Low-frequency multi-light-source terahertz human security imaging system

Abstract

The present application relates to the technical field of terahertz security imaging, in particular to a low-frequency multi-light-source terahertz human security imaging system, which is formed by the cooperation of an illumination module, an acquisition module and a processing module; the processing module generates an illumination parameter set, an acquisition parameter set and corresponding control instructions according to scanning task information, controls the illumination module to emit a low-frequency terahertz illumination beam, and controls the acquisition module to collect echo signals to generate terahertz imaging data, and then forms a security imaging result; the present application generates an illumination parameter set and an acquisition parameter set according to scanning task information by the processing module, and forms corresponding control instructions, so that the illumination module emits a low-frequency terahertz illumination beam and the acquisition module collects a low-frequency terahertz echo signal, which are matched; combined with multiple images and echo intensity data, a security imaging result is generated, and the imaging cooperation, integrity and reliability are improved.

Description

Technical Field

[0001] This invention relates to the field of terahertz security inspection imaging technology, and more specifically, to a low-frequency multi-source terahertz human body security inspection imaging system. Background Technology

[0002] Human body security imaging technology typically requires non-contact detection of target human body areas to obtain imaging results that reflect the contours of the human body surface, areas covered by clothing, and the distribution of suspicious attachments. Terahertz waves, located between microwaves and infrared waves, possess certain penetrating power and low photon energy, making them suitable for applications such as human body security checks, concealed object detection, and non-contact imaging. Low-frequency terahertz imaging exhibits good overall characteristics in terms of penetration power, imaging security, and adaptability to different human body areas, meeting the needs for scanning and imaging localized areas of the human body such as the chest, abdomen, waist, and legs.

[0003] Existing terahertz body scanning imaging solutions typically include terahertz illumination, echo reception, and image processing. By illuminating a human body area with a terahertz illumination beam and receiving the terahertz echo signal returned from that area, corresponding terahertz body images or echo intensity data can be obtained. This type of solution has already achieved body scanning imaging to a certain extent, providing a technological foundation for non-contact security checks.

[0004] However, in actual human body security checks, the target human body area is not a fixed planar structure. Different body parts vary in height, surface curvature, clothing thickness, degree of obstruction, and reflectivity. For example, the chest and abdomen area usually has a large positive illumination area, the sides of the waist are prone to oblique reflections and dark areas at the edges, and the leg area may have uneven echo intensity due to changes in posture. Accordingly, different scanning tasks have different requirements for parameters such as illumination frequency band, beam pointing, illumination power, illumination timing, sampling window, sampling frequency, receiver gain, and acquisition timing.

[0005] In the aforementioned scenarios, if lighting control and acquisition control are primarily executed according to fixed parameters, or if there is a lack of matching between lighting and acquisition parameters based on scanning task information, issues such as insufficient local lighting coverage, inadequate correspondence between echo sampling windows and lighting emission time windows, significant differences in echo intensity across different human body regions, and compromised imaging data integrity and stability may arise. This is particularly true in low-frequency, multi-source terahertz human body security imaging scenarios, where coordination between multiple illumination beams and multiple echo acquisitions is crucial in terms of time, space, and energy; otherwise, effective fusion of multiple terahertz human body images and echo intensity data can be compromised.

[0006] Therefore, this application proposes a low-frequency multi-source terahertz human body security inspection imaging system. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the present invention aims to provide a low-frequency multi-source terahertz human body security inspection imaging system.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A low-frequency multi-source terahertz human body security inspection imaging system includes:

[0010] It includes a lighting module, a data acquisition module, and a processing module;

[0011] The lighting module is used to emit low-frequency terahertz lighting beams toward the target human body area according to lighting control commands;

[0012] The acquisition module is used to acquire low-frequency terahertz echo signals returned from the target human body area according to the acquisition control command, and generate terahertz imaging data;

[0013] The processing module is communicatively connected to the lighting module and the acquisition module respectively. It is used to acquire the scanning task information corresponding to the target human body area, generate a set of lighting parameters and a set of acquisition parameters based on the scanning task information, generate lighting control commands based on the lighting parameter set, generate acquisition control commands based on the acquisition parameter set, and generate security inspection imaging results of the target human body area based on terahertz imaging data.

[0014] The lighting parameter set includes at least one of the following: lighting frequency band, beam pointing, lighting power, and lighting timing. The acquisition parameter set includes at least one of the following: sampling window, sampling frequency, receiving gain, and acquisition timing. The terahertz imaging data includes multiple terahertz human body images and echo intensity data.

[0015] In one embodiment, the processing module includes a scan task parsing unit, a parameter collaborative generation unit, and a control instruction generation unit.

[0016] In one embodiment, the scanning task parsing unit is used to extract the regional location, scanning range and imaging resolution requirements of the target human body region from the scanning task information.

[0017] The parameter co-generation unit is used to generate a set of matching illumination parameters and a set of acquisition parameters based on the region location, scanning range, and imaging resolution requirements.

[0018] The control command generation unit is used to convert the lighting parameter set into lighting control commands and the collected parameter set into collected control commands.

