Rapid pass-type millimeter wave security check image reconstruction method and related equipment

Abstract

The invention provides a rapid pass-type millimeter wave security check image reconstruction method and related equipment, and the method comprises the steps: transmitting a step frequency radio frequency signal, and collecting echo data of a calibration body at two different positions for phase calibration; collecting echo data of a to-be-detected target, dividing the echo data into a plurality of processing windows, and making preparations for generating corresponding sub-graphs for data of the processing windows; defining a three-dimensional space grid for each sub-image, and performing time domain backward projection calculation on the data of each processing window by using each three-dimensional space grid to obtain a plurality of complex images; according to the method, a plurality of images of each sub-image are converted into modulus value images, spatial transformation and projection processing are carried out on each modulus value image, and a two-dimensional security check image is output after fusion, so that the image quality and the recognition precision of the rapid pass-type millimeter wave security check image reconstruction method can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of millimeter-wave device control, and in particular to a fast-pass millimeter-wave security inspection image reconstruction method and related devices. Background Art

[0002] Currently, most of the body security inspections adopt contact or cooperative security inspections. This method not only requires touching the passenger's body parts, resulting in a poor passenger experience, but also has a slow passing speed and is difficult to meet the needs of rapid security inspections for large passenger flows.

[0003] Currently, there is a fast-pass millimeter-wave body security inspection system, which has the advantages of large passenger flow, high privacy, zero contact, high frame rate, etc., can meet the requirements of real-time imaging, and is applicable to security application fields such as subways, high-speed rails, customs, and public security checkpoints with high population density. However, the existing fast-pass millimeter-wave security inspection image reconstruction methods still need to be improved in terms of image quality and recognition accuracy. Summary of the Invention

[0004] The following is an overview of the subject matter described in detail in this document. This overview is not intended to limit the scope of protection of the claims.

[0005] Aiming to solve at least one of the technical problems existing in the prior art, embodiments of the present invention provide a fast-pass millimeter-wave security inspection image reconstruction method and related devices. By collecting the echo data of the cylindrical calibration body twice to complete the system phase correction, and using the multi-window processing method for the echo data of the target to be detected to complete the three-dimensional space grid division for back-projection imaging, the image quality and recognition accuracy of the fast-pass millimeter-wave security inspection image reconstruction method can be effectively improved.

[0006] To achieve the above object, a first aspect of an embodiment of the present invention provides a fast-pass millimeter-wave security inspection image reconstruction method, including: selecting a target frequency band, and transmitting a stepped-frequency RF signal according to the frequency band by a transmitting array element; collecting echo data of a calibration body at two different positions by a receiving array element for phase calibration; collecting echo data of a target to be detected by the receiving array element, dividing the echo data into multiple processing windows, and preparing to generate corresponding sub-images for the data of each processing window; defining a three-dimensional space grid for each sub-image, where the range and elevation grids in the three-dimensional space grid are divided based on a unified reference, and the azimuth grids are grouped and shared according to the azimuth coverage range of the processing window; performing time-domain back-projection calculation on the data of each processing window with each three-dimensional space grid to obtain a plurality of complex images; converting the complex images of each sub-image into modulus images, and performing spatial transformation and projection processing on each modulus image, and fusing and outputting a two-dimensional security inspection image.

[0007] In some embodiments, dividing the echo data into multiple processing windows to prepare for generating corresponding sub - graphs for the data of each processing window includes: dividing the echo data into four different sliding windows; respectively performing three - dimensional back - projection algorithm processing on the four sliding windows to obtain sub - graphs corresponding to the four sliding windows.

[0008] In some embodiments, defining a three - dimensional space grid for each of the sub - graphs includes: the four sub - graphs share distance - direction and elevation - direction space grid coordinates based on a preset division with the central zero point as the reference; sub - Figure 1 and sub - Figure 4 share horizontal - direction space grid coordinates, and sub - Figure 2 and sub - Figure 3 share horizontal - direction space grid coordinates.

[0009] In some embodiments, collecting echo data of a calibration body at two different positions by receiving array elements for phase calibration includes: placing a cylindrical calibration body at a first position, collecting first echo data of the cylindrical calibration body by the receiving array elements, and using the first echo data as the phase compensation for sliding window one and sliding window four among the four sliding windows; moving the cylindrical calibration body to a second position, collecting second echo data of the cylindrical calibration body by the receiving array elements, and using the second echo data as the phase compensation for sliding window two and sliding window three among the four sliding windows; realizing the phase calibration of the four sliding windows according to the phase compensation.

[0010] In some embodiments, performing time - domain back - projection calculation on the data of each processing window with each three - dimensional space grid to obtain multiple complex images includes: traversing the three - dimensional space grid corresponding to each sub - graph, calculating the phase delay of each grid point in the horizontal direction, elevation direction, and distance direction in sequence, where the phase delay is calculated based on the propagation path of the signal from the transmitting array element to the grid point and then to the receiving array element; constructing an ideal echo signal at each grid point, where the ideal echo signal is in complex form and contains amplitude and phase information; performing data processing on the ideal echo signal and the actually collected echo data after pulse compression to obtain the complex value of each grid point, and forming the complex image of the three - dimensional space grid by traversing all grid points.

[0011] In some embodiments, converting the complex images of each sub - graph into modulus images, performing spatial transformation and projection processing on each modulus image, and fusing and outputting a two - dimensional security inspection image includes: performing modulus calculation on the complex images of each sub - graph to obtain the modulus images of each sub - graph in the three - dimensional space grid; performing rotation bilinear interpolation processing on the modulus images of each sub - graph and performing maximum projection in the distance direction to obtain the two - dimensional images of each sub - graph; fusing the two - dimensional images of each sub - graph to output a two - dimensional security inspection image.

[0012] In some embodiments, performing rotational bilinear interpolation processing on the modulus images of the subgraphs and performing maximum projection in the range direction to obtain two-dimensional images of the subgraphs includes: determining the rotation angle, and determining the rotation direction and angle according to the azimuth array element distribution of the processing window corresponding to each subgraph, so that the rotated image is aligned with a preset projection direction; performing rotation processing on the modulus images of the subgraphs, calculating the modulus values of each pixel point in the rotated image by using the bilinear interpolation algorithm, performing maximum projection on the rotated modulus image, selecting the maximum value of the modulus values of each pixel point in the range direction along the range dimension, and using the maximum value as the modulus value of the pixel point in the two-dimensional image to obtain the two-dimensional images of the subgraphs.

