Two-dimensional real-time back-projection perception imaging method, device, equipment and medium

Abstract

The application discloses a kind of two-dimensional real-time back-projection perception imaging method, device, equipment and medium based on multi-antenna system, wherein, imaging method includes the following steps: S1, echo signal is carried out hilbert transform, frequency transfer and fast fourier transform, and echo signal spectrum is output;S2, echo signal and echo signal are carried out matched filtering, and time domain pulse compression result is output;S3, according to the imaging area and grid size calculation imaging area coordinate, according to pulse compression result, the corresponding pixel signal intensity of image is calculated by two-dimensional interpolation;Then the imaging area coordinate is used to calculate phase compensation factor, and the interpolation result is phase compensated, and finally two-dimensional range image is output and pre-stored;S4, to the multi-channel two-dimensional range image is windowed and superimposed, and two-dimensional SAR image is output in real time.The application is based on SAR simulation imaging method, can process multi-antenna echo signal, can improve angle resolution and realize real-time perception imaging of indoor environment.

Description

Technical Field

[0001] This invention belongs to the field of communication sensing imaging technology, specifically relating to a two-dimensional real-time back-projection sensing imaging method, device, computer equipment, and storage medium based on a multi-antenna system. Background Technology

[0002] Integrated Sensing and Communication (ISAC) is a technology that deeply integrates communication and environmental perception capabilities. By sharing spectrum resources and hardware platforms, it enables collaborative work between information transmission and target perception. Sensing imaging, as a key application, uses wireless signals to perform high-resolution modeling and real-time imaging of targets or environments, and is widely used in autonomous driving, smart cities, and the Industrial Internet of Things (IIoT). Currently, the main sensing imaging methods include: Synthetic Aperture Radar (SAR) simulated imaging, passive sensing imaging using communication signals, millimeter-wave imaging, and compressed sensing imaging. SAR simulated imaging utilizes radio waves reflected from targets to generate high-resolution images through signal processing, exhibiting all-weather adaptability and strong penetration capabilities; furthermore, its real-time imaging capabilities have been thoroughly validated in various application scenarios.

[0003] SAR simulation imaging methods are divided into frequency domain and time domain methods. Frequency domain imaging methods are computationally efficient and suitable for imaging wide scenes and large areas, especially in remote sensing and ground monitoring. However, they are sensitive to orbital errors and motion parameters, requiring precise geometric correction; otherwise, image distortion is likely to occur. In addition, they require linear apertures, making motion compensation difficult to embed, and the algorithms and correction steps are complex in some scenarios, increasing the difficulty of system development. Time domain imaging methods use point-by-point imaging, which can achieve high-resolution imaging and is not limited by the size of the imaging scene. They are suitable for parallel hardware processing using digital signal processing devices. However, point-to-point image reconstruction requires a large number of interpolation operations, resulting in a large computational load, especially as the target distance and processing area increase, the computational efficiency decreases significantly.

[0004] Back projection algorithm (BP) is a precise SAR imaging method. This algorithm back-projects the compressed echo signal of each pulse to the image domain and performs coherent accumulation of pixels to reconstruct a high-resolution imaging result. However, in actual imaging, the back projection imaging method requires two-dimensional interpolation for each pixel in the imaging area, which has high computational complexity and slow operation speed. It is especially inefficient when processing large amounts of data and is difficult to meet the requirements of high precision and real-time imaging on traditional processor platforms (such as ARM, x86, etc.).

[0005] Field-Programmable Gate Arrays (FPGAs) possess characteristics such as strong parallel processing capabilities, abundant hardware resources, high programmability, low latency, and low power consumption. They integrate multipliers, RAM, and various computational IP cores, enabling significant improvements in the real-time performance of imaging devices through parallel processing and pipelined techniques. Implementing back-projection imaging methods on FPGAs presents several technical challenges. First, back-projection imaging is a computationally intensive task; FFT and 2D interpolation require numerous multiplication and addition operations, making the rational allocation of computational resources a significant technical challenge in real-time imaging applications. Second, FPGAs have limited on-chip storage resources, and the large amount of data access required in back-projection imaging complicates storage management, necessitating efficient on-chip and off-chip data exchange and optimized caching mechanisms to reduce access latency. Furthermore, implementing back-projection imaging on FPGAs requires a trade-off between accuracy and latency to achieve real-time processing while maintaining computational accuracy. Summary of the Invention

[0006] The main objective of this invention is to overcome the shortcomings and deficiencies of existing technologies and provide a two-dimensional real-time back-projection sensing imaging method, apparatus, computer device, and storage medium based on a multi-antenna system. This two-dimensional real-time back-projection sensing imaging method achieves high-resolution SAR imaging by coherently superimposing antenna echo signals from different locations; furthermore, this two-dimensional real-time back-projection sensing imaging method can be deployed on an FPGA hardware platform and possesses real-time sensing imaging capabilities.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system is disclosed. The multi-antenna system includes K antennas, K single channels, one analog-to-digital converter (ADC), and one two-dimensional back-projection sensing imaging module. Each antenna corresponds to one single channel. Each antenna receives electromagnetic wave signals from different directions, achieving omnidirectional signal acquisition of the target space. The single channel corresponding to the antenna processes the received radio frequency (RF) signals, including signal amplification, frequency conversion, and filtering, converting weak high-frequency RF signals into intermediate frequency (IF) signals of suitable strength for subsequent processing. The ADC converts all analog IF signals output from the single channels into digital signals, realizing the conversion from the analog domain to the digital domain. The two-dimensional back-projection sensing imaging module uses the multiple digital signals output from the ADC to reconstruct a two-dimensional image of the target using the two-dimensional real-time back-projection sensing imaging method. The two-dimensional real-time back-projection sensing imaging method includes the following steps:

[0009] S1. Perform Hilbert transform, frequency shifting, and fast Fourier transform on the echo signal received by the analog-to-digital converter in the multi-antenna system, and output the echo signal spectrum.

[0010] S2. Perform matched filtering on the transmitted signal and the echo signal, and output the time-domain pulse compression result;

[0011] S3. A rectangular plane with a length of L meters and a width of W meters is used as the imaging area. This imaging area is uniformly divided into I×J grid units, where I and J represent the number of grid units in the length and width directions, respectively. The size of each grid unit is determined by L / I and W / J, corresponding to the unit size in the length and width directions, respectively. The imaging area coordinates are calculated based on the imaging area and grid size. The corresponding pixel signal intensity of the image is calculated by two-dimensional interpolation based on the pulse compression result. Then, the phase compensation factor is calculated using the imaging area coordinates. Phase compensation is performed on the interpolation result based on the phase compensation factor. Single-channel two-dimensional range images are output and pre-stored.

