have gradually been applied to data acquisition to further improve imaging system performance. As the array size increases, the efficiency of the data acquisition process can be gradually improved, and the mechanical errors can be gradually reduced. However, this approach introduces a problem: the echo preprocessing workflow and the accompanying imaging algorithms become increasingly complex. In contrast, using commercial millimeter-wave radars with small to medium-sized arrays to achieve synthetic aperture radar (SAP) provides a better balance between acquisition efficiency and processing complexity.

[0004] In traditional imaging algorithms, the Back Projection Algorithm (BPA) can achieve imaging by accumulating echo energy point by point and is highly adaptable, capable of imaging echoes from arbitrary antenna arrangements. However, when dealing with large scenes or high-resolution imaging requirements, it needs to traverse and accumulate energy point by point for each pixel, causing the computational load to increase exponentially with the size of the imaging scene, making it difficult to meet the needs of real-time imaging scenarios. In contrast, time-domain imaging algorithms based on Non-Uniform Fast Fourier Transform (NUFFT) are more suitable for the technical requirements of near-field millimeter-wave radar imaging. This algorithm maps the time-domain echo signal to the wavenumber domain through Fourier transform and stolt interpolation of the wavenumber domain signal to simplify the calculation, reducing the computational complexity from that of BPA. Down to ( It represents the number of pixels in one dimension of the imaging scene, greatly improving data processing efficiency and making it more suitable for application scenarios with high requirements for both imaging quality and real-time performance, such as near-field cloaked target detection.

[0005] It is worth noting that efficient imaging algorithms are typically only applicable to ideal uniformly sampled data. However, the raw data acquired by actual MIMO systems is in a multi-static mode. Therefore, for MIMO millimeter-wave imaging systems, the imaging process requires first performing a multi-static to single-static equivalent of the raw echo data, simulating monostatic radar transmission and reception through a virtual array phase center, thus enabling the application of imaging algorithms under uniform sampling conditions. In the near-field, the spherical wave propagation characteristics of electromagnetic waves break the assumption of parallel waves in the far-field. Therefore, when the target is located at the edge of the imaging scene, using the scene center as a reference for multi-static to single-static compensation of the echo signal is easily affected by range migration errors, resulting in target defocusing and distortion, severely limiting the application of millimeter-wave radar in target detection. Summary of the Invention

[0006] To address the aforementioned technical issues, this invention proposes a MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning. This method, through the coordinated processing of virtual array equivalence, multi-static to single-static conversion, phase error compensation, and scene block imaging, significantly improves the focusing effect of scene edge targets while suppressing sidelobe artifacts, achieving high-precision, low-distortion, and high-efficiency near-field three-dimensional imaging.

[0007] To achieve the above objectives, this invention provides a MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning, comprising: Acquire raw echo data; Based on the original echo data, construct an equivalent model of the MIMO radar virtual array; Based on the equivalent model of the MIMO radar virtual array, determine the spatial coordinates of the virtual antenna phase center; Based on the spatial coordinates, a multi-static to single-static data conversion model is established, and the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal through distance difference and phase difference compensation. Based on the intermediate frequency signal of the virtual array, the imaging area is divided into multiple sub-blocks according to the focusing constraints. After local imaging of each sub-block, a complete image is obtained by spatial coherent superposition.

[0008] Optionally, acquiring raw echo data includes: Based on the requirements of near-field imaging scenarios, a near-field MIMO millimeter-wave radar data acquisition platform is built by configuring a millimeter-wave radar transceiver module, a data acquisition card, and a motion control platform. Based on the hardware architecture of the data acquisition platform, the radar signal acquisition process and the platform movement process are synchronously controlled; Based on the acquisition timing after synchronization control, the raw echo data is acquired point by point through two-dimensional mechanical scanning and multi-antenna time-division multiplexing transceiver.