[0019] In one embodiment, the parameter co-generation unit is also used to establish a synchronous mapping relationship between the illumination timing and the acquisition timing, so that the low-frequency terahertz illumination beam transmission time window determined by the illumination timing matches the sampling window of the low-frequency terahertz echo signal, and adjusts the illumination power, beam pointing and receiving gain according to the synchronous mapping relationship.

[0020] In one embodiment, the lighting module includes a low-frequency terahertz source array, a beam modulation unit, and an lighting drive unit.

[0021] In one embodiment, the lighting driving unit is used to parse lighting control commands and drive a low-frequency terahertz source array to generate a low-frequency terahertz lighting beam.

[0022] The beam modulation unit is used to modulate the low-frequency terahertz illumination beam according to the beam direction and illumination power to form a controllable illumination field covering the target human body area.

[0023] In one embodiment, the acquisition module includes a low-frequency terahertz receiving array, a synchronization triggering unit, and a data reconstruction unit.

[0024] In one embodiment, the synchronization triggering unit is used to trigger the low-frequency terahertz receiving array to acquire low-frequency terahertz echo signals within the sampling window according to the acquisition control command.

[0025] The data reconstruction unit is used to perform time-domain alignment, intensity calibration, and image reconstruction on low-frequency terahertz echo signals, generating terahertz imaging data that includes multiple terahertz human images and echo intensity data.

[0026] In one embodiment, the processing module is further configured to evaluate the imaging quality of terahertz imaging data according to a human body partition template including multiple human body partitions, obtain imaging quality feedback corresponding to each human body partition, and correct the lighting parameter set and acquisition parameter set according to the imaging quality feedback, and iteratively update the lighting control command and acquisition control command.

[0027] In one embodiment, the processing module maps the human body partition template to multiple terahertz human body images and echo intensity data, calculates the image sharpness, echo signal-to-noise ratio, boundary integrity and illumination uniformity of each human body partition, and fuses them to generate the imaging quality feedback corresponding to each human body partition.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] This invention obtains scanning task information corresponding to the target human body area through a processing module, and generates a set of lighting parameters and a set of acquisition parameters based on the scanning task information. This enables a task matching relationship to be formed between the lighting frequency band, beam pointing, lighting power, lighting timing and the sampling window, sampling frequency, receiving gain and acquisition timing, which is beneficial to improving the coordination between low-frequency terahertz lighting beam transmission and low-frequency terahertz echo signal acquisition.

[0030] This invention generates illumination control commands and acquisition control commands through a processing module. The illumination module emits a low-frequency terahertz illumination beam toward the target human body region according to the illumination control commands, and the acquisition module acquires the low-frequency terahertz echo signal returned from the target human body region according to the acquisition control commands. This allows different target human body regions to obtain illumination coverage conditions and echo acquisition conditions adapted to the scanning task, which is beneficial to improving the stability and integrity of the human body region imaging process.

[0031] This invention generates terahertz imaging data, including multiple terahertz human body images and echo intensity data, through an acquisition module. The processing module then generates security inspection imaging results for the target human body area based on the terahertz imaging data. This allows the security inspection imaging results to take into account both image information and echo intensity information, which helps to improve the clarity, effectiveness, and reliability of the security inspection imaging results for the target human body area. Attached Figure Description

[0032] Figure 1 This is an overall structural block diagram of the low-frequency terahertz human body imaging system provided in an embodiment of the present invention;

[0033] Figure 2 This is a block diagram of the internal structure of the processing module;

[0034] Figure 3 This is a structural block diagram of the lighting module;

[0035] Figure 4 This is a structural block diagram of the data acquisition module;

[0036] Figure 5 This is a schematic diagram of the large lens assembly of a terahertz lens module.

[0037] Figure 6 This is a schematic diagram of the structure of a terahertz lens module. Detailed Implementation

[0038] Reference Figures 1 to 4 A low-frequency multi-source terahertz human body security inspection imaging system, comprising:

[0039] In this embodiment, the low-frequency terahertz source array includes A low-frequency terahertz light source group that can be driven independently. , No. The locations of the low-frequency terahertz light source groups are as follows: Low-frequency terahertz receiver arrays include The receiving sub-area, the first The location of each receiving sub-region is The target human body region is divided into: Personal body partitions, number Individual body partitions are recorded as The central position is The origin of the coordinate system is the midpoint of the intersection of the vertical plane at the center of the passage entrance and the ground. The width of the human body is The axis, the direction of travel of the channel is The axis is the height direction of the human body. Axis. The first. The operating frequency of the low-frequency terahertz light source group In to Within the range, among which, This refers to the speed at which electromagnetic waves propagate in air. The low-frequency terahertz light source group is arranged along the height of the human body, the width of the human body, or both sides of the passage. The light source group includes at least two of the following: an upper light source group, a middle light source group, a lower light source group, a left-side light source group, and a right-side light source group.