[0013] To achieve the above object, a second aspect of the embodiments of the present invention provides a fast-pass millimeter-wave security inspection device, including a pass-through folded array system placed opposite and parallel to both sides of a channel, the system including a transceiver array, the transceiver array including a plurality of transmitting array elements and a plurality of receiving array elements; the system is configured to: select a target frequency band, and transmit a stepped-frequency RF signal according to the frequency band through the transmitting array elements; collect echo data of a calibration object at two different positions through the receiving array elements for phase calibration; collect echo data of a target to be detected through the receiving array elements, divide the echo data into a plurality of processing windows, and prepare to generate corresponding subgraphs for the data of each processing window; define a three-dimensional space grid for each subgraph, the range and elevation grids in the three-dimensional space grid are divided based on a unified reference, and the horizontal grids are grouped and shared according to the azimuth coverage range of the processing window; perform time-domain back-projection calculation on the data of each processing window with each three-dimensional space grid to obtain a plurality of complex images; convert the complex images of each subgraph into modulus images, and perform spatial transformation and projection processing on each modulus image, and output a two-dimensional security inspection image after fusion.

[0014] To achieve the above object, a third aspect of the embodiments of the present invention provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor implementing the method described in the first aspect when executing the computer program.

[0015] To achieve the above object, a fourth aspect of the embodiments of the present invention provides a computer-readable storage medium, the computer-readable storage medium storing a computer program, and the computer program implementing the method described in the first aspect when executed by a processor.

[0016] The fast-pass millimeter-wave security inspection image reconstruction method proposed by the present invention has at least the following beneficial effects: By collecting the echo data of the cylindrical calibration body twice and performing phase compensation on four different sliding windows, the present invention can effectively complete the system phase correction, improve the accuracy and reliability of imaging, divide the echo data of the target to be detected into multiple processing windows, and prepare to generate corresponding sub-images for the data of each processing window, which can focus on the target information at different angles respectively, reduce the number of transceiver array elements in a single sliding window at the same time, and reduce the computational complexity of the back-projection algorithm; Further, a three-dimensional space grid is defined for each sub-image, where the range and elevation grids are divided based on a unified benchmark, and the horizontal grid is grouped and shared according to the azimuth coverage range of the processing window. This grid division method can reduce the number of horizontal grid points, reduce the calculation time of the three-dimensional grid back-projection algorithm, and improve the imaging efficiency; Further, for the data of each processing window, perform time-domain back-projection calculation with each three-dimensional space grid to obtain multiple complex images, then convert these complex images into modulus images, and perform spatial transformation and projection processing on the modulus images, and then fuse and output two-dimensional security inspection images, which can effectively solve the image blind area problem, improve the image quality, enhance the recognition ability of the target side position, and thus improve the accuracy and efficiency of security inspection. In summary, the method of the present invention can effectively improve the image quality, reduce the computational complexity, and enhance the recognition ability, and is applicable to the application scenario of millimeter-wave human body fast-pass security inspection, which can effectively improve the security inspection efficiency and the passenger experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flowchart of the fast-pass millimeter-wave security inspection image reconstruction method provided by an embodiment of the present invention.

[0018] Figure 2 is a side view of the first position of the cylindrical calibration provided by an embodiment of the present invention;

[0019] Figure 3 is a side view of the second position of the cylindrical calibration provided by an embodiment of the present invention;

[0020] Figure 4 is an array element layout diagram of a single array surface provided by another embodiment of the present invention;

[0021] Figure 5 is an array element layout diagram of the system array surface provided by another embodiment of the present invention;

[0022] Figure 6 is a schematic diagram of a human body passing through a fast channel provided by another embodiment of the present invention;

[0023] Figure 7 is an example flowchart of the fast-pass millimeter-wave security inspection image reconstruction method provided by an embodiment of the present invention;

[0024] Figure 8 It is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. Specific Embodiments

[0025] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0026] It should be noted that although functional module division is performed in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different module division in the device or a different order in the flowchart. Terms such as "first" and "second" in the specification, claims and the above-mentioned drawings are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used herein are only for the purpose of describing the embodiments of the present invention and are not intended to limit the present invention.

[0028] Most current body security inspections use contact or cooperative security inspections. This method not only requires touching the passenger's body parts, resulting in a poor passenger experience, but also has a slow passing speed and is difficult to meet the needs of rapid security inspections for large passenger flows. Currently, there is a fast-through millimeter-wave body security inspection system, which has the advantages of large passenger flow, high privacy, zero contact, high frame rate, etc., can meet the requirements of real-time imaging, and is suitable for security application fields such as subways, high-speed rails, customs, and public security checkpoints with high population density. However, the existing fast-through millimeter-wave security inspection image reconstruction methods still need to be improved in terms of image quality and recognition accuracy.

[0029] Based on this, aiming to solve at least one of the technical problems existing in the prior art, an embodiment of the present invention provides a fast-through millimeter-wave security inspection image reconstruction method and related equipment. By collecting the echo data of the cylindrical calibration body twice to complete the system phase correction, and using a multi-window processing method for the echo data of the target to be detected to complete the three-dimensional space grid division for back-projection imaging, the image quality and recognition accuracy of the fast-through millimeter-wave security inspection image reconstruction method can be effectively improved.

[0030] The following will further describe the embodiments of the present invention with reference to the accompanying drawings.

[0031] Please refer to Figure 1 , Figure 1It is a flowchart of a fast-pass millimeter-wave security inspection image reconstruction method provided by an embodiment of the present invention; To achieve the above object, a first aspect of the embodiments of the present invention proposes a fast-pass millimeter-wave security inspection image reconstruction method, including but not limited to the following steps:

[0032] Step S110, select a target frequency band, and transmit a stepped-frequency RF signal according to the frequency band through a transmitting array element;

[0033] Step S120, collect echo data of a calibration object at two different positions through a receiving array element for phase calibration;

[0034] Step S130, collect echo data of the target to be detected through a receiving array element, divide the echo data into multiple processing windows, and prepare to generate corresponding sub-images for the data of each processing window;

[0035] Step S140, define a three-dimensional space grid for each sub-image. The range and elevation grids in the three-dimensional space grid are divided based on a unified reference, and the horizontal grids are grouped and shared according to the azimuth coverage range of the processing window;

[0036] Step S150, perform time-domain backprojection calculation on the data of each processing window with each three-dimensional space grid to obtain a plurality of complex images;

[0037] Step S160, convert the complex images of each sub-image into modulus images, perform spatial transformation and projection processing on each modulus image, and output a two-dimensional security inspection image after fusion.

