A millimeter wave sparse array time domain fast image reconstruction method and system

By using sparse array time domain fast image reconstruction method in millimeter wave personal safety inspection technology, image reconstruction is carried out using equivalent array and symmetry principles, the problem of slow image reconstruction speed is solved, and efficient image processing is achieved, suitable for In the event of large passenger flow, the hardware cost is reduced.

CN114494473BActive Publication Date: 2024-11-22BRAINWARE TERAHERTZ INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210001908.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-01-04
Publication Date
2024-11-22
Estimated Expiration
2042-01-04

AI Technical Summary

Technical Problem

In the existing millimeter wave personal safety inspection technology, the two-dimensional sparse array imaging system has advantages in acquiring echo data speed, but the image reconstruction speed limits the rapid passage of human bodies, and the backward projection reconstruction algorithm is large in computing, resulting in hardware. The problem of increasing costs.

Method used

The time-domain image reconstruction method of millimeter wave sparse array is adopted to realize antenna beam scanning through spatially arranged two-dimensional sparse arrays, divide the imaging areas, calculate the oblique distance using the principle of equivalent arrays and symmetry, and perform the image reconstruction of the four sub-arrays. Processing sub-images in parallel reduces the amount of calculation and improves computing efficiency.

Benefits of technology

Without increasing hardware costs, the calculation speed of image reconstruction is significantly improved, and is suitable for large passenger flow and high-precision millimeter wave personal safety inspection, which improves traffic efficiency and reduces the calculation amount of image reconstruction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114494473B_ABST
    Figure CN114494473B_ABST
Patent Text Reader

Abstract

The present invention discloses a millimeter wave sparse array time domain fast image reconstruction method and system, which belongs to the field of millimeter wave personal safety inspection technology, and includes the following steps: S1: receiving sparse array echo signals; S2: subarray division; S3: imaging area division; S4: calculating skew moments; S5: four-subarray image reconstruction; S6: subimage flipping and sorting; S7: image splicing. The present invention reduces the computational complexity of the traditional back-projection reconstruction algorithm by synthesizing limited echo data, dividing a limited imaging grid area, and sharing a set of matched filter coefficients and interpolation coefficients, and properly balances the imaging quality and computational efficiency. Compared with the traditional back-projection reconstruction algorithm, the computational efficiency has been effectively improved, and it is worthy of being promoted and used.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of millimeter wave personal safety inspection, and in particular to a millimeter wave sparse array time domain fast image reconstruction method and system. Background Art

[0002] With the commercial application of millimeter wave human body security inspection imaging technology, the mainstream security inspection imaging systems are divided into one-dimensional dense array plane scanning system, one-dimensional dense array cylindrical scanning system, two-dimensional phased array imaging system, and two-dimensional sparse array imaging system. The imaging system of one-dimensional dense array scanning has a relatively slow process of acquiring echo data, and requires the human body to remain still for 1 to 2 seconds in a fixed scene, which limits the efficiency of passengers' passage and is only suitable for occasions with small security inspection traffic. The two-dimensional phased array imaging system and the two-dimensional sparse array imaging system have obvious advantages in the speed of acquiring human echo data. It only takes tens of milliseconds to obtain backscattered echo data covering the human body area, so it has the potential for human body rapid passage application. The two-dimensional sparse array imaging system has obvious advantages in the speed of acquiring echo data, so the bottleneck limiting the application of human body rapid passage lies in the speed of image reconstruction.

[0003] Due to the sparse characteristics of spatial distribution, the back-projection reconstruction algorithm is the best choice for sparse array configuration. The basic principle of the back-projection reconstruction algorithm is delayed phase compensation and cumulative calculation, so the computational complexity of this algorithm is large. In order to improve the computational efficiency of the back-projection reconstruction algorithm, either the processing capability of the hardware signal processing platform should be improved, or the computational efficiency of the back-projection reconstruction algorithm should be improved. Improving the processing capability of the hardware signal processing platform means increasing computing resources, control resources, and storage resources, which is accompanied by an increase in hardware costs. How to improve the existing processing capabilities and improve the computational efficiency of the back-projection reconstruction algorithm without increasing the hardware cost is an urgent problem to be solved. To this end, a millimeter-wave sparse array time-domain fast image reconstruction method is proposed. Summary of the invention

[0004] The technical problem to be solved by the present invention is: how to enhance the existing processing capabilities and improve the computational efficiency of the back-projection reconstruction algorithm without increasing the hardware cost, and provide a millimeter-wave sparse array time-domain fast image reconstruction method. This method is aimed at the application field of millimeter-wave personal safety inspection with large passenger flow and high precision. It is a fast millimeter-wave image reconstruction signal processing imaging method under the conditions of large passenger flow and no stop.

