Image reconstruction method based on millimeter wave human body security check imaging data processing

Abstract

The invention discloses an image reconstruction method based on millimeter wave human body security check imaging data processing, and belongs to the technical field of millimeter wave three-dimensional holographic imaging. According to the invention, on the basis of initial three-dimensional holographic imaging data, the effective three-dimensional imaging data of each block area is obtained by dividing the block areas, and the three-dimensional holographic imaging data can be obtained on the premise of retaining the information of human bodies and hidden articles. Additional interference signals such as useless backgrounds, clutters, reflection, noise interference and state drifting in front of, behind and on two sides of different heights and parts of a human body are removed or eliminated as far as possible, interference is directly eliminated and reduced from a data source, the image quality is improved from the source, and the detection and false alarm performance of equipment is greatly improved; after interference is eliminated and reduced on a data source, the weak target is further highlighted, the weak target detection rate of equipment is improved, the false alarm rate is reduced, and the detection performance bottleneck of the equipment is remarkably reduced by reducing background noise signals within the same range as the distance of a human body and through nonlinear correction.

Description

Technical Field

[0001] The present invention relates to the technical field of millimeter wave three-dimensional holographic imaging, and in particular to an image reconstruction method based on millimeter wave human body security inspection imaging data processing. Background Art

[0002] At present, active millimeter wave human body security inspection equipment has attracted much attention for its advantages of high resolution, high precision, strong target recognition and anti-interference ability, and is widely used in airports and other scenarios. Millimeter wave human body security inspection equipment mainly adopts the range migration (RMA) algorithm, phase shift migration (PSM) algorithm, error back propagation (BP) algorithm, etc. to realize the conversion from receiving echo signals from the device to three-dimensional holographic imaging data and display imaging images. Usually, the three-dimensional holographic imaging data obtained by echo signals are inevitably mixed with various interferences, including millimeter wave background or noise interference signals generated by the surrounding environment and other electronic devices at different heights and near different parts of the human body; interference is generated between the two arms and torso of the human body and between the two thighs due to multiple reflections of millimeter waves; during the long-term use of the equipment, the millimeter wave module devices generate some noise or interference in the security inspection area due to performance differences and temperature drift.

[0003] These backgrounds, clutter, reflections, noise interference, and state drift appear and exist in the initial three-dimensional holographic imaging data obtained by the algorithm. As the data source for later image processing and target recognition, these data cannot be retroactively removed by later image enhancement, noise reduction, filtering, contrast improvement, and clarity improvement. Although the millimeter wave security inspection equipment has strong anti-interference ability and can achieve personnel security inspection and identification of strong signal targets without being affected, these interferences will cause various invalid and false targets, seriously affecting the detection of weak signal targets, resulting in bottlenecks in weak target detection and false alarms, which is also one of the key performance bottlenecks of the equipment. It is difficult to significantly improve the performance of the equipment if a trade-off is required. Continuously optimizing and improving the target recognition performance of the equipment and improving the detection rate of weak targets will inevitably lead to a significant increase in the false alarm rate of the equipment. In order to ensure the availability of the equipment and reduce the false alarm rate, the detection rate of weak targets will decrease and the missed detection rate will increase significantly. This cannot guarantee the good use and wide application of the equipment. The above problems need to be solved urgently. To this end, an image reconstruction method based on millimeter wave human body security inspection imaging data processing is proposed. Summary of the invention

[0004] The technical problem to be solved by the present invention is: how to eliminate invalid and false targets from the source of initial three-dimensional holographic imaging data, highlight weak targets, significantly improve the target recognition detection rate of the device, reduce the false alarm rate, and greatly improve the upper limit of the device's detection performance. An image reconstruction method based on millimeter-wave human body security inspection imaging data processing is provided.