[0009] Optionally, determining the spatial coordinates of the virtual antenna phase center includes: ; ; ; in, For the first root transmitting antenna and the first The position of the virtual antenna equivalent to the root receiving antenna. and These represent the coordinates of the virtual antenna in the horizontal and vertical directions, respectively. and They represent the first The horizontal and vertical coordinates of the transmitting antenna, and They represent the first The horizontal and vertical coordinates of the transmitting antenna.

[0010] Optionally, based on the spatial coordinates, a multi-static to single-static data conversion model is established, and the real array element echo intermediate frequency signal is equivalent to a virtual array intermediate frequency signal through distance difference and phase difference compensation, including: The actual transmission and reception distance of the radar signal is obtained by comparing the spatial coordinates of the scattering point with the spatial coordinates of the actual transmitting and receiving antennas. The transmit / receive distance at the virtual antenna is obtained based on the spatial coordinates of the phase center of the virtual antenna and the spatial coordinates of the scattering point. Based on the actual transmit / receive distance and the transmit / receive distance at the virtual antenna, the difference in transmit / receive distance between the real array element and the virtual array element is derived. Based on the transmit / receive distance difference, the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal.

[0011] Optionally, based on the transmit / receive distance difference, converting the real array element echo intermediate frequency signal into a virtual array intermediate frequency signal includes: Based on the transmit / receive distance difference and the free space wavenumber, calculate the phase difference between the intermediate frequency signals received by the virtual array element and the real array element; Based on the phase difference, phase compensation is performed on the intermediate frequency signal of the real array element echo to obtain an intermediate frequency signal equivalent to that transmitted and received by a single station at the virtual array.

[0012] Optionally, based on the intermediate frequency signal of the virtual array, dividing the imaging region into multiple sub-blocks according to focusing constraints includes: Based on the intermediate frequency signal of the virtual array, and combined with the scene center coordinates of the imaging area and the virtual antenna position, the distance difference between the scene center point and the edge point to the virtual antenna is calculated. Based on the distance difference and the wavelength of the radar transmitted signal, determine the phase error caused by the reference point phase compensation; Based on a preset phase error threshold, focusing constraints are established to limit the maximum spatial scale of the sub-blocks.

[0013] Optionally, based on a preset phase error threshold, focusing constraints are established to limit the maximum spatial scale of the sub-blocks, including: Based on the phase error threshold and radar wavelength, a constraint relationship is established between the sub-block size, imaging distance, and operating wavelength; Based on the constraints, calculate the maximum allowable dimensions of the sub-block in the horizontal and vertical directions; Based on the maximum allowable size, the imaging region is divided into multiple non-overlapping sub-blocks.

[0014] Optionally, obtaining a complete image by locally imaging each sub-block and then spatially coherently superimposing it includes: Establish the wavenumber domain integral relationship between the reflectivity of the scattering point and the echo signal; Based on the wavenumber domain integral relationship, a two-dimensional fast Fourier transform is performed on the echo signal to map it to the wavenumber domain; Based on the non-uniform distribution characteristics of wavenumber components in the range direction, Stolt interpolation is performed on the wavenumber domain signal to correct phase error; The complete image is obtained based on the corrected relative error.

[0015] Optionally, based on the corrected relative error, the complete image includes: Based on the interpolated wavenumber domain signal, the range-oriented spectral conversion is completed through non-uniform fast Fourier transform; Based on the spectral conversion results in the range direction, a two-dimensional inverse fast Fourier transform is performed to reconstruct the reflectivity distribution of the scattering points in the sub-block region; Based on the reflectivity distribution and spatial location of the scattering points of each sub-block, the reflectivity distribution of the complete imaging region is obtained through spatial coherent superposition.

[0016] Compared with the prior art, the present invention has the following advantages and technical effects: 1. This invention uses MIMO virtual array equivalent and multi-static to single-static data conversion to accurately compensate for the distance and phase difference between real array elements and virtual array elements, eliminates the phase error caused by multi-station acquisition from the source, significantly suppresses imaging sidelobes and artifacts, and makes the imaging of the target center area more focused and the outline clearer.