[0038] Among them, the present invention can effectively complete system phase correction, improve the accuracy and reliability of imaging by collecting echo data of the cylindrical calibration body twice and performing phase compensation on four different sliding windows, divide the echo data of the target to be detected into multiple processing windows, prepare to generate corresponding sub-images for the data of each processing window, can focus on target information at different angles respectively, and at the same time reduce the number of transmitting and receiving array elements in a single sliding window, reducing the computational complexity of the back-projection algorithm; further, define a three-dimensional space grid for each sub-image, where the range and elevation grids are divided based on a unified benchmark, and the azimuth grids of the processing windows are grouped and shared according to the azimuth coverage range. The above grid division method can reduce the number of horizontal grid points, can reduce the computational time of the three-dimensional grid back-projection algorithm, and improve the imaging efficiency; further, for the data of each processing window, perform time-domain back-projection calculation with each three-dimensional space grid to obtain multiple complex images, then convert these complex images into modulus images, and perform spatial transformation and projection processing on the modulus images, and then fuse and output a two-dimensional security inspection image, which can effectively solve the problem of image blind spots, improve the image quality, enhance the recognition ability of the target side position, and thus improve the accuracy and efficiency of security inspection. In summary, the method of the present invention can effectively improve the image quality, reduce the computational complexity and enhance the recognition ability, is applicable to the application scenario of rapid millimeter-wave human body security inspection, and can effectively improve the security inspection efficiency and passenger experience.

[0039] In some embodiments, corresponding to step S110, first, a suitable target frequency band needs to be determined, and this frequency band should be selected according to specific security inspection requirements and system design to ensure that it can effectively penetrate obstacles such as clothing and detect hidden contraband. After the frequency band is selected, the transmitting array elements will transmit stepped-frequency RF signals according to this frequency band, and these signals will cover the target area, providing basic data for subsequent detection and imaging.

[0040] In some embodiments, corresponding to step S120, place the cylindrical calibration body at two different positions respectively, and the receiving array elements collect the echo data of the calibration body at these two positions respectively. By comparing and analyzing these two sets of echo data, the phase of the system can be accurately corrected, thereby improving the accuracy of subsequent imaging.

[0041] In some embodiments, corresponding to step S130, the receiving array elements collect the echo data of the target to be detected, and these data contain the scattering information of the target object. To effectively process these data, they are divided into multiple processing windows, and each window corresponds to a specific angular range. Then, perform the back-projection algorithm on the data of each processing window to generate corresponding sub-images. These sub-images will show the characteristics of the target object from different angles, providing a basis for subsequent image fusion and analysis.

[0042] In some embodiments, corresponding to step S140, a three-dimensional spatial grid is defined for each sub-image to represent the position and characteristics of the target object in space. The range and elevation grids are divided based on a unified reference to ensure spatial consistency among sub-images. The azimuth grid is grouped and shared according to the azimuth coverage range of the processing window, which can reduce the number of grid points, improve the calculation efficiency, and ensure the resolution and quality of the image.

[0043] In some embodiments, corresponding to step S150, time-domain back-projection calculations are performed on the data of each processing window using each three-dimensional spatial grid. Specifically, according to the propagation path of the signal from the transmitting array element to the target object and then back to the receiving array element, the phase delay of each grid point is calculated, and an ideal echo signal is constructed. These ideal echo signals are correlated with the actually collected echo data to obtain the complex value of each grid point. By traversing all grid points, multiple complex images are finally formed, and these images contain the amplitude and phase information of the target object, providing a rich data basis for subsequent image processing.

[0044] In some embodiments, corresponding to step S160, first, the complex image of each sub-image is converted into a modulus image, that is, the modulus of each complex number is calculated to obtain an image representing the scattering intensity of the target object. Then, spatial transformation and projection processing are performed on these modulus images, specifically including rotation bilinear interpolation processing and maximum projection. The rotation bilinear interpolation processing is used to correct the geometric distortion of the image to align it with the preset projection direction; the maximum projection is used to highlight the strongest scattering information of the target object at different distances. Finally, the two-dimensional images of each processed sub-image are fused to output the final two-dimensional security inspection image, which can clearly display the contour and characteristics of the target object and provide accurate detection results for security inspection personnel.

[0045] In some embodiments, the echo data is divided into multiple processing windows to prepare for generating corresponding sub - graphs for the data of each processing window, including: dividing the echo data into four different sliding windows; respectively performing three - dimensional back - projection algorithm processing on the four sliding windows to obtain sub - graphs corresponding to the four sliding windows; wherein, in this application, by dividing the echo data into multiple processing windows to prepare for generating corresponding sub - graphs for the data of each processing window, it specifically includes the following steps: First, divide the echo data into four different sliding windows, and each sliding window corresponds to a specific angular range. Through the above - mentioned method, the target information at different angles can be focused respectively, which is convenient for subsequent deep - learning identification of contraband. Second, respectively perform three - dimensional back - projection algorithm processing on the four sliding windows. Through the above - mentioned processing method, the number of transceiver array elements in a single sliding window can be effectively reduced, thereby reducing the computational complexity of the back - projection algorithm. Finally, obtain sub - graphs corresponding to the four sliding windows. These sub - graphs show the characteristics of the target object from different angles and provide a basis for subsequent image fusion and analysis. It can be understood that when the target to be detected passes through, the receiving array elements collect the echo data and divide it into a four - sliding - window mode. The echo data is divided into four sliding windows and then respectively undergoes subsequent three - dimensional back - projection algorithm processing to obtain four sub - graphs.

[0046] In some embodiments, a three - dimensional spatial grid is defined for each sub - graph, including: the four sub - graphs share the range - direction and elevation - direction spatial grid coordinates divided based on a preset center zero point; sub - Figure 1 and sub - Figure 4 share the azimuth - direction spatial grid coordinates, and sub - Figure 2 and sub - Figure 3 share the azimuth - direction spatial grid coordinates; wherein, the four sub - graphs share the range - direction and elevation - direction spatial grid coordinates divided based on a preset center zero point. The above - mentioned shared grid coordinate system can ensure the spatial consistency between different sub - graphs, which is convenient for subsequent image fusion and analysis. Second, sub - Figure 1 and sub - Figure 4 share the azimuth - direction spatial grid coordinates, and sub - Figure 2 and sub - Figure 3 share the azimuth - direction spatial grid coordinates. The above - mentioned grouped sharing method can reduce the number of grid points in the azimuth - direction, thereby reducing the computational complexity and improving the imaging efficiency. At the same time, the above - mentioned design can also ensure that each sub - graph has sufficient resolution within the azimuth coverage range of its corresponding processing window to accurately represent the characteristics of the target object. It can be understood that through the above - mentioned method of defining the three - dimensional spatial grid, it can effectively support subsequent time - domain back - projection calculations and generate high - quality complex images.