[0005] The present invention solves the above technical problems through the following technical solutions, and the present invention comprises the following steps:

[0006] S1: Receive sparse array echo signals

[0007] The spatially arranged two-dimensional sparse array is switched by the antenna array switch to realize the two-dimensional antenna beam scanning in space, covering the imaging target range. A single transmitting antenna is used to transmit electromagnetic wave signals, and all receiving antennas simultaneously receive the backscattered electromagnetic waves, which are down-converted to intermediate frequency signals, and then digitized by the multi-channel digital acquisition module. The digitized signal is demodulated into a digital I / Q signal through digital down-conversion. The obtained target backscattered three-dimensional echo signal is s(x Tx ,y Tx ,x Rx ,y Rx ,k),x Tx is the horizontal dimension of the transmitting array, y Tx is the vertical dimension of the transmitting array, x Rx is the horizontal dimension of the receiving array, y Rx is the vertical dimension of the receiving array, k = 2 × π × fre / c is the frequency scanning dimension, where fre is the frequency of the transmitted signal, fre∈[f min ,f max ];

[0008] S2: Sub-array division

[0009] The sixteen sub-arrays are divided into groups, and each group is composed of four overlapping sub-arrays. The resulting echo signal is s sub4 (x Tx ,y Tx ,x Rx ,y Rx ,k);

[0010] S3: Imaging area division

[0011] The imaging area is divided. According to the equivalent array concept, a single subarray transmitting element and four subarray receiving elements are grouped together in the four-subarray unit to obtain the sampling position coordinates of the equivalent array element. The obtained area is the imaging area division within the four-subarray unit, and a total of four imaging areas are divided.

[0012] S4: Calculate the skew moment

[0013] Calculate the slant range R of the backprojection reconstruction algorithm in the single-subarray transmitting and four-subarray receiving mode sub_Tx , R sub_Rx According to the symmetry principle of array spatial distribution, the slant ranges of the other single-subarray transmitting and four-subarray receiving modules within the four-subarray are consistent, and the matched filter Href(x sub_Tx ,y sub_Tx , x sub_Rx ,y sub_Rx, x′, y′, z′) are also consistent. According to the principle of relative coordinate system, the grid coordinate values ​​of the divided imaging sub-areas are consistent with the vertex of the transmitting single sub-array as the origin, and then the three-dimensional complex image σ of the back-projection reconstruction algorithm of the four imaging sub-areas is calculated. sub (x',y',z');

[0014] S5: Four-subarray image reconstruction

[0015] According to the principle of symmetry and relative coordinate system, the three-dimensional images of the four imaging sub-areas are symmetrically flipped and sorted, and the overlapping areas are accumulated to obtain the four-subarray module sub-images.

[0016] S6: Subgraph flipping and sorting

[0017] Repeat steps S3 to S5 until the sub-image calculations of the imaging areas of all four sub-array modules are completed;

[0018] S7: Image stitching

[0019] All three-dimensional complex sub-images are spliced ​​to obtain the final three-dimensional complex image, and then the maximum value projection of the distance dimension is performed to obtain a two-dimensional projection image of two azimuth dimensions.

[0020] Furthermore, in step S1, the two-dimensional sparse array is composed of sixteen equally spaced U-shaped sub-arrays, the frequency range of the signal emitted by the imaging system is 30 to 40 GHz, the number of transmitting array elements in each U-shaped sub-array is 46, and the spacing is Δx T =0.0068m, the number of receiving elements in the subarray is 46, and the spacing is Δy R =0.0068m.

[0021] Furthermore, when the imaging system is working, under the triggering of the timing generator of the entire system, the transmitting array elements are turned on in time division, and the receiving array elements are always turned on and receive the target backscattered echo signals at the same time.