[0040] Equipment geometric calibration data are denoted as These data are obtained from equipment installation calibration or factory calibration. Equipment geometric calibration data includes the coordinate origin. Positions of each low-frequency terahertz light source group Location of each receiving sub-area Geometric center position of low-frequency terahertz receiver array Relative positions of phased array elements The mounting attitude of the light source group, the mounting attitude of the receiving sub-region, and the element spacing. The geometric center position of the low-frequency terahertz receiving array. The location is determined based on the geometric center of each receiving sub-area. The processing module uses the equipment's geometric calibration data. Calculate the propagation path, beam direction, sampling window, and spatial location of image pixels.

[0041] The illumination module corresponds to the illumination emission section of the low-frequency multi-source terahertz human body security imaging system. Based on the illumination control commands generated by the processing module, it provides a low-frequency terahertz illumination beam to the target human body area. This module corresponds to the illumination frequency band, beam direction, illumination power, and illumination timing, undertaking tasks such as illumination energy output, illumination direction organization, and illumination rhythm coordination. During the human body security imaging process, the illumination module forms low-frequency terahertz illumination conditions facing the target human body area, providing the illumination basis for the low-frequency terahertz echo signal returned from the target human body area.

[0042] After receiving the lighting control command from the processing module, the lighting drive unit analyzes the lighting frequency band, lighting timing, beam pointing and lighting power in the lighting control command, and drives multiple low-frequency terahertz light sources in the low-frequency terahertz source array in groups according to the analysis results.

[0043] For example, when the target human body area includes the chest, abdomen, waist and legs, the low-frequency terahertz source array is divided into an upper light source group, a middle light source group and a lower light source group. Each light source group corresponds to a different human body height area. The illumination driving unit independently configures the start-up time, radiation frequency band and output power of each light source group according to the regional position and scanning range of the target human body area in the scanning task information. This allows each light source group to emit low-frequency terahertz illumination beams in different time windows, avoiding echo aliasing caused by multiple light sources irradiating at the same time, and improving the illumination coverage integrity of the curved parts, obstructed parts and edge parts of the human body.

[0044] Record any two lighting acquisition parameters within the same scheduling cycle and When corresponding to different light source groups or different human body zones Record the sequence number of another lighting acquisition parameter, whose lighting scheduling reference time satisfies the anti-aliasing constraint:

[0045]

[0046] in, and These are the light source group numbers from the two lighting acquisition parameter records. and These are the human body partition numbers from the two lighting data acquisition parameter records. It is the maximum value among the illumination emission durations or equivalent illumination integral times within the same scheduling cycle. It is the maximum value among the arrival delays of all echoes within the same scheduling period. It is the minimum arrival delay of all echoes within the same scheduling period. This is a timing protection interval. Within the same receiving sub-area, the sampling windows corresponding to different lighting acquisition parameter records do not overlap. If two sampling windows overlap in time, the two lighting acquisition parameter records use different lighting frequencies. Different modulation codes Or different receiving sub-regions The following distinctions are made: The interval between different lighting frequencies is no less than the sum of the receiving filter bandwidth and the frequency protection interval, and they are separated according to frequency domain filtering. Different modulation codes are separated according to correlation demodulation, and the cross-correlation peak value between modulation codes of different light source groups is lower than a preset cross-correlation threshold. Different receiving sub-regions are separated according to channel number.

[0047] The beam modulation unit adjusts the direction and shapes the energy of the low-frequency terahertz illumination beam output by the low-frequency terahertz source array according to the beam direction and illumination power determined in the illumination control command.

[0048] The beam modulation unit includes at least one of a phased array phase-shifting network, a quasi-optical lens group, and a mirror scanning mechanism. The low-frequency terahertz light source group points to the first Distance between individual body sections Based on the position of the light source group With the center position of the human body partition The Euclidean distance between them is determined. The low-frequency terahertz light source group points to the first Beam direction vector of individual body partition satisfy:

[0049]

[0050] When the beam modulation unit uses a phased array phase-shifting network, the illumination wavelength of the i-th low-frequency terahertz light source group is... Based on propagation speed With illumination frequency Confirmed, number The first low-frequency terahertz light source group Each array element faces the first Phase of individual body regions satisfy:

[0051]

[0052] Incident angle correction Take beam direction vector External normal vector of human body partition The absolute value of the dot product, the minimum angle of incidence correction The larger of the two. The first low-frequency terahertz light source group faces the first Initial illumination power for individual body zones satisfy:

[0053]

[0054] in, For array element sequence number, For the first The first low-frequency terahertz light source group The position vectors of each array element relative to the phase center of the light source group Basic lighting power, For the first Estimates of reflectance coefficients for individual body regions. To prevent zero constant, This is the correction amount for the angle of incidence. For the first The external normal vector of an individual body partition. This is the minimum incident angle correction amount. This is the lower limit of lighting power. This is the upper limit for lighting power. Indicates will Limited to to Within the range. When historical scan data is unavailable, the estimated reflectance coefficient for human body zones. Use the factory default value for reflection coefficient If historical scan data exists, update based on historical partition echo intensity. Human body partition external normal vector Estimated based on the human body outline bounding box, human body partition template, or the terahertz human body outline from the previous frame; unable to estimate. season .

