Material identification method, device, storage medium and electronic device

Through X-ray image processing and base material decomposition method, the accuracy of liquid material recognition in complex scenarios is solved, the accurate identification of multiple materials is achieved, and the accuracy and scope of application of recognition are improved.

CN118247586BActive Publication Date: 2024-11-22HANGZHOU RAYIN TECH CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410669880.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-28
Publication Date
2024-11-22
Estimated Expiration
2044-05-28

AI Technical Summary

Technical Problem

Existing sandwich-type double-layer detectors are difficult to accurately identify liquid materials in complex scenarios, especially in the case of occlusion, and cannot effectively distinguish liquid properties.

Method used

By acquiring X-ray images, high-energy values, low-energy values ​​and atomic numbers are determined, and the region of interest and the occlusion area are divided based on this information. The equivalent aliased material is decomposed into the occlusion material and unknown material by using the base material decomposition method. The unknown is calculated. The effective atomic number of the material, thereby determining the material category of the object to be tested.

Benefits of technology

The accurate identification of liquid materials is achieved in complex scenarios, and the recognition ability of common materials is improved. It is suitable for the identification of multiple materials. It does not rely on energy spectrum information and object thickness information, and is highly generalized.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118247586B_ABST
    Figure CN118247586B_ABST
Patent Text Reader

Abstract

The present application discloses a material identification method, device, storage medium and electronic device, including: obtaining an X-ray image of a package, determining the high energy value, low energy value and atomic number of each pixel on the X-ray image; determining the region of interest and the obstruction region on the X-ray image based on the high energy value, low energy value and atomic number of each pixel on the X-ray image; based on the high and low energy curves of the aliasing of two preset base materials, the atomic number of the pixels in the obstruction region, the high energy value, low energy value and atomic number of the pixels in the region of interest, using the base material decomposition method to decompose the equivalent aliasing material of the region of interest into a combination of the obstruction material and the unknown material in the obstruction region, and calculating the effective atomic number of the unknown material; based on the effective atomic number of the unknown material, determining the material category of the object to be tested. By using the present application, material identification can be easily realized in a complex environment with obstructions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to X-ray imaging technology, and in particular to a material identification method, device, storage medium and electronic device. Background Art

[0002] X-rays are one of the commonly used non-destructive testing methods in the field of security inspection. In order to improve the detection accuracy, a sandwich-type double-layer detector structure is widely used. Its main structure mainly includes a low-energy detector, a filter and a high-energy detector. By detecting the difference in the absorption rate of X-rays of different energies in materials with different atomic numbers, the atomic number and other information of the object to be tested can be accurately inferred, thereby realizing early warning of dangerous goods.

[0003] At present, the ability of sandwich-type double-layer detectors to distinguish common liquid properties is close to 100% when there are no other interferences, but other interferences often exist in actual complex scenes. Take subway station security inspection as an example, liquid containers are often accompanied by other packages, and finally, purses, clothes, notebooks and other objects are superimposed in the perspective image, causing the atomic number of the liquid to be tested to shift, making it impossible to directly determine the liquid properties. Summary of the invention

[0004] The present application provides a material identification method, device, storage medium and electronic device, which can conveniently realize material identification in a complex environment with obstructions.

[0005] To achieve the above objectives, this application adopts the following technical solutions:

[0006] A material identification method, comprising:

[0007] Acquire an X-ray image of the package, and determine a high energy value, a low energy value, and an atomic number of each pixel on the X-ray image;

[0008] Based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image, determine the region of interest and the obstruction region on the X-ray image; wherein the region of interest is the internal region of the object to be tested, and the obstruction region is the region where the background of the object to be tested is located in the region where the object to be tested is located;

[0009] Based on the preset high and low energy curves of the overlap of two base materials, the atomic number of the pixel points in the occluder area, the high energy value, low energy value and atomic number of the pixel points in the region of interest, the base material decomposition method is used to decompose the equivalent aliasing material of the region of interest into a combination of the occluder material and the unknown material in the occluder area, and the effective atomic number of the unknown material is calculated; wherein the equivalent aliasing material is an equivalent material corresponding to the effective atomic number of the pixel points in the region of interest, the first base material of the two base materials is an organic matter with an atomic number less than a first threshold value, the second base material of the two base materials is an inorganic matter with an atomic number greater than a second threshold value, and the high and low energy curves of the overlap of the two base materials are used to represent the relationship curve between the high energy value and the low energy value obtained after the X-ray passes through the two base materials superimposed together in various thickness combinations;

[0010] The material type of the object to be tested is determined based on the effective atomic number of the unknown material.

[0011] Preferably, the method of calculating the effective atomic number of the material to be tested includes:

[0012] Determine a first mapping relationship for mapping the atomic numbers of the two base materials to the effective atomic numbers of the shielding material in the shielding region, and map the high and low energy curves of the two base materials according to the first mapping relationship to obtain the high and low energy curves of the shielding material;

[0013] According to the first mapping relationship, mapping the high and low energy curves of the two base materials overlapped with the high and low energy curves of the first base material to obtain the high and low energy curves of the shielding material and the first base material overlapped;

[0014] Based on the high energy value and the low energy value of the pixel point in the region of interest, a second mapping relationship is obtained by base material decomposition, which maps the aliased high and low energy curves of the occluder material and the first base material and the high and low energy curves of the occluder material to the high and low energy curves of the equivalent aliased material, and the second mapping relationship is equivalent to a mapping relationship in which the atomic number of the first base material and the effective atomic number of the occluder material are mapped to the effective atomic number of the unknown material; wherein the effective atomic number of the occluder material is determined based on the atomic number of the pixel point in the occluder region;

[0015] Based on the atomic number of the first base material and the effective atomic number of the shielding material, the effective atomic number of the unknown material is determined according to the second mapping relationship.

[0016] Preferably, the method further comprises:

[0017] In the region of interest, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more sub-regions of interest;

[0018] In the occluder region, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more occluder sub-regions;

[0019] The calculating the effective atomic number of the unknown material comprises:

[0020] The sub-regions of interest and the occluder sub-regions are matched one by one, and for the currently matched first sub-region of interest and first occluder sub-region, the effective atomic number of the unknown material obtained by eliminating the occluder material of the first occluder sub-region from the equivalent aliasing material in the first sub-region of interest is calculated.

[0021] Preferably, determining the material type of the object to be tested based on the effective atomic number of the unknown material comprises:

[0022] For any sub-region of interest, excluding the occluder sub-regions that are not reasonably matched with the any sub-region of interest from all occluder sub-regions, and determining the region category of the any sub-region of interest based on the remaining occluder sub-regions;

[0023] If any of the sub-regions of interest is an unblocked region, the effective atomic number of each pixel in the region is used as the effective atomic number of the object to be measured in any of the sub-regions of interest;

[0024] If any of the sub-regions of interest is blocked and there is no remaining blocking object sub-region, then any of the sub-regions of interest is regarded as an invalid sub-region;

[0025] If any of the sub-regions of interest is blocked and there is only one remaining blocking object sub-region, the any sub-region of interest is used as a valid sub-region, and the effective atomic number of the unknown material determined when the blocking object sub-region is matched and combined with the any sub-region of interest is used as the effective atomic number of the object to be tested in the any sub-region of interest;

[0026] If any of the sub-regions of interest is occluded and there are multiple remaining occluding sub-regions, then any of the sub-regions of interest is taken as a valid sub-region, and for each matching combination of the remaining occluding sub-region and any of the sub-regions of interest, the high and low energy curves of the equivalent aliasing material are determined according to the second mapping relationship under the matching combination, and the degree of agreement between the high and low energy curves and the actual high energy values ​​and low energy values ​​of each pixel point in any of the sub-regions of interest is calculated; the occluding sub-region with the highest degree of agreement is selected, and the effective atomic number of the unknown material determined when the occluding sub-region is matched with any of the sub-regions of interest is taken as the effective atomic number of the object to be measured in any of the sub-regions of interest;

[0027] Traverse all sub-regions of interest. If there are unobstructed sub-regions of interest, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in each unobstructed sub-region of interest; if there are no unobstructed sub-regions of interest and there are valid sub-regions, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in all valid sub-regions.

[0028] Preferably, the method of determining the high and low energy curves of the overlap of the two base materials comprises:

[0029] A wedge block made of a first base material and a wedge block made of a second base material are stacked and placed; wherein the heights of the wedge blocks of the first base material and the second base material are M and N respectively, and the two wedge blocks are placed with the inclined surfaces facing upwards, the inclined surface direction of the second base material is parallel to the collimation slit of the X-ray source, and the inclined surface direction of the first base material is perpendicular to the inclined surface direction of the second base material;

[0030] X-rays are irradiated on two overlapping wedge-shaped blocks, and the detector collects an aliased image; wherein the data in the aliased image includes high energy values ​​and low energy values ​​under all combinations of a first base material thickness of x and a second base material thickness of y, x is a uniformly sampled value between 0 and M, and y is a uniformly sampled value between 0 and N;

[0031] X-rays are respectively irradiated on the wedge-shaped blocks of the first base material and the wedge-shaped blocks of the second base material which are placed separately, and the detector collects the image of the first base material and the image of the second base material respectively; wherein the data in the image of the first base material includes the high energy value and the low energy value of the first base material with a thickness of x, and the data in the image of the second base material includes the high energy value and the low energy value of the second base material with a thickness of y;

[0032] Positionally matching the first base material image and the second base material image with the aliasing image, and determining that any pixel position in the aliasing image matches the first pixel position in the first base material image and the second pixel position in the second base material image; wherein the thickness of the first base material at any pixel position in the aliasing image is the same as the thickness of the first base material at the first pixel position in the first base material image, and the thickness of the second base material at any pixel position in the aliasing image is the same as the thickness of the second base material at the second pixel position in the second base material image;

[0033] Establishing a first corresponding relationship, used to indicate that the low energy value of the first pixel position and the low energy value of the second pixel position correspond to the low energy value of any pixel position;

[0034] Establishing a second corresponding relationship, used to indicate that the high energy value of the first pixel position and the high energy value of the second pixel position correspond to the high energy value of any pixel position;

[0035] Using all the first corresponding relationships, a first functional relationship for mapping the second low energy value and the third low energy value to the first low energy value is determined by two-dimensional function fitting;

[0036] Using all the second corresponding relationships, a second functional relationship for mapping the second low energy value and the third low energy value to the first high energy value is determined by two-dimensional function fitting;

[0037] Based on the first functional relationship and the second functional relationship, high and low energy curves of the two base materials overlap are generated.