[0022] Furthermore, in step S2, the position coordinate of the equivalent array element is x c =(x Tx +x Rx ) / 2,y c =(y Tx +y Rx ) / 2,z c =0, the sub-areas for internal imaging of the four sub-arrays are divided according to the coverage area of ​​the equivalent array element position coordinates.

[0023] Furthermore, in step S4, the slant range expression of the single-subarray transmitting and four-subarray receiving mode is calculated as:

[0024] Furthermore, in step S4, the matched filter calculation expression is:

[0025] Href(x sub_Tx ,y sub_Tx ,x sub_Rx ,y sub_Rx ,x',y',z')=exp(jk min (R sub_Tx +R sub_Rx ))

[0026] Among them, k min =2×π×f min / c,f min is the minimum frequency of the transmitted signal, and c is the speed of electromagnetic waves in a vacuum.

[0027] Furthermore, the obtained three-dimensional complex image of the imaging sub-area is:

[0028]

[0029] in, Inverse Fast Fourier Transform in frequency scan dimension k, It is the spatial sampling interval after IFFT upsampling in the frequency scan dimension.

[0030] The present invention also provides a millimeter wave sparse array time domain fast image reconstruction system, which uses the above reconstruction method to reconstruct an image, including:

[0031] The echo signal receiving module is used to transmit electromagnetic wave signals through a single transmitting antenna. All receiving antennas receive backscattered electromagnetic waves at the same time, down-convert them to intermediate frequency signals, and then digitize the analog signals through the multi-channel digital acquisition module. The digitized signals are demodulated into digital I / Q signals through digital down-conversion. The obtained target backscattered three-dimensional echo signal is s(x Tx ,y Tx ,x Rx ,y Rx ,k);

[0032] The subarray division module is used to divide the sixteen subarrays into groups, with four subarrays overlapping as one group. The resulting echo signal is s sub4 (x Tx ,y Tx ,x Rx ,y Rx ,k);

[0033] The imaging area division module is used to divide the imaging area. According to the equivalent array concept, a single subarray transmitting element and four subarray receiving elements are grouped together in the four-subarray unit to obtain the sampling position coordinates of the equivalent array element. The obtained area is the imaging area division within the four-subarray unit, and a total of four imaging areas are divided.

[0034] Slant moment calculation module, used to calculate the slant distance R of the back-projection reconstruction algorithm in the single-subarray transmitting and four-subarray receiving mode sub_Tx , R sub_Rx , and calculate the three-dimensional complex image σ of the back-projection reconstruction algorithm of the four imaging sub-areas sub (x',y',z');

[0035] The image reconstruction module is used to symmetrically flip and sort the three-dimensional images of the four imaging sub-areas according to the principle of symmetry and relative coordinate system, and accumulate the overlapping areas to obtain the four-subarray module sub-images.

[0036] A flipping and sorting module is used to repeat steps S3 to S5 until the sub-image calculations of the imaging areas of all four sub-array modules are completed;

[0037] An image stitching module is used to stitch all the three-dimensional complex sub-images to obtain a final three-dimensional complex image, and then perform maximum projection of the distance dimension to obtain a two-dimensional projection image of two azimuth dimensions;

[0038] The control processing module is used to issue instructions to each module to complete the corresponding actions;

[0039] The echo signal receiving module, the sub-array division module, the imaging area division module, the skew moment calculation module, the image reconstruction module, the flipping and sorting module, and the image stitching module are all electrically connected to the central processing module.