[0055] For example, when the target human body area is the area on both sides of the waist, the beam modulation unit modulates the low-frequency terahertz illumination beam from the central light source group into a left oblique incidence beam and a right oblique incidence beam, respectively, and appropriately increases the illumination power in the waist contour edge area and appropriately decreases it in the flat front area to form a controllable illumination field covering the target human body area. This controllable illumination field has a zoned beam direction and power distribution within the illumination range, which can adapt to the reflection differences of different parts of the human body and reduce the impact of local dark areas and overexposed areas on the acquisition of low-frequency terahertz echo signals.

[0056] The acquisition module corresponds to the echo reception and data formation section of the low-frequency multi-source terahertz human body security inspection imaging system. It receives low-frequency terahertz echo signals returned from the target human body area according to acquisition control commands and forms terahertz imaging data. This module corresponds to the sampling window, sampling frequency, receiving gain, and acquisition timing, undertaking tasks such as echo signal reception, echo intensity acquisition, and the formation of multiple terahertz human body images. The data acquired by the acquisition module includes multiple terahertz human body images and echo intensity data, providing a data source for the processing module to generate security inspection imaging results of the target human body area.

[0057] After receiving the acquisition control command, the synchronous triggering unit analyzes the sampling window, sampling frequency, receiving gain and acquisition timing in the acquisition control command, and performs partitioned triggering of the low-frequency terahertz receiving array according to the location distribution of the target human body area.

[0058] No. Sampling frequency of each receiving sub-region The settings are based on the effective bandwidth of the received echo and the range resolution sampling requirements. The effective bandwidth of the received echo of each receiving sub-cell is denoted as . , No. The distance resolution interval corresponding to the individual body partition is denoted as Distance resolution interval Based on target imaging resolution Conversion factor with distance Multiply to obtain, The calibration coefficient is greater than zero and not greater than one. The effective bandwidth of the demodulated beat frequency signal under continuous wave or frequency-modulated continuous wave systems is denoted as... Sampling frequency Not less than the effective bandwidth of the received echo Twice as fast as the transmission speed Divide by twice the distance resolution interval The obtained frequency, and in continuous wave or FM continuous wave mode, is not lower than the effective bandwidth of the demodulated beat frequency signal. Twice as much.

[0059] No. The receiving sub-area faces the first Initial receive gain of individual body partition Set based on the estimated echo amplitude. Assume the current lighting acquisition parameters are recorded. Middle and the first The receiving sub-region and the first The light source group number for individual body partition matching is , No. The receiving sub-region to the first The distance from the center of an individual body section is , No. The channel response calibration coefficient for each receiving sub-region is: , No. The standard deviation of the background noise in each receiving sub-region is: The target reference amplitude is The estimated echo amplitude is Then the following conditions are met:

[0060]

[0061]

[0062] in, Location of receiving sub-area With the center position of the human body partition The Euclidean distance between them is determined. This is the lower limit of the receive gain. This is the upper limit of the receive gain. Obtained from the receive channel response test. Acquired from an empty background. Determined by the equipment's factory calibration.

[0063] For example, when the target human body area includes the chest, abdomen, waist, and legs, the low-frequency terahertz receiving array is divided into an upper receiving area, a middle receiving area, and a lower receiving area. Each receiving area corresponds to a different range of human body height. The synchronization triggering unit controls each receiving area to receive the low-frequency terahertz echo signal returned from the target human body area within the corresponding sampling window according to the acquisition timing. The start time and duration of the sampling window are matched with the illumination timing to reduce the time overlap between the return signals from different illumination beams. For areas such as the sides of the waist, the armpit edges, or areas with many clothing folds, the synchronization triggering unit increases the receiving gain of the corresponding receiving area and extends the sampling window of that area to enhance the effective acquisition of weak echo areas.

[0064] After receiving the low-frequency terahertz echo signal output by the low-frequency terahertz receiving array, the data reconstruction unit performs time-domain alignment, intensity calibration and image reconstruction on the echo signals obtained from different receiving areas.

[0065] Let the first The receiving sub-region in the first The raw echo sequence obtained within the individual body partition sampling window is The sampling point number, This represents the total number of sampling points. (The first...) The background echo sequence of each receiving sub-region is denoted as ,Depend on The frame was obtained by averaging the empty field data collected when no human was passing through, point by point at the same sampling location. The echo sequence was then calibrated for intensity. satisfy:

[0066]

[0067] The background echo sequence is updated when the device is started, when the empty field condition lasts for a preset time, when the background acquisition cycle is reached, when the ambient temperature changes beyond a preset temperature threshold, or when the ambient humidity changes beyond a preset humidity threshold. The number of frames captured for an empty background. For the first Channel response calibration coefficients for each receiving sub-region For the first The receiving sub-area faces the first The receiver gain of the individual human body partition, To prevent zero constant, echo intensity data includes peak intensity, mean intensity, or integral intensity; this embodiment uses integral intensity. Let the image pixel point... The corresponding set of sampling points is Based on pixels The propagation path and sampling window are determined, and the set of receiving sub-regions is... For the first The fusion weight of the receiving sub-region, then the... Individual body partitions at pixels Echo intensity data at the location satisfy:

[0068]

[0069] Among them, time-domain alignment unifies the time reference of multiple low-frequency terahertz echo signals according to the start time of each sampling window, the propagation path of the illumination beam and the position of the receiving channel; intensity calibration normalizes the echo amplitude of each receiving channel according to the receiving gain, channel response differences and background echo intensity; and image reconstruction maps the echo signals after time-domain alignment and intensity calibration into multiple terahertz human images and generates echo intensity data simultaneously.