[0017] 2. This invention proposes a scene segmentation imaging strategy based on phase focusing constraints. Phase compensation is performed using local sub-blocks as reference centers, which effectively overcomes the distance migration error caused by near-field spherical wave propagation. It completely solves the problems of defocusing, deformation, and positional shift of scene edge targets in traditional global imaging, and significantly improves the geometric fidelity and size characterization accuracy of targets.

[0018] 3. This invention adopts a NUFFT-based temporal fast imaging algorithm and combines it with block processing to further reduce computational complexity. While maintaining high-resolution imaging performance, it significantly improves data processing efficiency. Compared with the traditional back projection algorithm, the amount of computation is greatly reduced, making it more suitable for near-field real-time imaging and engineering deployment needs.

[0019] 4. This invention has stronger robustness in low signal-to-noise ratio environments, and the imaging results have lower image entropy and higher contrast. It is less affected by noise interference and can stably output high-quality images under harsh conditions such as complex backgrounds and penetrating imaging, making it more widely applicable.

[0020] 5. This invention can be directly applied to practical scenarios such as concealed target detection, non-destructive testing of materials, and security inspection. It can achieve clear imaging and accurate identification of metallic targets and defect structures inside non-metallic media. It is highly practical and has broad application prospects, providing an efficient and feasible technical solution for near-field MIMO millimeter-wave radar imaging systems. Attached Figure Description

[0021] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a connection architecture diagram of the imaging data acquisition system according to an embodiment of the present invention; Figure 2 This is an equivalent schematic diagram of the virtual antenna in an embodiment of the present invention; Figure 3 This is a schematic diagram of a MIMO imaging system according to an embodiment of the present invention; Figure 4 This is a schematic diagram of block imaging according to an embodiment of the present invention; Figure 5 This is a flowchart of the MIMO millimeter-wave radar near-field SAR imaging method based on scene block division according to an embodiment of the present invention; Figure 6 This is an optical image of the target being tested according to an embodiment of the present invention. Detailed Implementation

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0024] This embodiment proposes a MIMO millimeter-wave radar near-field SAR imaging method based on scene block division, such as... Figure 5 As shown, the specific steps include: Acquire raw echo data; Based on the original echo data, construct an equivalent model of the MIMO radar virtual array; Based on the equivalent model of the MIMO radar virtual array, determine the spatial coordinates of the virtual antenna phase center; Based on the spatial coordinates, a multi-static to single-static data conversion model is established, and the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal through distance difference and phase difference compensation. Based on the intermediate frequency signal of the virtual array, the imaging area is divided into multiple sub-blocks according to the focusing constraints. After local imaging of each sub-block, a complete image is obtained by spatial coherent superposition.

[0025] Specifically, step 1: Build a near-field MIMO millimeter-wave radar data acquisition platform and complete the acquisition of raw echo data through two-dimensional mechanical scanning and multi-antenna transceiver. Step 2: Construct an equivalent model of the MIMO radar virtual array, use the midpoint approximation to determine the position of the virtual antenna phase center and derive the coordinate calculation method; Step 3: Establish a multi-static to single-static data conversion model, introduce a distance difference and phase difference compensation mechanism, and convert the real array element echo intermediate frequency signal into a virtual array intermediate frequency signal to eliminate the phase error caused by multi-station acquisition; Step 4: Divide the imaging region into multiple sub-blocks according to the focusing constraints, call the time-domain imaging algorithm based on non-uniform fast Fourier transform for each sub-block to perform local imaging, and then obtain the complete image through spatial coherence superposition.

[0026] Furthermore, the acquisition of raw echo data includes: Based on the requirements of near-field imaging scenarios, a near-field MIMO millimeter-wave radar data acquisition platform is built by configuring a millimeter-wave radar transceiver module, a data acquisition card, and a motion control platform. Based on the hardware architecture of the data acquisition platform, the radar signal acquisition process and the platform movement process are synchronously controlled; Based on the acquisition timing after synchronization control, the raw echo data is acquired point by point through two-dimensional mechanical scanning and multi-antenna time-division multiplexing transceiver.