[0040] Compared with the prior art, the present invention has the following advantages: the millimeter-wave sparse array time-domain rapid image reconstruction method is applied to human body security inspection occasions with large passenger flow and high traffic efficiency, and a millimeter-wave Ka-band two-dimensional sparse array configuration is designed, and a time-domain rapid imaging method is designed for this array configuration. Compared with the known back-projection reconstruction algorithm, the computational efficiency is greatly improved, and the spatial distribution characteristics of the U-shaped array are mainly adopted, and the equivalent array method, the symmetry principle, and the relative coordinate system principle are fully utilized to reasonably divide the echo data. When calculating the slant distance from the transmitting and receiving antenna array elements to the imaging area, only one calculation is required, and other module units can use the same group of slant distance data, which corresponds to the same group of matching filters and interpolation coefficients. This In terms of improving the computational efficiency of image reconstruction, the matched filter coefficients and interpolation coefficients can be stored in a coefficient file during real-time calculation, and automatically loaded into the hardware signal processing platform when the entire system is turned on. The coefficients can be called from the memory during image reconstruction. The overall structure of the algorithm proposed in the present invention has greater parallelism, and the data between the entire four-subarray imaging modules are independent of each other, and there is no data mutual dependence. At the same time, inside the four-subarray imaging module, a single subarray transmits and four subarrays receive modes also have parallelism, which is particularly suitable for parallel processing using FPGA, DSP, and GPU hardware structures. The present invention can speed up the computational speed of image reconstruction without increasing the cost of the hardware signal processing platform. It is a millimeter-wave image reconstruction method with high engineering application value and is worthy of popularization. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the process of the millimeter wave sparse array time domain fast image reconstruction method of the present invention;

[0042] Figure 2 2 is a schematic diagram of the distribution of millimeter wave two-dimensional sparse array elements in an embodiment of the present invention, wherein the horizontal dotted lines represent the transmitting array elements and the vertical solid lines represent the receiving array elements;

[0043] Figure 3 is the internal division of the imaging area of ​​the four-subarray module in the embodiment of the present invention, wherein: Figure 3 (a) represents the equivalent array coverage of all receiving subarrays of the first transmitting subarray, and the shaded area represents the equivalent array coverage. Figure 3 (b) represents the equivalent array coverage of all receiving subarrays of the second transmitting subarray. The shaded area represents the equivalent array coverage. Figure 3 (c) represents the equivalent array coverage of all receiving sub-arrays of the third transmitting sub-array. The shaded area represents the equivalent array coverage. Figure 3 (d) represents the equivalent array coverage of all receiving sub-arrays of the fourth transmitting sub-array. The shaded area represents the equivalent array coverage. Figure 3 (e) is a schematic diagram showing the relative coordinate system when the first transmitting sub-array is working, and other transmitting sub-arrays are similarly configured when working;

[0044] Figure 4 is a schematic diagram of sixteen subarrays divided into four subarray modules in an embodiment of the present invention, and the shaded portion is the imaging area division when four receiving subarrays of a single transmitting subarray are working, wherein: Figure 4 (a)~ Figure 4 (f) is a schematic diagram of the imaging area division when the first emission sub-array to the sixth emission sub-array work in sequence, and the sixteenth emission sub-array works until the sixteenth emission sub-array finishes working. DETAILED DESCRIPTION

[0045] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0046] This embodiment provides a technical solution: the millimeter wave two-dimensional sparse array configuration in this embodiment is as follows: Figure 2 As shown, the horizontal dotted line represents the transmitting array element, and the vertical solid line represents the receiving array element. The entire two-dimensional array consists of sixteen square-shaped sub-arrays, and the spacing between the transmitting array elements is Δx T =0.0068m, the receiving array element spacing is Δy R =0.0068m, the number of transmitting array elements is 736, the number of receiving array elements is 736, the number of transmitting array elements in a single sub-array is 46, and the number of receiving array elements is 46. Figure 2 The arrows shown indicate the working order of the transmitting array. The transmitting array switches from left to right in sequence, and the receiving array elements simultaneously receive the backscattered echo of the target. The echo signal demodulated by the millimeter wave digital intermediate frequency receiver is s(x Tx ,y Tx ,x Rx ,y Rx ,k),x Tx is the horizontal dimension of the transmitting array, y Tx is the vertical dimension of the transmitting array, x Rx is the horizontal dimension of the receiving array, y Rx is the vertical dimension of the receiving array, k is the frequency scanning dimension, k = 2 × π × fre / c is the frequency scanning dimension, where fre is the frequency of the transmitted signal, fre∈[f min ,f max ], f min is the minimum frequency of the transmitted signal, f max is the maximum frequency of the transmitted signal.