[0070] During time-domain alignment, the data reconstruction unit maps sampling points to a unified time axis according to the start time of the sampling window and the sampling frequency. Let the first... The number of pixels in the image of a personal body partition is pixel The corresponding human body spatial location is , No. Each receiving sub-region and pixel The matched sampling point number is The current lighting acquisition parameter record shows the light source group number as follows: Then the following conditions are met:

[0071]

[0072] No. Terahertz human images of individual body regions satisfy:

[0073]

[0074] in, Indicates rounding down. For pixels Corresponding human body spatial position To participate in the The set of receiving subregions for individual body partition reconstruction. For the first The fusion weight of each receiving sub-region For the first In each receiving sub-region, the pixel point Matched intensity calibration echo value.

[0075] For example, if there are metallic foreign objects or highly reflective attachments in the waist area, the echo intensity data of the corresponding area will show local enhancement, and the corresponding position will show a stable bright outline in multiple terahertz human body images. Based on this, the security inspection imaging results can obtain clearer imaging information of the human body area.

[0076] The processing module corresponds to the task parsing, parameter generation, command formation, and imaging processing parts of the low-frequency multi-source terahertz human body security inspection imaging system. Based on the scanning task information corresponding to the target human body area, it coordinates the control of the illumination and acquisition modules. It generates illumination and acquisition parameter sets based on the scanning task information, and forms illumination and acquisition control commands accordingly. The processing module also combines terahertz imaging data to complete image processing and result generation for the target human body area. The security inspection imaging results reflect the imaging situation of the target human body area under low-frequency terahertz illumination and echo acquisition conditions.

[0077] The scanning task information is formed from at least one of the following: channel entrance trigger signal, human contour detection results, manually selected scanning mode, and results from the previous terahertz imaging round. The scanning task information corresponding to the individual body partition is recorded as follows: The scanning task information includes the task number, the human body partition number, and the human body partition bounding box. Human body zone center position Scanning coverage and target imaging resolution And scan priority. Human body partition bounding box. It consists of the minimum and maximum coordinates along the x, y, and z axes. The center of the individual body section It is determined by the midpoint between the minimum and maximum coordinates of the human body partition bounding box in the x, y, and z axes. The distance from the human body partition to the geometric center of the low-frequency terahertz receiving array is determined by the Euclidean distance between the center of the human body partition and the geometric center of the low-frequency terahertz receiving array.

[0078] After receiving the scanning task information corresponding to the target human body region, the scanning task parsing unit parses the region location, scanning range and imaging resolution requirements in the scanning task information. The region location includes the distribution coordinates of the target human body region in the height and width directions of the human body. The scanning range includes the chest, abdomen, waist, legs or a combination of the above regions. The imaging resolution requirements include the imaging detail requirements of the human body contour edges, clothing-covered areas and suspected high reflectivity areas.

[0079] For example, when the target human body area is the sides of the waist and the front of the abdomen, the scanning task analysis unit extracts the oblique coverage area of ​​the sides of the waist, the frontal coverage area of ​​the abdomen, and the high imaging resolution requirements at the edge contours. The parameter co-generation unit generates a matching set of illumination parameters and a set of acquisition parameters based on the above-mentioned area location, scanning range, and imaging resolution requirements. The illumination parameter set includes illumination frequency band, beam pointing, illumination power, and illumination timing. The acquisition parameter set includes sampling window, sampling frequency, receiving gain, and acquisition timing. A synchronous mapping relationship is established between the illumination timing and the acquisition timing, so that the low-frequency terahertz illumination beam transmission time window determined by the illumination timing corresponds to and matches the sampling window of the low-frequency terahertz echo signal. A narrower beam pointing and higher illumination power are configured in the waist edge area, and a wider beam pointing and more stable illumination power are configured in the front of the abdomen area. The corresponding sampling window, sampling frequency, and receiving gain are synchronously matched.

[0080] The parameter co-generation unit organizes the lighting parameter set and the acquisition parameter set into a lighting acquisition parameter record. , Record the sequence number for each lighting acquisition parameter. Record each lighting acquisition parameter record. This corresponds to one human body zone, one low-frequency terahertz light source group, and one low-frequency terahertz receiver sub-region. The illumination acquisition parameter record includes the human body zone number. Light source group number Receive sub-area sequence number Lighting frequency Illumination modulation code Beam direction vector Lighting power Illumination emission time window, sampling window, sampling frequency Receiver gain And acquisition timing. The illumination emission time window is determined by the illumination scheduling reference time. and illumination emission duration or equivalent illumination integral time Confirmed; the sampling window is determined by the start time of the sampling window. and the end time of the sampling window Sure.