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

[0006] Step S1: The millimeter wave device collects millimeter wave echo signals and obtains the initial three-dimensional holographic imaging data in the security inspection area through algorithm calculation, wherein the initial three-dimensional holographic imaging data range is the human body width direction w∈[0,W], the human body height direction h∈[0,H], and the distance direction between the millimeter wave device and the human body d∈[D min ,D max ];

[0007] Step S2: Divide the security inspection area into N different sub-areas according to the security inspection personnel and the inspection requirements, and the sub-areas have no overlap or omission in the security inspection area;

[0008] Step S3: Calculate the front surface distances of the human body at three heights in each block area respectively, and then comprehensively determine the front surface distance d of the human body in the block area. n , where n∈[1,N];

[0009] Step S4: according to the difference between the front and back sides of the human body and the gender, intercept the effective three-dimensional imaging data in each block area through the distance of the front surface of the human body;

[0010] Step S5: splicing the 3D imaging data of each block area according to the position of the security inspection area when it was intercepted, and filling the missing part in the human body distance direction with zeros at the end to obtain valid 3D imaging data in the complete security inspection area;

[0011] Step S6: Select background in the unmanned area on both sides of the security inspection area, take the average and then expand it into three-dimensional background data with the same dimension as the imaging data;

[0012] Step S7: subtract the three-dimensional background data of the same dimension from the corresponding position of the intercepted effective three-dimensional imaging data, and then map it into corresponding two-dimensional data along the human body distance direction, and reconstruct and display the millimeter wave human body security inspection image.

[0013] Furthermore, in step S2, the security inspection area is divided into blocks as follows:

[0014] Starting from the height 0 of the security inspection area, the height range of the first block area is 0 to 0.35m, the height range of the last block area is 1.55m to H, and the heights of the remaining intermediate block areas are all in the range of 0.06 to 0.2m, which is divided into N block areas in total; according to actual conditions, the wrists, forearms, elbows and upper arms on both sides of the human body trunk can be further separated from the block areas of the trunk to form separate block areas.

[0015] Furthermore, in step S3, the distance d between the front surface of the human body and the millimeter wave device is n The calculation method is:

[0016] S31: In each block area n, take the lowest height h of this block area n respectively nmin 、Maximum height h nmax and the average height h of the block area n nmid The two-dimensional imaging data corresponding to the three heights are obtained by summing each set of two-dimensional data along the width direction of the human body to obtain one-dimensional data corresponding to the distance of the millimeter wave device at the three heights, where h nmid =(h nmin +h nmax ) / 2;

[0017] S32: Calculate the maximum value of each height one-dimensional data. When this maximum value is within the set distance range and reaches the set threshold V threshold When , the distance corresponding to the maximum value is considered to be the distance d between the front surface of the human body and the millimeter wave device at this height. body , otherwise the height is considered to be in the imaging background area;

[0018] S33: Take the distance d between the front surface of the human body and the millimeter wave device at three heights in the block area n body The minimum value among them is taken as the distance d between the front surface of the human body and the millimeter wave device in the block area n n ; If any height is the imaging background area, then this height is not included in the calculation; if all three heights are the imaging background area, then d n =(D min +D max ) / 2.

[0019] Furthermore, in step S4, the method for acquiring effective three-dimensional imaging data of the block area n is:

[0020] For the block area with a front height of ≤0.85m or ≥1.55m and the block area with a back height of ≤0.85m or ≥1.25m, the distance between the block area and the millimeter wave device d∈[d n -a,d n +b] range, where a ranges from 3 to 5 cm, b ranges from 11 to 15 cm, and for other height block areas, a ranges from 3 to 5 cm, and b ranges from 13 to 17 cm.

[0021] Furthermore, in the step S5, the method for splicing the three-dimensional imaging data of the block areas is as follows: in the width direction of the human body, all the block areas are aligned with 0 as the starting point; in the height direction of the human body, all the block areas are spliced ​​to the corresponding height positions according to their respective height ranges when intercepted; in the distance direction of the human body, all the block areas are aligned with the data closest to the millimeter wave device. When the distance dimension is insufficient, the missing data of the remaining block areas are filled with 0 using the maximum value of the distance dimension.