[0047] For the echo signal s(xTx ,y Tx ,x Rx ,y Rx ,k) According to the symmetry characteristics, the sixteen sub-arrays are divided into groups of four sub-arrays according to the overlapping rule, and the obtained signal is s sub4 (x Tx ,y Tx ,x Rx ,y Rx ,k), according to Figure 3 As shown, for the four sub-array signals s sub (x Tx ,y Tx ,x Rx ,y Rx ,k) Further sub-array division is performed, which is divided into a single transmitting array and four receiving array sub-aperture patterns. The resulting signal is right Perform equivalent phase center processing, and the equivalent array coordinates obtained are x c =(x Tx +x Rx ) / 2,y c =(y Tx +y Rx ) / 2,z c =0, where x c ∈[x' min ,x' max ],y c ∈[y' min ,y' max ], by x c and c The range and imaging area range of the human body in the distance dimension z'∈[z' min ,z' max ] Determine the spatial coverage of the equivalent array Ω(x',y',z'), where the specific method of determining the spatial coverage is to calculate: x' min =0, x' max =2×(NT sub / 2)×Δx T , y' min =0,y' max =2×(NR sub / 2)×Δy R , z' min =0.2m, z' max =0.6m.

[0048] According to the range of the equivalent array coverage area Ω(x',y',z'), as Figure 3As shown, the gray shaded area is gridded with the vertex of the transmitting array as the origin, and the coordinates of the division are (x', y', z'). The slant distances from the transmitting array element and the receiving array element to the equivalent array space coverage area Ω(x', y', z') are calculated as follows: The calculated matched filter coefficient is Href(x sub_Tx ,y sub_Tx ,x sub_Rx ,y sub_Rx ,x',y',z')=exp(jk min (R sub_Tx +R sub_Rx )), where k min =2×π×f min / c,f min is the minimum frequency of the transmitted signal, c≈3×10 8 is the speed of electromagnetic waves in vacuum, according to R sub_Rc =(R sub_Tx +R sub_Rx ) / 2 Calculate the distance compression domain signal The interpolation coefficients in the distance dimension, For subi (x Tx ,y Tx ,x Rx ,y Rx ,k) Inverse Fast Fourier Transform in frequency scan dimension k, lr = R sub_Rc / taor+1, l2=l1+1, taor=c / (2×B) / ((Nf0-1) / (Nf-1)) is the spatial sampling interval after IFFT upsampling in the frequency scanning dimension, Nf is the number of scanning points in the frequency dimension, Nf0 is the number of IFFT points, B=f max -f min is the bandwidth of the transmitted signal.

[0049] The echo signal Perform inverse fast Fourier transform of Nf0 points in the frequency scanning dimension to obtain the range compression domain signal Calculate the range compression domain signal The interpolation sampling points The calculated three-dimensional complex image of the imaging sub-area is:

[0050]

[0051] According to the principle of symmetry and relative coordinate system, the three-dimensional sub-images σ of the four imaging sub-areas are sub(x', y', z') are flipped and sorted, and the image formed by the relative coordinate system is flipped up and down and left and right to convert from the relative coordinate system to the absolute coordinate system. The flipped images are then spliced ​​and sorted at corresponding positions in the entire imaging field of view, and the overlapping areas are accumulated to obtain the image in the four-subarray mode. The images of the other four sub-array modules are calculated in parallel until the three-dimensional complex images σ of the imaging areas corresponding to all the four sub-array modules are calculated. sub4 (x',y',z') calculation completed.

[0052] For all three-dimensional complex sub-images σ sub4 (x', y', z') are spliced ​​to obtain the final three-dimensional complex image σ(x', y', z'), and the amplitude signal σ is extracted amp (x', y', z') = abs(σ(x', y', z')), perform the maximum projection of the distance dimension, and calculate the two-dimensional projection image σ of the two azimuth dimensions amp (x',y'). For the two-dimensional amplitude image σ amp (x', y') is used for target detection, recognition or image processing, and then sent to the display end for image display.