[0038] Preferably, the region of interest is determined by: compressing the boundary of the region of the object to be measured inwardly, and the region defined by the compressed boundary is the region of interest;

[0039] The method of determining the occluder region is: expanding the boundary of the object region outward, subtracting the area defined by the boundary before the expansion from the total area defined by the boundary after the expansion, and using the remaining area as the occluder region.

[0040] A material identification device, comprising: an image processing unit, a region division unit, an atomic number calculation unit and a material identification unit;

[0041] The image processing unit is used to obtain an X-ray image of the package and determine the high energy value, low energy value and atomic number of each pixel on the X-ray image;

[0042] The region division unit is used to determine the region of interest and the obstruction region on the X-ray image based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image; wherein the region of interest is the internal region of the object to be detected, and the obstruction region is the adjacent region of the object to be detected;

[0043] The atomic number calculation unit is used to decompose the equivalent aliasing material of the region of interest into a combination of the occluder material and the unknown material of the occluder region based on the preset high and low energy curves of the aliasing of two base materials, the atomic number of the pixel points in the occluder region, the high energy value, the low energy value and the atomic number of the pixel points in the region of interest by using the base material decomposition method, and calculate the effective atomic number of the unknown material; wherein the equivalent aliasing material is the equivalent material corresponding to the effective atomic number of the pixel points in the region of interest, the first base material of the two base materials is an organic matter with an atomic number less than a first threshold value, the second base material of the two base materials is an inorganic matter with an atomic number greater than a second threshold value, and the high and low energy curves of the aliasing of the two base materials are used to represent the relationship curve between the high energy value and the low energy value obtained after the X-ray passes through the two base materials superimposed together in various thickness combinations;

[0044] The material identification unit is used to determine the material category of the object to be tested based on the effective atomic number of the unknown material.

[0045] Preferably, in the atomic number calculation unit, the method of calculating the effective atomic number of the material to be tested includes:

[0046] Determine a first mapping relationship for mapping the atomic numbers of the two base materials to the effective atomic numbers of the shielding material in the shielding region, and map the high and low energy curves of the two base materials according to the first mapping relationship to obtain the high and low energy curves of the shielding material;

[0047] According to the first mapping relationship, mapping the high and low energy curves of the two base materials overlapped with the high and low energy curves of the first base material to obtain the high and low energy curves of the shielding material and the first base material overlapped;

[0048] Based on the high energy value and the low energy value of the pixel point in the region of interest, a second mapping relationship is obtained by base material decomposition, which maps the aliased high and low energy curves of the occluder material and the first base material and the high and low energy curves of the occluder material to the high and low energy curves of the equivalent aliased material, and the second mapping relationship is equivalent to a mapping relationship in which the atomic number of the first base material and the effective atomic number of the occluder material are mapped to the effective atomic number of the unknown material; wherein the effective atomic number of the occluder material is determined based on the atomic number of the pixel point in the occluder region;

[0049] Based on the atomic number of the first base material and the effective atomic number of the shielding material, the effective atomic number of the unknown material is determined according to the second mapping relationship.

[0050] Preferably, the area division unit is further used for:

[0051] In the region of interest, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more sub-regions of interest;

[0052] In the occluder region, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more occluder sub-regions;

[0053] In the atomic number calculation unit, calculating the effective atomic number of the unknown material includes:

[0054] The sub-regions of interest and the occluder sub-regions are matched one by one, and for the currently matched first sub-region of interest and first occluder sub-region, the effective atomic number of the first unknown material obtained by eliminating the occluder material of the first occluder sub-region from the equivalent aliasing material in the first sub-region of interest is calculated.

[0055] Preferably, in the atomic number calculation unit, determining the material category of the object to be tested based on the effective atomic number of the unknown material includes:

[0056] For any sub-region of interest, excluding the occluder sub-regions that are not reasonably matched with the any sub-region of interest from all occluder sub-regions, and determining the region category of the any sub-region of interest based on the remaining occluder sub-regions;

[0057] If any of the sub-regions of interest is an unblocked region, the effective atomic number of each pixel in the region is used as the effective atomic number of the object to be measured in any of the sub-regions of interest;

[0058] If any of the sub-regions of interest is blocked and there is no remaining blocking object sub-region, then any of the sub-regions of interest is regarded as an invalid sub-region;

[0059] If any of the sub-regions of interest is blocked and there is only one remaining blocking object sub-region, the any sub-region of interest is used as a valid sub-region, and the effective atomic number of the unknown material determined when the blocking object sub-region is matched and combined with the any sub-region of interest is used as the effective atomic number of the object to be tested in the any sub-region of interest;

[0060] If any of the sub-regions of interest is occluded and there are multiple remaining occluding sub-regions, then any of the sub-regions of interest is taken as a valid sub-region, and for each matching combination of the remaining occluding sub-region and any of the sub-regions of interest, the high and low energy curves of the equivalent aliasing material are determined according to the second mapping relationship under the matching combination, and the degree of agreement between the high and low energy curves and the actual high energy values ​​and low energy values ​​of each pixel point in any of the sub-regions of interest is calculated; the occluding sub-region with the highest degree of agreement is selected, and the effective atomic number of the unknown material determined when the occluding sub-region is matched with any of the sub-regions of interest is taken as the effective atomic number of the object to be measured in any of the sub-regions of interest;

[0061] Traverse all sub-regions of interest. If there are unobstructed sub-regions of interest, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in each unobstructed sub-region of interest; if there are no unobstructed sub-regions of interest and there are valid sub-regions, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in all valid sub-regions.

[0062] Preferably, in the material identification unit, the method of determining the high and low energy curves of the overlap of the two base materials includes:

[0063] A wedge block made of a first base material and a wedge block made of a second base material are stacked and placed; wherein the heights of the wedge blocks of the first base material and the second base material are M and N respectively, and the two wedge blocks are placed with the inclined surfaces facing upwards, the inclined surface direction of the second base material is parallel to the collimation slit of the X-ray source, and the inclined surface direction of the first base material is perpendicular to the inclined surface direction of the second base material;

[0064] X-rays are irradiated on two overlapping wedge-shaped blocks, and the detector collects an aliased image; wherein the data in the aliased image includes high energy values ​​and low energy values ​​under all combinations of a first base material thickness of x and a second base material thickness of y, x is a uniformly sampled value between 0 and M, and y is a uniformly sampled value between 0 and N;

[0065] X-rays are respectively irradiated on the wedge-shaped blocks of the first base material and the wedge-shaped blocks of the second base material which are placed separately, and the detector collects the image of the first base material and the image of the second base material respectively; wherein the data in the image of the first base material includes the high energy value and the low energy value of the first base material with a thickness of x, and the data in the image of the second base material includes the high energy value and the low energy value of the second base material with a thickness of y;

[0066] Positionally matching the first base material image and the second base material image with the aliasing image, and determining that any pixel position in the aliasing image matches the first pixel position in the first base material image and the second pixel position in the second base material image; wherein the thickness of the first base material at any pixel position in the aliasing image is the same as the thickness of the first base material at the first pixel position in the first base material image, and the thickness of the second base material at any pixel position in the aliasing image is the same as the thickness of the second base material at the second pixel position in the second base material image;

[0067] Establishing a first corresponding relationship, used to indicate that the low energy value of the first pixel position and the low energy value of the second pixel position correspond to the low energy value of any pixel position;

[0068] Establishing a second corresponding relationship, used to indicate that the high energy value of the first pixel position and the high energy value of the second pixel position correspond to the high energy value of any pixel position;

[0069] Using all the first corresponding relationships, a first functional relationship for mapping the second low energy value and the third low energy value to the first low energy value is determined by two-dimensional function fitting;

[0070] Using all the second corresponding relationships, a second functional relationship for mapping the second low energy value and the third low energy value to the first high energy value is determined by two-dimensional function fitting;

[0071] Based on the first functional relationship and the second functional relationship, high and low energy curves of the two base materials overlap are generated.

[0072] Preferably, in the area division unit,

[0073] The method of determining the region of interest is: compressing the boundary of the region of the object to be measured inwardly, and the region defined by the compressed boundary is the region of interest;

[0074] The method of determining the occluder region is: expanding the boundary of the object region outward, subtracting the area defined by the boundary before the expansion from the total area defined by the boundary after the expansion, and using the remaining area as the occluder region.

[0075] A computer-readable storage medium stores computer instructions, which can implement any of the above-mentioned material identification methods when executed by a processor.

[0076] An electronic device comprising at least a computer-readable storage medium and a processor;

[0077] The processor is used to read executable instructions from the computer-readable storage medium and execute the instructions to implement any of the material identification methods described above.