[0022] Furthermore, in step S6, the three-dimensional background data acquisition method is:

[0023] S61: Calculate the background data value at each height h Among them, the human body width ranges w∈[0,w0] and w∈[W-w0,W] are both in the background area without people;

[0024] S62: Back at each height h h They are respectively expanded into two-dimensional data of the same dimension as the effective imaging data, and constitute three-dimensional background data as a whole.

[0025] Furthermore, in step S7, the specific processing process is as follows:

[0026] S71: Subtract each point in the intercepted effective three-dimensional imaging data from the corresponding position in the three-dimensional background data;

[0027] S72: when the difference is a negative number, the point is set to 0, and effective three-dimensional holographic imaging data after removing the background is obtained;

[0028] S73: Mapping the three-dimensional data into corresponding two-dimensional data along the direction of human body distance, reconstructing and displaying to obtain a millimeter wave human body security inspection image.

[0029] Furthermore, in the step S73, there are four ways to map the three-dimensional data into corresponding two-dimensional data, namely, mapping the maximum value of the one-dimensional data in the direction of human body distance to the security inspection area, taking the standard deviation of the one-dimensional data in the direction of human body distance to the security inspection area, taking the maximum value in the direction of human body distance and the three points before and after and mapping them to the security inspection area, taking the maximum value in the direction of human body distance and the seven points in the surrounding space and mapping them to the security inspection area.

[0030] Furthermore, the specific process of taking the maximum value of the human body distance direction and the 7 points in the surrounding space and mapping them to the security inspection area is as follows:

[0031] S7311: Calculate the data S of each point of the two-dimensional imaging image from the effective three-dimensional holographic imaging data after removing the background ij When , first calculate and determine the maximum value s at this position along the direction of human body distance ijk ,k∈[dn -a,d n +b], where d n The distance between the front surface of the human body and the millimeter wave device in the block area of ​​the position;

[0032] S7312: When s ijk When it is not a boundary point of three-dimensional data, the corresponding two-dimensional imaging image point S ij Calculated as:

[0033] S ij =σ×s ijk +s (i-1)jk +s (i+1)jk +s i(j-1)k +s i(j+1)k +s ij(k-1) +s ij(k+1)

[0034] Among them, σ is the weight, σ≥1; when s ijk When it is the boundary point of the three-dimensional data, the points outside the three-dimensional data range among the above 7 points are marked with s ijk Instead of calculating S ij .

[0035] Furthermore, in step S73, when reconstructing the two-dimensional data, nonlinear correction is performed on each point of data, and the specific process is as follows:

[0036] S7321: Get the maximum value S of all imaging data points max , minimum value S min , calculate the cut-off point data before correction: S m =S min +m×(S max -S min ), corrected cut-off point data: S n =S min +n×(S max -S min ), where m∈(0,1),n∈(0,1),m <n;

[0037] S7322: Calculate the corrected data of all data points in two stages:

[0038] For the pre-corrected data S∈[S min ,S m ), the corrected data S′=S min +n×(SS min ) / m, and S′∈[S min ,S n );

[0039] For the pre-corrected data S∈[S m ,S max], the corrected data S′=S min +n×(S max -S min )+(1-n)×(SS m ) / (1-m), and S′∈[S n ,S max ].

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] 1. Based on the initial 3D holographic imaging data, by dividing the block area, obtain the effective 3D imaging data of each block area, and remove or eliminate the useless background, clutter, reflection, noise interference, state drift and other additional interference signals in front, behind and on both sides of the human body at different heights and parts as much as possible while retaining the information of the human body and hidden objects. Eliminate and reduce interference directly from the source of the data, improve the image quality from the source, and thus greatly improve the detection and false alarm performance of the equipment.