[0053] In summary, the millimeter-wave sparse array time-domain fast image reconstruction method of the above embodiment divides a total of sixteen sub-arrays of 4*4 into a mode of four sub-arrays of 2*2 according to the equivalent array distribution and antenna illumination characteristics. The four sub-arrays overlap with each other. Due to the limitation of the antenna beam width, the antenna beam width of the receiving array is limited. Therefore, the received target effective echo is limited to a certain area in terms of spatial distribution. Within the four sub-arrays, according to the mode of single sub-array transmission and four sub-array reception, the equivalent array element spatial distribution position of the sparse array is calculated according to the equivalent array method and set as the main effective echo area. It is considered that the echoes in the remaining areas are weak and contribute less to the reconstructed image. Therefore, when dividing the imaging grid area, the echo data corresponding to the single-subarray transmitting and four-subarray receiving mode only needs to divide the spatial area of ​​the equivalent array distribution; when calculating the four-subarray sub-image, a relative coordinate system is set. According to the symmetric characteristics of the subarray spatial distribution, the transmit and receive slant ranges of different single-subarray transmitting and four-subarray receiving modes are repeated. Therefore, the entire algorithm only needs to calculate the slant range of the single-subarray transmitting and four-subarray receiving. The matched filter coefficients and interpolation coefficients in the corresponding back-projection algorithm are also the same set of coefficients. At the same time, all the sub-images of the four-subarray modules can be calculated in parallel. In this way, by synthesizing limited echo data, dividing limited imaging grid areas, and sharing a set of matched filter coefficients and interpolation coefficients, the calculation amount of the traditional back-projection reconstruction algorithm is reduced, and an appropriate balance is made in terms of imaging quality and calculation efficiency. Compared with the traditional back-projection reconstruction algorithm, the calculation efficiency has been effectively improved, and it is worthy of being promoted and used.

[0054] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A millimeter-wave sparse array time-domain fast image reconstruction method, It is characterized in that The following steps are involved: S1: Receive sparse array echo signals The spatially arranged two-dimensional sparse array is switched by the antenna array switch to realize the two-dimensional antenna beam scanning in space, covering the imaging target range. A single transmitting antenna is used to transmit electromagnetic wave signals, and all receiving antennas simultaneously receive the backscattered electromagnetic waves, which are down-converted to intermediate frequency signals, and then digitized by the multi-channel digital acquisition module. The digitized signal is demodulated into a digital I / Q signal through digital down-conversion. The obtained target backscattered three-dimensional echo signal is s(x Tx ,y Tx , x Rx ,y Rx , k), x Tx is the horizontal dimension of the transmitting array, y Tx is the vertical dimension of the transmitting array, x Rx is the horizontal dimension of the receiving array, y Rx is the vertical dimension of the receiving array, k = 2 × π × fre / c is the frequency scanning dimension, where fre is the frequency of the transmitted signal, fre∈[f min , f max ]; S2: Sub-array division The sixteen sub-arrays are divided into groups, and each group is composed of four overlapping sub-arrays. The resulting echo signal is s sub4 (x Tx ,y Tx , x Rx ,y Rx , k); S3: Imaging area division The imaging area is divided. According to the equivalent array concept, a single subarray transmitting element and four subarray receiving elements are grouped together in the four-subarray unit to obtain the sampling position coordinates of the equivalent array element. The obtained area is the imaging area division within the four-subarray unit, and a total of four imaging areas are divided. S4: Calculate the skew moment Calculate the slant range R of the backprojection reconstruction algorithm in the single-subarray transmitting and four-subarray receiving mode sub_Tx , R sub_Px According to the symmetry principle of array spatial distribution, the slant ranges of the other single-subarray transmitting and four-subarray receiving modules within the four-subarray are consistent, and the matched filter Href(x sub_Tx ,y sub_Tx , x sub_Rx ,y sub_Rx , x, y, z) are also consistent. According to the principle of relative coordinate system, the grid coordinate values ​​of the divided imaging sub-areas are consistent with the vertex of the transmitting single sub-array as the origin, and then the three-dimensional complex image σ of the back-projection reconstruction algorithm of the four imaging sub-areas is calculated. sub (x′, y′, z′); S5: Four-subarray image reconstruction According to the principle of symmetry and relative coordinate system, the three-dimensional images of the four imaging sub-areas are symmetrically flipped and sorted, and the overlapping areas are accumulated to obtain the four-subarray module sub-images. S6: Subgraph flipping and sorting Repeat steps S3 to S5 until the sub-image calculations of the imaging areas of all four sub-array modules are completed; S7: Image stitching All three-dimensional complex sub-images are spliced ​​to obtain the final three-dimensional complex image, and then the maximum value projection of the distance dimension is performed to obtain a two-dimensional projection image of two azimuth dimensions.