[0012] S4. Window and overlay multiple single-channel two-dimensional range images to output a two-dimensional SAR image in real time.

[0013] Furthermore, step S1 is performed as follows:

[0014] S11. Perform Hilbert transform on the echo signal sequence received by the analog-to-digital converter to realize the conversion of the signal from the real domain to the complex domain, so as to extract the amplitude and phase information of the signal in the frequency domain analysis, and provide a basis for subsequent frequency shifting and Fourier transform.

[0015] S12. The echo signal sequence after Hilbert transform is frequency shifted to the intermediate frequency band. The calculation formula is as follows:

[0016] in, Let be the intermediate frequency echo signal sequence of the k-th antenna. Let K be the echo signal sequence of the k-th antenna, k=1,2,…,K. It is the intermediate frequency. This represents the sampling frequency, where n is the number of sampling points. This is an imaginary number. This step simplifies the subsequent spectrum analysis process, reduces high-frequency noise interference, and improves the accuracy and efficiency of frequency domain processing;

[0017] S13. Perform a Fast Fourier Transform on the frequency-shifted echo signal sequence to output the spectrum of the echo signal sequence. This step converts the echo signal from the time domain to the frequency domain, providing necessary spectral information for subsequent imaging and signal processing;

[0018] Furthermore, step S2 is as follows:

[0019] S21. Perform Hilbert transform, frequency shifting, windowing, and fast Fourier transform on the transmitted signal sequence, and then store the spectrum sequence of the transmitted signal in random access memory. This step avoids repetitive calculations and redundant processing, effectively reducing computational resource consumption and lowering the system's data processing latency;

[0020] S22. Perform a complex conjugate dot product of the spectral sequence of the transmitted signal and the spectral sequence of the echo signal to achieve matched filtering. The calculation formula is as follows:

[0021]

[0022] in, and The real and imaginary parts of the transmitted signal spectrum sequence, and Let the real and imaginary parts of the echo signal spectrum sequence be represented. This is an imaginary number. This step fully utilizes the correlation in the frequency domain to accurately extract the features of the echo signal, improving the accuracy and stability of target detection;

[0023] S23. Perform an inverse fast Fourier transform on the frequency-domain pulse compression data obtained by matched filtering, and simultaneously calculate the echo index. Output the time-domain pulse compression result and echo index and store them in random access memory. This step converts the frequency-domain pulse compression data to the time domain for processing and adopts a time-domain back-projection imaging algorithm. Compared with back-projection imaging in the frequency domain, this significantly reduces computational complexity and latency, and improves the real-time processing capability of the system.

[0024] Furthermore, step S3 is as follows:

[0025] S31. Calculate the imaging region coordinates based on the imaging region size and grid size; then, based on the antenna position and the imaging region coordinates, calculate the grid delay from each coordinate point to different antennas, using the following formula:

[0026]

[0027] in, Let i be the grid delay from the i-th row and j-th column grid point in the imaging region to the k-th antenna. and Let x and y be the x-coordinates of the grid point in the i-th row and j-th column of the imaging region, i = 1, 2, ..., I, j = 1, 2, ..., J, k = 1, 2, ..., K. and Let x and y be the x and y coordinates of the k-th antenna;

[0028] S32. Based on the grid delay of the coordinate points on the imaging region, read the adjacent echo index and corresponding pulse compression result closest to the grid delay from the random access memory, and then calculate the signal intensity of the corresponding pixel in the image through two-dimensional interpolation. The calculation formula is as follows:

[0029]

[0030] For the k-th antenna, The signal intensity corresponding to the grid point in the i-th row and j-th column of the imaging region. The grid delay is the grid delay corresponding to the grid point in the i-th row and j-th column of the imaging region. and For the nearest neighboring echo with the closest grid delay, and The echo index corresponding to the k-th antenna and The pulse compression results are obtained. This step accurately obtains the delay relationship between each grid point in the imaging area and the antenna, providing accurate delay information for subsequent signal processing;

[0031] S33. Calculate the phase compensation factor based on the grid delay of the coordinate points in the imaging area. The calculation formula is as follows:

[0032]

[0033] For the k-th antenna, It is the phase compensation factor corresponding to the grid point in the i-th row and j-th column of the imaging region. It is the carrier frequency. It is the intermediate frequency. This represents the grid delay corresponding to the grid point in the i-th row and j-th column of the imaging region. This step uses a two-dimensional interpolation algorithm to accurately estimate the signal strength of each grid point in the imaging region, ensuring a smooth transition and high accuracy in imaging quality, and avoiding distortion or blurring caused by data discreteness.

[0034] The pixel signal strength output in step S32 is adjusted using a phase compensation factor. Perform phase compensation and output the two-dimensional range image of the k-th antenna.

[0035] Furthermore, in step S4, a Taylor window function of length K is set for each single channel corresponding to the K antennas. The two-dimensional range image output in step S3 is then windowed and accumulated to obtain a high-precision two-dimensional SAR image. This step effectively reduces the influence of sidelobes and the interference of spurious signals by windowing the signal of each antenna through Taylor windowing. By accumulating the signals from multiple channels, the angular resolution is effectively enhanced, avoiding the resolution deficiency problem that may be caused by a single antenna signal, making the final two-dimensional SAR image more detailed and realistic.

[0036] Furthermore, the echo signal sequences of each antenna are input sequentially; for each antenna's echo signal sequence, steps S1 to S3 are executed cyclically; in step S4, the current antenna's two-dimensional range image is multiplied by the windowing coefficient, then added to the historical two-dimensional range image, and the result is stored in random access memory; after a frame of image data is determined, a two-dimensional SAR image is output. The above steps, by serializing the processing of single-channel signals and processing only single-channel data at a time, significantly reduce the system's demand for computing resources and storage space; the cyclic processing method ensures efficient utilization of large-scale antenna data while avoiding the hardware resource pressure caused by high parallel processing, thus effectively optimizing computational efficiency.

