A low-radiation X-ray imaging recognition method based on cosmic muon image

Abstract

The application discloses a low-radiation X-ray imaging recognition method based on a cosmic muon image, and steps include: S1, generating a cosmic muon imaging feature map; S2, recognizing a density abnormal area in the cosmic muon imaging feature map; S3, according to the recognition result of step S2, performing X-ray single-photon imaging on the density abnormal area of an imaging object; and S4, based on a wavelet transform method, fusing the cosmic muon imaging feature map and the X-ray single-photon imaging to obtain a fusion image capable of clearly displaying a defective position and shape; the method is suitable for large-range high-precision nondestructive imaging detection, can solve the radiation problem of current nondestructive imaging while ensuring nondestructive imaging, and forms a new method of low-dose large-range high-precision nondestructive detection imaging.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nondestructive testing, in particular to a low-radiation X-ray imaging recognition method based on a cosmic muon image. BACKGROUND

[0002] As a unique natural particle, cosmic muons have super strong penetration ability and can easily penetrate through complex medium layers in closed spaces. This characteristic makes cosmic muons have great potential in the detection application of closed spaces and complex medium scenes. However, current muon imaging schemes mostly rely on long-time accumulation detection and data processing, which greatly limits the imaging speed and accuracy of cosmic muon imaging technology in real-time monitoring.

[0003] In the field of nondestructive testing, X-ray single photon counting imaging technology uses high-sensitivity semiconductor detectors such as cadmium telluride or zinc telluride to accurately count each absorbed X-ray photon. The charge signal generated by each photon is amplified and digitized to record its energy and position information. These information is stored in a pixel array, each pixel corresponds to a small area of the detector. Through back projection or iterative reconstruction algorithm, high-resolution and low-noise image reconstruction is realized by using these independent photon counting data, so as to accurately detect the small defects, foreign matters or structural problems of the material. However, the process of using X-rays for nondestructive testing of objects usually produces a large dose of radiation, which not only harms the operator's body, but also restricts the use of technology. Therefore, for the requirement of large-scale high-precision nondestructive imaging detection, the radiation dose generated by X-ray single photon counting imaging cannot be ignored, which not only causes certain loss to the measured object, but also causes great harm to the detection personnel.

[0004] Therefore, it is necessary to design a nondestructive testing scheme that can effectively reduce the radiation dose of X-ray imaging while ensuring the accuracy of nondestructive imaging detection. SUMMARY

[0005] The purpose of the present application is to provide a low-radiation X-ray imaging recognition method based on a cosmic muon image, which solves the above technical problems.

[0006] To this end, the technical scheme of the present application is as follows:

[0007] A low-radiation X-ray imaging recognition method based on a cosmic muon image, the steps are as follows:

[0008] S1, laying two liquid scintillator muon detectors on the upper and lower of the closed block containing the imaging object, collecting the real track information of each muon entering the closed block and the real track information of each muon exiting the closed block, to generate a cosmic muon imaging feature map;

[0009] S2, identify the density anomaly area in the cosmic muon imaging feature map, and the specific steps are as follows:

[0010] S201, divide the closed block into a voxel grid, find the position points where each muon is scattered in the imaging object and the voxels to which the position points belong according to the real track information of each muon entering and exiting the closed block;

[0011] S202, set k-nearest neighbors, and calculate the average distance between each position point and its k-nearest neighbors;

[0012] S203, calculate the LOF score of each position point, and determine whether it is abnormal according to the set threshold;

[0013] S204, set the threshold of the abnormal position points contained in the abnormal voxels, and determine the abnormal voxels in the imaging object according to the judgment result of step S203;

[0014] S205, according to the position coordinates of the abnormal voxels, mark the density anomaly area on the cosmic muon imaging feature map generated by step S1;

[0015] S3, according to the recognition result of step S2, perform X-ray single photon imaging on the density anomaly area of the imaging object;

[0016] S4, based on the wavelet transform method, fuse the cosmic muon imaging feature map and the X-ray single photon imaging to obtain a fusion image clearly showing the defect position and shape of the imaging object in the closed block.

