A fast hierarchical kirchhoff imaging method for MIMO array radar

Abstract

The application discloses a fast multi-stage Kirchhoff imaging method for MIMO array radar, which realizes wide adaptation from small-scale array to large-scale array through dynamic multi-stage, spatial down-conversion based on real-time estimation of local geometry and multi-stage fusion strategy based on phase alignment, guarantees high imaging quality and significantly improves calculation efficiency. The method can stably generate high-quality images in a complex target scene, is superior to comparative algorithms in spatial resolution, shape fidelity and sidelobe suppression level and exhibits excellent robustness.

Description

Technical Field

[0001] This invention belongs to the field of radar imaging technology, specifically relating to a fast graded Kirchhoff imaging method for MIMO array radar. Background Technology

[0002] Near-field radar imaging technology has significant application value in fields such as non-destructive testing and security screening. To achieve high-resolution imaging, large-aperture arrays are required, while real-time imaging demands high-speed data acquisition. MIMO arrays synthesize large virtual apertures through sparse arraying and combine them with electronic scanning, resolving the aforementioned contradiction at the hardware level, but shifting the computational bottleneck to the image reconstruction stage.

[0003] Existing imaging algorithms are mainly divided into wavenumber domain algorithms and time-domain algorithms. Wavenumber domain algorithms (such as the distance migration algorithm RMA) are computationally efficient, but their performance is heavily dependent on the uniformity and regularity of the array. In general non-uniform, ultra-sparse MIMO arrays, the imaging quality will be significantly degraded due to wavenumber domain spectral aliasing and distortion. Time-domain algorithms (such as the Kirchhoff migration algorithm KMA) are universally applicable to array structures, but their computational complexity is linearly related to the number of pixels and array elements in the imaging domain. This results in a heavy computational burden when imaging large-scale arrays or large scenes, making it difficult to meet real-time requirements.

[0004] Some studies have attempted to introduce the Fast Frame Back Projection (FFBP) concept from Synthetic Aperture Radar (SAR) into near-field MIMO imaging, proposing the Fast Factorized Kirchhoff Transfer Algorithm (FFKMA). This algorithm reduces computation by recursively partitioning the subarray and leveraging the reduced bandwidth of the sub-images. However, existing FFKMA algorithms have limitations: firstly, their speedup benefits are based on the premise that "the subarray aperture is much smaller than the full array," and the speedup gains are not significant in small-scale MIMO systems with limited virtual apertures; secondly, the algorithm relies on complex pre-computation steps (such as Helmholtz-Hodge decomposition) to optimize spectral compression, lacking flexibility and ease of deployment in scenarios with frequent array configuration changes or embedded platforms.

[0005] Therefore, designing a fast imaging method that can maintain the universality of time-domain algorithm arrays, achieve computational efficiency close to that of wavenumber domain algorithms, adapt to arrays of different sizes, and be easy to implement is a technical problem that urgently needs to be solved. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, this invention provides a fast graded Kirchhoff imaging method for MIMO array radar. Through dynamic graded imaging, spatial down-conversion based on real-time local geometry estimation, and a multi-level fusion strategy using phase alignment, it achieves broad adaptability to arrays ranging from small to large scales, significantly improving computational efficiency while maintaining high imaging quality. The method of this invention can stably generate high-quality images in complex target scenarios, outperforming comparative algorithms in spatial resolution, shape fidelity, and sidelobe suppression, demonstrating excellent robustness.

[0007] The technical solution adopted by this invention to solve its technical problem is as follows: Step 1: Dynamic array sorting; Recursively divide the complete MIMO array into Subarray structure;