[0047] In some embodiments, the spatial grid coordinates of each of the four sub - graphs are divided. The four sub - graphs share the range - direction and elevation - direction spatial grid coordinates divided based on a preset center zero point. Sub - Figure 1 and sub - Figure 4Share the horizontal spatial grid coordinates, sub- Figure 2 and sub- Figure 3 share the horizontal spatial grid coordinates. The four sub-figures and the four sliding windows of the echo data are in a corresponding relationship, that is, the echo data of sliding window 1, after the time-domain back-projection algorithm, obtains sub- Figure 1 , and so on. The echo data of sliding window 4, after the time-domain back-projection algorithm, obtains sub- Figure 4 . The three-dimensional spatial grid coordinate expression is:

[0048] X = [X mid - Sampx·Numx / 2, X mid + Sampx·Numx / 2]

[0049] Y = [Y mid - Sampy·Numy / 2, Y mid + Sampy·Numy / 2]

[0050]

[0051] Sampz = (Z max - Z min ) / Numz

[0052] where Sampx = 7.5mm is the spatial sampling grid size in the range direction, Numx = 80 is the number of sampling data in the range direction, X mid = 0 is the center point value in the range direction, Sampy = 6.5mm is the spatial sampling grid size in the elevation direction, Numy = 256 is the number of sampling data in the elevation direction, Y mid = 0 is the center point value in the elevation direction, X = [-0.3m, 0.3m] is the spatial grid value range in the range direction, Y = [-0.832m, 0.832m] is the spatial grid value range in the elevation direction, - Sampx·Numx / 2 = -0.3m is the minimum value of the range direction value range, Sampx·Numx / 2 = 0.3m is the maximum value of the range direction value range, - Sampy·Numy / 2 = -0.832m is the minimum value of the elevation direction value range, Sampy·Numy / 2 = 0.832m is the maximum value of the elevation direction value range; Z is the spatial grid value range in the horizontal direction, Z min is the minimum value of the horizontal direction value range, Z max is the maximum value of the horizontal direction value range, Sampz = 7.5mm is the spatial sampling grid size in the horizontal direction, Numz = 128 is the number of sampling data in the horizontal direction, sub- Figure 1 and sub- Figure 4 share the horizontal spatial grid value range, sub- Figure 2 and sub- Figure 3 share the horizontal spatial grid value range.

[0053] It should be noted that the sub- Figure 1 and the sub- Figure 4 share the horizontal spatial grid value range. Z min = 0, Z max = 0.96m. The sub- Figure 2 and the sub- Figure 3 share the horizontal spatial grid value range. Z min = 0.35m, Z max = 1.31m. When the target passes through the security inspection channel, the sliding window 1 corresponds to the sub- Figure 1 area for back-projection imaging calculation, and so on. The sliding window 4 corresponds to the sub- Figure 4 area for back-projection imaging calculation, that is, different sliding window areas correspond to different sub-graph area imaging calculations. The horizontal area of interest is calculated quickly. The advantage is to reduce the number of horizontal grid points and reduce the calculation time of the three-dimensional grid back-projection algorithm. Using GPU acceleration or an imaging calculation chip can meet real-time imaging.

[0054] Please refer to Figures 2 to 3 , Figure 2 which is the side view of the first position of the cylinder calibration provided by an embodiment of the present invention, Figure 3 and Figure 3 is the side view of the second position of the cylinder calibration provided by an embodiment of the present invention; it should be noted that the side view of the first position of the cylinder calibration provided by the embodiment of the present application can be referred to Figure 2 as shown. In Figure 2 , a plurality of transmitting array elements and receiving array elements are combined into a circular through-folded array surface around both sides of the cylinder calibration body. The horizontal field of view is 1.31m, and the elevation field of view is 1.7m. The cylinder calibration can calibrate the antenna modules of the sliding window 1 and the sliding window 4 at the same time; the side view of the second position of the cylinder calibration can be referred to Figure 3 as shown, and the cylinder calibration can calibrate the antenna modules of the sliding window 2 and the sliding window 3 at the same time.

[0055] In some embodiments, echo data of a calibration object at two different positions are collected by receiving array elements for phase calibration, which specifically includes the following steps: First, place the cylindrical calibration object at the first position, and collect the first echo data of the cylindrical calibration object through the receiving array elements. The purpose of this step is to obtain the echo characteristics of the calibration object at a specific position for subsequent phase compensation. Then, use the first echo data as the phase compensation for sliding window one and sliding window four among the four sliding windows. This means that by analyzing the first echo data, the phase characteristics of the system in these two sliding windows can be determined and corresponding adjustments can be made to ensure the accuracy of subsequent imaging. Next, move the cylindrical calibration object to the second position, and collect the second echo data of the cylindrical calibration object through the receiving array elements. The purpose of this step is to obtain the echo characteristics of the calibration object at another specific position to further improve the phase calibration of the system. Then, use the second echo data as the phase compensation for sliding window two and sliding window three among the four sliding windows. By analyzing the second echo data, the phase characteristics of the system in these two sliding windows can be determined and corresponding adjustments can be made to ensure the accuracy of subsequent imaging.

[0056] Through the above steps, the phase calibration inside the system can be effectively completed. This phase calibration method can accurately compensate for the phase differences of the system in different sliding windows, improving the accuracy and reliability of imaging. At the same time, by placing the cylindrical calibration object at two different positions, the phase characteristics of the system can be comprehensively calibrated to ensure the imaging quality at different angles and positions.

[0057] In some embodiments, it should be noted that for the first position of the cylindrical calibration, the antenna modules of sliding window one and sliding window four can be calibrated simultaneously. The first combination mode of the sliding window one antenna module can include array 1 and array 2, or the second combination mode can include array 1, array 2, and array 6. The first combination mode of the sliding window four antenna module can include array 5 and array 6, or the second combination mode can include array 5, array 6, and array 1. For the second position of the cylindrical calibration, the antenna modules of sliding window two and sliding window three can be calibrated simultaneously. Among them, the first combination mode of the sliding window three antenna module can include array 2 and array 3, or the second combination mode can include array 2, array 3, and array 4. The first combination mode of the sliding window three antenna module can include array 4 and array 5, or the second combination mode can include array 3, array 4, and array 5. That is, collecting the echo data of the cylinder at two positions is fast and convenient to complete the calibration of the system.