[0078] It can be seen from the above technical solution that in the present application, an X-ray image of the package is obtained, the high energy value, low energy value and effective atomic number of each pixel on the X-ray image are determined, and based on this information, the region of interest and the obstruction region are found from the X-ray image, thereby distinguishing the internal area of ​​the object to be tested excluding the boundary and the adjacent area near the boundary, thereby facilitating the identification of the material category of the object to be tested. Next, based on the high and low energy curves of the two base materials, the base material decomposition method is used to decompose the equivalent aliased material of the area of ​​interest into a combination of the occluder material and the unknown material in the occluder area, and the effective atomic number of the unknown material is calculated; in this process, the high and low energy curves of the two base materials are used to give the relationship between the high energy value and the low energy value of the two base materials under various thickness combinations, so that the high and low energy curves imply various thickness combinations of the two base materials, and there is no need to introduce thickness information separately when decomposing the equivalent aliased material. At the same time, combined with the selection of the base materials, it is ensured that the two base materials are organic and inorganic respectively, and the closed interval with the atomic number values ​​of the two base materials as the boundary value can cover the equivalent atomic number values ​​corresponding to the combination of two common materials, so as to ensure that the equivalent aliased material can be decomposed into any two material combinations, which can also ensure that the equivalent aliased material can be decomposed according to the specified combination, and the effective atomic number of the unknown material can be obtained. Finally, the material category of the object to be tested is determined based on the effective atomic number of the unknown material, so that the material of the object to be tested can be effectively identified from complex scenes. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 A schematic diagram of the basic process of the material identification method in complex scenarios in this application;

[0080] Figure 2 This is a schematic diagram of a specific process of a material identification method in a specific embodiment of the present application;

[0081] Figure 3a , Figure 3b and Figure 3c They are the front view, top view and side view of the wedge block respectively;

[0082] Figure 4 This is a schematic diagram of superimposing data calibration of two base materials in a specific embodiment of the present application;

[0083] Figure 5 This is a basic flow chart of the material identification method in a specific embodiment of the present application;

[0084] Figure 6 It is a basic structural diagram of the material identification device in this application;

[0085] Figure 7This is a schematic diagram of the basic structure of the electronic device provided in this application. DETAILED DESCRIPTION

[0086] In order to make the objectives, technical means and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings.

[0087] In order to realize the recognition of liquid materials in complex scenes, some people proposed to select the less obscured liquid area a and its adjacent obstruction area b, and input the RGB values, high and low energy values ​​and atomic numbers of the two areas into a pre-trained deep neural network, and obtain the liquid ratio after processing by the neural network, and then calculate the liquid equivalent atomic number; others proposed to use X-ray energy spectrum to calculate the energy spectrum hardening effect of different objects or base material decomposition by table lookup, calculate the liquid atomic number, and identify the liquid material. However, the above methods either need to add a large number of training sets for training deep neural networks when adding new materials, or need to introduce energy spectrum information and object thickness information, and the scope of application is relatively limited. The present application provides a material recognition method in complex scenes, which implies various thickness combination information of two base materials through the high and low energy curves of the two base materials overlapped by preset, and based on the high and low energy curves of the overlapped, the base material decomposition is used to decompose the material of the obscured area of ​​the object to be tested into the surrounding obstruction material and the unknown material, thereby calculating the atomic number of the unknown material, and identifying the material of the object to be tested based on the atomic number. By rationally selecting the base material, it can be ensured that various combinations of common materials are covered, so that the material identification method of the present application has a strong generalization, and can be used not only for liquid material identification, but also for the identification of other common materials.

[0088] Figure 1 This is a basic flow chart of the material identification method in complex scenarios in this application. Figure 1 As shown, the method includes:

[0089] Step 101, obtaining an X-ray image of the package, and determining the high energy value, low energy value and atomic number of each pixel on the X-ray image.

[0090] The specific processing of this step can be implemented in the existing manner.

[0091] Step 102, based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image, determine the region of interest and the obstruction region on the X-ray image.

[0092] This step finds the region of interest and the occluder region on the X-ray image. The region of interest is the internal region of the object to be tested, that is, the internal part of the object to be tested after removing the boundary; the occluder region is the adjacent region of the object to be tested, that is, the region outside the boundary of the object to be tested and adjacent to the object to be tested.

[0093] In specific implementation, the area of ​​the object to be detected on the X-ray image, that is, the area where material identification is required, can be determined based on the high energy value (i.e., high energy gray value), low energy value (i.e., low energy gray value) and atomic number of each pixel point on the X-ray image. For example, the area of ​​the bottle body can be used for liquid material identification. The area of ​​the object to be detected can be determined, for example, by a pre-trained neural network.

[0094] After the region of the object to be tested is determined, the boundary of the region of the object to be tested is compressed inwardly, and the region defined by the compressed boundary is the region of interest; for example, the boundary of the region of the object to be tested may be eroded inwardly by a number of pixels to obtain the region of interest;

[0095] After determining the area of ​​the object to be measured, the boundary of the area of ​​the object to be measured is expanded outward, and the total area defined by the boundary after expansion is subtracted from the area defined by the boundary before expansion, and the remaining area is used as the occlusion area; for example, the boundary of the area of ​​the object to be measured can be expanded outward by several pixels, and the expanded area is used as the occlusion area.

[0096] Step 103, based on the preset high and low energy curves of the aliasing of two base materials, the effective atomic number of the pixels in the occluder area, the high energy value, low energy value and effective atomic number of the pixels in the area of ​​interest, the base material decomposition method is used to decompose the equivalent aliasing material of the area of ​​interest into a combination of the occluder material and the unknown material in the occluder area, and the effective atomic number of the unknown material is calculated.

[0097] The high and low energy curves of two overlapping base materials are introduced in this application to represent the relationship curve between the high energy value and the low energy value obtained after the X-ray passes through the two base materials superimposed together in various thickness combinations. As described above, the high energy value and the low energy value corresponding to the high and low energy curves of the two overlapping base materials are the superposition of two base materials of different thickness combinations. Through the reasonable selection of the base materials, it can be ensured that the overlapping high and low energy curves cover a continuous range of high and low energy values, and can ensure that the high and low energy values ​​corresponding to various thickness combinations of two common materials are covered. Specifically, the two base materials are an organic substance (hereinafter referred to as the first base material) and an inorganic substance (hereinafter referred to as the second base material), the atomic number of the first base material is less than the first threshold value, and the atomic number of the second base material is greater than the second threshold value, so that the closed interval of the atomic number values ​​of the two base materials as the boundary value can cover the equivalent atomic number values ​​corresponding to the combination of the two common materials. For example, when distinguishing the combination of two materials with atomic numbers of 6-10, the first base material can be polyethylene with an atomic number of 5.5, and the second base material can be aluminum (Al) with an atomic number of 13.

[0098] Through the high and low energy curves of the above two base materials, the base material decomposition method can be used to decompose the equivalent aliased material of the region of interest into a combination of the occluder material and the unknown material in the occluder region. Based on this decomposition process, combined with the effective atomic number of the pixel points in the occluder region, the high energy value, low energy value and effective atomic number of the pixel points in the region of interest, the effective atomic number of the unknown material can be calculated. The equivalent aliased material mentioned here is the equivalent material represented by the effective atomic number of the region of interest.

[0099] Step 104, determining the material type of the object to be tested based on the effective atomic number of the unknown material.

[0100] At this point, the basic process of the material identification method in complex scenarios in this application is completed.

[0101] In the above process, considering that different parts of the object to be tested may be obstructed by more than one material or obstructions of different thicknesses of the same material, or the object to be tested may be partially obstructed by the unobstructed part, therefore, both the region of interest and the obstruction region can usually be divided into multiple sub-regions, and each sub-region of interest and obstruction sub-region is combined and matched. In step 103, the effective atomic number of the unknown material in the sub-region of interest under the corresponding combination is calculated. Accordingly, in step 104, various matching combinations of the sub-region of interest and the obstruction sub-region are screened, and then based on the effective atomic number of the unknown material under the screened corresponding combination, the material category of the object to be tested is determined.

[0102] As mentioned above, using the material identification method of the present application, the superposition of different common materials can be effectively decomposed, and the decomposition process does not require the calculation of energy spectrum information and thickness information of obstructions, nor does it require the introduction of a large number of neural network training sample sets, which can easily realize material identification in complex occlusion environments.

[0103] The following is a specific implementation of the material identification method of the present application. Before introducing the material identification method of the specific embodiment, the method for determining the high and low energy curves of the two base materials used in the aforementioned step 103 will first be introduced. The determination process can be achieved by calibrating the security detector using the two base materials. Here, PE and Al are used as two base materials for illustration. The specific process is as follows Figure 2 As shown, including:

[0104] Step 201 : measuring and generating a base material image of a single base material.

[0105] To achieve calibration, you need to prepare a wedge block made of the first base material and a wedge block made of the second base material, that is, prepare a wedge block of PE and a wedge block of Al. The two wedge blocks are placed with the inclined surface facing upward, so the bottom surfaces of the two wedge blocks are squares of the same size, and the heights can be the same or different. In this example, it is assumed that the three-view drawing of the wedge block is as follows Figure 3a~3c As shown, the height of the PE wedge block is 15 cm and the bottom is a square with a side length of 20 cm; the height of the Al wedge block is 8 cm and the bottom is a square with a side length of 20 cm.