[0037] In a second aspect, the present invention provides a two-dimensional real-time back-projection sensing imaging device for performing the above-described two-dimensional real-time back-projection sensing imaging method, the two-dimensional real-time back-projection sensing imaging device comprising:

[0038] The echo signal processing module is used to perform Hilbert transform, frequency shifting, and fast Fourier transform on the echo signal received by the analog-to-digital converter in the multi-antenna system, and output the echo signal spectrum.

[0039] The signal matched filtering module is used to perform matched filtering on the transmitted signal and the echo signal, and output the time-domain pulse compression result;

[0040] The phase compensation module is used to take a rectangular plane with a length of L meters and a width of W meters as the imaging area. This imaging area is uniformly divided into I×J grid cells, where I and J represent the number of grid cells in the length and width directions, respectively. The size of each grid cell is determined by L / I and W / J, corresponding to the cell size in the length and width directions, respectively. The imaging area coordinates are calculated based on the imaging area and grid size. The corresponding pixel signal intensity of the image is calculated by two-dimensional interpolation based on the pulse compression result. Then, the phase compensation factor is calculated using the imaging area coordinates. The phase compensation factor is used to perform phase compensation on the interpolation result, and single-channel two-dimensional range images are output and pre-stored.

[0041] The 2D SAR image output module is used to window and overlay multiple single-channel 2D range images to output a 2D SAR image in real time.

[0042] Thirdly, the present invention provides a computer device including a processor and a memory for storing a processor-executable program, wherein when the processor executes the program stored in the memory, it implements the above-described two-dimensional real-time back-projection sensing imaging method.

[0043] Fourthly, the present invention provides a storage medium storing a program, which, when executed by a processor, implements the above-described two-dimensional real-time back-projection sensing imaging method.

[0044] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0045] 1. This invention can process multi-antenna echo sequences and introduces windowing processing in pulse compression and coherent accumulation operations, which effectively suppresses sidelobe interference during the imaging process while improving angular resolution.

[0046] 2. This invention can simultaneously perform grid delay calculation, phase compensation factor calculation, two-dimensional interpolation, and phase compensation operations. The phase compensation factor calculation and two-dimensional interpolation operations can be executed in parallel, and each operation employs a pipelined structure, significantly accelerating imaging speed. Through parallel design and pipelined mechanisms, data processing efficiency is effectively improved.

[0047] 3. This invention employs a coherent accumulation-based data processing algorithm to transform serial multi-antenna data into a parallel processing format. It requires only one RF transceiver link and on-chip storage resources to complete the rapid processing of multi-channel data, effectively reducing memory access latency and resource consumption, lowering hardware costs, and significantly improving imaging speed. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart of the two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system in Embodiment 1 of the present invention;

[0050] Figure 2 This is a block diagram of the parallel pipelined processing in Embodiment 1 of the present invention;

[0051] Figure 3 This is a Python simulation result diagram of the back-projection sensing imaging in Embodiment 1 of the present invention;

[0052] Figure 4This is a diagram showing the back-projection sensing imaging result implemented by FPGA in Embodiment 1 of the present invention;

[0053] Figure 5 This is a flowchart of the two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system implemented under limited resource conditions in Embodiment 2 of the present invention;

[0054] Figure 6 This is a block diagram of parallel pipelined processing under limited resource conditions in Embodiment 2 of the present invention;

[0055] Figure 7 This is a structural block diagram of the two-dimensional real-time back-projection sensing imaging device based on a multi-antenna system in Embodiment 3 of the present invention;

[0056] Figure 8 This is a structural block diagram of the computer device in Embodiment 4 of the present invention. Detailed Implementation

[0057] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort are within the scope of protection of the present application.

[0058] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0059] Example 1

[0060] like Figure 1As shown, this embodiment provides a two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system. The multi-antenna system includes K antennas, K single channels, one analog-to-digital converter (ADC), and one two-dimensional back-projection sensing imaging module. Each antenna corresponds to one single channel. Each antenna receives electromagnetic wave signals from different directions. The single channel corresponding to the antenna processes the received radio frequency (RF) signals, including signal amplification, frequency conversion, and filtering, converting the high-frequency RF signals into intermediate frequency (IF) signals. The ADC converts the analog IF signals output from all single channels into digital signals. The two-dimensional back-projection sensing imaging module uses the multiple digital signals output from the ADC to reconstruct a two-dimensional image of the target using the two-dimensional real-time back-projection sensing imaging method, including the following steps:

[0061] S0. Determine the system parameters of the back-projection imaging method module. For multi-antenna systems, parameters need to be determined, including the operating clock, input and output data types and lengths, sampling frequency, intermediate frequency, carrier frequency, receiving antenna coordinates, number of antennas, and transmitted signal sequence type and length. This specifically includes the following steps:

[0062] S01. Set the operating clock to 100MHz and the sampling frequency... The intermediate frequency is 2.4576 GHz. The carrier frequency is 614.4MHz. It is 7.425 GHz with a signal bandwidth of 400 MHz;

[0063] S02. The echo signal output from the ADC module is the input to the back-projection imaging method. The input to this module is a real number sequence with a length of 4096 points and a data type of 16-bit signed fixed-point numbers; the transmitted signal sequence is a real number sequence with a length of 2472 points and a data type of 16-bit signed fixed-point numbers. In this embodiment, the imaging area has a length L of 4 meters and a width W of 4 meters. This area is uniformly divided into I×J grid cells through grid partitioning, where the number of grid cells I in the length direction is 300 and the number of grid cells J in the width direction is 300. The output SAR image size is... The module outputs data with a length of 90,000 points and a data type of 32-bit signed floating-point number.

[0064] S1. Perform Hilbert transform, frequency shifting, and fast Fourier transform on the echo signal received by the analog-to-digital converter to output the echo signal spectrum. This includes the following steps:

[0065] S11. Input the echo signal sequence output by the ADC module. The data type of this sequence is a 16-bit signed fixed-point number with a data length of 4096.

[0066] S12. Zero-padded to the echo signal sequence in step S11, 4096 zero-value data points are added to the end of the sequence to extend the sequence length to 8192 points.

[0067] S13. In MATLAB, use FDATool to design Hilbert filter coefficients. Set the response type to Hilbert, the filter order to 32, the sampling frequency to 2.4576 GHz, and the filter passband range to 61.44 MHz to 1167.36 MHz. This filter can handle signals with a frequency range of 414.4 MHz to 814.4 MHz in this embodiment. Use FDATool to automatically generate 32nd-order Hilbert filter coefficients and convert them into 16-bit signed fixed-point numbers, then output them to a COE file.