[0042] 2. After eliminating and reducing interference at the source of the data, by reducing the background noise signal within the same distance as the human body and through nonlinear correction, weak targets are further highlighted, the weak target detection rate of the device is improved, the false alarm rate is reduced, and the detection performance bottleneck of the device is significantly reduced.

[0043] 3. There is no need to make any changes to the existing millimeter wave body security inspection equipment hardware. The process is simple and easy to understand, the process is easy to implement, and it consumes little system resources, so it can be quickly applied and promoted. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flowchart of an image reconstruction method based on millimeter wave human body security inspection imaging data processing in Embodiment 1 of the present invention;

[0045] Figure 2 It is a schematic diagram of the block area division method in the first embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of a block area division method in Embodiment 2 of the present invention;

[0047] Figure 4 is the security inspection imaging image of the initial three-dimensional holographic imaging data in the fourth embodiment of the present invention;

[0048] Figure 5 It is a security inspection imaging diagram after data processing and nonlinear correction of the initial three-dimensional holographic imaging data in the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0049] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0050] Embodiment 1

[0051] like Figure 1 As shown, this embodiment provides a technical solution: an image reconstruction method based on millimeter wave human body security inspection imaging data processing, comprising the following steps:

[0052] Step S1: Initial 3D holographic imaging data acquisition

[0053] For millimeter-wave echo data, the initial three-dimensional holographic imaging data of the equipment inspection area is obtained through the Fourier transform, matched filtering, distance dimension integration and other processes of the phase shift migration (PSM) algorithm, where the three-dimensional data range is w∈[0,1] in the width direction of the human body, h∈[0,2] in the height direction of the human body, and d∈[0.1,0.6] in the distance direction between the millimeter-wave device and the human body.

[0054] Step S2: Security inspection area division

[0055] Starting from the security inspection area height 0, the height range of the first block area is 0 to 0.35m, the height range of the last block area is 1.55 to 2m, and the height of the remaining middle block areas is 0.1m. There are 14 block areas in total, such as Figure 2 shown.

[0056] Step S3: The distance d between the front surface of the human body and the millimeter wave device n calculate

[0057] In each block area n, take the lowest height h of this block area n nmin 、Maximum height h nmax and the average height h of the block area n nmid (i.e. h nmid =(h nmin +h nmax ) / 2) The two-dimensional imaging data corresponding to the three heights are summed along the width of the human body to obtain the one-dimensional data corresponding to the distance of the millimeter wave device at the three heights. The maximum value of the one-dimensional data at each height is calculated. When this maximum value is within the set distance range and reaches the set threshold V threshold When , the distance corresponding to the maximum value is considered to be the distance d between the front surface of the human body and the millimeter wave device at this height. body , otherwise the height is considered to be in the imaging background area.

[0058] In the block areas 2 to 13, three heights of the distance d between the front surface of the human body and the millimeter wave device are selected. body The minimum value among them is taken as the distance d between the front surface of the human body and the millimeter wave device in this block area. n In the block areas 1 and 14, some heights are imaging background areas and are not included in the calculation. The remaining heights are taken as the distance d between the front surface of the human body and the millimeter wave device. body The minimum value among them is taken as the distance d between the front surface of the human body and the millimeter wave device in this block area. n .

[0059] Step S4: Acquisition of effective three-dimensional imaging data of block area n

[0060] The initial 3D holographic imaging data obtained is the frontal data of the human body. For the block areas 1 to 6 and 14, the distance between the block area and the millimeter wave device d∈[d n -0.04,d n +0.13], for block areas 7 to 13, the distance between the block area and the millimeter wave device d∈[d n -0.04,d n +0.15] range of three-dimensional data.