[0106] In step 201, base material images of two single materials are measured respectively, and the image data includes high and low energy information of the two single materials at different thicknesses. The high and low energy information below refers to high energy grayscale values ​​and low energy grayscale values.

[0107] Specifically, X-rays are used to irradiate PE wedge blocks, and the high energy values ​​and low energy values ​​when passing through PE materials of different thicknesses are collected by the detector; then X-rays are used to irradiate Al wedge blocks, and the high energy values ​​and low energy values ​​when passing through PE materials of different thicknesses are collected by the detector.

[0108] Step 202, fitting high and low energy curves of a single base material.

[0109] The high and low energy curves of PE at different thicknesses collected in step 201 are fitted to map any low energy value of PE within the effective range to a high energy value. Various existing fitting methods, such as Gaussian regression, can be used.

[0110] The high and low energy curves of Al at different thicknesses collected in step 201 are fitted to map any low energy value of Al within the effective range to a high energy value. Various existing fitting methods, such as Gaussian regression, can be used.

[0111] Step 203 , measuring aliased images of two base materials at different thickness combinations.

[0112] Place the two wedge-shaped blocks on top of each other on the security inspection equipment. Figure 4 As shown in the figure, 1 represents the X-ray source, 2 represents the collimation slit, 3 represents the optical path of the X-ray photon, 4 represents the aluminum wedge block, 5 represents the fixed tooling, 6 represents the PE wedge block, 7 represents the supporting structure such as the security inspection equipment transmission device, 8 represents the low-energy detector, and 9 represents the high-energy detector. The X-rays released by the X-ray emitting component become fan-shaped beams after passing through the collimation slits, irradiating the two wedge blocks. In order to avoid the different thicknesses of objects passing through different optical paths due to non-parallel beams, such as Figure 4As shown, the slope direction of Al is parallel to the collimation slit of the radiation source, and the slope direction of PE is perpendicular to the slope direction of Al, driving the conveyor belt of the wedge block to move perpendicular to the plane of the screen to ensure that the X-ray light path detected by the same detector pixel passes through the same thickness of Al. After the above arrangement of the base material and the data collection of the detector, the detector collects and generates an aliased image. The data of the aliased image almost contains low energy values ​​and high energy values ​​under any combination of aluminum 0-8 cm and PE 0-15 cm.

[0113] Step 204, fitting high and low energy curves of PE and Al aliasing.

[0114] When X-rays pass through two materials, Al and PE, and are received by the detector, the path length of the light path in the two wedge blocks, i.e., the thickness information, can usually be calculated using precise geometric positions. The amount of calculation required to calculate the thickness information is very large. This application proposes a solution that directly uses the low energy values ​​of the two base materials to map to the high and low energy values ​​after superposition. In this processing method, there is no need to directly calculate the thickness. Instead, the low energy values ​​of the two single base materials represent the thickness information, and the two low energy values ​​are mapped to the high and low energy values ​​of the two materials after superposition according to the thickness corresponding to the corresponding low energy values.

[0115] As described in step 203, the wedge blocks of the two base materials are fixed according to the above-mentioned tooling, and the signals received by the same pixel in the detector are extended in the direction of belt movement. The light paths corresponding to these signals pass through the same thickness of Al, while the thickness of PE increases from 0 to the maximum height of the wedge block. Therefore, different pixel positions on the perspective view that passes through the two wedge blocks at the same time represent different thickness combinations of the two base materials.

[0116] By performing position matching on the base material images (Al image and PE image) of the single base material obtained in step 201 and the aliased image simultaneously passing through the two wedge blocks obtained in step 202, the low energy value L of the single Al and PE at each pixel position of the aliased image can be obtained. Al and L PE And high energy value H Al and H PE That is to say, for a certain pixel position of the aliased image, the pixel position A corresponds to a thickness combination of Al and PE. Assuming that the PE thickness is x and the Al thickness is y, the first position of the PE image and the second position of the Al image can be matched according to the position matching of the image, wherein the PE thickness corresponding to the first position is the same as the PE thickness corresponding to the pixel position A in the aliased image, and the Al thickness corresponding to the second position of the Al image is the same as the Al thickness corresponding to the pixel position A in the aliased image. At the same time, the pixel position in the aliased image also corresponds to the superimposed low energy value L Al+PE and high energy value H Al+PE, thus, the first corresponding relationship can be established to represent L PE and L Al Corresponding to L Al+PE , and then establish a second corresponding relationship to represent L PE and L Al Corresponding to H Al+PE As described above, for each position of the aliased image, that is, corresponding to each thickness combination, a corresponding first correspondence relationship and a second correspondence relationship can be established. The specific position matching process can be implemented in an existing manner.

[0117] Next, based on all the first correspondences, we can fit L PE and L Al Mapped to the corresponding L Al+PE The first functional relationship of L can be fitted based on all the second corresponding relationships. PE and L Al Mapped to the corresponding H Al+PE The second functional relationship of .

[0118] Specifically, the above two mapping relationships are essentially two-dimensional function fittings, and the variable coefficient decomposition method can be used to reduce the difficulty of fitting and reduce the impact of statistical noise. Taking the establishment of the first mapping relationship as an example, in more detail, first, the data of all pixels can be divided into different low energy values ​​L Al Group the same low energy values ​​into the same group, and transform the two-dimensional data into a set of several one-dimensional functions; then traverse each group L Al , the independent variable is L PE , the output is aliased to the L Al Low energy data L Al+PE , whose coefficients {a0, a1,…, a m}; Next, after traversing, we will get several groups of coefficients with the same order a i The set composed is L Al The function can be further fitted with polynomials to obtain the coefficients {b i0 , b i1 ,…,b in}, effectively compressing the base material superposition model into m×n coefficients, that is, obtaining the first functional relationship mentioned above. The second functional relationship can be obtained through the same processing. Using the above two functional relationships, based on any input L Al and L PE , can determine the high energy value H after superposition Al+PE and low energy value L Al+PE , and based on the corresponding high energy value H Al+PE and low energy value L Al+PE The combination of values ​​generates high and low energy curves of the two base materials.

[0119] At this point, the high and low energy curves of the two base materials overlap are generated, that is, the corresponding relationship curve between the high energy value and the low energy value of the two base materials under various thickness combinations.

[0120] Next, the specific implementation of the material identification method in the present application is described through a specific embodiment. In the specific embodiment, liquid material identification in a security inspection scenario is used as an example for description, and the two base materials are still selected as Al and PE. Figure 5 This is a basic flow chart of the material identification method in the specific embodiment of this application. Figure 5 As shown, the liquid material identification method process includes:

[0121] Step 501, obtain an X-ray image of the package, and determine the high energy value, low energy value and effective atomic number of each pixel on the X-ray image.

[0122] The specific processing of this step can be implemented in the existing manner.

[0123] Specifically, the high energy value and low energy value of each pixel point can be corrected by full load and background correction to eliminate the inconsistency between detectors, and then the high and low energy curves of the fitted single base material obtained by the above step 202 are used to calculate the atomic number corresponding to each pixel, output the atomic number map and reduce noise through a specific filtering method (such as BM3D).

[0124] When calculating the atomic number corresponding to a pixel A, the low energy value L of pixel A can be determined first. A and high energy value H A , then L Al and L PE All are set to L A Equal, according to the high and low energy curves of the two base materials, the corresponding H Al and H PE , then based on H A , H Al and H PE The numerical relationship is determined by H Al and H PE To H A The mapping relationship is equivalent to the atomic number Z of Al. Al and the atomic number Z of PE PE To the atomic number Z of pixel A A The mapping relationship between the high and low energy values ​​of different materials is consistent with the relationship between the atomic numbers, so the above mapping relationship can be equivalent. Therefore, using this mapping relationship, based on Z Al and Z PE , we can calculate Z A .

[0125] Step 502, based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image, determine the region of interest and the obstruction region on the X-ray image.

[0126] In this step, the liquid area to be tested is first determined. Specifically, the high and low energy grayscale images and the filtered atomic number image obtained in the above step 501 can be input into a trained deep learning neural network (such as R-CNN) model to locate the liquid container, that is, the liquid area to be tested. Then, the smallest rectangle that can cover the above area can be found and expanded to a certain number (for example, 20) pixels around it as a sub-image. The following processing is performed on the sub-image to accelerate the calculation.

[0127] In order to eliminate the edge noise effect, the precise boundary of the liquid container located by the neural network model is expanded outward by several pixels (for example, 10 pixels), and the expanded area is inverted as the "occluder area", which contains the pure occluder to be selected; the precise boundary of the liquid container is eroded inward by several pixels (for example, 10 pixels), and the remaining area after corrosion is used as the "region of interest", which contains the liquid to be tested with the occluder to be removed.

[0128] Step 503: divide the region of interest into one or more sub-regions of interest, and divide the occluder region into one or more occluder sub-regions.

[0129] In the region of interest, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more sub-regions of interest;

[0130] In the obstruction region, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more obstruction sub-regions.

[0131] The region growing algorithm can be used for sub-region division. Taking the division of the region of interest as an example, the region of interest is divided into several subsets according to the high and low energy values ​​and atomic number of the image through the region growing algorithm, where the data within each subset can be considered to be similar in superposition, that is, the high and low energy values ​​and atomic number of each pixel within the subset are similar. When using the region growing algorithm, the seed point can be located on the boundary of the region of interest. If an unclassified pixel is found to be adjacent to a classified pixel and the difference between the high energy value, low energy value and atomic number of the classified pixel is less than the threshold, the unclassified pixel is merged into the subset where the classified pixel is located, and a new round of search is started. If there are still unclassified pixels after the process is completed, the above process is repeated until all pixels are classified.