[0068] S14. Call the FIR IP core in Xilinx IP Catalog. Import the filter coefficients of the IP core using the coe file from step S13. Set the filter type to Hilbert, the clock frequency to 100MHz, the input data type to 16-bit signed fixed-point numbers, the quantization type of the filter coefficients to Maximize Dynamic Range, and the data type of the real and imaginary parts of the output data to 16-bit signed integers.

[0069] S15. Input the echo signal sequence that was zero-filled in step S12 into the FIR IP core to complete the Hilbert transform and output a complex sequence of 16-bit signed integers.

[0070] S16. Call the Complex Multiplier IP core in the Xilinx IP Catalog, input the complex sequence of the echo signal and the frequency shift coefficient sequence from step S25, and calculate the following formula:

[0071] in, Let be the intermediate frequency echo signal sequence of the k-th antenna. Let K be the echo signal sequence of the k-th antenna, k=1,2,…,K. It is the intermediate frequency. This represents the sampling frequency, where n is the number of sampling points. The input is an imaginary number, and the output data type is a 32-bit signed integer, which is a complex number sequence of length 8192.

[0072] S17. Call the FFT IP core in Xilinx IP Catalog. Set the IP core working mode to Pipelined, Streaming I / O data stream mode, the processing length to 8192 points, the input data type to 32-bit signed fixed-point complex number, and the output data type to 32-bit signed fixed-point complex number.

[0073] S18. Input the complex sequence from step S16 into the FFT IP core. The IP core is configured to perform FFT, complete the Fast Fourier Transform, and output an echo signal spectrum sequence with a length of 8192 points and a data type of 32-bit signed fixed-point complex number.

[0074] S2. Perform matched filtering on the transmitted signal and the echo signal to output the time-domain pulse compression result, which includes the following steps:

[0075] S21. In this embodiment, the transmitted signal sequence is a frequency-modulated continuous wave (FMCW) with a bandwidth of 400MHz. A Hilbert filter is designed using MATLAB's FDATool, and a Hilbert transform is performed on the transmitted signal. MATLAB is then used to shift the frequency of the transmitted signal. The specific calculation formula is as follows:

[0076]

[0077] in, For transmitting signal sequence, It is the intermediate frequency. This represents the sampling frequency, where n is the number of sampling points. The value is an imaginary number. The Taylorwin function in MATLAB is called to window the transmitted signal sequence after frequency shifting. The FFT function in MATLAB is called to perform a Fast Fourier Transform on the windowed transmitted signal sequence. The output is a complex sequence with a length of 8192 points and a data type of 32-bit signed fixed-point number. This sequence represents the spectrum of the transmitted signal sequence.

[0078] S22. Call the Complex Multiplier IP core in the Xilinx IP Catalog, negative the imaginary part of the transmitted signal spectrum from step S31 and input it into the IP core, then input the echo signal spectrum sequence from step S18 into the IP core, and perform a complex conjugate dot product of the transmitted signal spectrum sequence and the echo signal spectrum sequence to achieve matched filtering. The calculation formula is as follows:

[0079]

[0080] in, and The real and imaginary parts of the transmitted signal spectrum sequence, and Let the real and imaginary parts of the echo signal spectrum sequence be represented. If the value is imaginary, the floating-point IP core in the Xilinx IP Catalog is called to convert the data type of the pulse compression result to a 32-bit signed floating-point number.

[0081] S23. Calculate the echo index in MATLAB using the following formula:

[0082]

[0083] Where N is the length of the echo signal spectrum sequence (8192), and n is the number of sampling points. It is the sampling frequency; then the sequence is converted into a 16-bit signed fixed-point number and converted into a COE file; in the FPGA, the BlockMemory Generator IP core in the Xilinx IP Catalog is called, the memory type is set to single-port ROM, and the echo index sequence is stored in the single-port ROM;

[0084] S24. Input the complex sequence from step S22 into the FFT IP core. The IP core is configured to perform IFFT and output a 32-bit signed floating-point number to complete the inverse fast Fourier transform. Call the Block Memory Generator IP core in the Xilinx IP Catalog and set the memory type to single-port RAM to store the pulse compression result into the single-port RAM.

[0085] S3. Calculate the imaging region grid coordinate parameters based on the imaging region and grid size. Calculate the imaging region grid delay and phase compensation coefficients based on the imaging region grid coordinate parameters. Perform two-dimensional interpolation on the imaging region based on the pulse compression results and compensate for the phase. Figure 2 As shown, step S3 includes mesh delay calculation, phase compensation factor calculation, two-dimensional interpolation, and phase compensation. In this embodiment, step S3 is processed in parallel pipelined mode, specifically including the following steps:

[0086] S31 In this embodiment, the receiving antenna is a uniform linear array antenna (ULA), with a number of antennas K of 64. The geometric center of the antenna array is used as the origin of the coordinate system, and the arrangement direction of the antenna array is defined as the Y-axis direction; the imaging area size is 4m. 4m, SAR image resolution is ;

[0087] S32. MATLAB is used to calculate the X and Y parameters of the imaging area coordinates. In this embodiment, the arrangement direction of the antenna array is defined as the Y-axis direction. For the k-th antenna, k=1,2,…,64, the X parameter calculation formula is:

[0088]

[0089] Where i and j are the row and column numbers corresponding to the grid points in the imaging region, i=1,2,…,300, j=1,2,…,300; the formula for calculating the Y parameter is:

[0090]

[0091] Where i and j are the row and column numbers corresponding to the grid points in the imaging region, i=1,2,…,300, j=1,2,…,300. It is the y-coordinate of the k-th antenna element, k=1,2,…,64;

[0092] Convert the X and Y parameters to 32-bit signed floating-point numbers; to facilitate parallel pipelined processing, such as... Figure 2 As shown, the X parameter is stored in the Flip-Flop resource in the FPGA array structure with a data length of 300; the Y parameter is stored in the single-port ROM with a data length of 19200.