2. According to the millimeter wave sparse array time domain fast image reconstruction method of claim 1, Features: In the step S1, the two-dimensional sparse array is composed of sixteen U-shaped sub-arrays arranged at equal intervals, and the frequency band of the signal emitted by the imaging system is in the range of 30 to 40 GHz.

3. According to claim 2, a millimeter wave sparse array time domain fast image reconstruction method, Features: When the imaging system is working, under the triggering of the timing generator of the whole system, the transmitting array elements are turned on in time division, and the receiving array elements are always turned on and receive the target backscattered echo signals at the same time.

4. The millimeter wave sparse array time domain fast image reconstruction method according to claim 2, Features: In step S2, the position coordinates of the equivalent array element are x c =(x Tx +x Rx ) / 2,y c =(y Tx +y Rx ) / 2,z c =0, the sub-areas for internal imaging of the four sub-arrays are divided according to the coverage area of ​​the equivalent array element position coordinates.

5. The millimeter wave sparse array time domain fast image reconstruction method according to claim 4, Features: In step S4, the calculated slant range expression of the single-subarray transmitting and four-subarray receiving mode is:

6. The millimeter wave sparse array time domain fast image reconstruction method according to claim 5, Features: In step S4, the matched filter calculation expression is: Href(x sub_Tx ,y sub_Tx ,x sub_Rx ,y sub_Rx ,x',y',z')=exp(jk min (R sub_Tx +R sub_Rx )) Among them, k min =2×π×f min / c,f min is the minimum frequency of the transmitted signal, and c is the speed of electromagnetic waves in a vacuum.

7. The millimeter wave sparse array time domain fast image reconstruction method according to claim 6, Features: In step S4, the obtained three-dimensional complex image of the imaging sub-region is: in, for Inverse Fast Fourier Transform in frequency scan dimension k, lr = R sub_Rc / taor+1, l2=l1+1,taor=c / (2×B) / ((Nf 0 -1) / (Nf-1)) is the spatial sampling interval after IFFT upsampling in the frequency scanning dimension.

8. A millimeter-wave sparse array time-domain fast image reconstruction system, It is characterized in that Reconstructing an image using the reconstruction method according to any one of claims 1 to 7, comprising: The echo signal receiving module is used to transmit electromagnetic wave signals through a single transmitting antenna. All receiving antennas receive backscattered electromagnetic waves at the same time, down-convert them to intermediate frequency signals, and then digitize the analog signals through the multi-channel digital acquisition module. The digitized signals are demodulated into digital I / Q signals through digital down-conversion. The obtained target backscattered three-dimensional echo signal is s(x Tx ,y Tx ,x Rx ,y Rx ,k); The subarray division module is used to divide the sixteen subarrays into groups, with four subarrays overlapping as one group. The resulting echo signal is s sub4 (x Tx ,y Tx ,x Rx ,y Rx ,k); The imaging area division module is used to divide the imaging area. According to the equivalent array concept, a single subarray transmitting element and four subarray receiving elements are grouped together in the four-subarray unit to obtain the sampling position coordinates of the equivalent array element. The obtained area is the imaging area division within the four-subarray unit, and a total of four imaging areas are divided. Slant moment calculation module, used to calculate the slant distance R of the back-projection reconstruction algorithm in the single-subarray transmitting and four-subarray receiving mode sub_Tx , R sub_Rx , and calculate the three-dimensional complex image σ of the back-projection reconstruction algorithm of the four imaging sub-areas sub (x',y',z'); The image reconstruction module is used to symmetrically flip and sort the three-dimensional images of the four imaging sub-areas according to the principle of symmetry and relative coordinate system, and accumulate the overlapping areas to obtain the four-subarray module sub-images. A flipping and sorting module is used to repeat steps S3 to S5 until the sub-image calculations of the imaging areas of all four sub-array modules are completed; An image stitching module is used to stitch all the three-dimensional complex sub-images to obtain a final three-dimensional complex image, and then perform maximum projection of the distance dimension to obtain a two-dimensional projection image of two azimuth dimensions; The control processing module is used to issue instructions to each module to complete the corresponding actions; The echo signal receiving module, the sub-array division module, the imaging area division module, the skew moment calculation module, the image reconstruction module, the flipping and sorting module, and the image stitching module are all electrically connected to the central processing module.