[0058] In some embodiments, for the data of each processing window, time-domain backprojection calculation is performed on each three-dimensional spatial grid to obtain a plurality of complex images, including: traversing the three-dimensional spatial grids corresponding to each sub-image, and sequentially calculating the phase delays of each grid point in the horizontal direction, pitch direction, and range direction. The phase delays are calculated based on the propagation path of the signal from the transmitting array element to the grid point and then to the receiving array element; constructing the ideal echo signals at each grid point according to the phase delays. The ideal echo signals are in complex form and contain amplitude and phase information; performing data processing on the ideal echo signals and the actually acquired echo data after pulse compression to obtain the complex values of each grid point, and forming a complex image of the three-dimensional spatial grid by traversing all grid points.

[0059] It can be understood that in this application, by traversing the three-dimensional spatial grids corresponding to each sub-image, all grid points in each sub-image are processed one by one, and the phase delays of each grid point in the horizontal direction, pitch direction, and range direction are sequentially calculated. The calculation of the phase delays is based on the propagation path of the signal from the transmitting array element to the grid point and then from the grid point to the receiving array element; specifically, when the signal propagates in space, due to the different path lengths, the phase will change. By calculating this phase change, the propagation characteristics of the signal in different directions can be determined.

[0060] Furthermore, according to the calculated phase delays, the ideal echo signals at each grid point are constructed. The ideal echo signals are represented in complex form and contain amplitude and phase information. The amplitude information reflects the intensity of the signal, while the phase information reflects the phase change of the signal. By integrating the phase delay information into the ideal echo signals, the propagation of the signal in space can be more accurately simulated.

[0061] Furthermore, data processing is performed on the constructed ideal echo signals and the actually acquired echo data after pulse compression. Specifically, by performing correlation processing on the ideal echo signals and the actual echo data, the complex values of each grid point can be obtained. These complex values contain the scattering characteristics of the target object at the corresponding grid points and are the key data for imaging. Then, by traversing all grid points, a complex image of the three-dimensional spatial grid is formed, and the complex values of each grid point calculated previously are integrated to generate a complete complex image of the three-dimensional spatial grid. These complex images contain detailed information of the target object in the three-dimensional space and provide a data basis for subsequent image processing and analysis.

[0062] In some embodiments, each of the four sub - graphs is processed by a three - dimensional back - projection algorithm. Based on the time - domain back - projection algorithm, all complex numerical values of the three - dimensional space grids of each sub - graph are traversed. The order of traversing all three - dimensional space grids of each sub - graph is horizontal direction, pitch direction, and range direction. The phase delay from the space grid nodes to the antenna elements is calculated in sequence, and the ideal echo signal at the space grid points is constructed according to the ideal phase delay. The ideal echo is correlated with the echo after pulse compression to obtain the three - dimensional grid space imaging result Sar3DImag(m,n,p), where Sar3DImag(m,n,p) is the complex numerical value of the three - dimensional grid space image, m is the horizontal index value, n is the pitch index value, and p is the range index value.

[0063] It should be noted that when the back - projection algorithm is processed for each sub - graph, they are not correlated with each other and can be processed in parallel to obtain four sets of three - dimensional space grid data, saving calculation time.

[0064] In some embodiments, for each of the four sub - graphs of the three - dimensional space grid complex images, the modulus is obtained for each to get the modulus images of the four three - dimensional space grids. Figure 3 The expression is as follows:

[0065] Sar3DImagAbs(m,n,p) = abs(Sar3DImag(m,n,p))

[0066] Where abs is to obtain the modulus value of the complex three - dimensional space grid image, and Sar3DImagAbs(m,n,p) is the modulus value of the three - dimensional space grid image.

[0067] It should be noted that since each sub - graph is a complex image of a three - dimensional space grid, the modulus value processing needs to be performed separately for each to obtain the modulus images of the four three - dimensional space grids. Figure 3

[0068] In some embodiments, the complex images of each sub - graph are converted into modulus images, and spatial transformation and projection processing are performed on each modulus image, and a two - dimensional security inspection image is output after fusion, including: performing modulus processing on the complex images of each sub - graph to obtain the modulus images of each sub - graph in the three - dimensional space grid; performing rotation bilinear interpolation processing on the modulus images of each sub - graph and performing maximum value projection on the range to obtain the two - dimensional images of each sub - graph; fusing the two - dimensional images of each sub - graph to output a two - dimensional security inspection image.

[0069] It can be understood that when performing modulus processing on the complex images of each sub - graph to obtain the modulus images of each sub - graph in the three - dimensional space grid, the complex images are converted into modulus images for subsequent processing. The modulus image represents the amplitude information of the signal. Through modulus processing, the signal intensity distribution of each sub - graph in the three - dimensional space grid can be obtained.

[0070] In some embodiments, the modulus images of each sub-image are subjected to rotational bilinear interpolation processing, and maximum value projection is performed on the distance to obtain the two-dimensional images of each sub-image. Specifically, first, the rotation angle is determined. According to the azimuth array element distribution of the processing window corresponding to each sub-image, the rotation direction and angle are determined so that the rotated image is aligned with the preset projection direction. Then, the modulus images of each sub-image are rotated, and the bilinear interpolation algorithm is used to calculate the modulus values of each pixel point in the rotated image. The bilinear interpolation algorithm can effectively handle the change of pixel point positions during image rotation, ensuring the smoothness and continuity of the image. Furthermore, maximum value projection is performed on the rotated modulus image. Along the distance dimension, the maximum value of the modulus values of each pixel point in the distance direction is selected, and the maximum value is used as the modulus value of the pixel point in the two-dimensional image, thereby obtaining the two-dimensional images of each sub-image.

[0071] It can be understood that by fusing the two-dimensional images of each sub-image obtained previously, the final two-dimensional security inspection image is generated. That is, through the fusion process, the information of each sub-image can be comprehensively utilized to improve the quality and resolution of the image, providing clear and accurate detection results for security inspection personnel.

[0072] In some embodiments, the modulus images of each sub-image are subjected to rotational bilinear interpolation processing, and maximum value projection is performed on the distance direction to obtain the two-dimensional images of each sub-image, including: determining the rotation angle, according to the azimuth array element distribution of the processing window corresponding to each sub-image, determining the rotation direction and angle so that the rotated image is aligned with the preset projection direction; rotating the modulus images of each sub-image, using the bilinear interpolation algorithm to calculate the modulus values of each pixel point in the rotated image, performing maximum value projection on the rotated modulus image, along the distance direction dimension, selecting the maximum value of the modulus values of each pixel point in the distance direction, and using the maximum value as the modulus value of the pixel point in the two-dimensional image to obtain the two-dimensional images of each sub-image.