[0093] S33. Perform 8 parallel grid delay calculation operations. For the k-th antenna, k=1,2,…,64, calculate the grid delay between each grid point and the receiving antenna element. The specific implementation steps of this operation are as follows:

[0094] S331. For the k-th antenna, k=1,2,…,64, retrieve the parameters corresponding to the grid in the i-th row and j-th column from the X-parameter array and the Y-parameter RAM. and , i=1,2,…,300, j=1,2,…,300;

[0095] S332. Calculate the mesh delay using the following formula:

[0096]

[0097] Where k is the antenna number, k=1,2,…,64, i and j are the row and column numbers corresponding to the grid points in the imaging region, and c is the speed of light, with a value of ;

[0098] S34. Perform 8 parallel two-dimensional interpolation operations and phase compensation factor calculation operations. The two-dimensional interpolation operation can calculate the signal intensity of the corresponding pixel in the image. The specific implementation steps of this operation are as follows:

[0099] Based on the grid delay in step S33, the adjacent echo index closest to the grid delay and the corresponding pulse compression result are read from RAM. Then, the signal intensity of the corresponding pixel in the image is calculated by two-dimensional interpolation. The calculation formula is as follows:

[0100]

[0101] For the k-th antenna, The signal intensity corresponding to the grid point in the i-th row and j-th column of the imaging region. The grid delay is the grid delay corresponding to the grid point in the i-th row and j-th column of the imaging region. and For the nearest neighboring echo with the closest grid delay, and The echo index corresponding to the k-th antenna and The result of pulse compression;

[0102] The phase compensation factor calculation operation is based on the grid delay of the coordinate points on the imaging area. The calculation formula is as follows:

[0103] For the k-th antenna, It is the phase compensation factor corresponding to the grid point in the i-th row and j-th column of the imaging region. It is the carrier frequency. It is the intermediate frequency. It is the grid delay corresponding to the grid point in the i-th row and j-th column of the imaging region. It is an imaginary number;

[0104] S35. Perform 8 phase compensation operations in parallel. Use the compensation parameters calculated by the phase compensation factor calculation operation to perform phase compensation on the pixel signal intensity output by the two-dimensional interpolation module, and output the two-dimensional range image of the k-th antenna. The data length is 90000 and the data type is 32-bit signed floating-point number.

[0105] S4. In the FPGA, call the Block Memory Generator IP core in the Xilinx IP Catalog, set the memory type to dual-port RAM, set the RAM width to 64 bits, and the depth to 90000, to temporarily store the windowed two-dimensional range image; for K antennas, set a Taylor window function of length K, and accumulate the windowed two-dimensional range image output in step S3 to obtain a high-precision two-dimensional SAR image.

[0106] The following are the test environment and specific parameter settings for this embodiment:

[0107] Indoor division The imaging area is a corner reflector placed in the center of the imaging area. The receiving antenna in this embodiment is a uniform linear array antenna. The geometric center of the antenna array is taken as the origin of the coordinate system. The arrangement direction of the antenna array is defined as the Y-axis direction, the antenna spacing is half a wavelength, and the direction perpendicular to the arrangement direction of the antenna array is defined as the X-axis direction. The transmitting antenna is placed at the origin of the imaging area coordinate system.

[0108] The specific parameter is set as follows: sampling frequency =2.4576GHz, intermediate frequency =614.4MHz, carrier frequency =7.425GHz, signal bandwidth B=400MHz, number of antennas ;

[0109] A two-dimensional back-projection imaging module is deployed on an FPGA. It transmits a 400MHz bandwidth FMCW signal via a transmitting antenna, with a signal sequence length of 2472 points. Simultaneously, the antenna and corresponding single channel are switched sequentially, and the 64-channel echo signal is input serially to the two-dimensional back-projection imaging module. After generating two-dimensional SAR image data, the module transmits it to a PC via Ethernet for reconstructing a two-dimensional environmental image.

[0110] Figure 3 and Figure 4 The Python simulation results and FPGA implementation results of the two-dimensional back projection imaging method are presented respectively. Figure 3 To acquire echo signals from a multi-antenna system, a two-dimensional back-projection imaging algorithm was implemented using Python, and the results were obtained through simulation testing. Figure 4 The result is obtained by executing the two-dimensional back projection imaging method on the FPGA according to the specific steps of the invention embodiment.

[0111] A correct imaging result should show that the energy of the SAR image is concentrated at the corner reflector in the center of the imaging area. Figure 3 In the simulation, Python results show that energy is accurately focused onto the corner reflector at the center of the imaging region, indicating high imaging accuracy and precise positioning in the simulation test; Figure 4 The results implemented on the FPGA also demonstrate energy focusing, consistent with theoretical expectations.

[0112] By comparison Figure 3 and Figure 4It can be observed that both imaging methods concentrate energy at the corner reflector in the center of the imaging area, indicating that the FPGA-implemented two-dimensional back-projection imaging method meets the theoretical requirements in terms of accuracy and positioning. Furthermore, this method not only accurately perceives and reconstructs the environment within the imaging area but also meets the requirements of real-time imaging, further verifying its reliability and practicality in real-world applications. Example 2

[0113] like Figure 5 As shown, this embodiment provides a two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system under resource-constrained conditions. Compared to Embodiment 1, this embodiment optimizes the design of the grid delay parameters, eliminating the need to calculate the nearest adjacent echo index and perform division operations during two-dimensional interpolation. Furthermore, it eliminates the need for a single-port ROM to store the echo index data. Compared to Embodiment 1, this method saves the hardware resources required for eight 32-bit signed floating-point subtractors, eight 32-bit signed floating-point dividers, and a single-port ROM with a length of 8192 and a depth of 32 bits. This embodiment specifically includes the following steps:

[0114] S0. Determine the system parameters of the back projection imaging method module. The system parameters include the module running clock, input and output data types and lengths, sampling frequency, intermediate frequency, carrier frequency, receiving antenna coordinates, number of antennas, and type and length of transmitted signal sequence, etc. For specific implementation, refer to step S0 of Embodiment 1.

[0115] S1. Perform Hilbert transform, frequency shifting and fast Fourier transform on the echo signal received by the analog-to-digital converter to output the echo signal spectrum. For specific implementation steps, refer to step S1 of Example 1.