[0017] Further, the specific implementation steps of step S1 are as follows:

[0018] S101, according to the real track information of each muon entering and exiting the closed block, obtain the real muon scattering angle data of the closed block;

[0019] S102, construct a forward model to obtain the theoretical muon scattering angle data of the closed block by simulating the muon flux distribution;

[0020] S103, calculate the optimal solution of the density distribution of the closed block by the MLSD inversion algorithm;

[0021] S104, project the optimal solution of the density distribution of the closed block onto a two-dimensional plane and assign color values to generate a cosmic muon imaging feature map.

[0022] Furthermore, step S102 is implemented using Geant4 software; wherein, the modeling models of the closed block and the imaging object are consistent with reality, and the model of the object to be imaged is set to have a uniform theoretical density value, and the remaining space in the closed block is set to be filled with air; a liquid scintillator muon detector is set in the center of the upper and lower paths of the closed block model, and the scattering data collection area is made to completely cover the closed block model, and the direction of the cosmic emission muons is set to a random direction.

[0023] Furthermore, the specific processing steps of step S103 are:

[0024] S1031, assigning an initial density value to each voxel in the voxel grid divided into the closed block

[0025] S1032. Define the objective function F: , where △ θ i For the i The actual scattering angle of a muon, △ θ' i For the i The theoretical scattering angle of muons, N is the total number of muons;

[0026] S1033. Using the least squares method, with minimizing the objective function F as the optimization goal, the optimal density distribution result of the closed block is obtained through continuous iterative solution.

[0027] Furthermore, in step S201, the method for determining the location point where any muon is scattered is as follows:

[0028] ① When the intersection of the extended line of the incident trajectory of the muon and the extended line of its outgoing trajectory is inside the imaging object, the intersection is the location where the muon is scattered, and correspondingly, the voxel containing the location is the voxel to which the location belongs;

[0029] ② When the intersection of the extended line of the incident track of the muon and the extended line of its outgoing track is outside the imaging object or there is no intersection, the intersection of the midpoint of the extended line of the incident track and the perpendicular line of the extended line of the outgoing track within the imaging object is taken as the position point where the muon is scattered. Correspondingly, the voxel containing this position point is the voxel to which this position point belongs.

[0030] Furthermore, in step S2, the k value of step S202 is set to 12-15, the threshold of step S203 is set to 1.5-1.8, and the threshold of the number of abnormal location points of step S204 is set to 4-6.

[0031] Furthermore, the specific implementation steps of step S3 are as follows:

[0032] S301, an X-ray source and a single photon counting camera are respectively set up on the upper and lower roads of the closed block and at locations corresponding to density anomaly areas of the imaging object;

[0033] S302, capturing low-energy photons emitted by the X-ray source and penetrating the enclosed block and the imaging object using a single-photon counting camera to obtain the arrival position and energy of each photon;

[0034] S303 , using a direct projection method to map the photon data onto a two-dimensional plane, and generate a color image reflecting the X-ray single photon counts.

[0035] Furthermore, the specific implementation steps of step S303 are as follows:

[0036] S3031, mapping the photon arrival position data obtained in step S302 into a two-dimensional array, where the two-dimensional array consists of arrival positions and photon counts;

[0037] S3032: Using the arrival position of the photons as pixel coordinates, the photon counts are mapped to color intervals according to the maximum and minimum values ​​of the photon counts and the set RGB value range to generate a color image.

[0038] Furthermore, the specific implementation steps of step S4 are:

[0039] S401, selecting a wavelet basis function to perform multi-level decomposition on the two images;

[0040] S402, fusing the frequency sub-bands of each level decomposed from the two images respectively;

[0041] S403 , using an inverse wavelet transform function to reconstruct the fused frequency sub-bands at all levels to obtain a fused image.