[0008] in, This represents the total number of sent and received pairs. This is the minimum channel number threshold for the lowest-level subarray; go through The second division eventually forms A bottom-level subarray; the bottom-level subarray is the subarray generated by the last binary search; Step 2: Sub-image reconstruction; For each bottom-level subarray, Kirchhoff's transfer algorithm (KMA) is used for imaging to obtain a set of bottom-level sub-images; Step 3: Real-time calculation of the Spatial Down Conversion (SDC) factor; For each bottom-level subarray, the SDC factor is calculated in real time based on local geometric relationships, specifically as follows: Computational subarray Geometric center position ; According to the center of the imaging domain With geometric center position The relative direction determines the local wave vector. ; Then construct the SDC factor The phase field , Represents the two-dimensional coordinate variables on the imaging plane. Represents the imaginary unit. e It is a natural constant; Step 4: Multi-level sub-image fusion; Starting from the lowest-level sub-image, merge images level by level upwards; merge two levels of sub-images. and First, the global phase difference is calculated using normalized cross-correlation. and to Perform phase compensation: , This represents the sub-image after phase compensation; then, the phase-aligned sub-images are coherently superimposed to obtain the next-level sub-image. Step 5: Repeat step 4 until the images are merged to the top layer to obtain the final full-resolution radar image.

[0009] Furthermore, the global phase difference Δφ is calculated using the following formula:

[0010] in, express The complex conjugate of , angle(·) represents the phase angle of taking the complex number.

[0011] An electronic device includes: a processor and a memory; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to enable the electronic device to perform the above-described fast fractional Kirchhoff imaging method.

[0012] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described fast graded Kirchhoff imaging method.

[0013] A chip includes a processor for calling and running a computer program from a memory, causing a device equipped with the chip to perform the aforementioned fast fractional Kirchhoff imaging method.

[0014] A computer program product includes a computer storage medium storing a computer program, the computer program including instructions executable by at least one processor, which, when executed by the at least one processor, implement the above-described fast fractional Kirchhoff imaging method.

[0015] The beneficial effects of this invention are as follows: 1. Wide array adaptability: By introducing a dynamic tiering mechanism based on the minimum channel number threshold, the algorithm can adaptively adjust the tiering level, thereby effectively adapting to different application scenarios from small-scale arrays with limited virtual channels to large-scale arrays, overcoming the limitation of the original FFKMA algorithm having insignificant acceleration benefits in small-scale arrays.

[0016] 2. High Imaging Quality: Employing a cross-correlation-based phase alignment fusion strategy, this invention effectively compensates for phase errors between sub-images caused by near-field wavefront curvature, achieving precise alignment and fusion of complex domain data. Compared to traditional geometric interpolation fusion methods, this invention significantly improves image contrast, better preserves the sharpness of target edges and contours, and suppresses sidelobes and background artifacts.

[0017] 3. High computational efficiency and easy deployment: By utilizing local spectral analysis to compress sub-image bandwidth, the number of sampling points required for KMA reconstruction is significantly reduced. Complex global optimization steps such as Helmholtz-Hodge decomposition are abandoned, and real-time calculation of the SDC factor based on local geometry is adopted, reducing pre-computation complexity and storage overhead. The overall algorithm complexity can be reduced to... Magnitude ( (Total number of pixels in the imaging domain), far lower than that of traditional KMA. It meets the requirements for real-time imaging and is easier to implement and deploy in engineering.

[0018] 4. Through imaging experiments on a lying human target, the method of the present invention can stably generate high-quality images under a 4-transmitter 12-receiver array configuration and complex target scenarios. It outperforms the comparison algorithm in terms of spatial resolution, shape fidelity and sidelobe suppression level, demonstrating excellent robustness. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention.

[0020] Figure 2 This is a schematic diagram of the experimental system and array configuration of an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of a supine human body detection scenario according to an embodiment of the present invention.

[0022] Figure 4 The following is an embodiment of the present invention for a supine human body scene, where (a) is the first-order imaging result of FFKMA and (b) is the first-order imaging result of the method of the present invention; Figure 5 The following is an example of an embodiment of the present invention for a supine human body scene, wherein (a) is the second-order fusion imaging result of FFKMA, and (b) is the second-order fusion imaging result of the method of the present invention; Figure 6 The following is an example of an embodiment of the present invention for a supine human body scene, wherein (a) is an FFKMA imaging result image and (b) is an imaging result image of the method of the present invention. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0024] This invention provides a fast graded Kirchhoff imaging method for MIMO array radar. This method achieves broad adaptability to arrays ranging from small to large scales through dynamic graded imaging, spatial down-conversion based on real-time local geometry estimation, and a multi-level fusion strategy using phase alignment. It significantly improves computational efficiency while maintaining high imaging quality.