[0073] Among them, to determine the rotation angle, according to the azimuth array element distribution of the processing window corresponding to each sub-image, the rotation direction and angle are determined so that the rotated image is aligned with the preset projection direction, which can ensure that the image can be consistent with the preset projection direction after rotation, facilitating subsequent image processing and analysis; further, the modulus images of each sub-image are rotated, and the bilinear interpolation algorithm is used to calculate the modulus of each pixel point in the rotated image. The bilinear interpolation algorithm can effectively calculate the modulus of the rotated pixel points during the image rotation process, ensuring the smoothness and continuity of the image and avoiding holes and distortions during the image rotation process; further, the maximum value projection is performed on the rotated modulus image. Along the range dimension, the maximum value of the modulus of each pixel point in the range direction is selected, and the maximum value is used as the modulus of the pixel point in the two-dimensional image, thereby obtaining the two-dimensional images of each sub-image to highlight the strongest scattering information of the target at different ranges and improve the contrast and resolution of the image.

[0074] In some embodiments, for the complex images of the three-dimensional space grids of the four sub-images, the modulus is obtained for each to get the modulus images of the four Figure 3 dimensional space grids. The expression is:

[0075] Sar3DImagAbs(m,n,p) = abs(Sar3DImag(m,n,p))

[0076] where abs is to obtain the modulus of the complex three-dimensional space grid image, and Sar3DImagAbs(m,n,p) is the modulus of the three-dimensional space grid image.

[0077] It should be noted that each sub-image is a complex image of a three-dimensional space grid, and the modulus value processing needs to be performed separately to obtain the modulus images of the four Figure 3 dimensional space grids.

[0078] In some embodiments, for the four Figure 3 dimensional images, maximum value projection is performed after rotation and interpolation. For the four Figure 3 dimensional images, rotation is performed with bilinear interpolation, and maximum value projection is performed on the range direction. The corresponding expression is:

[0079] Sar3DImagAbsNew(m,n,p) = Sar3DImagAbs(m,n,p)·R·Interpolated_pixel

[0080]

[0081] Interpolated_pixel = (1 - dx)·(1 - dy)·pixel_x1y1 + dx·(1 - dy)·pixel_x2y1 + (1 - dx)·dy·pixel_x1y2 + dx·dy·pixel_x2y2

[0082] Sar2DImag(m,n) = max(Sar3DImagAbsNew(m,n,p), 3)

[0083] Wherein, Sar3DImagAbs(m,n,p) is the modulus value of the three-dimensional spatial grid image, R is the rotation matrix, α = 45° is the rotation angle, Interpolated_pixel is bilinear interpolation, Sar3DImagAbsNew(m,n,p) is the modulus value of the three-dimensional image after spatial grid rotation interpolation, pixel_x1y1, pixel_x2y1, pixel_x1y2, pixel_x2y2 are the modulus values of the four nearest spatial grids in the horizontal and pitch directions respectively, dx and dy are the relative distances between the target value and the nearest grid value in the horizontal and pitch directions respectively, max(Sar3DImagAbsNew(m,n,p), 3) is the projection of the maximum value of the third dimension of Sar3DImagAbsNew(m,n,p), Sar2DImag(m,n) is the two-dimensional projection map, m is the horizontal index value, n is the pitch index value, and p is the range index value.

[0084] It should be noted that when the four sub - graphs perform three - dimensional data rotation, the rotation direction should be parallel to the direction connecting the maximum and minimum values of the azimuth array elements of the sliding window, that is, sub - Figure 1 sub - Figure 3 , rotate 45 degrees clockwise, sub - Figure 2 sub - Figure 4 , rotate 45 degrees counterclockwise, which is beneficial to the recognition of the target side position and solves the problem of unrecognizable blind areas. Through bilinear interpolation processing, the image is weighted and smoothed by the weights of the neighboring regions, avoiding holes in some regions during the rotation of the three - dimensional image, ensuring better image equalization and higher image quality.

[0085] Please refer to Figures 4 to 6 , Figure 4 which is the layout diagram of the single - array - element arrangement provided by another embodiment of the present invention, Figure 5 which is the layout diagram of the system - array - element arrangement provided by another embodiment of the present invention, Figure 6It is a schematic diagram of the human body passing through the express lane provided by another embodiment of the present invention; it can be understood that the fast-pass millimeter-wave security inspection image reconstruction method provided by the embodiments of the present invention is applied to a through-type millimeter-wave human body security inspection system. In the through-type millimeter-wave human body security inspection system, the through-type folded array system is placed opposite and parallel on both sides of the lane. The through-type human body security inspection system includes a transceiver array, and the transceiver array includes a plurality of transmitting array elements and a plurality of receiving array elements. Among them, Figure 4 is a layout diagram of a single array element, and the array element layout is in a folded manner; Figure 5 is a layout diagram of the system array elements, and the layout of the two array elements is in a folded manner; Based on the Figure 6 schematic diagram shown, the schematic diagram of the two opposite array surfaces can be referred to Figure 5 shown. The middle channel between the two array surfaces is the channel for the human body to pass quickly, and the arrow indicates the passing route.

[0086] In some embodiments, referring to Figure 6 , in the through-type millimeter-wave human body security inspection system, the entrances or exits of the two array surfaces of the through-type folded array system are separated by about 0.7 m and are placed opposite each other, and the human body passes through according to the indicated route in the figure.

[0087] In some embodiments, since the millimeter-wave method system of the present invention does not depend on the frequency band, the frequency band to be used can be selected as 10:14 GHz, where the frequency is divided into nFreq = 64 points, and the center wavelength λ of the frequency band c = c / f c = 0.025 m. The spacing of the transmitting antenna array elements is set to satisfy d ≤ 0.75λ c , and the spacing of the transmitting and receiving array elements is set to be d ant = 18 mm. The number of transmitting array elements in a single sub-array is 16, and the number of receiving array elements is 16; a total of 36 sub-arrays are required for a single-sided array, and a total of 72 sub-arrays are required for both sides. The total number of transmitting array elements is N t = 1152, and the total number of receiving array elements is N r = 1152.