[0042] Furthermore, in step S401, the wavelet basis function adopts the db4 wavelet function and is set to the third-level wavelet decomposition; in step S402, the fusion method adopts the weighted average method, and the feature matching method adopts the nearest neighbor matching method.

[0043] Compared with the existing technology, this low-radiation X-ray imaging recognition method based on cosmic muon images is suitable for large-scale, high-precision non-destructive imaging detection. On the basis of using cosmic muons to achieve large-scale, dose-free transmission imaging, it effectively identifies abnormal areas and combines with low-radiation X-ray imaging recognition to obtain a detection image with high-precision imaging of the local defect location of the imaging object. While ensuring non-destructive imaging, it solves the radiation problem of current non-destructive imaging, forming a new method of low-dose, large-scale, high-precision non-destructive testing imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Flow chart of the low-radiation X-ray imaging recognition method based on the muon image of the present application;

[0045] Figure 2 Schematic diagram of the liquid scintillator muon detector laid out in two ways of upper and lower outside the closed block in which the imaging object is built in, in step S101 of the embodiment of the present application;

[0046] Figure 3 Schematic diagram of the forward model constructed by using Geant4 software in step S102 of the embodiment of the present application;

[0047] Figure 4 The muon imaging feature map drawn in step S1 of the embodiment of the present application;

[0048] Figure 5 Schematic diagram of the density abnormal area in the muon imaging feature map identified in step S2 of the embodiment of the present application;

[0049] Figure 6 Schematic diagram of the arrangement of the X-ray source and the single photon counting camera outside the closed block in which the imaging object is built in, in step S3 of the embodiment of the present application;

[0050] Figure 7 Schematic diagram of the X-ray single photon counting image obtained in step S3 of the embodiment of the present application;

[0051] Figure 8 Schematic diagram of the fusion image obtained in step S4 of the embodiment of the present application. DETAILED DESCRIPTION

[0052] The present application will be further described below in conjunction with the drawings and specific embodiments, but the following embodiments are by no means any limitation on the present application.

[0053] Reference Figure 1 The specific implementation steps of the low-radiation X-ray imaging recognition method based on the muon image are described as follows.

[0054] S1, obtain the muon imaging feature map of the imaging object in the closed block.

[0055] The specific implementation steps of step S1 are described as follows.

[0056] S101, lay out the liquid scintillator muon detector 3 in two ways of upper and lower outside the closed block 1 in which the imaging object 2 is built in, to collect the muon scattering data of the closed block; see Figure 2In the actual scene setting, the imaging object 2 is a lead alloy block with uniform density distribution, and there is an irregular E-shaped crack in the B part of the imaging object 2, that is, the defect identification target to be realized by non-destructive imaging in this embodiment.

[0057] In this step, in order to meet the imaging needs of the closed space, the liquid scintillator muon detector 3 is laid in two ways of upper and lower to obtain the scattering data of muons before and after passing through the closed block 1 respectively; at the same time, in order to ensure the continuity of the scattering imaging data on the imaging result, the liquid scintillator muon detectors 3 located in the upper and lower ways are laid in a linear and equidistant manner with multiple (3 in this embodiment), and the interval between the adjacent two liquid scintillator muon detectors 3 located in the same way is determined according to the detection range of the liquid scintillator muon detector 3, so as to ensure that the scattering data acquisition area has a certain spatial overlap. Since the muon scattering imaging is realized by accumulation of muon scattering data, therefore, in this step, the collection time length of the muon scattering data is set to 2h.

[0058] In this step S101, the muon scattering data to be obtained are the real track information of each muon entering the closed block 1 X i and the real track information of each muon passing out of the closed block 1 X i ’ , i is the number of muons; and further, the angle change caused by scattering during the passage of each muon through the closed block 1, that is, the real scattering angle △ θ i .

[0059] S102, constructing a forward model to obtain the theoretical muon scattering data of the closed block by simulating the muon flux distribution.