[0025] This invention provides a fast Kirchhoff imaging method for MIMO array radar, the core process of which is as follows: Figure 1 As shown, the specific implementation steps are as follows: 1. Initialization and parameter settings: Configure the positions of the transmit elements of the MIMO array. and the position of the receiving array element Set the range of the imaging area D. And determine the discretized grid, These represent the horizontal and vertical coordinates within the imaging region plane, respectively. Indicates the imaging area in Minimum boundary of direction, Indicates the imaging area in Maximum boundary of direction, Indicates the imaging area in Minimum boundary of direction, The image area is indicated. Maximum directional boundary. Setting the radar operating frequency band. and the number of frequency sampling points , These represent the lowest and highest frequencies of the radar transmitted signal, respectively. Based on application requirements, set the minimum number of channels threshold for the lowest-level subarray. .

[0026] 2. Dynamic order division and subarray partitioning: According to the formula Calculate the fractional order ,in This represents the total number of transmit and receive channels. These represent the total number of transmitting elements and the total number of receiving elements, respectively. The entire array is treated as the top-level (Level-L) subarray, and then a binary search is recursively performed until a result is generated. A Level-1 subarray. Ensure that the virtual aperture profiles of the subarrays do not overlap during partitioning.

[0027] 3. Calculate the SDC factor and sampling grid: for each subarray at each level. Perform the following calculations: a. Calculate the geometric center of the subarray .

[0028] b. Based on the center of the imaging domain and The vector difference, combined with the mean wavenumber Calculate the local wave vector The direction.

[0029] c. According to the formula Generate the SDC factor matrix of the imaging grid corresponding to the subarray.

[0030] d. The local wavenumber support domain based on this subarray (from the formula) and (Define), calculate its local bandwidth , This allows us to determine the number of compressed sampling points. And generate an optimized non-uniform sampling grid. , The wavenumber corresponding to the highest operating frequency of the radar system. Indicates the half-aperture angle of the subarray in the x-direction. This represents the component of the wave vector in the x-direction; this step can be completed offline.

[0031] 4. Level-1 Sub-image Reconstruction: For each Level-1 sub-array, its corresponding frequency domain or time domain echo data is reconstructed. In its pre-calculated optimized sampling grid The Kirchhoff migration KMA core formula is used for imaging to obtain low-resolution sub-images. , These represent the position coordinates of the transmitting element, the position coordinates of the receiving element, and the two-way propagation time of the signal from transmission to the target and back to reception, respectively. The KMA formula involves the two-way delay. The focusing and coherent superposition, Indicates the distance from the imaging point to the transmitting array element distance, This indicates the distance from the imaging point to the receiving array element. distance, It represents the speed of light.

[0032] 5. Multi-level phase alignment and fusion: a. Starting from Level-1, combine two child images belonging to the same parent node. and (as obtained in the previous step) Pair them up.

[0033] b. Resample the two sub-images onto the sampling grid of their parent node's sub-image (usually using bilinear interpolation).

[0034] c. Calculate the global phase difference between the two images using the formula. .

[0035] d. For the image Apply phase compensation: .

[0036] e. Coherently superimpose the two phase-aligned images: This yields the sub-image of its parent node (Level-2).

[0037] f. Repeat step ae, merging upwards level by level until reaching the highest level (Level-L). The result at this point... This is the final full-resolution radar image. That is, target reflectivity The estimate.

[0038] Example: To verify the effectiveness of the method of this invention, a stepped-frequency continuous-wave (SFCW) MIMO near-field radar imaging system (operating frequency band 2-6 GHz) was built. An imaging experiment was conducted on a lying human target using a 4-transmitter, 12-receiver array configuration, and the results were compared with the classic Fast Factorized Kirchhoff Transfer Algorithm (FFKMA). Figure 2 and Figure 3 As shown.