[0088] Please refer to Figure 7 , Figure 7 is a flow chart of an example of the fast-pass millimeter-wave security inspection image reconstruction method provided by an embodiment of the present invention; in some embodiments, the fast-pass millimeter-wave security inspection image reconstruction method includes the following steps:

[0089] First, in step S710, a target frequency band is selected, and the transmitting array elements transmit stepped-frequency RF signals according to this frequency band. This step is the basis of the entire millimeter-wave security inspection system. By selecting an appropriate frequency band, it is ensured that the system can effectively transmit and receive millimeter-wave signals. Next, in step S720, cylindrical calibration bodies are placed at two different positions respectively, and the receiving array elements collect the corresponding calibration body echo data respectively. The purpose of this step is to perform phase calibration inside the system. By collecting the calibration body echo data at different positions, it is ensured that the system can accurately correct the phase and improve the accuracy of subsequent imaging. Then, in step S730, when the target to be detected passes through the security inspection channel, the receiving array elements collect the echo data of the target and divide these echo data into four sliding window modes. By dividing the echo data into multiple processing windows, this step can focus on the target information at different angles respectively, facilitating subsequent deep learning to identify prohibited items, and at the same time reducing the number of transmitting and receiving array elements in a single sliding window and reducing the computational amount of the back-projection algorithm. In step S740, three-dimensional space grids are defined for the four sub-images respectively. Among them, the range dimension and elevation dimension grids are divided based on a unified benchmark, while the azimuth dimension grids are grouped and shared according to the azimuth coverage range of each processing window. This grid division method can reduce the number of azimuth grid points and further reduce the calculation time of the three-dimensional grid back-projection algorithm, improving the imaging efficiency. Next, in step S750, time-domain back-projection calculations are performed on the three-dimensional space grids of the four sub-images respectively to obtain multiple complex images. By traversing all the three-dimensional space grids of each sub-image, this step calculates the phase delay in turn, constructs an ideal echo signal, and performs correlation processing with the actually collected echo data after pulse compression to form a complete three-dimensional space grid complex image. In step S760, the complex images of the four sub-images are converted into modulus images. By performing modulus calculation on the complex images of each sub-image, this step obtains the modulus images of each sub-image in the three-dimensional space grid, preparing for subsequent spatial transformation and projection processing. Then, in step S770, rotation bilinear interpolation processing is performed on the modulus images of the four sub-images, and maximum projection is performed in the range dimension to obtain the two-dimensional images of each sub-image. By determining the rotation angle, this step calculates the modulus of each pixel point in the rotated image using the bilinear interpolation algorithm and selects the maximum value along the range dimension, highlighting the strongest scattering information of the target at different ranges. Finally, in step S780, the two-dimensional images of the four sub-images are output respectively.

[0090] It can be understood that the above steps output the processed two-dimensional images, completing the entire millimeter-wave security inspection image reconstruction process, providing high-quality image data for subsequent image recognition and analysis, and effectively enhancing the system's ability to identify the side positions of targets, thus improving the accuracy and efficiency of security inspection.

[0091] Second aspect, a second aspect of the embodiments of the present invention provides a fast-pass millimeter-wave security inspection device, including a pass-through folded array system placed opposite and parallel on both sides of a channel. The system includes a transceiver array, and the transceiver array includes a plurality of transmitting array elements and a plurality of receiving array elements. The system is configured to: select a target frequency band, and transmit stepped-frequency RF signals according to the frequency band through the transmitting array elements; collect echo data of a calibration object at two different positions through the receiving array elements for phase calibration; collect echo data of a target to be detected through the receiving array elements, divide the echo data into a plurality of processing windows, and prepare to generate corresponding sub-images for the data of each processing window; define a three-dimensional space grid for each sub-image, where the range direction and elevation direction grids in the three-dimensional space grid are divided based on a unified reference, and the horizontal direction grid is grouped and shared according to the azimuth coverage range of the processing window; perform time-domain back projection calculation on the data of each processing window with each three-dimensional space grid to obtain a plurality of complex images; convert the complex images of each sub-image into modulus images, and perform spatial transformation and projection processing on each modulus image, and output a two-dimensional security inspection image after fusion.

[0092] Third aspect, the embodiments of the present invention further provide an electronic device, which includes: a memory, a processor, a program stored on the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory. When the program is executed by the processor, it realizes the above-mentioned fast-pass millimeter-wave security inspection image reconstruction method. The electronic device can be any intelligent terminal including a tablet computer, an in-vehicle computer, etc.

[0093] Please refer to Figure 8 , Figure 8 which schematically shows the hardware structure of an electronic device in another embodiment. The electronic device includes:

[0094] A processor 801, which can be implemented in the form of a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided by the embodiments of the present invention;

[0095] A memory 802, which can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM), etc. The memory 802 can store an operating system and other application programs. When implementing the technical solutions provided by the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 802, and the processor 801 is used to call and execute the fast-pass millimeter-wave security inspection image reconstruction method of the embodiments of the present invention;

[0096] An input / output interface 803 for implementing information input and output;

[0097] A communication interface 804 for implementing communication interaction between this device and other devices, which can achieve communication through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.);

[0098] A bus 805 for transmitting information between various components of the device (such as a processor 801, a memory 802, an input / output interface 803, and a communication interface 804);

[0099] Among them, the processor 801, the memory 802, the input / output interface 803, and the communication interface 804 achieve communication connections with each other inside the device through the bus 805.

[0100] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium. The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above-mentioned fast-through millimeter-wave security inspection image reconstruction method.

[0101] As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely set relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above-mentioned network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0102] The embodiments described in the embodiments of the present invention are for more clearly illustrating the technical solutions of the embodiments of the present invention, and do not constitute a limitation on the technical solutions provided by the embodiments of the present invention. Those skilled in the art can know that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present invention are equally applicable to similar technical problems.

[0103] In summary, the present invention proposes a fast-through millimeter-wave security inspection image reconstruction method and system. The millimeter-wave image reconstruction method and system include collecting echo data of a cylindrical calibration body twice, compensating phases of four different sliding windows to complete system phase correction; using a four-sliding-window processing method for the echo data of the target to be detected, combining the common grid division in the pitch direction and the range direction, and the horizontal grid division is based on the range of the azimuth direction of interest to complete the four sub- Figure 3Perform back-projection imaging through three-dimensional space grid division; further, through the design of the three-dimensional space grid image rotation bilinear interpolation method, solve the image blind area problem, with higher image quality, providing strong support for subsequent image recognition. This solution is applicable to the fast real-time imaging of the two-dimensional electronic scanning system, and is more suitable for the application scenario of rapid passing of millimeter-wave human bodies, which can effectively improve the image quality and recognition accuracy of the rapid passing millimeter-wave security inspection image reconstruction method.

[0104] Those skilled in the art can understand that Figures 1 to 8 the technical solutions shown in do not constitute a limitation to the embodiments of the present invention, and may include more or fewer steps than those shown, or combine certain steps, or different steps.