[0081] The parameter co-generation unit calculates the echo arrival delay based on the location of the low-frequency terahertz light source group, the center location of the human body partition, and the location of the receiving sub-region. The first low-frequency terahertz light source group was... After individual body part reflection, it reaches the first Echo arrival delay of each receiving sub-region satisfy:

[0082]

[0083] Echo arrival delay after channel delay calibration Echo arrival delay Add the Channel delay calibrated for each receiving sub-region The sampling window start and end times satisfy the following:

[0084]

[0085]

[0086] in, For Euclidean distance, For the first Channel delay calibration of each receiving sub-region For the pre-protection time, This is the post-protection time. When the system uses continuous wave, frequency-modulated continuous wave, or integrating imaging, the sampling window is set around the arrival time of the echo after the channel delay calibration, serving as the demodulation integration window or frame acquisition window.

[0087] The control command generation unit converts the lighting parameter set into lighting control commands and the acquisition parameter set into acquisition control commands. After the low-frequency terahertz lighting beam and the low-frequency terahertz echo signal complete the lighting and acquisition according to the synchronous mapping relationship, terahertz imaging data including multiple terahertz human body images and echo intensity data is obtained.

[0088] The human body partitioning template uses the bounding box of the human body's outer contour in the terahertz human body image as the mapping reference. The template divides the body according to the normalized position of pixels within the height direction of the bounding box. The normalized position is determined by the pixel height, the lower boundary of the bounding box's height, and the upper boundary of the height. Pixels with a normalized height between 0.75 and 1 are assigned to the head and shoulders region, those between 0.45 and 0.75 to the chest and abdomen region, those between 0.30 and 0.45 to the waist region, and those between 0 and 0.30 to the legs region. When auxiliary human body contour information exists, the bounding box is corrected based on externally input contour information, channel positioning information, the previous frame's terahertz contour information, or manually configured templates. When auxiliary human body contour information does not exist, the bounding box of the human body in the terahertz human body image is used as the mapping reference. The template boundary of the human body partitioning is obtained by mapping the human body partitioning template to the coordinates of the current terahertz human body image according to the normalized height.

[0089] The human body partitioning template is mapped to multiple terahertz human body images and echo intensity data according to multiple human body partitions such as head and shoulder area, chest and abdomen area, waist area, and leg area. The image sharpness, echo signal-to-noise ratio, boundary integrity and illumination uniformity are calculated for each human body partition, and the images are fused to generate the corresponding imaging quality feedback for each human body partition.

[0090] No. The image clarity of a personal body partition is denoted as The echo signal-to-noise ratio is denoted as Boundary integrity is denoted as Illumination uniformity is denoted as The target area is denoted as The target area is the area covered by the human body partition template; the background area is denoted as... The background area includes an empty background area, the extended neighborhood of a human body area, or the background acquisition area when no human body is passing through. The average echo intensity of the target area... The standard deviation of background noise is calculated based on the mean of echo intensity data within the target area. Image sharpness is calculated based on the standard deviation of echo intensity data within the background area. Using partitioned terahertz human body images Calculation of Laplace response variance; echo signal-to-noise ratio Calculated using the average echo intensity of the target area and the standard deviation of background noise; boundary integrity. Calculations were made using the lengths of the detection boundary and the template boundary; illumination uniformity. Using echo intensity data The mean and standard deviation are calculated to satisfy:

[0091]

[0092]

[0093]

[0094]

[0095] in, For variance calculation, For the first Laplacian response of terahertz human images in individual human body regions For the first The detection boundary length for individual human body regions is obtained from terahertz human images through gradient operators, threshold segmentation, or contour tracking. For the first The template boundary length of the individual human body partition is obtained by mapping the human body partition template to the coordinates of the current terahertz human body image. For the first Mean value of echo intensity data for individual body regions. For the first Standard deviation of individual body region echo intensity data.

[0096] Image sharpness, echo signal-to-noise ratio, boundary integrity, and illumination uniformity are linearly normalized according to their respective upper and lower normalization limits, and restricted to the range of 0 to 1. The normalization results are denoted as follows: and . No. Imaging quality feedback of individual body regions satisfy:

[0097]

[0098] in, and These are the fusion weights for image sharpness, echo signal-to-noise ratio, boundary integrity, and illumination uniformity, respectively, and the sum of the four fusion weights is one. In a set of calibration examples, Quality threshold Set the signal-to-noise ratio threshold to 0.70. Pick Boundary integrity threshold Set the threshold value to 0.80 for lighting uniformity. Take 0.75.

[0099] Equipment parameters are determined based on factory calibration samples, on-site background data, receiver channel response test data, environmental status records, and historical scan data. Operational data is generated based on current scan task information, equipment geometric calibration data, lighting acquisition parameter records, and low-frequency terahertz receiver array output data. The thresholds, weights, and calibration examples mentioned above are not limited to unique, fixed values.

[0100] For example, when the echo signal-to-noise ratio of the waist region is lower than that of the chest and abdomen region and the boundary integrity is insufficient, the illumination power, beam pointing, receiving gain and sampling window corresponding to the waist region are corrected. The updated illumination control command and acquisition control command continue to participate in the next round of target human body region imaging. The security inspection imaging results are formed under the joint constraints of multiple terahertz human body images and echo intensity data.