[0061] Step S5: Stitching of 3D imaging data of divided regions

[0062] In the direction of human body width, all block areas are aligned with 0 as the starting point; in the direction of human body height, all block areas are spliced ​​to the corresponding height positions according to their respective height ranges when intercepted; in the direction of human body distance, all block areas are aligned with the data closest to the millimeter wave device, and block areas 1 to 6 and 14 are filled with data 0 in the 0.02m distance range at the maximum distance.

[0063] Step S6: Acquisition of 3D background data

[0064] Calculate the background data value at each height h Among them, the human body width range w∈[0,0.08] and w∈[0.92,1] are both in the background area without people. h They are respectively expanded into two-dimensional data of the same dimension as the effective imaging data, and constitute three-dimensional background data as a whole.

[0065] Step S7: Subtract the 3D background data and reconstruct the millimeter wave human body security inspection image

[0066] Subtract each point in the intercepted effective three-dimensional imaging data from the corresponding position in the three-dimensional background data. When the difference is negative, the point is taken as 0 to obtain the effective three-dimensional holographic imaging data after removing the background. Then, the maximum value in the three-dimensional data is mapped into the corresponding two-dimensional data along the direction of human body distance, and the millimeter wave human body security inspection image is obtained by reconstructing the display.

[0067] In this embodiment, by dividing the block area, according to the different front surface distances of the human body in each block area, the effective three-dimensional imaging data in each block area is intercepted respectively, and then spliced, and after deducting the background signal data, the millimeter wave human body security inspection image is reconstructed. While retaining the information of the human body and hidden objects, this embodiment removes the background, clutter, noise interference, reflection, device or module drift and other interferences in front, behind and on both sides of the human body at different heights and parts, reduces the noise interference within the same range as the human body distance, directly eliminates and reduces the noise interference from the source of the three-dimensional data, fundamentally reduces and eliminates various stray and interference information mixed in the later image processing and target recognition, and obtains a security inspection imaging result that is closer to the truth, thereby significantly improving the detection rate of hidden objects, and greatly reducing the false alarm rate caused by invalid and false interference.

[0068] In this embodiment, by eliminating interference and background from the source of three-dimensional data, the upper limit of device detection performance that can be achieved by later image processing and target recognition is significantly improved, and the detection performance bottleneck of the device is greatly reduced.

[0069] Embodiment 2

[0070] This embodiment is improved on the basis of the first embodiment, and the same parts as the above technical solution will not be repeated here. This embodiment improves the method of dividing the block area, such as Figure 3 As shown. According to the actual structural characteristics of the human body, the arms are generally smaller than the torso in the direction of the human body distance, and the security personnel's arms often exceed the front surface of the human body, exceed the back surface of the human body, and the elbows are bent. Therefore, from the block area of ​​the human torso, the wrist and lower forearm block area, as well as the upper forearm, elbow and lower upper arm block area are added on both sides, and a total of 18 block areas are divided. These arm block areas are separated from the torso block area, which can more accurately calculate the front surface distance of the arm and more effectively obtain their three-dimensional imaging data. At the same time, it is also more targeted to eliminate noise and interference, and the detection and false alarm performance of the equipment is further improved.

[0071] Embodiment 3

[0072] This embodiment is a further improvement on the first embodiment. The same parts as the above technical solution will not be repeated here. This embodiment improves the method of mapping three-dimensional data into two-dimensional data. From the effective three-dimensional holographic imaging data after removing the background, calculate the data S of each point of the two-dimensional imaging image ij When , first calculate and determine the maximum value s at this position along the direction of human body distance ijk ,k∈[d n -a,d n +b], where d n is the distance between the front surface of the human body and the millimeter wave device in the location block area. ijk When it is not a boundary point of three-dimensional data, the corresponding two-dimensional imaging image point S ij Calculated as:

[0073] S ij =σ×s ijk +s (i-1)jk +s (i+1)jk +s i(j-1)k +s i(j+1)k +s ij(k-1) +s ij(k+1)

[0074] Among them, σ is the weight, σ≥1, which can be adjusted according to the data situation and actual needs. ijk When it is the boundary point of the three-dimensional data, the points outside the three-dimensional data range among the above 7 points are marked with s ijk Instead of calculating S ij .