[0105] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0106] Those of ordinary skill in the art will understand that all or some of the steps in the methods disclosed above, and the corresponding systems, can be implemented as software, firmware, hardware, and their appropriate combinations. Some physical components or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include a computer-readable storage medium (or non-transitory medium) and a communication medium (or transitory medium). As is well known to those of ordinary skill in the art, the term computer-readable storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. The computer-readable storage medium includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, as is well known to those of ordinary skill in the art, a communication medium typically contains computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and can include any information delivery medium.

[0107] The preferred embodiments of the embodiments of the present invention have been described above with reference to the accompanying drawings, and thus do not limit the scope of the rights of the embodiments of the present invention. Any modifications, equivalent replacements, and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present invention shall be within the scope of the rights of the embodiments of the present invention.

Claims

1. A fast through-type millimeter-wave security inspection image reconstruction method, characterized in that, Comprising: Select a target frequency band, and transmit a stepped-frequency RF signal according to the frequency band through a transmitting array element; Collect echo data of a calibration object at two different positions through a receiving array element for phase calibration; Collect echo data of a target to be detected through the receiving array element, divide the echo data into a plurality of processing windows, and prepare to generate corresponding subgraphs for the data of each processing window; Define a three-dimensional space grid for each of the subgraphs, wherein the range and elevation grids in the three-dimensional space grid are divided based on a unified reference, and the azimuth grids are grouped and shared according to the azimuth coverage range of the processing window; Perform time-domain back-projection calculation on the data of each processing window with each three-dimensional space grid to obtain a plurality of complex images; Convert the complex images of each subgraph into modulus images, perform spatial transformation and projection processing on each modulus image, and output a two-dimensional security inspection image after fusion.

2. The fast-pass millimeter-wave security inspection image reconstruction method according to claim 1, wherein The step of dividing the echo data into a plurality of processing windows and preparing to generate corresponding subgraphs for the data of each processing window includes: Divide the echo data into four different sliding windows; Perform three-dimensional back-projection algorithm processing on the four sliding windows respectively to obtain subgraphs corresponding to the four sliding windows.

3. The rapid-pass millimeter-wave security inspection image reconstruction method according to claim 2, characterized in that, The step of defining a three-dimensional space grid for each of the subgraphs includes: The four subgraphs share range and elevation space grid coordinates divided based on a preset center zero point; Subgraph 1 and subgraph 4 share azimuth space grid coordinates, and subgraph 2 and subgraph 3 share azimuth space grid coordinates.

4. The fast-pass millimeter-wave security inspection image reconstruction method according to claim 2, characterized in that The step of collecting echo data of a calibration object at two different positions through a receiving array element for phase calibration includes: Place a cylindrical calibration object at a first position, collect first echo data of the cylindrical calibration object through a receiving array element, and use the first echo data as the phase compensation for sliding window 1 and sliding window 4 in the four sliding windows; Move the cylindrical calibration object to a second position, collect second echo data of the cylindrical calibration object through the receiving array element, and use the second echo data as the phase compensation for sliding window 2 and sliding window 3 in the four sliding windows; Realize the phase calibration of the four sliding windows according to the phase compensation.

5. The rapid-pass millimeter-wave security inspection image reconstruction method according to claim 1, wherein The step of performing time-domain back-projection calculation on the data of each processing window with each three-dimensional space grid to obtain a plurality of complex images includes: Traverse the three-dimensional space grid corresponding to each subgraph, and sequentially calculate the phase delays of each grid point in the azimuth, elevation, and range directions. The phase delays are calculated based on the propagation path of the signal from the transmitting array element to the grid point and then to the receiving array element; Construct an ideal echo signal at each grid point according to the phase delay. The ideal echo signal is in complex form and contains amplitude and phase information; Perform data processing on the ideal echo signal and the actually collected echo data after pulse compression to obtain the complex values of each grid point. By traversing all grid points, a complex image of the three-dimensional space grid is formed.

6. The rapid-pass millimeter-wave security inspection image reconstruction method according to claim 1, wherein The step of converting the complex images of each subgraph into modulus images, performing spatial transformation and projection processing on each modulus image, and outputting a two-dimensional security inspection image after fusion includes: Modulus processing is performed on the complex images of each of the sub - images to obtain the modulus images of each of the sub - images in the three - dimensional space grid; Rotational bilinear interpolation processing is performed on the modulus images of each of the sub - images, and maximum value projection is performed in the range direction to obtain the two - dimensional images of each of the sub - images; The two - dimensional images of each of the sub - images are fused to output a two - dimensional security inspection image.

7. The fast-pass millimeter-wave security inspection image reconstruction method according to claim 6, wherein The rotational bilinear interpolation processing is performed on the modulus images of each of the sub - images, and maximum value projection is performed in the range direction to obtain the two - dimensional images of each of the sub - images, including: Determine the rotation angle. According to the azimuth array element distribution of the processing window corresponding to each of the sub - images, determine the rotation direction and angle so that the rotated image is aligned with the preset projection direction; Perform rotation processing on the modulus images of each of the sub - images, and use the bilinear interpolation algorithm to calculate the modulus values of each pixel point in the rotated image; Perform maximum value projection on the rotated modulus image. Along the range - dimension, select the maximum value of the modulus values of each pixel point in the range direction, and use the maximum value as the modulus value of the pixel point in the two - dimensional image to obtain the two - dimensional images of each of the sub - images.

8. A fast-pass millimeter-wave security inspection device, characterized in that It includes a through - type folded array system placed parallel to each other on both sides of the channel. The system includes a transceiver array, and the transceiver array includes a plurality of transmit array elements and a plurality of receive array elements; The system is used for: Select a target frequency band, and the transmit array elements transmit stepped - frequency RF signals according to the frequency band; collect echo data of the calibration body at two different positions through the receive array elements for phase calibration; collect echo data of the target to be detected through the receive array elements, divide the echo data into multiple processing windows to prepare for generating corresponding sub - images for each of the processing windows; define a three - dimensional space grid for each of the sub - images. The range - direction and elevation - direction grids in the three - dimensional space grid are divided based on a unified reference, and the horizontal - direction grids are grouped and shared according to the azimuth coverage range of the processing window; perform time - domain back - projection calculation on the data of each of the processing windows with each three - dimensional space grid to obtain a plurality of complex images; convert the complex images of each of the sub - images into modulus images, and perform spatial transformation and projection processing on each of the modulus images, and fuse them to output a two - dimensional security inspection image.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the fast through - type millimeter - wave security inspection image reconstruction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, There is a computer program stored. When the computer program is executed by a processor, it implements the fast through - type millimeter - wave security inspection image reconstruction method according to any one of claims 1 to 7.

Images