[0132] Of course, other methods can also be used for sub-region division, as long as the differences between high and low energy values ​​and atomic numbers of different pixels in the same sub-region after division are within the set range.

[0133] Step 504 , matching the sub-region of interest and the obstruction sub-region one by one, and determining the effective atomic number of the unknown material in the sub-region of interest under the corresponding matching combination.

[0134] In the present application, the sub-regions of interest and the obstruction sub-regions are matched one by one to calculate the effective atomic number of the unknown material. Compared with the existing processing method of matching the sub-region of interest only with the adjacent obstruction sub-region, the matching method of the present application is more comprehensive.

[0135] Taking a matching combination of an interested sub-region and an occluder sub-region as an example, the calculation method of the effective atomic number of an unknown material is explained. The specific processing can include the following four steps:

[0136] Step 1, calculating the effective atomic number of the shielding material, and determining the mapping relationship between the atomic numbers of PE and Al and the effective atomic number of the shielding material;

[0137] The atomic numbers of all pixels have been calculated through the aforementioned step 501, including the atomic numbers of all pixels in the current occluder sub-region, and the effective atomic number of the occluder material in the current occluder sub-region is calculated based on these atomic numbers. Since the atomic numbers of the pixels in the region are relatively close when the occluder sub-region is divided, it can be considered that there is only one material in the same occluder sub-region, so it is necessary to calculate an atomic number value based on the atomic numbers of all pixels as the effective atomic number of the occluder material in the current occluder region. The specific calculation method can be to calculate the mean or median of the atomic numbers of all pixels, which can be set as needed in actual implementation.

[0138] After determining the effective atomic number of the shielding material, based on the quantitative relationship between the atomic numbers of PE and Al and the effective atomic number of the shielding material, a mapping relationship of the atomic numbers of PE and Al to the effective atomic number of the shielding material is determined, which is called the first mapping relationship.

[0139] Step 2, determining the high and low energy curves of the shielding material;

[0140] According to the first mapping relationship determined in step 1, the high and low energy curves of PE and the high and low energy curves of Al determined in the above calibration process are mapped to the high and low energy curves of the shielding material.

[0141] Step 3, according to the first mapping relationship, mapping the high and low energy curves of PE and Al aliasing and the high and low energy curves of PE to obtain the high and low energy curves of the obstruction material and PE aliasing;

[0142] In the present application, the mapping relationship between the high and low energy curves of PE and Al aliasing and the high and low energy curves of PE is defined as the high and low energy curves of the obstruction material and PE aliasing (i.e., the high and low energy curves of the obstruction material + PE aliasing), which is the same as the first mapping relationship. Figure 2 The high and low energy curves generated by the process shown represent the relationship between the high energy value and the low energy value under different thickness combinations of PE and Al. The value range of the high and low energy values ​​of the aliased high and low energy curves combined with the high and low energy curves of PE can cover the value range of the high and low energy values ​​of various common obstruction materials and PE in various thickness combinations. At the same time, through aliasing experiments on various materials, it is found that the high and low energy curves of the obstruction material and PE aliased according to the above-mentioned first mapping relationship are well fitted with the high and low energy values ​​after aliasing obtained by actual measurement. Therefore, the high and low energy curves of the obstruction material and PE aliased can be directly mapped according to the first mapping relationship.

[0143] Step 4, decomposing the equivalent aliasing material of the sub-region of interest into a combination of the occluder material and the unknown material of the occluder sub-region by the basis material decomposition method, and calculating the effective atomic number of the unknown material.

[0144] Assuming that the sub-region of interest is just blocked by the occluder material of the occluder sub-region, the equivalent aliasing material of the sub-region of interest can be regarded as the occluder material and the liquid to be tested in the sub-region of interest superimposed with a certain thickness. Here, the liquid to be tested is called unknown material. In other words, the sub-region of interest can be regarded as the occluder material + unknown material.

[0145] The occluder material +PE and the occluder material in the occluder sub-region are respectively used as two base materials (recorded as base material combination C), and a mapping relationship (hereinafter referred to as the second mapping relationship) from the base material combination C to the equivalent aliasing material of the sub-region of interest is determined by the base material decomposition method, which is consistent with the second mapping relationship determined by H in the above step 501. Al and H PE To H AThe mapping relationship is the same. Through observation, it is found that the two base materials are occluder material + PE (hereinafter referred to as base material 1) and occluder material (hereinafter referred to as base material 2), among which base material 2 can also be regarded as occluder material + occluder material, and the equivalent aliasing material is occluder material + unknown material. It can be seen that the three materials all include occluder material. If the occluder material is removed from the three materials, the linear relationship between the three materials after removal remains unchanged, and the three materials after removing the occluder material are PE (corresponding to base material 1), occluder material (corresponding to base material 2) and unknown material (corresponding to equivalent aliasing material). That is to say, the second mapping relationship of high and low energy values ​​of base material 1 and base material 2 mapped to high and low energy values ​​of equivalent aliasing material can be equivalent to the mapping relationship of high and low energy values ​​of PE and high and low energy values ​​of occluder material mapped to high and low energy values ​​of unknown material, which can also be equivalent to the third mapping relationship of atomic number of PE and effective atomic number of occluder material mapped to effective atomic number of unknown material, so the third mapping relationship becomes a known mapping relationship.

[0146] According to the third mapping relationship, the atomic number of PE and the effective atomic number of the obstruction material are mapped to the effective atomic number of the unknown material. The third mapping relationship, the atomic number of PE and the effective atomic number of the obstruction material are all known, so the effective atomic number of the unknown material can be calculated.

[0147] In the calculation process of this step, each pixel point in the sub-region of interest has its corresponding equivalent aliasing material. Therefore, the above processing process needs to be performed once for each pixel point, and the effective atomic number of the unknown material at the pixel point position can be obtained for each pixel point. Based on the effective atomic number of the unknown material at these pixel points, the classification atomic number of the sub-region of interest can be calculated, which is used for the classification of the sub-region of interest in the next step. Among them, the way to calculate the classification atomic number can be to take the mean or median of the atomic numbers of all unknown materials.

[0148] Step 505 : For all matching combinations of the sub-regions of interest and the occluder sub-regions, unreasonable combinations are screened.

[0149] Among them, unreasonable combinations may include but are not limited to the following combinations: the difference between the effective atomic number of the occluder and the effective atomic number of the unknown material in the matching combination is too large (for example, the difference is greater than the set threshold), the low energy value of the occluder sub-region is greater than the low energy value of the sub-region of interest, etc.

[0150] Step 506: for each sub-region of interest A, determine the region category of the sub-region of interest A based on screening the remaining occluder sub-regions.

[0151] In this application, the categories of sub-regions of interest include the following four:

[0152] 1. The sub-region of interest is an unoccluded region, for example, the occluder material in the remaining occluder sub-regions in the sub-region of interest matching combination is air or the occluder is judged to be a thin object (such as a plastic bag, etc.) through a specific threshold;

[0153] 2. The sub-region of interest is an occluded region and there are no remaining occluder sub-regions, that is, the sub-region of interest can be determined to be occluded based on the effective atomic number and high and low energy values, but there are no remaining occluder sub-regions after screening in step 505, for example, the effective atomic number of the occluder material exceeds 15, the occluder contour is completely located in the region where the object to be measured is located, or it is difficult to find an occluder sub-region with a pixel number exceeding a preset value due to the uneven occluder, etc.;

[0154] 3. The sub-region of interest is blocked and after screening in step 505, there is one blocking object sub-region left;

[0155] 4. The sub-region of interest is blocked and after screening in step 505, a plurality of blocking object sub-regions remain.

[0156] Step 507: Determine the effective atomic number of the liquid to be tested in each sub-region of interest according to the region category of each sub-region of interest.

[0157] If the sub-region of interest is the aforementioned first category, the effective atomic number of each pixel in the sub-region of interest is used as the effective atomic number of the object to be measured in the sub-region of interest;

[0158] If the sub-region of interest is of the aforementioned second category, the sub-region of interest is set as an invalid sub-region;

[0159] If the sub-region of interest is the aforementioned third category, the sub-region of interest is taken as a valid sub-region, and the effective atomic number of the unknown material at each pixel position determined when the remaining occluder sub-region after screening is matched and combined with the sub-region of interest is taken as the effective atomic number of the object to be tested in the sub-region of interest;

[0160] If the sub-region of interest is the aforementioned fourth category, the sub-region of interest is taken as a valid sub-region, and for each remaining matching combination of the occluder sub-region and the sub-region of interest, the high and low energy curves of the equivalent aliasing material are determined according to the aforementioned second mapping relationship under the matching combination, and the degree of consistency between the high and low energy curves and the actual high energy values ​​and low energy values ​​of each pixel point in the sub-region of interest is calculated; the occluder sub-region with the highest degree of consistency is selected, and the effective atomic number of the unknown material at each pixel position determined when the occluder sub-region is matched with the sub-region of interest is used as the effective atomic number of the object to be measured in the sub-region of interest.

[0161] Step 508: Determine the material type of the liquid to be tested according to the effective atomic number of the liquid to be tested in each sub-region of interest.

[0162] All sub-regions of interest are traversed. If there are unobstructed sub-regions of interest, the material category of the object to be tested is determined based on the effective atomic numbers of the liquids to be tested in each unobstructed sub-region of interest. Specifically, the mean or median of the effective atomic numbers of the liquids to be tested in these sub-regions of interest can be taken as the final atomic number, and the material category is determined based on this. If there are no unobstructed sub-regions of interest and there are valid sub-regions, the material category of the object to be tested is determined based on the effective atomic numbers of the objects to be tested in all valid sub-regions. Specifically, the mean or median of the effective atomic numbers of the liquids to be tested in these sub-regions of interest can be taken as the final atomic number, and the material category is determined based on this. If there are no unobstructed sub-regions of interest and no valid sub-regions, material recognition failure information is output.