[0075] In this embodiment, each point data in the two-dimensional imaging image uses the weighted sum of the seven points in the surrounding space with the maximum value along the human body distance direction at that position, so as to eliminate or reduce the influence of abnormal interference such as background noise and discrete isolated points and singular points in the surrounding environment as much as possible, highlight the information of the security personnel themselves and hidden objects, and the detection results of the equipment are more accurate and reliable, with a lower false alarm rate.

[0076] Embodiment 4

[0077] This embodiment is a further improvement on the first embodiment. The same parts as the above technical solutions will not be repeated here. This embodiment provides a technical solution: nonlinear correction is performed on each point of imaging data, and small and weak signal data are stretched to a larger data range. The nonlinear correction steps are as follows:

[0078] Get the maximum value S of all imaging data points max , minimum value S min , calculate the cut-off point data before correction: S m =S min +m×(S max -Smin ), corrected cut-off point data: S n =S min +n×(S max -S min ), where m∈(0,1),n∈(0,1),m <n;

[0079] Calculate the corrected data of all data points in two stages:

[0080] For the pre-corrected data S∈[S min ,S m ), the corrected data S′=S min +n×(SS min ) / m, and S′∈[S min ,S n );

[0081] For the pre-corrected data S∈[S m ,S max ], the corrected data S′=S min +n×(S max -S min )+(1-n)×(SS m ) / (1-m), and S′∈[S n ,S max ].

[0082] In this embodiment, m is 0.3 and n is 0.5, that is, the original larger 70% data is uniformly compressed to the new larger 50% data range, and the original smaller 30% data is uniformly stretched and enlarged to the new smaller 50% data range. The order of the size of the data remains unchanged, but the smaller data is closer to the larger data than before correction. In the actual correction process, the two-stage correction of this embodiment can also be improved to multi-stage correction.

[0083] like Figure 4 As shown in FIG. , it is the security inspection imaging diagram of the initial three-dimensional holographic imaging data, such as Figure 5 The figure shows the security inspection image after the initial three-dimensional holographic imaging data has been processed and nonlinearly corrected. In the initial data, multiple invalid and false targets existing on both sides of the human calf, between the thighs, between the torso and the arm, under the armpit, and around the real objects are eliminated from the data source, and the real weak target hidden under the right armpit of the security inspector is highlighted, and the strong target is more accurate and clear, and is accurately and effectively identified.

[0084] When the overall background interference is significantly reduced or eliminated, the method of this embodiment can make weak targets with smaller signals more obvious and prominent, the detection of weak targets can be further effectively improved, and the performance of millimeter wave body security inspection equipment will be significantly improved.

[0085] To summarize, the image reconstruction method based on millimeter-wave human body security inspection imaging data processing in the above-mentioned embodiment processes the acquired initial three-dimensional holographic imaging data of the security inspection area, directly removes the background, clutter, noise interference, reflection, device or module drift and other interference in front, behind and on both sides of the human body at different heights and parts from the source of the security inspection imaging data, eliminates or reduces the noise interference within the same range as the human body distance, and has no effect on the security inspectors and the hidden objects themselves, so that the subsequent image processing and target recognition can obtain more real and accurate security inspection images, and the security inspection results are more accurate. The detection rate and false alarm rate performance upper limits of the equipment can be greatly improved, and there will be no detection and false alarm bottlenecks caused by various invalid and false target interferences, which will affect the use and promotion of the equipment; by improving the method of mapping three-dimensional data into two-dimensional imaging data and performing nonlinear correction, the abnormal interference influence of discrete isolated points and singular points can be eliminated, weak targets with smaller signals can be highlighted, the imaging quality of the equipment can be further improved, and the detection performance of weak targets of the security inspection equipment can be greatly improved; there is no need to make any changes to the hardware of the existing millimeter wave human body security inspection equipment, and noise and interference can be eliminated from the data source through data processing to improve the imaging quality, thereby achieving more accurate security inspections, greatly improving the detection rate, reducing the false alarm rate, and greatly improving the upper limit of the equipment detection performance; the method flow of the present invention is simple and easy to understand, the process is easy to implement, consumes little system resources, and can be quickly applied and promoted.