[0101] The processing module follows the image quality feedback. echo signal-to-noise ratio Boundary integrity and lighting uniformity The illumination power, receiver gain, and sampling window are iteratively corrected. Let the... In the first round of imaging The first low-frequency terahertz light source group faces the first The lighting power of each individual body zone is , No. The receiving sub-area faces the first The receiver gain of the individual body partition is , No. The first light source group and the first The receiving sub-region corresponds to the first The sampling window start time for individual body partitions is The lighting power update step size is The receive gain update step size is Overexposure suppression step size is The sampling window correction step size is The measured echo arrival delay is Calculate the echo arrival delay as , No. In the first round of imaging The maximum echo intensity of an individual body region is Then the following conditions are met:

[0102]

[0103]

[0104]

[0105] in, For the first In the first round of imaging Imaging quality feedback for individual body regions For the first In the first round of imaging The echo signal-to-noise ratio of individual body regions, Obtained from the peak position of the intensity-calibrated echo sequence or the position of the relevant peak. and Determined by factory calibration or on-site calibration data. When the beam direction vector is below the boundary integrity threshold Angle correction is performed based on the center position of the dark area at the boundary of the human body partition; Below the illumination uniformity threshold and When the overexposure intensity exceeds the overexposure threshold, reduce the illumination power of the light source group corresponding to the overexposed sub-region and increase the illumination power of the light source group corresponding to the dark area. All human body partition quality feedback reaches the quality threshold, and the number of iterations reaches the maximum number of iterations. The change in quality feedback between two adjacent rounds is less than the threshold for change in quality feedback. When any of the conditions is met, the processing module ends the parameter correction and outputs the security inspection imaging results.

[0106] Security inspection imaging results include terahertz human body images, echo intensity data, zoned imaging quality feedback, and abnormal area markings. High echo threshold, low echo threshold, local contrast threshold, and boundary abrupt change threshold are determined based on corresponding threshold coefficients, zoned echo intensity statistics, factory calibration samples, or historical scan data. Local contrast is calculated based on the difference between the echo intensity of a pixel and the average echo intensity of its neighboring area; boundary gradient magnitude is calculated based on the gradient magnitude of the terahertz human body image at the corresponding pixel. A pixel echo intensity higher than the high echo threshold is marked as a high echo anomaly; a pixel echo intensity lower than the low echo threshold is marked as a low echo depression anomaly; local contrast exceeding the local contrast threshold is marked as a local contrast anomaly; and boundary gradient magnitude exceeding the boundary abrupt change threshold is marked as a boundary abrupt change anomaly. Pixels The abnormal region is marked as .

[0107] When confirming consecutive frames, count the current frame and the frames before the current frame. The number of times anomaly regions are marked within a frame. The number of anomaly region markings reaches the consecutive frame confirmation threshold. At that time, it was confirmed that the corresponding pixel belonged to an abnormal region. The number of frames already captured was insufficient to meet the required number of consecutive frames. At this time, already acquired frames are used for confirmation, and the valid consecutive frame confirmation threshold is set to a value no greater than the number of already acquired frames. Abnormal areas after consecutive frame confirmation are marked as... Security screening imaging results include terahertz images of the human body in various body sections. Echo intensity data Continuous frame confirmation of abnormal region markers and image quality feedback .

[0108] Reference Figures 5 to 6 The terahertz lens module mainly consists of six core parts: lens barrel body, optical lens assembly, rotation positioning mechanism, large lens group, axial sliding adjustment mechanism, and locking assembly. The large lens group is coaxially assembled as a whole, and its shape and size are compatible with traditional fixed lens modules, so it can directly replace the original lens.

[0109] 1. Optical lens assembly and rotation adjustment structure

[0110] The optical lens is mounted inside a ring-shaped lens mount. The outer circumference of the lens mount is uniformly machined with anti-slip serrations. The lens mount and the inner wall of the lens barrel are fitted with a clearance fit, allowing for 360° rotation around the central axis of the lens barrel, enabling continuous lens adjustment. A miniature adjustment window is provided on the side wall of the lens barrel, with a locking screw corresponding to the position of the lens mount. After the external rotation of the lens mount completes the angle adjustment, tightening the locking screw locks the lens mount, preventing lens displacement during equipment operation. The lens rotation adjustment range can be controlled within ±30° according to usage requirements, meeting optical path correction needs.

[0111] 2. Large lens assembly and axial distance adjustment structure

[0112] The large lens assembly is mounted on a sliding base. Guide ridges are located on the outer side of the sliding base, and corresponding guide grooves are formed on the inner wall of the lens barrel. The sliding base can slide linearly along the guide grooves, thereby changing the relative distance between the large lens assembly and the front optical lens, achieving continuous focal length adjustment. The operator can rotate the lens barrel to move the camera horizontally. After adjustment, scale lines are marked to accurately record the adjustment position, facilitating later reuse and batch calibration.

[0113] 3. Locking and limiting structure

[0114] 1. Lens locking inside the lens assembly: Multi-point elastic set screws are used to distribute force evenly, avoid lens deformation under pressure, and ensure stable optical performance.