[0116] S2. Perform matched filtering on the transmitted signal and the echo signal to output the time-domain pulse compression result. The specific implementation steps are as follows:

[0117] S21. In this embodiment, the transmitted signal sequence is a frequency-modulated continuous wave (FMCW) with a bandwidth of 400MHz. A Hilbert filter is designed using MATLAB's FDATool, and a Hilbert transform is performed on the transmitted signal. MATLAB is then used to shift the frequency of the transmitted signal. The specific calculation formula is as follows:

[0118] in, For transmitting signal sequence, It is the intermediate frequency. This represents the sampling frequency, where n is the number of sampling points. The value is an imaginary number. The Taylorwin function in MATLAB is called to window the transmitted signal sequence after frequency shifting. The FFT function in MATLAB is called to perform a Fast Fourier Transform on the windowed transmitted signal sequence. The output is a complex sequence with a length of 8192 points and a data type of 32-bit signed fixed-point number. This sequence represents the spectrum of the transmitted signal sequence.

[0119] S22. Call the Complex Multiplier IP core in the Xilinx IP Catalog, negative the imaginary part of the transmitted signal spectrum from step S31) and input it into the IP core. Then input the echo signal spectrum sequence from step S18 into the IP core. Perform a complex conjugate dot product on the transmitted signal spectrum sequence and the echo signal spectrum sequence to achieve matched filtering. The calculation formula is as follows:

[0120]

[0121] in, and The real and imaginary parts of the transmitted signal spectrum sequence, and Let the real and imaginary parts of the echo signal spectrum sequence be represented. If the value is imaginary, the floating-point IP core in the Xilinx IP Catalog is called to convert the data type of the pulse compression result to a 32-bit signed floating-point number.

[0122] S3. Calculate the imaging region grid coordinate parameters based on the imaging region and grid size. Then, design imaging region grid delay and phase compensation coefficients suitable for resource-constrained conditions based on the imaging region grid coordinate parameters. Finally, perform two-dimensional interpolation and phase compensation on the imaging region based on the pulse compression results, and output a single-channel two-dimensional range image. Figure 6 The diagram shown illustrates the parallel pipelined processing of step S3 in this embodiment. This step includes mesh delay calculation, phase compensation factor calculation, two-dimensional interpolation, and phase compensation operations, specifically including the following steps:

[0123] S31, refer to step S31 of Example 1;

[0124] S32. Use MATLAB to calculate the X and Y parameters of the imaging area coordinates. In this example, the arrangement direction of the antenna array is defined as the Y-axis direction. For the k-th antenna, k=1,2,…,64, the X parameter calculation formula is:

[0125]

[0126] Where i and j are the row and column numbers corresponding to the grid points in the imaging region, i=1,2,…,300, j=1,2,…,300; the formula for calculating the Y parameter is:

[0127]

[0128] Where i and j are the row and column numbers corresponding to the grid points in the imaging region, i=1,2,…,300, j=1,2,…,300. It is the y-coordinate of the k-th antenna element, k=1,2,…,64;

[0129] S33. Based on the X and Y parameters obtained in step S32, design the imaging region grid delay parameters DX and DY suitable for resource-constrained conditions. For the k-th antenna, k=1,2,…,64, the calculation formula for DX is as follows:

[0130]

[0131] Where DX is the abscissa parameter of the imaging region grid delay applicable under resource-constrained conditions, and X is the abscissa parameter of the imaging region grid coordinates obtained in step S32. It is the sampling frequency;

[0132] The formula for calculating DY is as follows:

[0133]

[0134] Where DY is the ordinate parameter of the imaging region grid delay applicable under resource-constrained conditions, and Y is the ordinate parameter of the imaging region grid coordinates obtained in step S32. It is the sampling frequency;

[0135] Combine DX parameters and D The parameters are converted to 32-bit signed floating-point numbers; this is to facilitate parallel pipelined processing, such as... Figure 6 As shown, the DX parameter is stored in the Flip-Flop resource in the FPGA array structure with a data length of 300; the DY parameter is stored in the single-port ROM with a data length of 19200.

[0136] S34. Perform 8 parallel grid delay calculation operations. For the k-th antenna, k=1,2,…,64, calculate the grid delay between each grid point and the receiving antenna element under resource constraints. The specific implementation steps of this operation are as follows:

[0137] S341. For the k-th antenna, k=1,2,…,64, retrieve the parameters corresponding to the grid in the i-th row and j-th column from the DX parameter array and the DY parameter RAM. and , i=1,2,…,300, j=1,2,…,300;

[0138] S342. Calculate the grid delay under constrained resource conditions. The calculation formula is as follows:

[0139]

[0140] Where k is the antenna number, k=1,2,…,64, and i and j are the row and column numbers corresponding to the grid points in the imaging region, i=1,2,…,300, j=1,2,…,300. This refers to the grid latency under resource-constrained conditions, where c is the speed of light, and the value is... ;

[0141] S35. Perform 8 parallel two-dimensional interpolation operations and phase compensation factor calculation operations. The two-dimensional interpolation operation can calculate the signal intensity of the corresponding pixel in the image. The specific implementation steps of this operation are as follows:

[0142] Based on the grid delay under constrained resource conditions in step S41 The formula for calculating the echo index is as follows:

[0143]

[0144] The pulse compression results corresponding to addresses m and m+1 are retrieved from the RAM storing the pulse compression sequence. Then, the signal intensity of the corresponding pixel in the image is calculated by two-dimensional interpolation under limited resource conditions. The calculation formula is as follows:

[0145]

[0146] For the k-th antenna, The signal intensity corresponding to the grid point in the i-th row and j-th column of the imaging region. Let k be the grid delay corresponding to the grid point in the i-th row and j-th column of the imaging region, where i = 1, 2, ..., 300, j = 1, 2, ..., 300, and k = 1, 2, ..., 64. and For the k-th antenna, store the pulse compression results at addresses m and m+1 in the RAM that contain the pulse compression sequence;

[0147] S36. Grid delay under constrained resource conditions Calculate the phase compensation factor using the following formula:

[0148]

[0149] For the k-th antenna, It is the phase compensation factor corresponding to the grid point in the i-th row and j-th column of the imaging region, where i=1,2,…,300, j=1,2,…,300, k=1,2,…,64. It is the carrier frequency. It is the intermediate frequency. It is the sampling frequency. It is the grid delay corresponding to the grid point in the i-th row and j-th column of the imaging region under limited resources. It is an imaginary number;

[0150] S37. Perform 8 phase compensation operations in parallel. Use the compensation parameters calculated by the phase compensation factor calculation operation to perform phase compensation on the pixel signal intensity output by the two-dimensional interpolation module, and output the two-dimensional range image of the k-th antenna. The data length is 90000 and the data type is 32-bit signed floating-point number.