[0039] Image quality comparison is attached. Figures 4-6 As shown, the experimental results indicate that: In human body imaging, the reconstructed human body contours using the method of this invention are more continuous and sharper, and background clutter and artifacts are significantly suppressed. Figure 4 (a) Figure 4 (b) Figure 5 (a) Figure 5 (b) Figure 6 (a) Figure 6 (b)). In scenarios with a large field of view and strong near-field effect, such as a supine human body, the imaging results of the method of this invention can clearly distinguish the torso and limbs, while the FFKMA results produce obvious ghosting in the far region ( Figure 6 (a) Figure 6 (b) Figure 4 In (a), from left to right, are: the first-order FFKMA imaging result of subarray 1, the first-order FFKMA imaging result of subarray 2, the first-order FFKMA imaging result of subarray 3, and the first-order FFKMA imaging result of subarray 4. Figure 4 In (b), from left to right, are: the first-order imaging result of the present invention method of subarray 1, the first-order imaging result of the present invention method of subarray 2, the first-order imaging result of the present invention method of subarray 3, and the first-order imaging result of the present invention method of subarray 4. Figure 5In (a), from left to right are: the FFKMA second-order fusion imaging result of synthetic array 1 and the FFKMA second-order fusion imaging result of synthetic array 2; Figure 5 In (b), from left to right, are: the second-order fusion imaging result of the present invention by the synthetic array 1, and the second-order fusion imaging result of the present invention by the synthetic array 2.

[0040] In terms of computational efficiency, thanks to the use of dynamic gradation and local SDC factor calculation, FCWKM achieves significantly lower running time than traditional KMA while maintaining imaging quality, and is comparable to or even better than FFKMA, especially with smaller arrays.

[0041]

[0042] In summary, the method proposed in this invention achieves a good balance between imaging quality and computational efficiency, and is suitable for near-field radar imaging applications of sparse MIMO arrays that require both real-time performance and imaging accuracy.

Claims

1. A fast graded Kirchhoff imaging method for MIMO array radar, characterized in that, Includes the following steps: Step 1: Dynamic array sorting; Recursively divide the complete MIMO array into Subarray structure; in, This represents the total number of sent and received pairs. This is the minimum channel number threshold for the lowest-level subarray; go through The second division eventually forms A bottom-level subarray; the bottom-level subarray is the subarray generated by the last binary search; Step 2: Sub-image reconstruction; For each bottom-level subarray, Kirchhoff's transfer algorithm (KMA) is used for imaging to obtain a set of bottom-level sub-images; Step 3: Real-time calculation of the Spatial Down Conversion (SDC) factor; For each bottom-level subarray, the SDC factor is calculated in real time based on local geometric relationships, specifically as follows: Computational subarray Geometric center position ; According to the center of the imaging domain With geometric center position The relative direction determines the local wave vector. ; Then construct the SDC factor The phase field , Represents the two-dimensional coordinate variables on the imaging plane. Represents the imaginary unit. e It is a natural constant; Step 4: Multi-level sub-image fusion; Starting from the lowest-level sub-image, merge images level by level upwards; merge two levels of sub-images. and First, the global phase difference is calculated using normalized cross-correlation. and to Perform phase compensation: , This represents the sub-image after phase compensation; then, the phase-aligned sub-images are coherently superimposed to obtain the next-level sub-image. Step 5: Repeat step 4 until the images are merged to the top layer to obtain the final full-resolution radar image.

2. The fast graded Kirchhoff imaging method for MIMO array radar according to claim 1, characterized in that, The global phase difference Δφ is calculated using the following formula: in, express The complex conjugate of , angle(·) represents the phase angle of taking the complex number.

3. An electronic device, characterized in that, include: Processor and memory; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to cause the electronic device to perform the method as described in any one of claims 1 to 2.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 2.

5. A chip, characterized in that, include: A processor for retrieving and running a computer program from memory, causing a device on which the chip is mounted to perform the method as described in any one of claims 1 to 2.

6. A computer program product, characterized in that, The computer program product includes a computer storage medium storing a computer program, the computer program including instructions executable by at least one processor, which, when executed by the at least one processor, implement the method as described in any one of claims 1 to 2.

Images