[0027] Specifically, the near-field MIMO millimeter-wave radar data acquisition platform built in step 1 is as follows: like Figure 1 As shown, the experimental system mainly consists of a millimeter-wave radar transceiver module IWR1642BOOST, a data acquisition card DCA1000 EVM, a stepper motor controller FMC4030, two stepper motors, two sliding module platforms, and a personal computer. The millimeter-wave radar uses a 2-transmit, 4-receive antenna configuration, responsible for generating and receiving FMCW signals; the data acquisition card is responsible for signal acquisition and preprocessing operations such as mixing; the stepper motor controller, stepper motors, and sliding modules together form a motion control platform for achieving precise positioning and movement of the radar within the scanning plane. To ensure the consistency and reliability of data acquisition, the system implements strict synchronous control of radar signal acquisition and platform movement.

[0028] The system workflow is as follows: The PC first sends a running signal to the stepper motor controller via the platform control bus. After a period of time, once the platform's running speed reaches the set constant speed, it then sends a running signal to the radar sensor and data acquisition card via the transceiver control bus to initiate data acquisition. The sensor's operating path is as follows: Figure 3 As shown by the red path, the radar transmits a linear frequency modulated signal at each sampling location and receives echo signals from the target scene. After being acquired and processed by the data acquisition card, the signals are uploaded to the PC via an Ethernet data bus. By sampling the entire scanning area point by point, spatial observation data covering the imaging area can be obtained. These data, in the two-dimensional spatial dimension and the distance dimension, together constitute a three-dimensional data cube.

[0029] Specifically, in step 2, constructing the equivalent model of the MIMO radar virtual array involves: like Figure 2 As shown, the radar sensor employs a MIMO configuration with separate transmit and receive operations, and the array includes two transmit antennas and four receive antennas. In a spatial Cartesian coordinate system, the radar is located on a plane. Inside, The area is scanned, where the radar sensor's first... root transmitting antenna and the first The location of the root receiving antenna can be represented by coordinates as follows: (1); (2); in, and They represent the first The horizontal and vertical coordinates of the transmitting antenna, and They represent the first The horizontal and vertical coordinates of the transmitting antenna.

[0030] Assuming the orthogonality between radar transmitting antennas is achieved through time-division multiplexing (TDM), each transmitting and receiving antenna can be mapped to a transmit-receive pair. In near-field imaging scenarios, if the distance from the target scattering point to the transmit-receive antenna array is much greater than the interval between the transmitting and receiving antennas, a midpoint approximation can be derived, i.e., there exists an equivalent phase center between each pair of transmit and receive antennas. Assuming that at a certain sampling position, the radar signal is generated by the first... The first transmitting antenna emits the signal, and it is received by the second... If the data received by the receiving antenna is equivalent to completing a single-station transmission and reception operation at the phase center, then each pair of transmitting and receiving antennas can be represented by a virtual transceiver at the equivalent phase center. Different virtual transceivers do not overlap with each other, forming a set of virtual transmission and reception arrays.

[0031] like Figure 3 As shown, the original array can be equivalent to a virtual array of 8 antennas, the positions of which are determined by... and Constraints are imposed by the first root transmitting antenna and the first The position of the virtual antenna equivalent to the root receiving antenna can be represented as: (3); in: (4); (5); in, and These represent the coordinates of the virtual antenna in the horizontal and vertical directions, respectively.

[0032] Furthermore, based on the spatial coordinates, a multi-static to single-static data conversion model is established, and the real array element echo intermediate frequency signal is equivalent to a virtual array intermediate frequency signal through distance difference and phase difference compensation, including: The actual transmission and reception distance of the radar signal is obtained by comparing the spatial coordinates of the scattering point with the spatial coordinates of the actual transmitting and receiving antennas. The transmit / receive distance at the virtual antenna is obtained based on the spatial coordinates of the phase center of the virtual antenna and the spatial coordinates of the scattering point. Based on the actual transmit / receive distance and the transmit / receive distance at the virtual antenna, the difference in transmit / receive distance between the real array element and the virtual array element is derived. Based on the transmit / receive distance difference, the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal.