[0151] S4. In the FPGA, call the Block Memory Generator IP core in the Xilinx IP Catalog, set the memory type to dual-port RAM, set the RAM width to 64 bits, and the depth to 90000, to temporarily store the windowed two-dimensional range image; for K antennas, set a Taylor window function of length K, and accumulate the windowed two-dimensional range image output in step S3) to obtain a high-precision two-dimensional SAR image.

[0152] Example 4

[0153] Reference Figure 7 This embodiment provides a two-dimensional real-time back-projection sensing imaging device based on a multi-antenna system. The device includes, in sequence, an echo signal processing module 701, a signal matched filtering module 702, a phase compensation module 703, and a two-dimensional SAR image output module 704, wherein:

[0154] The echo signal processing module 701 is used to perform Hilbert transform, frequency shifting and fast Fourier transform on the echo signal received by the analog-to-digital converter in the multi-antenna system, and output the echo signal spectrum.

[0155] The signal matching filter module 702 is used to perform matching filtering on the transmitted signal and the echo signal, and output the time-domain pulse compression result.

[0156] The phase compensation module 703 is used to take a rectangular plane with a length of L meters and a width of W meters as the imaging area. The imaging area is uniformly divided into I×J grid cells by a grid, where I and J represent the number of grid cells in the length and width directions, respectively. The size of each grid cell is determined by L / I and W / J, which correspond to the cell size in the length and width directions, respectively. The imaging area coordinates are calculated based on the imaging area and grid size. The corresponding pixel signal intensity of the image is calculated by two-dimensional interpolation based on the pulse compression result. Then, the phase compensation factor is calculated using the imaging area coordinates. The phase compensation factor is used to perform phase compensation on the interpolation result. The single-channel two-dimensional range image is output and pre-stored.

[0157] The 2D SAR image output module 704 is used to window and overlay multiple single-channel 2D range images to output a 2D SAR image in real time.

[0158] The content of the above method embodiments is applicable to this system embodiment, and the specific functions implemented in this system embodiment are the same as those in the above method embodiments.

[0159] Example 4

[0160] This embodiment provides a computer device, which can be a computer, such as... Figure 8 As shown, the system is connected via a system bus 801 to a processor 802, a memory, an input device 803, a display 804, and a network interface 805. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium 806 and internal memory 807. The non-volatile storage medium 806 stores the operating system, computer programs, and a database. The internal memory 807 provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. When the processor 802 executes the computer programs stored in the memory, it implements the two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system proposed in Embodiment 1. The two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system includes the following steps:

[0161] S1. Perform Hilbert transform, frequency shifting, and fast Fourier transform on the echo signal received by the analog-to-digital converter in the multi-antenna system, and output the echo signal spectrum.

[0162] S2. Perform matched filtering on the transmitted signal and the echo signal, and output the time-domain pulse compression result;

[0163] S3. A rectangular plane with a length of L meters and a width of W meters is used as the imaging area. This imaging area is uniformly divided into I×J grid units, where I and J represent the number of grid units in the length and width directions, respectively. The size of each grid unit is determined by L / I and W / J, corresponding to the unit size in the length and width directions, respectively. The imaging area coordinates are calculated based on the imaging area and grid size. The corresponding pixel signal intensity of the image is calculated by two-dimensional interpolation based on the pulse compression result. Then, the phase compensation factor is calculated using the imaging area coordinates. Phase compensation is performed on the interpolation result based on the phase compensation factor. Single-channel two-dimensional range images are output and pre-stored.

[0164] S4. Window and overlay multiple single-channel two-dimensional range images to output a two-dimensional SAR image in real time.

[0165] Example 5

[0166] This embodiment provides a storage medium, which is a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system proposed in Embodiment 1 above. The two-dimensional real-time back-projection sensing imaging method based on a multi-antenna system includes the following steps:

[0167] S1. Perform Hilbert transform, frequency shifting, and fast Fourier transform on the echo signal received by the analog-to-digital converter in the multi-antenna system, and output the echo signal spectrum.

[0168] S2. Perform matched filtering on the transmitted signal and the echo signal, and output the time-domain pulse compression result;

[0169] S3. A rectangular plane with a length of L meters and a width of W meters is used as the imaging area. This imaging area is uniformly divided into I×J grid units, where I and J represent the number of grid units in the length and width directions, respectively. The size of each grid unit is determined by L / I and W / J, corresponding to the unit size in the length and width directions, respectively. The imaging area coordinates are calculated based on the imaging area and grid size. The corresponding pixel signal intensity of the image is calculated by two-dimensional interpolation based on the pulse compression result. Then, the phase compensation factor is calculated using the imaging area coordinates. Phase compensation is performed on the interpolation result based on the phase compensation factor. Single-channel two-dimensional range images are output and pre-stored.

[0170] S4. Window and overlay multiple single-channel two-dimensional range images to output a two-dimensional SAR image in real time.

[0171] The storage medium described in this embodiment can be a disk, optical disk, computer memory, random access memory (RAM), USB flash drive, portable hard drive, etc.

[0172] It should be noted that, for the sake of simplicity, the aforementioned method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously.