[0060] In this step S102, the interaction between cosmic muons and the imaging object 2 is simulated by the constructed forward model, so as to obtain the theoretical muon scattering data, which serves as the data basis for further comparison with the real muon scattering data measured in step S101 and for calculating the actual density distribution of the imaging object 2.

[0061] In this embodiment, the forward model construction and simulation are implemented using Geant4 software; specifically, the forward model construction includes: first, modeling the closed block 1 and the imaging object 2 located in the closed block 1 in the software, and keeping the geometric models of the two consistent with the actual scene state in step S1, specifically, the closed block model is a hollow shell model, and the imaging object is a solid model composed of a number of voxels; then, according to the lead alloy block properties of the physical imaging object, the model of the object to be imaged is set to have a uniformly distributed theoretical lead alloy density value, and the remaining space inside the closed block model is set to be filled with air; finally, a liquid scintillator muon detector is set in the center of the upper and lower paths of the closed block model, and it is ensured that the scattering data acquisition area completely covers the closed block model, and the direction of the cosmic emission muons is set to a random direction; Figure 3 Shown is a schematic diagram of the forward model constructed using Geant4 software (top view).

[0062] Then, Geant4 software can simulate the theoretical muon flux distribution based on the above model construction conditions, and then obtain theoretical muon flux data; specifically, the same as step S101, the theoretical muon flux data is: the theoretical track information of each muon penetrating the closed block 1 Y i and theoretical track information of passing through closed block 1 Y i ’ ,in, i is the number of the muon; further calculations yield the following: the theoretical angle change caused by scattering when each muon passes through the closed block, i.e. the theoretical scattering angle △ θ' i .

[0063] S103, calculating the optimal solution of the density distribution of the imaging object by using the MLSD inversion algorithm.

[0064] The specific implementation steps of step S103 are described as follows.

[0065] S1031 , dividing the closed block 1 into a voxel grid, and assigning an initial density value to each voxel; wherein the division accuracy of the voxel grid is determined according to actual needs to meet the resolution requirement of two-dimensional imaging.

[0066] In this embodiment, the initial density distribution of each voxel is set to be consistent with the theoretical density distribution set in the forward model of step S102, that is, the initial density of the voxel corresponding to the imaging object is the theoretical density of the lead alloy block, and the initial density of the voxel corresponding to other parts in the closed block is the air density.

[0067] S1032. Define the objective function F as: the actual scattering angle of each muon △ θi and the theoretical scattering angle △ θ' i The sum of the squares of the differences between them is used to utilize the correlation between the muon flux distribution and the closed block density distribution to deduce the optimal solution of the closed block density distribution.

[0068] Specifically, the expression of the objective function F is:

[0069] ,

[0070] Where, θ i For the i The actual scattering angle of a muon, △ θ' i For the i The theoretical scattering angle of muons, N is the total number of muons.

[0071] S1033. Using the least squares method, with minimizing the objective function F as the optimization goal, the optimal density distribution result of the closed block is obtained through continuous iterative solution.

[0072] In this embodiment, the density distribution range of the closed block obtained by solution is 1.1 kg / m 3 (approximately the density of air) ~8.4×10 3 kg / m 3 , which is consistent with the actual scenario.

[0073] S104. Assign color values ​​to the density distribution of the closed block obtained in step S103 and project it onto a two-dimensional plane to generate a cosmic muon imaging feature map.

[0074] The specific implementation steps of step S104 are described as follows.

[0075] S1041, projecting the density distribution result of the closed block obtained in step S103 into a density distribution result of a two-dimensional plane;

[0076] In this embodiment, the two-dimensional plane is specifically a horizontal plane, and the density distribution result of each pixel on the two-dimensional plane is the density accumulation result of the longitudinal voxels.

[0077] S1042. Based on the minimum density value and the maximum density value in the density distribution result of the two-dimensional plane, the density value of each pixel on the two-dimensional plane is normalized to the interval of [0, 1].