[0033] Furthermore, based on the aforementioned transmit / receive distance difference, converting the real array element echo intermediate frequency signal into a virtual array intermediate frequency signal includes: Based on the transmit / receive distance difference and the free space wavenumber, calculate the phase difference between the intermediate frequency signals received by the virtual array element and the real array element; Based on the phase difference, phase compensation is performed on the intermediate frequency signal of the real array element echo to obtain an intermediate frequency signal equivalent to that transmitted and received by a single station at the virtual array.

[0034] Specifically, step 3: Establish a multi-static to single-static data conversion model to compensate for phase errors caused by multi-station acquisition; Assume the coordinates of the scattering point in the imaging region are Therefore, the actual transmission and reception distance of the radar signal can be expressed as: (6); in, (7); (8); The transmit / receive distance at the corresponding virtual antenna can be expressed as: (9); Now, this embodiment introduces two distance differences to represent the actual distance between the transmitting and receiving antennas. (10); (11); Based on the spatial distribution of the transmitting and receiving antennas, it can be known that... , Substituting equations (10) and (11) into equations (2) and (3) respectively, we obtain the following set of expressions: (12); Substituting equations (7), (8), and (12) into equation (6), we get: (13); According to the Taylor series expansion formula, in A third Taylor expansion of the above equation yields: (14); Ignoring the cubic term in equation (14), we can obtain an approximate relationship between the transmission and reception distances of the real array elements and the virtual array elements: (15); Therefore, the difference in transmission and reception distance between virtual array elements and real array elements can be expressed as: (16); Assuming the signal transmitted by the millimeter-wave radar is a linear frequency modulated continuous wave signal, then the... The transmitted signal from the root transmitting antenna can be represented as: (17); in, Indicates the starting frequency of the transmitted signal. This represents the frequency modulation slope of the radar. (From the...) The reflected signal from a single scattering point received by the root receiving antenna can be expressed as: (18); in, express Reflectance at the scattering point This represents the speed of electromagnetic waves in a vacuum. After mixing the two signals and ignoring the residual video phase in the formula, the corresponding intermediate frequency signal expression is: (19); According to equation (15), there is a difference in transmission and reception distance between the virtual array elements and the real array elements. Therefore, the IF signals received at the virtual array elements and the real array elements have the following phase difference: (20); in, represents the wavenumber in free space. The phase difference mentioned above is compensated to... After that, you can get: (twenty one); in, This corresponds to the IF signal at the virtual array location. In actual imaging work, the real echo data is composed of the superposition of reflected signals from all spatial scattering points in the imaging scene. The beat frequency signal received by the virtual array can be further represented as: (twenty two); Before performing image reconstruction on the equivalent echo data, they need to be rearranged into three-dimensional data blocks according to the spatial distribution of the virtual array. Assume that the radar sensor scans the horizontal and vertical directions a certain number of times during the imaging process. and The data obtained by scanning a row horizontally each time can be represented as: (twenty three); in, Indicates the first The received data of the line, The first in the horizontal direction At the receiving location, by the first root transmitting antenna and the first The intermediate frequency signal received by the virtual array elements formed by the root receiving antennas. Then, each row of received data is arranged according to its spatial position. In this embodiment, the complete echo data of the entire scanning array can be obtained: (twenty four); Therefore, the original multi-channel echo data can be reconstructed into a regular three-dimensional data block form according to their spatial position relationship, thus completing the equivalent approximation of the observation data from multi-static to equivalent single-static data.

[0035] Furthermore, based on the intermediate frequency signal of the virtual array, the imaging region is divided into multiple sub-blocks according to focusing constraints, including: Based on the intermediate frequency signal of the virtual array, and combined with the scene center coordinates of the imaging area and the virtual antenna position, the distance difference between the scene center point and the edge point to the virtual antenna is calculated. Based on the distance difference and the wavelength of the radar transmitted signal, determine the phase error caused by the reference point phase compensation; Based on a preset phase error threshold, focusing constraints are established to limit the maximum spatial scale of the sub-blocks.