[0115] 2. Lens assembly locking: After the focus is adjusted, tighten the locking screws by hand to fix the lens assembly in the designated position of the adjustment slot and lock the lens spacing.

[0116] 3. End-to-end limiting: Limiting rings are installed at the front and rear ends inside the lens barrel to limit the maximum adjustment range of the lens mount and large lens group, preventing parts from falling off or causing optical damage due to over-adjustment.

[0117] 4. Assembly and Working Principle

[0118] Overall assembly sequence: optical lens assembly → lens barrel body → large lens group → axial sliding base → rotation positioning mechanism → rotating lens.

[0119] Workflow:

[0120] ① After installing the equipment on site, observe the image. Based on the image offset, loosen the lens locking screws, rotate the lens mount to adjust the optical lens angle, and correct the incident light path until the imaging field of view is complete and the image is centered.

[0121] ② Based on the detection distance and clarity requirements, loosen the locking nut, rotate it to move the large lens group axially, change the lens group spacing, adjust the focal length, and optimize the image clarity;

[0122] ③ After all parameters are adjusted, tighten all locking parts in sequence to lock the structure, and the equipment can then operate stably for a long time.

[0123] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A low-frequency multi-source terahertz human body security inspection imaging system, characterized in that, include: It includes a lighting module, a data acquisition module, and a processing module; The lighting module is used to emit low-frequency terahertz lighting beams toward the target human body area according to lighting control commands; The acquisition module is used to acquire low-frequency terahertz echo signals returned from the target human body area according to the acquisition control command, and generate terahertz imaging data; The processing module includes a scan task parsing unit, a parameter collaborative generation unit, and a control command generation unit; The scanning task parsing unit is used to extract the regional location, scanning range, and imaging resolution requirements of the target human body area from the scanning task information. The parameter co-generation unit is used to generate a set of matching illumination parameters and a set of acquisition parameters based on the region location, scanning range, and imaging resolution requirements. The control command generation unit is used to convert the lighting parameter set into lighting control commands and the collected parameter set into collected control commands; The parameter co-generation unit is also used to establish a synchronous mapping relationship between the illumination timing and the acquisition timing, so that the low-frequency terahertz illumination beam transmission time window determined by the illumination timing matches the sampling window of the low-frequency terahertz echo signal, and adjusts the illumination power, beam pointing and receiving gain according to the synchronous mapping relationship. The processing module is communicatively connected to the lighting module and the acquisition module respectively. It is used to acquire the scanning task information corresponding to the target human body area, generate a set of lighting parameters and a set of acquisition parameters based on the scanning task information, generate lighting control commands based on the lighting parameter set, generate acquisition control commands based on the acquisition parameter set, and generate security inspection imaging results of the target human body area based on terahertz imaging data. The lighting parameter set includes at least one of the following: lighting frequency band, beam pointing, lighting power, and lighting timing. The acquisition parameter set includes at least one of the following: sampling window, sampling frequency, receiving gain, and acquisition timing. The terahertz imaging data includes multiple terahertz human body images and echo intensity data.

2. The low-frequency multi-source terahertz human body security inspection imaging system according to claim 1, characterized in that, The lighting module includes a low-frequency terahertz source array, a beam modulation unit, and a lighting drive unit.

3. The low-frequency multi-source terahertz human body security inspection imaging system according to claim 2, characterized in that, The lighting drive unit is used to parse lighting control commands and drive the low-frequency terahertz source array to generate a low-frequency terahertz lighting beam. The beam modulation unit is used to modulate the low-frequency terahertz illumination beam according to the beam direction and illumination power to form a controllable illumination field covering the target human body area.

4. The low-frequency multi-source terahertz human body security inspection imaging system according to claim 3, characterized in that, The acquisition module includes a low-frequency terahertz receiving array, a synchronization triggering unit, and a data reconstruction unit.

5. The low-frequency multi-source terahertz human body security inspection imaging system according to claim 4, characterized in that, The synchronous triggering unit is used to trigger the low-frequency terahertz receiving array to acquire low-frequency terahertz echo signals within the sampling window according to the acquisition control command; The data reconstruction unit is used to perform time-domain alignment, intensity calibration, and image reconstruction on low-frequency terahertz echo signals, generating terahertz imaging data that includes multiple terahertz human images and echo intensity data.

6. A low-frequency multi-source terahertz human body security inspection imaging system according to any one of claims 1-5, characterized in that, The processing module is also used to evaluate the imaging quality of terahertz imaging data according to a human body partition template that includes multiple human body partitions, obtain imaging quality feedback corresponding to each human body partition, and correct the lighting parameter set and acquisition parameter set according to the imaging quality feedback, and iteratively update the lighting control command and acquisition control command.

7. A low-frequency multi-source terahertz human body security inspection imaging system according to claim 6, characterized in that, The processing module maps the human body partition template to multiple terahertz human body images and echo intensity data, calculates the image sharpness, echo signal-to-noise ratio, boundary integrity and illumination uniformity of each human body partition, and fuses them to generate the corresponding imaging quality feedback for each human body partition.

Images