[0163] At this point, the liquid material identification method process in this embodiment ends.

[0164] The material identification method in the above-mentioned present application can map two specific base materials to the superposition of any two materials with effective atomic numbers between 5 and 15, and has extremely strong generalization, especially for the problem of misjudgment of unknown materials caused by material stacking in complex scenes; at the same time, the recognition accuracy is greatly improved through the one-by-one matching processing of the occluded sub-region and the sub-region of interest, and is suitable for most scenes, especially expanding the scope of application of liquid identification in the security inspection field; compared with other liquid material identification methods, the material identification method of the present application does not need to measure energy spectrum information and object thickness information, nor does it need to add a large number of training sets when new materials are added in deep learning; at the same time, the base materials Al and PE selected in the calibration process of the present application method cover a large dynamic range, and the generalization performance is good for thinner and thicker objects.

[0165] The above is the specific implementation of the material identification method in this application. This application also provides a material identification device, Figure 6 Figure 1 is a schematic diagram of the basic structure of the device. Figure 6 As shown, the device includes: an image processing unit, a region division unit, an atomic number calculation unit and a material identification unit.

[0166] The image processing unit is used to obtain an X-ray image of the package and determine the high energy value, low energy value and atomic number of each pixel on the X-ray image;

[0167] A region division unit, used to determine a region of interest and an obstruction region on the X-ray image based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image; wherein the region of interest is an internal region of the object to be detected, and the obstruction region is an adjacent region of the object to be detected;

[0168] an atomic number calculation unit, for decomposing the equivalent aliased material of the region of interest into a combination of the occluder material and the unknown material in the occluder region by using a base material decomposition method based on the preset high and low energy curves of the aliasing of two base materials, the atomic number of the pixel points in the occluder region, the high energy value, the low energy value and the atomic number of the pixel points in the region of interest, and calculating the effective atomic number of the unknown material; wherein the equivalent aliased material is the equivalent material corresponding to the effective atomic number of the pixel points in the region of interest, the first base material of the two base materials is an organic matter having an atomic number less than a first threshold value, the second base material of the two base materials is an inorganic matter having an atomic number greater than a second threshold value, and the high and low energy curves of the aliasing of the two base materials are used to represent a relationship curve between the high energy value and the low energy value obtained after the X-ray passes through the two base materials superimposed together in various combinations of different thicknesses;

[0169] The material identification unit is used to determine the material category of the object to be tested based on the effective atomic number of the unknown material.

[0170] Optionally, in the atomic number calculation unit, the method of calculating the effective atomic number of the material to be tested includes:

[0171] Determine a first mapping relationship for mapping the atomic numbers of the two base materials to the effective atomic numbers of the shielding material in the shielding region, and map the high and low energy curves of the two base materials according to the first mapping relationship to obtain the high and low energy curves of the shielding material;

[0172] According to the first mapping relationship, high and low energy curves of the two base materials overlap and the high and low energy curves of the first base material are mapped to obtain high and low energy curves of the obstruction material and the first base material overlap;

[0173] Based on the high energy value and the low energy value of the pixel point in the region of interest, a second mapping relationship is obtained by decomposing the aliased high and low energy curves of the occluder material and the first base material and the high and low energy curves of the occluder material to the high and low energy curves of the equivalent aliased material, and the second mapping relationship is equivalent to a mapping relationship from the atomic number of the first base material and the effective atomic number of the occluder material to the effective atomic number of the unknown material; wherein the effective atomic number of the occluder material is determined based on the atomic number of the pixel point in the occluder region;

[0174] Based on the atomic number of the first base material and the effective atomic number of the shielding material, the effective atomic number of the unknown material is determined according to the second mapping relationship.

[0175] Optionally, the area division unit may be further configured to:

[0176] In the region of interest, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more sub-regions of interest;

[0177] In the occluder region, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more occluder sub-regions;

[0178] In the atomic number calculation unit, calculating the effective atomic number of the unknown material includes:

[0179] The sub-regions of interest and the occluder sub-regions are matched one by one, and for the currently matched first sub-region of interest and first occluder sub-region, the effective atomic number of the first unknown material obtained by eliminating the occluder material of the first occluder sub-region from the equivalent aliasing material in the first sub-region of interest is calculated.

[0180] Optionally, in the atomic number calculation unit, determining the material category of the object to be tested based on the effective atomic number of the unknown material includes:

[0181] For any sub-region of interest, excluding the occluder sub-regions that are not reasonably matched with any sub-region of interest from all occluder sub-regions, and determining the region category of any sub-region of interest based on the remaining occluder sub-regions;

[0182] If any sub-region of interest is an unblocked region, the effective atomic number of each pixel in the region is used as the effective atomic number of the object to be measured in any sub-region of interest;

[0183] If any sub-region of interest is occluded and there is no remaining occluding sub-region, any sub-region of interest is regarded as an invalid sub-region;

[0184] If any sub-region of interest is blocked and there is only one remaining blocking object sub-region, then any sub-region of interest is taken as a valid sub-region, and the effective atomic number of the unknown material determined when the blocking object sub-region is matched and combined with any sub-region of interest is taken as the effective atomic number of the object to be tested in any sub-region of interest;

[0185] If any sub-region of interest is occluded and there are multiple remaining occluding sub-regions, any sub-region of interest is taken as a valid sub-region, and for each matching combination of the remaining occluding sub-region and any sub-region of interest, the high and low energy curves of the equivalent aliasing material are determined according to the second mapping relationship under the matching combination, and the degree of agreement between the high and low energy curves and the actual high energy values ​​and low energy values ​​of each pixel point in any sub-region of interest is calculated; the occluding sub-region with the highest degree of agreement is selected, and the effective atomic number of the unknown material determined when the occluding sub-region is matched with any sub-region of interest is taken as the effective atomic number of the object to be measured in any sub-region of interest;

[0186] All sub-regions of interest are traversed. If there are unobstructed sub-regions of interest, the material category of the object to be tested is determined based on the effective atomic number of the object to be tested in each unobstructed sub-region of interest; if there are no unobstructed sub-regions of interest and there are valid sub-regions, the material category of the object to be tested is determined based on the effective atomic number of the object to be tested in all valid sub-regions.

[0187] Optionally, in the material identification unit, a method of determining high and low energy curves of two base materials overlapped includes:

[0188] A wedge block made of a first base material and a wedge block made of a second base material are stacked and placed; wherein the heights of the wedge blocks of the first base material and the second base material are M and N respectively, and the two wedge blocks are placed with the inclined surfaces facing upwards, the inclined surface direction of the second base material is parallel to the collimation slit of the X-ray source, and the inclined surface direction of the first base material is perpendicular to the inclined surface direction of the second base material;

[0189] X-rays are irradiated on two overlapping wedge-shaped blocks, and the detector collects an aliased image; wherein the data in the aliased image includes high energy values ​​and low energy values ​​under all combinations of a first base material thickness of x and a second base material thickness of y, x is a uniformly sampled value between 0 and M, and y is a uniformly sampled value between 0 and N;

[0190] X-rays are respectively irradiated on the wedge-shaped blocks of the first base material and the wedge-shaped blocks of the second base material which are placed separately, and the detector collects the image of the first base material and the image of the second base material respectively; wherein the data in the image of the first base material includes the high energy value and the low energy value of the first base material with a thickness of x, and the data in the image of the second base material includes the high energy value and the low energy value of the second base material with a thickness of y;

[0191] Positionally matching the first base material image and the second base material image with the aliased image, and determining that any pixel position in the aliased image matches the first pixel position in the first base material image and the second pixel position in the second base material image; wherein the thickness of the first base material at any pixel position in the aliased image is the same as the thickness of the first base material at the first pixel position in the first base material image, and the thickness of the second base material at any pixel position in the aliased image is the same as the thickness of the second base material at the second pixel position in the second base material image;

[0192] Establishing a first corresponding relationship for indicating that the low energy value at the first pixel position and the low energy value at the second pixel position correspond to the low energy value at any pixel position;

[0193] Establishing a second corresponding relationship for indicating that the high energy value of the first pixel position and the high energy value of the second pixel position correspond to the high energy value of any pixel position;

[0194] Using all the first corresponding relationships, a first functional relationship for mapping the second low energy value and the third low energy value to the first low energy value is determined by two-dimensional function fitting;

[0195] Using all the second corresponding relationships, a second functional relationship for mapping the second low energy value and the third low energy value to the first high energy value is determined by two-dimensional function fitting;

[0196] Based on the first functional relationship and the second functional relationship, high and low energy curves of two base materials overlap are generated.

[0197] Optionally, in the area division unit,

[0198] The method of determining the region of interest is: compressing the boundary of the region of the object to be measured inwardly, and the region defined by the compressed boundary is the region of interest;

[0199] The method for determining the occluder region is: the boundary of the object region is expanded outward, the total area defined by the boundary after the expansion is subtracted from the area defined by the boundary before the expansion, and the remaining area is used as the occluder region.

[0200] The present application also provides a computer-readable storage medium, which stores instructions, and when the instructions are executed by a processor, the steps in the material identification method described above can be executed. In practical applications, the computer-readable medium can be included in each device / apparatus / system in the above embodiments, or it can exist independently without being assembled into the device / apparatus / system. Instructions are stored in the computer-readable storage medium, and when the instructions stored in the computer-readable storage medium are executed by a processor, the steps in the material identification method described above can be executed.