[0036] Furthermore, based on a preset phase error threshold, focusing constraints are established to limit the maximum spatial scale of the sub-blocks, including: Based on the phase error threshold and radar wavelength, a constraint relationship is established between the sub-block size, imaging distance, and operating wavelength; Based on the constraints, calculate the maximum allowable dimensions of the sub-block in the horizontal and vertical directions; Based on the maximum allowable size, the imaging region is divided into multiple non-overlapping sub-blocks.

[0037] Furthermore, the complete image is obtained by locally imaging each sub-block and then spatially coherently superimposing the images, including: Establish the wavenumber domain integral relationship between the reflectivity of the scattering point and the echo signal; Based on the wavenumber domain integral relationship, a two-dimensional fast Fourier transform is performed on the echo signal to map it to the wavenumber domain; Based on the non-uniform distribution characteristics of wavenumber components in the range direction, Stolt interpolation is performed on the wavenumber domain signal to correct phase error; The complete image is obtained based on the corrected relative error.

[0038] Furthermore, based on the corrected relative error, the complete image is obtained, including: Based on the interpolated wavenumber domain signal, the range-oriented spectral conversion is completed through non-uniform fast Fourier transform; Based on the spectral conversion results in the range direction, a two-dimensional inverse fast Fourier transform is performed to reconstruct the reflectivity distribution of the scattering points in the sub-block region; Based on the reflectivity distribution and spatial location of the scattering points of each sub-block, the reflectivity distribution of the complete imaging region is obtained through spatial coherent superposition.

[0039] Specifically, step 4: Divide the imaging area into multiple sub-blocks according to the focusing constraints, and then coherently overlay them after imaging one by one using the imaging algorithm; In near-field large-scale imaging scenarios, significant focus degradation occurs at scene edges. This is primarily because electromagnetic waves propagate as spherical waves in the near field. When the target is far from the center of the imaging area, phase compensation using a unified reference point cannot accurately describe the differences in propagation paths at different scattering points, leading to defocusing and geometric distortion of the target in the edge region. Therefore, it is necessary to further investigate imaging optimization strategies suitable for large-scale scenes within the near-field imaging framework.

[0040] To reduce the impact of reference point selection on image quality, the entire imaging area can be divided into several sub-regions, and independent phase compensation and imaging reconstruction can be performed in each sub-region, thereby reducing the impact of phase error accumulation on the focusing performance of edge targets.

[0041] Assume the scene center coordinates of the current imaging area are The position of any point in the same plane is Then located at The virtual antenna at that location is positioned relative to the two points as follows: (25); (26); It is clear that there is a difference between the two formulas above: (27); Based on the correspondence between the spatial propagation distance of radar waves and the degree of hysteresis in the vibration state, the phase difference generated during the entire transmission and reception process can be calculated as follows: (28); in, The wavelength of the radar transmitted signal. When the difference between the true phase and the compensated phase exceeds... At this time, the focusing effect will be significantly affected, which can be achieved by the following expression: Apply constraints: (29); In most near-field imaging systems, the transmitted signal is emitted perpendicularly, with the signal beam perpendicular to the radar array. This means that scattering is primarily affected by the emitted signal along the normal direction at the scattering location. The contributions from other emission locations gradually decrease as they deviate from the scattering direction. In the previous section, an equivalent representation was performed for the transceiver virtual array, where the contribution of the receiving element to scattering is the same as that of the transmitting element. Therefore, to calculate the maximum size of the scene partition, the coordinates of the virtual element position C that contributes the most to the target point within the scene are: Then equation (29) can be simplified to: (30); make Substituting into equation (25) and simplifying, we get: (31); This formula shows that the scene segmentation scale is mainly constrained by both the imaging distance and the operating wavelength. Therefore, in this embodiment, the data can be segmented into scenes according to formula (31). Figure 4 As shown, after dividing the data into scene segments, local imaging is performed according to their spatial locations. After imaging each data block, they are coherently superimposed according to their spatial locations to obtain a complete image.