[0078] S1043. Substitute the normalized density value of each pixel on the two-dimensional plane into a color mapping formula to obtain the color value of each pixel on the two-dimensional plane; wherein the color mapping formula is:

[0079] C = round(a x 255) + 1,

[0080] In the formula, C is the color index of the pixel, a is the normalized density value of the pixel, and round(·) represents the rounding of the calculation result to a positive integer.

[0081] S1045, based on the two-dimensional plane coordinate system of the closed block 1 on the horizontal plane, a two-dimensional cosmogenic muon imaging feature map is generated, which is also referred to as a cosmogenic muon image.

[0082] In this embodiment, this step can be realized by MATLAB software, and the specific operation is as follows: 1) call the jet function to generate a 256x3 matrix representing the jet color mapping; wherein each row of the matrix represents a color, the first column is the red component, the second column is the green component, and the third column is the blue component; 2) create a density matrix, and the density values in the matrix are the normalized density results of each pixel on the two-dimensional plane; 3) set the color mapping formula: C = round(a x 255) + 1, a is substituted into the normalized density value of the pixel, and the integer interval is converted to [1, 256] to correspond to the 256 rows of the color mapping table, and each row represents a specific color; 4) call the ind2rgb function to convert the index image to an RGB image.

[0083] After the above steps, the density distribution results of each pixel on the two-dimensional plane are mapped into color distribution results; specifically, if the pixel color value is a red hue, it means that the pixel corresponds to a low density interval; if the pixel color value is a green hue, it means that the pixel corresponds to a medium density interval; and if the pixel color value is a blue hue, it means that the pixel corresponds to a high density interval.

[0084] In this embodiment, the two-dimensional plane coordinate system of the closed block 1 takes the center point of the block as the origin, and the length direction is the x-axis direction and the width direction is the y-axis direction. As shown in Figure 4 The cosmogenic muon imaging feature map generated in the scenario of this embodiment is shown.

[0085] S2, identify the density abnormal area in the cosmogenic muon imaging feature map.

[0086] The specific implementation steps of this step S2 are described as follows.

[0087] S201, find the position point in the imaging object at which each muon is scattered and the voxel to which the position point belongs according to the real track information of each muon entering the closed block 1 X i (i.e. incident track ） and the real track information of each muon out of the closed block 1 X i ’ (i.e. exit track) in the imaging object.

[0088] Specifically, in the actual processing, the determination method of the position point of any muon scattering is as follows:

[0089] ① When the intersection point of the incident track extension line and the exit track extension line of the muon is inside the imaging object, the intersection point is the position point of the muon scattering, and correspondingly, the voxel containing the position point is the voxel to which the position point belongs.

[0090] ② When the intersection point of the incident track extension line and the exit track extension line of the muon is outside the imaging object or there is no intersection point, the perpendicular intersection point of the incident track extension line and the exit track extension line in the line segment inside the imaging object is taken as the position point of the muon scattering, and correspondingly, the voxel containing the position point is the voxel to which the position point belongs.

[0091] S202, set k neighbors, and calculate the average distance between each position point obtained by step S201 and its k neighbors.

[0092] In this embodiment, k is set to 15, and the calculation step of this step S202 is: for each position point, the Euclidean distance between the position point and other position points is calculated respectively; according to k being 15, the 15 nearest neighbor position points of the specified position point are set as the nearest neighbor position points of the specified position point; according to the Euclidean distance results calculated in advance, the nearest 15 neighbor position points of the specified position point are found as the neighbor set N k (p); then, the average distance of each position point and its k neighbor position points is calculated, and the calculation formula is:

[0093] ,

[0094] In the formula, d(p, o) is the Euclidean distance from the specified position point p to any nearest neighbor position point o, and k is the number of nearest neighbor position points.