[0201] According to the embodiments disclosed in the present application, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof, but it is not used to limit the scope of protection of the present application. In the embodiments disclosed in the present application, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus, or device.

[0202] Figure 7 This application also provides an electronic device. Figure 7 As shown, it shows a schematic diagram of the structure of the electronic device involved in the embodiment of the present application, specifically:

[0203] The electronic device may include a processor 701 with one or more processing cores, a memory 702 with one or more computer-readable storage media, and a computer program stored in the memory and executable on the processor. When the program in the memory 702 is executed, a material identification method may be implemented.

[0204] Specifically, in practical applications, the electronic device may further include components such as a power supply 703, an input / output unit 704, etc. Those skilled in the art may understand that Figure 7 The structure of the electronic device shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0205] Processor 701 is the control center of the electronic device. It uses various interfaces and lines to connect various parts of the entire electronic device. It executes various functions of the server and processes data by running or executing software programs and / or modules stored in memory 702 and calling data stored in memory 702, thereby monitoring the electronic device as a whole.

[0206] The memory 702 can be used to store software programs and modules, that is, the above-mentioned computer-readable storage medium. The processor 701 executes various functional applications and data processing by running the software programs and modules stored in the memory 702. The memory 702 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the server, etc. In addition, the memory 702 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 702 may also include a memory controller to provide the processor 701 with access to the memory 702.

[0207] The electronic device also includes a power supply 703 for supplying power to various components, which can be logically connected to the processor 701 through a power management system, so as to manage charging, discharging, power consumption and other functions through the power management system. The power supply 703 can also include any components such as one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, and power status indicators.

[0208] The electronic device may also include an input and output unit 704, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, and optical signal inputs related to user settings and function control. The input unit output 704 may also be used to display information input by the user or information provided to the user and various graphical user interfaces, which may be composed of graphics, text, icons, videos, and any combination thereof.

[0209] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A material identification method, characterized in that: include: Acquire an X-ray image of the package, and determine a high energy value, a low energy value, and an atomic number of each pixel on the X-ray image; Based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image, determine the region of interest and the occluder region on the X-ray image; wherein the region of interest is the internal region of the object to be tested, and the occluder region is the region remaining after the total region defined by the boundary after the boundary of the object to be tested is expanded outwards minus the region defined by the boundary before the expansion; Based on the preset high and low energy curves of the overlap of two base materials, the atomic number of the pixel points in the occluder area, the high energy value, low energy value and atomic number of the pixel points in the region of interest, the base material decomposition method is used to decompose the equivalent aliasing material of the region of interest into a combination of the occluder material and the unknown material in the occluder area, and the effective atomic number of the unknown material is calculated; wherein the equivalent aliasing material is an equivalent material corresponding to the effective atomic number of the pixel points in the region of interest, the first base material of the two base materials is an organic matter with an atomic number less than a first threshold value, the second base material of the two base materials is an inorganic matter with an atomic number greater than a second threshold value, and the high and low energy curves of the overlap of the two base materials are used to represent the relationship curve between the high energy value and the low energy value obtained after the X-ray passes through the two base materials superimposed together in various thickness combinations; Determining the material category of the object to be tested based on the effective atomic number of the unknown material; The method of calculating the effective atomic number of the unknown material includes: Determine a first mapping relationship for mapping the atomic numbers of the two base materials to the effective atomic numbers of the shielding material in the shielding region, and map the high and low energy curves of the two base materials according to the first mapping relationship to obtain the high and low energy curves of the shielding material; Mapping the high and low energy curves of the two base materials overlapped and the high and low energy curves of the first base material into a mapping relationship of the high and low energy curves of the shielding material and the first base material overlapped, and determining it to be the same as the first mapping relationship; according to the first mapping relationship, mapping the high and low energy curves of the two base materials overlapped and the high and low energy curves of the first base material to obtain the high and low energy curves of the shielding material and the first base material overlapped; The aliased material of the shielding material and the first base material is regarded as one base material, and the shielding material is regarded as another base material. Based on the high energy value and the low energy value of the pixel points in the region of interest, a second mapping relationship is obtained by base material decomposition, which maps the aliased high and low energy curves of the shielding material and the first base material and the high and low energy curves of the shielding material to the high and low energy curves of the equivalent aliasing material, and the second mapping relationship is equivalent to a mapping relationship in which the atomic number of the first base material and the effective atomic number of the shielding material are mapped to the effective atomic number of the unknown material; wherein the effective atomic number of the shielding material is determined based on the atomic number of the pixel points in the shielding region; Based on the atomic number of the first base material and the effective atomic number of the shielding material, the effective atomic number of the unknown material is determined according to the second mapping relationship.

2. The method according to claim 1, characterized in that The method further comprises: In the region of interest, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more sub-regions of interest; In the occluder region, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more occluder sub-regions; The calculating the effective atomic number of the unknown material comprises: Match the sub-region of interest and the obstruction sub-region one by one, and calculate the effective atomic number of the unknown material in the sub-region of interest under the corresponding matching combination; The step of determining the material category of the object to be tested based on the effective atomic number of the unknown material comprises: For any sub-region of interest, exclude the occluder sub-region that is unreasonable in matching with the any sub-region of interest from all occluder sub-regions, and determine the region category of the any sub-region of interest based on the remaining occluder sub-regions; wherein, based on the comparison of the effective atomic number of the unknown material in the sub-region of interest and the effective atomic number of the occluder material in the occluder sub-region, and / or based on the comparison of the low energy value of the pixel point in the sub-region of interest and the low energy value of the pixel point in the occluder sub-region, it is determined whether the matching between the sub-region of interest and the occluder sub-region is unreasonable; If any of the sub-regions of interest is an unblocked region, the effective atomic number of each pixel in the region is used as the effective atomic number of the object to be measured in any of the sub-regions of interest; If any of the sub-regions of interest is blocked and there is no remaining blocking object sub-region, then any of the sub-regions of interest is regarded as an invalid sub-region; If any of the sub-regions of interest is blocked and there is only one remaining blocking object sub-region, the any sub-region of interest is used as a valid sub-region, and the effective atomic number of the unknown material determined when the blocking object sub-region is matched and combined with the any sub-region of interest is used as the effective atomic number of the object to be tested in the any sub-region of interest; If any of the sub-regions of interest is occluded and there are multiple remaining occluding sub-regions, then any of the sub-regions of interest is taken as a valid sub-region, and for each matching combination of the remaining occluding sub-region and any of the sub-regions of interest, the high and low energy curves of the equivalent aliasing materials in the any of the sub-regions of interest under the corresponding matching combination are determined according to the second mapping relationship under the matching combination, and the degree of agreement between the high and low energy curves and the actual high energy values ​​and low energy values ​​of each pixel point in the any of the sub-regions of interest is calculated; the occluding sub-region with the highest degree of agreement is selected, and the effective atomic number of the unknown material determined when the occluding sub-region is matched with any of the sub-regions of interest is used as the effective atomic number of the object to be measured in any of the sub-regions of interest; Traverse all sub-regions of interest. If there are unobstructed sub-regions of interest, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in each unobstructed sub-region of interest; if there are no unobstructed sub-regions of interest and there are valid sub-regions, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in all valid sub-regions.

3. The method according to claim 1, characterized in that The method of determining the high and low energy curves of the overlap of the two base materials comprises: A wedge block made of a first base material and a wedge block made of a second base material are stacked and placed; wherein the heights of the wedge blocks of the first base material and the second base material are M and N respectively, and the two wedge blocks are placed with the inclined surfaces facing upwards, the inclined surface direction of the second base material is parallel to the collimation slit of the X-ray source, and the inclined surface direction of the first base material is perpendicular to the inclined surface direction of the second base material; X-rays are irradiated on two overlapping wedge-shaped blocks, and the detector collects an aliased image; wherein the data in the aliased image includes high energy values ​​and low energy values ​​under all combinations of a first base material thickness of x and a second base material thickness of y, x is a uniformly sampled value between 0 and M, and y is a uniformly sampled value between 0 and N; X-rays are respectively irradiated on the wedge-shaped blocks of the first base material and the wedge-shaped blocks of the second base material which are placed separately, and the detector collects the image of the first base material and the image of the second base material respectively; wherein the data in the image of the first base material includes the high energy value and the low energy value of the first base material with a thickness of x, and the data in the image of the second base material includes the high energy value and the low energy value of the second base material with a thickness of y; Positionally matching the first base material image and the second base material image with the aliasing image, and determining that any pixel position in the aliasing image matches the first pixel position in the first base material image and the second pixel position in the second base material image; wherein the thickness of the first base material at any pixel position in the aliasing image is the same as the thickness of the first base material at the first pixel position in the first base material image, and the thickness of the second base material at any pixel position in the aliasing image is the same as the thickness of the second base material at the second pixel position in the second base material image; Establishing a first corresponding relationship, used to indicate that the low energy value of the first pixel position and the low energy value of the second pixel position correspond to the low energy value of any pixel position; Establishing a second corresponding relationship, used to indicate that the low energy value of the first pixel position and the low energy value of the second pixel position correspond to the high energy value of any pixel position; Determine, by two-dimensional function fitting, a first functional relationship for mapping the low energy value of the first base material image and the low energy value of the second base material image to the low energy value of the aliased image using all the first corresponding relationships; Determine, by two-dimensional function fitting, a second functional relationship for mapping the low energy value of the first base material image and the low energy value of the second base material image to the high energy value of the aliased image using all the second corresponding relationships; Based on the first functional relationship and the second functional relationship, high and low energy curves of the two base materials overlap are generated.