[0042] Substituting equations (19) and (20) into equation (22) and expanding, we get: (32); Based on the dispersion relation, the second half of the above equation can be rewritten as: (33); In the formula: , , and Representing wavenumbers respectively Wavenumber components in space. Substituting the amplitude term from equation (33) into equation (32) and ignoring it, we obtain the following integral: (34); Based on the definitions of Fast Fourier Transform and Inverse Fast Fourier Transform, equation (34) can be rewritten as follows: (35); In the formula: and Let these represent the Fast Fourier Transform (FFT) and Inverse Fast Fourier Transform (IFFT) operators, respectively, due to reflectivity. This is the object to be restored in this embodiment. Therefore, the reconstruction formula can be obtained through the properties of Fourier transform in this embodiment: (36); Due to the scanning method, the wavenumber components... exist The range direction is non-uniform, therefore a stop interpolation is needed before applying the range-direction FFT to correct the phase error caused by this non-uniformity. Therefore, the final reconstruction formula should be expressed as: (37); In the formula: This represents the non-uniform fast Fourier transform operator, which means that the transform object is subjected to a stolt interpolation before the fast Fourier transform operation.

[0043] Table 1 illustrates the core process of this block imaging algorithm: Table 1 To verify the effectiveness and feasibility of this invention in practical near-field imaging scenarios, the radar sensor was configured in a 2-transmit, 4-receive MIMO mode. Multi-channel collaborative operation expanded the equivalent virtual array, thereby improving data acquisition efficiency and spatial sampling capability. Under this configuration, the radar system sampled point-by-point along the scanning plane. Figure 6 Data was collected from the target shown. The raw data was then imaged using traditional methods, while the data after multi-static to single-static conversion was imaged using both traditional methods and the proposed method.

[0044] Without the multi-static to single-static equivalent conversion, both the 2D and 3D imaging results obtained using the original imaging method exhibit significant focus degradation. The target energy shows a diffuse distribution, the main lobe is not concentrated, and there are strong side lobes and artifacts. This is mainly due to the spatial differences between the multi-channel observation data, leading to inconsistent phase information and affecting the coherent superposition effect. In contrast, after the multi-static to single-static equivalent conversion, the imaging results are improved compared to the unconverted case. The target outline gradually becomes clearer, and the energy concentration is improved. This indicates that the equivalent conversion corrects the geometric differences between the multi-channel data to some extent, enhancing the phase consistency during the imaging process. However, slight defocusing and incomplete structural details are still observed in some target areas, indicating that traditional methods still have certain limitations when processing complex multi-channel data. Furthermore, the imaging results using the block imaging method proposed in this embodiment show that the target's main lobe is more concentrated, side lobes and artifacts are effectively suppressed, structural details in the 2D image are clearer, and the target outline in the 3D reconstruction result is more complete and continuous. This embodiment has significant advantages in improving multi-channel data consistency and enhancing imaging focusing performance.

[0045] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations 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. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning, characterized in that, include: Acquire raw echo data; Based on the original echo data, construct an equivalent model of the MIMO radar virtual array; Based on the equivalent model of the MIMO radar virtual array, determine the spatial coordinates of the virtual antenna phase center; Based on the spatial coordinates, a multi-static to single-static data conversion model is established, and the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal through distance difference and phase difference compensation. Based on the intermediate frequency signal of the virtual array, the imaging area is divided into multiple sub-blocks according to the focusing constraints. After local imaging of each sub-block, a complete image is obtained by spatial coherent superposition.

2. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 1, characterized in that, The acquisition of raw echo data includes: Based on the requirements of near-field imaging scenarios, a near-field MIMO millimeter-wave radar data acquisition platform is built by configuring a millimeter-wave radar transceiver module, a data acquisition card, and a motion control platform. Based on the hardware architecture of the data acquisition platform, the radar signal acquisition process and the platform movement process are synchronously controlled; Based on the acquisition timing after synchronization control, the raw echo data is acquired point by point through two-dimensional mechanical scanning and multi-antenna time-division multiplexing transceiver.

3. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 1, characterized in that, Determining the spatial coordinates of the virtual antenna phase center includes: ； ； ； in, For the first root transmitting antenna and the first The position of the virtual antenna equivalent to the root receiving antenna. and These represent the coordinates of the virtual antenna in the horizontal and vertical directions, respectively. and They represent the first The horizontal and vertical coordinates of the transmitting antenna, and They represent the first The horizontal and vertical coordinates of the transmitting antenna.

4. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 1, characterized in that, Based on the spatial coordinates, a multi-static to single-static data conversion model is established, and the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal through distance difference and phase difference compensation, including: The actual transmission and reception distance of the radar signal is obtained by comparing the spatial coordinates of the scattering point with the spatial coordinates of the actual transmitting and receiving antennas. The transmit / receive distance at the virtual antenna is obtained based on the spatial coordinates of the phase center of the virtual antenna and the spatial coordinates of the scattering point. Based on the actual transmit / receive distance and the transmit / receive distance at the virtual antenna, the difference in transmit / receive distance between the real array element and the virtual array element is derived. Based on the transmit / receive distance difference, the real array element echo intermediate frequency signal is equivalent to the virtual array intermediate frequency signal.

5. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 4, characterized in that, Based on the aforementioned transmit / receive distance difference, the equivalent conversion of the real array element echo intermediate frequency signal to the virtual array intermediate frequency signal includes: Based on the transmit / receive distance difference and the free space wavenumber, calculate the phase difference between the intermediate frequency signals received by the virtual array element and the real array element; Based on the phase difference, phase compensation is performed on the intermediate frequency signal of the real array element echo to obtain an intermediate frequency signal equivalent to that transmitted and received by a single station at the virtual array.

6. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 1, characterized in that, Based on the intermediate frequency signal of the virtual array, the imaging region is divided into multiple sub-blocks according to the focusing constraints, including: Based on the intermediate frequency signal of the virtual array, and combined with the scene center coordinates of the imaging area and the virtual antenna position, the distance difference between the scene center point and the edge point to the virtual antenna is calculated. Based on the distance difference and the wavelength of the radar transmitted signal, determine the phase error caused by the reference point phase compensation; Based on a preset phase error threshold, focusing constraints are established to limit the maximum spatial scale of the sub-blocks.

7. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 6, characterized in that, Based on a preset phase error threshold, focusing constraints are established to limit the maximum spatial scale of the sub-blocks, including: Based on the phase error threshold and radar wavelength, a constraint relationship is established between the sub-block size, imaging distance, and operating wavelength; Based on the constraints, calculate the maximum allowable dimensions of the sub-block in the horizontal and vertical directions; Based on the maximum allowable size, the imaging region is divided into multiple non-overlapping sub-blocks.

8. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 1, characterized in that, The complete image is obtained by locally imaging each sub-block and then spatially coherently superimposing the images, including: Establish the wavenumber domain integral relationship between the reflectivity of the scattering point and the echo signal; Based on the wavenumber domain integral relationship, a two-dimensional fast Fourier transform is performed on the echo signal to map it to the wavenumber domain; Based on the non-uniform distribution characteristics of wavenumber components in the range direction, Stolt interpolation is performed on the wavenumber domain signal to correct phase error; The complete image is obtained based on the corrected relative error.

9. The MIMO millimeter-wave radar near-field SAR imaging method based on scene block partitioning according to claim 1, characterized in that, Based on the corrected relative error, the complete image includes: Based on the interpolated wavenumber domain signal, the range-oriented spectral conversion is completed through non-uniform fast Fourier transform; Based on the spectral conversion results in the range direction, a two-dimensional inverse fast Fourier transform is performed to reconstruct the reflectivity distribution of the scattering points in the sub-block region; Based on the reflectivity distribution and spatial location of the scattering points of each sub-block, the reflectivity distribution of the complete imaging region is obtained through spatial coherent superposition.

Images