[0095] In this step, the average distance result of each position point is taken as the measure of local density. When the calculation result of the average distance is smaller, it represents that the local density of the region is higher; on the contrary, when the calculation result of the average distance is larger, it represents that the local density of the region is lower.

[0096] S203, calculate the LOF score of each position point, and judge whether it is abnormal according to the set threshold.

[0097] Specifically, the calculation formula of the LOF score is:

[0098] ,

[0099] In the formula, the average distance of the specified position point and its k neighbor position points, the average distance of one neighbor position point o of the specified position point and its k neighbor position points.

[0100] In this embodiment, the threshold value of the LOF score is generally set to 1.5; when the LOF score of any position point is higher than the threshold value, the position point is determined as an abnormal position point; otherwise, when the LOF score of any position point is lower than or equal to the threshold value, the position point is determined as a normal position point.

[0101] S204, set the threshold value of the abnormal position points contained in the abnormal voxel, and determine the abnormal voxel in the imaging object according to the judgment result of step S203.

[0102] In this embodiment, the voxel with the number of abnormal position points > 5 is set as the abnormal voxel.

[0103] S205, according to the position coordinates of the abnormal voxel, mark the density abnormal area on the primordial muon imaging feature map generated by step S1.

[0104] In this embodiment, as shown in Figure 5 The density abnormal area marked in the primordial muon imaging feature map after step S2 (i.e. the red frame marked area) is shown in the figure; it can be seen from the figure that the imaging accuracy of the primordial muon cannot meet the more accurate imaging recognition of the abnormal area.

[0105] S3, X-ray single photon imaging of the density abnormal area of the imaging object identified by step S2.

[0106] Since the imaging quality of the primordial muon is limited by the number of cosmic muons within the zenith angle range, the imaging accuracy can only reach centimeter level within a period of time, which cannot meet the requirements of ultra-high precision non-destructive imaging detection such as local crack; therefore, in order to realize accurate imaging recognition of the abnormal area, after the primordial muon imaging, the X-ray single photon counting imaging of the abnormal area of the imaging object is continued, so as to obtain more accurate density abnormal distribution position through the fusion processing of the two kinds of images.

[0107] The specific implementation steps of this step S3 are described as follows.

[0108] S301, referring to Figure 6 X-ray source 4 and single photon counting camera 5 are respectively arranged on the upper and lower roads of closed block 1 and correspond to the density abnormal area position of imaging object 2.

[0109] The X-ray source should ensure that its energy and intensity meet the current detection requirements. Specifically, the selection of the X-ray source should ensure that its energy can penetrate closed blocks and imaging objects, while avoiding information loss due to excessive penetration, thereby meeting the requirements of detection depth and resolution. The intensity selection of the X-ray source must ensure that sufficient signal strength is obtained within the detection time, avoiding errors or low detection efficiency caused by too weak signals, and adapting to the requirements of detection accuracy and speed. Single-photon counting cameras are used to detect X-rays, and their selection should ensure that the resolution and sensitivity meet the experimental requirements. In actual applications, the relative position and angle of the X-ray source and the detector should be adjusted according to the location of the density anomaly area of ​​the imaging object to obtain optimal imaging.

[0110] S302 , capturing low-energy photons emitted by the X-ray source and penetrating the enclosed block and the imaging object by a single-photon counting camera 5 to obtain photon data, namely, the arrival position and energy of each photon.

[0111] The attenuation process of the above photons in the object can be described as:

[0112] I det = I 0 e -ux ,

[0113] Where, I det is the photon intensity received by the detector, I 0 is the initial photon intensity, u is the linear attenuation coefficient of the object, x is the distance that the photon travels through the detection area.

[0114] S303 , using a direct projection method, mapping the arrival position of each photon onto a two-dimensional plane to generate an X-ray single photon imaging reflecting the X-ray single photon count.

[0115] The specific processing steps of step S303 are described as follows.