4. The method according to claim 1, characterized in that: The region of interest is determined by compressing the boundary of the region of the object to be measured inwardly, and the region defined by the compressed boundary is the region of interest.

5. A material identification device, characterized in that: include: An image processing unit, a region division unit, an atomic number calculation unit, and a material identification unit; The image processing unit is used to obtain an X-ray image of the package and determine the high energy value, low energy value and atomic number of each pixel on the X-ray image; The region division unit is used to determine the region of interest and the obstruction region on the X-ray image based on the high energy value, low energy value and atomic number of each pixel point on the X-ray image; wherein the region of interest is the internal region of the object to be tested, and the obstruction region is the region remaining after the total region defined by the boundary after the boundary of the object to be tested region is expanded outwards minus the region defined by the boundary before the expansion; The atomic number calculation unit is used to decompose the equivalent aliasing material of the region of interest into a combination of the occluder material and the unknown material of the occluder region based on the preset high and low energy curves of the aliasing of two base materials, the atomic number of the pixel points in the occluder region, the high energy value, the low energy value and the atomic number of the pixel points in the region of interest by using the base material decomposition method, and calculate the effective atomic number of the unknown material; wherein the equivalent aliasing material is the equivalent material corresponding to the effective atomic number of the pixel points in the region of interest, the first base material of the two base materials is an organic matter with an atomic number less than a first threshold value, the second base material of the two base materials is an inorganic matter with an atomic number greater than a second threshold value, and the high and low energy curves of the aliasing of the two base materials are used to represent the relationship curve between the high energy value and the low energy value obtained after the X-ray passes through the two base materials superimposed together in various thickness combinations; The material identification unit is used to determine the material category of the object to be tested based on the effective atomic number of the unknown material; Wherein, in the atomic number calculation unit, the method of calculating the effective atomic number of the unknown material includes: Determine a first mapping relationship for mapping the atomic numbers of the two base materials to the effective atomic numbers of the shielding material in the shielding region, and map the high and low energy curves of the two base materials according to the first mapping relationship to obtain the high and low energy curves of the shielding material; Mapping the high and low energy curves of the two base materials overlapped and the high and low energy curves of the first base material into a mapping relationship of the high and low energy curves of the shielding material and the first base material overlapped, and determining it to be the same as the first mapping relationship; according to the first mapping relationship, mapping the high and low energy curves of the two base materials overlapped and the high and low energy curves of the first base material to obtain the high and low energy curves of the shielding material and the first base material overlapped; The aliased material of the shielding material and the first base material is regarded as one base material, and the shielding material is regarded as another base material. Based on the high energy value and the low energy value of the pixel points in the region of interest, a second mapping relationship is obtained by base material decomposition, which maps the aliased high and low energy curves of the shielding material and the first base material and the high and low energy curves of the shielding material to the high and low energy curves of the equivalent aliasing material, and the second mapping relationship is equivalent to a mapping relationship in which the atomic number of the first base material and the effective atomic number of the shielding material are mapped to the effective atomic number of the unknown material; wherein the effective atomic number of the shielding material is determined based on the atomic number of the pixel points in the shielding region; Based on the atomic number of the first base material and the effective atomic number of the shielding material, the effective atomic number of the unknown material is determined according to the second mapping relationship.

6. The device according to claim 5, characterized in that The area division unit is further used for: In the region of interest, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more sub-regions of interest; In the occluder region, sub-region division is performed based on the high energy value, low energy value and / or effective atomic number of each pixel point to obtain one or more occluder sub-regions; In the atomic number calculation unit, calculating the effective atomic number of the unknown material includes: Match the sub-region of interest and the obstruction sub-region one by one, and calculate the effective atomic number of the unknown material in the sub-region of interest under the corresponding matching combination; The step of determining the material category of the object to be tested based on the effective atomic number of the unknown material comprises: For any sub-region of interest, exclude the occluder sub-region that is unreasonable in matching with the any sub-region of interest from all occluder sub-regions, and determine the region category of the any sub-region of interest based on the remaining occluder sub-regions; wherein, based on the comparison of the effective atomic number of the unknown material in the sub-region of interest and the effective atomic number of the occluder material in the occluder sub-region, and / or based on the comparison of the low energy value of the pixel point in the sub-region of interest and the low energy value of the pixel point in the occluder sub-region, it is determined whether the matching between the sub-region of interest and the occluder sub-region is unreasonable; If any of the sub-regions of interest is an unblocked region, the effective atomic number of each pixel in the region is used as the effective atomic number of the object to be measured in any of the sub-regions of interest; If any of the sub-regions of interest is blocked and there is no remaining blocking object sub-region, then any of the sub-regions of interest is regarded as an invalid sub-region; If any of the sub-regions of interest is blocked and there is only one remaining blocking object sub-region, the any sub-region of interest is used as a valid sub-region, and the effective atomic number of the unknown material determined when the blocking object sub-region is matched and combined with the any sub-region of interest is used as the effective atomic number of the object to be tested in the any sub-region of interest; If any of the sub-regions of interest is occluded and there are multiple remaining occluding sub-regions, then any of the sub-regions of interest is taken as a valid sub-region, and for each matching combination of the remaining occluding sub-region and any of the sub-regions of interest, the high and low energy curves of the equivalent aliasing materials in the any of the sub-regions of interest under the corresponding matching combination are determined according to the second mapping relationship under the matching combination, and the degree of agreement between the high and low energy curves and the actual high energy values ​​and low energy values ​​of each pixel point in the any of the sub-regions of interest is calculated; the occluding sub-region with the highest degree of agreement is selected, and the effective atomic number of the unknown material determined when the occluding sub-region is matched with any of the sub-regions of interest is used as the effective atomic number of the object to be measured in any of the sub-regions of interest; Traverse all sub-regions of interest. If there are unobstructed sub-regions of interest, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in each unobstructed sub-region of interest; if there are no unobstructed sub-regions of interest and there are valid sub-regions, determine the material category of the object to be tested based on the effective atomic number of the object to be tested in all valid sub-regions.

7. The device according to claim 5, characterized in that In the material identification unit, a method of determining the high and low energy curves of the overlap of the two base materials includes: A wedge block made of a first base material and a wedge block made of a second base material are stacked and placed; wherein the heights of the wedge blocks of the first base material and the second base material are M and N respectively, and the two wedge blocks are placed with the inclined surfaces facing upwards, the inclined surface direction of the second base material is parallel to the collimation slit of the X-ray source, and the inclined surface direction of the first base material is perpendicular to the inclined surface direction of the second base material; X-rays are irradiated on two overlapping wedge-shaped blocks, and the detector collects an aliased image; wherein the data in the aliased image includes high energy values ​​and low energy values ​​under all combinations of a first base material thickness of x and a second base material thickness of y, x is a uniformly sampled value between 0 and M, and y is a uniformly sampled value between 0 and N; X-rays are respectively irradiated on the wedge-shaped blocks of the first base material and the wedge-shaped blocks of the second base material which are placed separately, and the detector collects the image of the first base material and the image of the second base material respectively; wherein the data in the image of the first base material includes the high energy value and the low energy value of the first base material with a thickness of x, and the data in the image of the second base material includes the high energy value and the low energy value of the second base material with a thickness of y; Positionally matching the first base material image and the second base material image with the aliasing image, and determining that any pixel position in the aliasing image matches the first pixel position in the first base material image and the second pixel position in the second base material image; wherein the thickness of the first base material at any pixel position in the aliasing image is the same as the thickness of the first base material at the first pixel position in the first base material image, and the thickness of the second base material at any pixel position in the aliasing image is the same as the thickness of the second base material at the second pixel position in the second base material image; Establishing a first corresponding relationship, used to indicate that the low energy value of the first pixel position and the low energy value of the second pixel position correspond to the low energy value of any pixel position; Establishing a second corresponding relationship, used to indicate that the low energy value of the first pixel position and the low energy value of the second pixel position correspond to the high energy value of any pixel position; Determine, by two-dimensional function fitting, a first functional relationship for mapping the low energy value of the first base material image and the low energy value of the second base material image to the low energy value of the aliased image using all the first corresponding relationships; Determine, by two-dimensional function fitting, a second functional relationship for mapping the low energy value of the first base material image and the low energy value of the second base material image to the high energy value of the aliased image using all the second corresponding relationships; Based on the first functional relationship and the second functional relationship, high and low energy curves of the two base materials overlap are generated.

8. The device according to claim 5, characterized in that In the area division unit, The region of interest is determined by compressing the boundary of the region of the object to be measured inwardly, and the region defined by the compressed boundary is the region of interest.

9. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instructions are executed by a processor, the material identification method described in any one of claims 1 to 4 can be implemented.

10. An electronic device, characterized in that: The electronic device includes at least a computer-readable storage medium and also includes a processor; The processor is used to read executable instructions from the computer-readable storage medium and execute the instructions to implement the material identification method described in any one of claims 1 to 4 above.

Citation Information

Patent Citations

  • Material composition detection from effective atomic number computation

    CN101598686A

  • Base material decomposition method, device and computer readable storage medium

    CN109959669A

  • Intelligent security inspection method for effectively improving prohibited goods recognition rate

    CN112162324A

  • Method and apparatus for substance identification

    US20090129544A1

  • Dual-energy ray imaging methods and systems

    US20170309043A1