[0086] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. An image reconstruction method based on millimeter wave human body security inspection imaging data processing, characterized in that: The following steps are involved: Step S1: The millimeter wave device collects millimeter wave echo signals and obtains the initial three-dimensional holographic imaging data in the security inspection area through algorithm calculation, wherein the initial three-dimensional holographic imaging data range is the human body width direction w∈[0,W], the human body height direction h∈[0,H], and the distance direction between the millimeter wave device and the human body d∈[D min ,D max ]; Step S2: Divide the security inspection area into N different sub-areas according to the security inspection personnel and the inspection requirements, and the sub-areas have no overlap or omission in the security inspection area; Step S3: Calculate the front surface distances of the human body at three heights in each block area respectively, and then comprehensively determine the front surface distance d of the human body in the block area. n , where n∈[1,N]; Step S4: according to the difference between the front and back sides of the human body and the gender, intercept the effective three-dimensional imaging data in each block area through the distance of the front surface of the human body; Step S5: splicing the 3D imaging data of each block area according to the position of the security inspection area when it was intercepted, and filling the missing part in the human body distance direction with zeros at the end to obtain valid 3D imaging data in the complete security inspection area; Step S6: Select background in the unmanned area on both sides of the security inspection area, take the average and then expand it into three-dimensional background data with the same dimension as the imaging data; Step S7: subtract the three-dimensional background data of the same dimension from the corresponding position of the intercepted effective three-dimensional imaging data, and then map it into corresponding two-dimensional data along the human body distance direction, and reconstruct and display the millimeter wave human body security inspection image.

2. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 1 is characterized in that: In step S2, the security inspection area is divided into blocks as follows: Starting from the height 0 of the security inspection area, the height range of the first block area is 0 to 0.35m, the height range of the last block area is 1.55m to H, and the heights of the remaining intermediate block areas are all in the range of 0.06 to 0.2m, and a total of N block areas are divided.

3. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 1 is characterized in that: In step S3, the distance between the front surface of the human body and the millimeter wave device is d n The calculation method is: S31: In each block area n, take the lowest height h of this block area n respectively nmin 、Maximum height h nmax and the average height h of the block area n nmid The two-dimensional imaging data corresponding to the three heights are obtained by summing each set of two-dimensional data along the width direction of the human body to obtain one-dimensional data corresponding to the distance of the millimeter wave device at the three heights, where h nmid =(h nmin +h nmax ) / 2; S32: Calculate the maximum value of each height one-dimensional data. When this maximum value is within the set distance range and reaches the set threshold V threshold When , the distance corresponding to the maximum value is considered to be the distance d between the front surface of the human body and the millimeter wave device at this height. body , otherwise the height is considered to be in the imaging background area; S33: Take the distance d between the front surface of the human body and the millimeter wave device at three heights in the block area n body The minimum value among them is taken as the distance d between the front surface of the human body and the millimeter wave device in the block area n n ; If any height is the imaging background area, then this height is not included in the calculation; if all three heights are the imaging background area, then d n =(D min +D max ) / 2.

4. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 1 is characterized in that: In step S4, the method for acquiring effective three-dimensional imaging data of the block area n is: For the block area with a front height of ≤0.85m or ≥1.55m and the block area with a back height of ≤0.85m or ≥1.25m, the distance between the block area and the millimeter wave device d∈[d n -a,d n +b] range, where a ranges from 3 to 5 cm, b ranges from 11 to 15 cm, and for other height block areas, a ranges from 3 to 5 cm, and b ranges from 13 to 17 cm.

5. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 1 is characterized in that: In step S5, the method for splicing the three-dimensional imaging data of the block areas is as follows: in the width direction of the human body, all the block areas are aligned with 0 as the starting point; in the height direction of the human body, all the block areas are spliced ​​to the corresponding height positions according to their respective height ranges when intercepted; in the distance direction of the human body, all the block areas are aligned with the data closest to the millimeter wave device. When the distance dimension is insufficient, the missing data of the remaining block areas are filled with 0 using the maximum value of the distance dimension.

6. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 1 is characterized in that: In step S6, the three-dimensional background data acquisition method is: S61: Calculate the background data value at each height h Among them, the human body width ranges w∈[0,w0] and w∈[W-w0,W] are both in the background area without people; S62: Back at each height h h They are respectively expanded into two-dimensional data of the same dimension as the effective imaging data, and constitute three-dimensional background data as a whole.

7. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 1 is characterized in that: In step S7, the specific processing process is as follows: S71: Subtract each point in the intercepted effective three-dimensional imaging data from the corresponding position in the three-dimensional background data; S72: when the difference is a negative number, the point is set to 0, and effective three-dimensional holographic imaging data after removing the background is obtained; S73: Mapping the three-dimensional data into corresponding two-dimensional data along the direction of human body distance, reconstructing and displaying to obtain a millimeter wave human body security inspection image.

8. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 7 is characterized in that: In step S73, there are four ways to map three-dimensional data into corresponding two-dimensional data, namely, mapping the maximum value of the one-dimensional data in the direction of human body distance to the security inspection area, taking the standard deviation of the one-dimensional data in the direction of human body distance to the security inspection area, taking the maximum value in the direction of human body distance and the three points before and after and mapping them to the security inspection area, taking the maximum value in the direction of human body distance and the seven points in the surrounding space and mapping them to the security inspection area.

9. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 8 is characterized in that: The specific process of taking the maximum value of the human body distance direction and the 7 points in the surrounding space and mapping them to the security inspection area is as follows: S7311: Calculate the data S of each point of the two-dimensional imaging image from the effective three-dimensional holographic imaging data after removing the background ij When , first calculate and determine the maximum value s at this position along the direction of human body distance ijk ,k∈[d n -a,d n +b], where d n The distance between the front surface of the human body and the millimeter wave device in the block area of ​​the position; S7312: When s ijk When it is not a boundary point of three-dimensional data, the corresponding two-dimensional imaging image point S ij Calculated as: S ij =σ×s ijk +s (i-1)jk +s (i+1)jk +s i(j-1)k +s i(j+1)k +s ij(k-1) +s ij(k+1) Among them, σ is the weight, σ≥1; when s ijk When it is the boundary point of the three-dimensional data, the points outside the three-dimensional data range among the above 7 points are marked with s ijk Instead of calculating S ij .

10. The image reconstruction method based on millimeter wave human body security inspection imaging data processing according to claim 7 is characterized in that: In step S73, when reconstructing the two-dimensional data, nonlinear correction is performed on each point of data. The specific process is as follows: S7321: Get the maximum value S of all imaging data points max , minimum value S min , calculate the cut-off point data before correction: S m =S min +m×(S max -S min ), corrected cut-off point data: S n =S min +n×(S max -S min ), where m∈(0,1),n∈(0,1),m <n; S7322: Calculate the corrected data of all data points in two stages: For the pre-corrected data S∈[S min ,S m ), the corrected data S′=S min +n×(SS min ) / m, and S′∈[S min ,S n ); For the pre-corrected data S∈[S m ,S max ], the corrected data S′=S min +n×(S max -S min )+(1-n)×(SS m ) / (1-m), and S′∈[S n ,S max ].

Images