[0116] S3031, mapping the photon arrival position data obtained in step S302 into a two-dimensional array, specifically, one element in the two-dimensional array is the arrival position, and the other element is the photon count;

[0117] S3032. Visualize the two-dimensional array into a color image to obtain X-ray single-photon imaging that intuitively displays the photon count difference value.

[0118] In the embodiment, the arrival position of the photon is taken as the pixel coordinate; according to the maximum and minimum values of the photon count and the RGB assignment range, the photon count is mapped into the color interval, and then a color image is generated, see Figure 7 ; wherein, based on the photon count range of 20-2000, the RGB range of the color interval is set to (0, 255, 0)-(255, 0, 255), and the closer the color is to green, the lower the photon count is.

[0119] S4, based on the wavelet transform method, the cosmogenic muon imaging feature map obtained from step S1 and the X-ray single photon imaging obtained from step S3 are fused to realize detailed local imaging of the density abnormal region based on the cosmogenic muon imaging feature map, and a fusion image clearly showing the defect position and shape of the imaging object in the closed block is obtained.

[0120] The specific implementation steps of this step S4 are:

[0121] S401, input the two images into the Matlab software, select the wavelet base function to perform multi-level decomposition on the original image; in the embodiment, the wavelet base function is preferably db4 wavelet function, and is set to three-level wavelet decomposition;

[0122] S402, fuse the frequency subbands of each level of the two images respectively; in the embodiment, the fusion method is preferably weighted average method, and the feature matching method is nearest neighbor matching method;

[0123] S403, use the wavelet inverse transform function to reconstruct the fused frequency subbands of each level to obtain the fusion image, as shown in Figure 8 .

[0124] As shown in Figure 8 , the fusion image obtained through the above steps S1-S4 can display the image consistent with the actual imaging object as shown in Figure 2 , and at the same time accurately identify the density abnormal region, through the fusion of X-ray imaging and the cosmogenic muon imaging feature map, the existence of the density abnormal region in the B part of the imaging object can be clearly determined, and the irregular cracks in the density abnormal region are in the shape of E. The imaging result provides effective image support for the subsequent processing mode of the local defect problem of the imaging object.

[0125] In conclusion, the low-radiation X-ray imaging and identification method based on the muon image of the universe in the application not only utilizes the muon imaging to reserve the overall rough image of the target object, but also performs local imaging on the abnormal area through X-ray imaging, effectively performs high-precision imaging on the loss details in the muon imaging, meets the demand of large-scale, fast and non-destructive detection imaging of the target object, has the advantage of low-dose radiation, and has a good application and promotion prospect.

Claims

1. A low-radiation X-ray imaging recognition method based on cosmic muon images, characterized in that: Here are the steps: S1. Lay two liquid scintillator muon detectors, one on top and one on the bottom, on the closed block with the built-in imaging object to collect the true track information of each muon entering and exiting the closed block to generate the cosmic muon imaging feature map; S2. Identify the density anomaly area in the cosmic muon imaging feature map. The specific steps are as follows: S201, dividing the closed block into a voxel grid, and finding the location point where each muon is scattered and the voxel to which the location point belongs in the imaged object based on the true track information of each muon entering and exiting the closed block in step S1; S202, setting k nearest neighbors and calculating the average distance between each location point and multiple location points of its k nearest neighbors; S203, calculating the LOF score of each location point, and judging whether it is abnormal based on the set threshold; S204, setting a threshold value of abnormal position points contained in the abnormal voxels, and determining the abnormal voxels in the imaged object according to the judgment result of step S203; S205, marking the density abnormality area on the cosmic muon imaging feature map generated in step S1 according to the position coordinates of the abnormal voxels; S3. Performing X-ray single photon imaging on the density abnormality region of the imaging object according to the recognition result of step S2; S4. Based on the wavelet transform method, the cosmic muon imaging feature map and the X-ray single photon imaging are fused to obtain a fused image that clearly shows the position and shape of the defects of the imaging object in the closed block.

2. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 1 is characterized in that: The specific implementation steps of step S1 are as follows: S101, obtaining real muon scattering angle data of the closed block based on the real track information of each muon entering the closed block and the real track information of each muon exiting the closed block; S102, constructing a forward model to obtain theoretical muon scattering angle data of a closed block by simulating muon flux distribution; S103, calculating the optimal solution of density distribution of the closed block by using the MLSD inversion algorithm; S104. Project the optimal solution of the density distribution of the closed block onto a two-dimensional plane and assign color values ​​to generate a cosmic muon imaging feature map.

3. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 2 is characterized in that: Step S102 is implemented using Geant4 software; the modeling of the enclosed block and the imaging object is consistent with reality, the model of the object to be imaged is set to have a uniform theoretical density value, and the remaining space in the enclosed block is set to be filled with air; a liquid scintillator muon detector is set in the center of the upper and lower paths of the enclosed block model, and the scattering data collection area is made to completely cover the enclosed block model, and the direction of the cosmic emission muons is set to random.

4. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 2 is characterized in that: The specific processing steps of step S103 are: S1031, assigning an initial density value to each voxel in the voxel grid divided into the closed block S1032. Define the objective function F: where △ θ i For the i The actual scattering angle of a muon, △ θ' i For the i The theoretical scattering angle of muons, N is the total number of muons; S1033. Using the least squares method, with minimizing the objective function F as the optimization goal, the optimal density distribution result of the closed block is obtained through continuous iterative solution.

5. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 1 is characterized in that: In step S201, the method for determining the location point where any muon scatters is as follows: ① When the intersection of the extended line of the incident trajectory of the muon and the extended line of its outgoing trajectory is inside the imaging object, the intersection is the location where the muon is scattered, and correspondingly, the voxel containing the location is the voxel to which the location belongs; ② When the intersection of the extended line of the incident track of the muon and the extended line of its outgoing track is outside the imaging object or there is no intersection, the intersection of the midpoint of the extended line of the incident track and the perpendicular line of the extended line of the outgoing track within the imaging object is taken as the position point where the muon is scattered. Correspondingly, the voxel containing this position point is the voxel to which this position point belongs.

6. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 1 is characterized in that: In step S2, the k value of step S202 is set to 12-15, the threshold of step S203 is set to 1.5-1.8, and the threshold of the number of abnormal location points of step S204 is set to 4-6.

7. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 1 is characterized in that: The specific implementation steps of step S3 are as follows: S301, an X-ray source and a single photon counting camera are respectively set up on the upper and lower roads of the closed block and at locations corresponding to density anomaly areas of the imaging object; S302, capturing low-energy photons emitted by the X-ray source and penetrating the enclosed block and the imaging object using a single-photon counting camera to obtain the arrival position and energy of each photon; S303 , using a direct projection method to map the photon data onto a two-dimensional plane, and generate a color image reflecting the X-ray single photon counts.

8. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 7 is characterized in that: The specific implementation steps of step S303 are as follows: S3031, mapping the photon arrival position data obtained in step S302 into a two-dimensional array, where the two-dimensional array consists of arrival positions and photon counts; S3032: Using the arrival position of the photons as pixel coordinates, the photon counts are mapped to color intervals according to the maximum and minimum values ​​of the photon counts and the set RGB value range to generate a color image.

9. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 1, characterized in that: The specific implementation steps of step S4 are: S401, selecting a wavelet basis function to perform multi-level decomposition on two images; S402, fusing the frequency sub-bands of each level decomposed from the two images respectively; S403 , using an inverse wavelet transform function to reconstruct the fused frequency sub-bands at all levels to obtain a fused image.

10. The low-radiation X-ray imaging recognition method based on cosmic muon images according to claim 9, characterized in that: In step S401, the wavelet basis function adopts the db4 wavelet function and is set to the third-level wavelet decomposition; in step S402, the fusion method adopts the weighted average method, and the feature matching method adopts the nearest neighbor matching method.

Images