[0173] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0174] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for two-dimensional real-time back-projection perception imaging based on a multi-antenna system, applied to a multi-antenna system, the multi-antenna system comprising K antennas, K single channels, one analog-to-digital converter and one two-dimensional back-projection perception imaging module, each antenna corresponding to one single channel; each antenna is used for receiving electromagnetic wave signals from different directions; the single channel corresponding to the antenna processes the received radio frequency signals, including signal amplification, frequency conversion and filtering, converting high-frequency radio frequency signals into intermediate frequency signals; the analog-to-digital converter converts the analog intermediate frequency signals output by all single channels into digital signals; the two-dimensional back-projection perception imaging module uses the multi-channel digital signals output by the analog-to-digital converter to reconstruct a two-dimensional image of the target through a two-dimensional real-time back-projection perception imaging method, characterized in that, The two-dimensional real-time back-projection perception imaging method comprises the following steps: S1, performing Hilbert transform, frequency shifting and fast Fourier transform on the echo signal received by the analog-to-digital converter in the multi-antenna system to output the echo signal spectrum; S2, performing matched filtering on the transmission signal and the echo signal to output the time-domain pulse compression result; S3, taking a rectangular plane with a length of L meters and a width of W meters as an imaging area, the imaging area is evenly divided into I*J grid units by a grid, I and J represent the number of grids in the length and width directions respectively, the size of each grid unit is determined by L / I and W / J, which respectively correspond to the unit size in the length and width directions, calculating the imaging area coordinates according to the imaging area and the grid size, calculating the image pixel signal intensity according to the pulse compression result through two-dimensional interpolation; then calculating the phase compensation factor by using the imaging area coordinates, performing phase compensation on the interpolation result according to the phase compensation factor, and respectively outputting and pre-storing the single-channel two-dimensional range image; the step S3 process is as follows: S31, calculating the imaging area coordinates according to the imaging area size and the grid size; subsequently, based on the position of the antenna and the coordinates of the imaging area, calculating the grid delay of each coordinate point to different antennas, and the calculation formula is as follows: wherein, is the grid delay from the imaging area's i-th row j-th column grid point to the k-th antenna, and is the horizontal coordinate and vertical coordinate of the imaging area's i-th row j-th column grid point, i = 1, 2, …, I, j = 1, 2, …, J, k = 1, 2, …, K, and is the horizontal coordinate and vertical coordinate of the k-th antenna; S32, reading the adjacent echo index and corresponding pulse compression result closest to the grid delay from the random access memory according to the grid delay of the coordinate point on the imaging area, and then calculating the image corresponding pixel signal intensity through two-dimensional interpolation The calculation formula is as follows: wherein, for the kthantenna, is the signal strength corresponding to the grid point in the i th row and j th column of the imaging region, is the grid delay corresponding to the grid point in the i th row and j th column of the imaging region, and is the index of the nearest neighbor echo to the grid delay, and is the pulse compression result of the kthantenna for the corresponding echo index and . S33, calculating the phase compensation factor according to the grid delay of the coordinate point on the imaging area, and the calculation formula is as follows: wherein, for the kth antenna, is a phase compensation factor corresponding to the grid point of the i th row and the j th column of the imaging area, is a carrier frequency, is an intermediate frequency frequency, is a grid delay corresponding to the grid point of the i th row and the j th column of the imaging area; The pixel signal intensity output in step S32 is multiplied by a phase compensation factor Phase compensation is performed, and a two-dimensional range profile of the kth antenna is output. S4, windowing and superimposing a plurality of single-channel two-dimensional range images to output a two-dimensional SAR image in real time.

2. The method of claim 1, wherein, The step S1 process is as follows: S11, performing Hilbert transform on the echo signal sequence received by the analog-to-digital converter to realize the conversion of the signal from the real number domain to the complex number domain; S12, the echo signal sequence after the Hilbert transform is frequency shifted, and the echo signal frequency is shifted to the intermediate frequency band, and the calculation formula is as follows: wherein is a sequence of intermediate frequency return signals for the kth antenna, is a sequence of return signals for the kth antenna, k = 1, 2, …, K, is an intermediate frequency, is a sampling frequency, n is a sampling point, is an imaginary number; S13, performing fast Fourier transform on the echo signal sequence after frequency shifting to output the spectrum of the echo signal sequence.

3. The method of claim 1, wherein, The step S2 process is as follows: S21, performing Hilbert transform, frequency shifting, windowing and fast Fourier transform on the transmission signal sequence, and then storing the spectrum sequence of the transmission signal into the random storage memory; S22, performing complex conjugate point multiplication on the spectrum sequence of the transmission signal and the spectrum sequence of the echo signal to realize matched filtering, and the calculation formula is as follows: wherein and are the real and imaginary parts of the transmitted signal spectrum sequence, and are the real and imaginary parts of the echo signal spectrum sequence, is an imaginary number; S23, performing inverse fast Fourier transform on the frequency domain pulse compression data obtained by matched filtering, and simultaneously calculating the echo index, outputting the time-domain pulse compression result and the echo index and storing them into the random access memory.

4. The method of claim 1, wherein, In the step S4, a Taylor window function with a length of K is set for the single channel corresponding to K antennas, the two-dimensional range image output in the step S3 is windowed and accumulated to obtain a high-precision two-dimensional SAR image.

5. The method of claim 1, wherein, The echo signal sequence of each antenna is input in sequence; for the echo signal sequence of each antenna, steps S1 to S3 are executed in a loop; In step S4, the two-dimensional range image of the current antenna is multiplied by the windowing coefficient, and then added to the historical two-dimensional range image, and the result is stored in the random access memory; after a frame of image data is completed, a two-dimensional SAR image is output.

6. A two-dimensional real-time back-projection perception imaging system for performing the two-dimensional real-time back-projection perception imaging method based on a multi-antenna system according to any one of claims 1 to 5, characterized in that, The two-dimensional real-time back-projection perception imaging system comprises: An echo signal processing module is configured to perform Hilbert transform, frequency shifting and fast Fourier transform on echo signals received by the analog-to-digital converter in the multi-antenna system, and output echo signal spectrum; A signal matched filter module is configured to perform matched filtering on the transmitted signal and the echo signal, and output time-domain pulse compression results; A phase compensation module is configured to use a rectangular plane with a length of L meters and a width of W meters as an imaging area, divide the imaging area into I×J grid units by uniform grid division, where I and J represent the number of grids in the length and width directions, respectively, the size of each grid unit is determined by L / I and W / J, respectively, corresponding to the size of the unit in the length and width directions, calculate imaging area coordinates according to the imaging area and the grid size, calculate the pixel signal intensity of the image according to the pulse compression results by two-dimensional interpolation, then calculate the phase compensation factor using the imaging area coordinates, perform phase compensation on the interpolation results according to the phase compensation factor, and output single-channel two-dimensional range images and prestore them, respectively; A two-dimensional SAR image output module is configured to perform windowing and superposition on a plurality of single-channel two-dimensional range images, and output two-dimensional SAR images in real time.

7. A computer device comprising a processor and a memory for storing a processor executable program, characterized in that, The processor executes the program stored in the memory, and the two-dimensional real-time back-projection perception imaging method of any one of claims 1 to 5 is implemented.

8. A storage medium storing a program, characterized by comprising: The program is executed by the processor, and the two-dimensional real-time back-projection perception imaging method of any one of claims 1 to 5 is implemented.

Images