Method for consistency evaluation of security x-ray data

Abstract

The application provides a security X-ray data consistency evaluation method, comprising the following steps: dividing a scanning area perpendicular to the conveying belt running direction of a first security inspection machine and a second security inspection machine into multiple areas; obtaining, for each area, a first high-low energy data set and a second high-low energy data set of calibration objects with different materials and thicknesses in the area; determining, based on the first high-low energy data set and the second high-low energy data set of calibration objects with the same material and thickness in the area, a first data consistency coefficient corresponding to the material and thickness in the area; determining, based on the first data consistency coefficients of each thickness, a second data consistency coefficient corresponding to the material in the area; determining, based on the second data consistency coefficients of each material, a third data consistency coefficient of the area; and determining, based on the third data consistency coefficients of all areas, a target data consistency coefficient. The application can effectively evaluate the differences between the security X-ray data output by different security inspection machines.

Description

Technical Field

[0001] This invention relates to the field of security inspection data processing technology, and in particular to a method for consistency assessment of security inspection X-ray data. Background Technology

[0002] In the field of public safety, such as public transportation and cultural and museum venues, scanning and inspecting passengers' or tourists' luggage and belongings with security screening machines can effectively improve the overall security of the environment. Currently, commonly used dual-energy security screening machines output data including underlying high-energy data, low-energy data, and equivalent atomic number maps generated from the high- and low-energy data, as well as pseudo-color images (RGB images) displayed on the screen interface. Among them, the pseudo-color image of the security screening machine assigns specific colors to substances with different densities and equivalent atomic numbers to help security personnel quickly identify the category of items. Therefore, the color of the pseudo-color image is particularly important for judging the attributes of items.

[0003] However, in practical applications, in addition to the differences in coloring algorithms among manufacturers causing color deviations in pseudo-color images, even for the same model of security inspection machine from the same manufacturer, differences in manufacturing processes and hardware components such as X-ray sources and detectors can lead to deviations in the high-energy and low-energy data collected at the underlying level. This results in differences between the equivalent atomic number map and the pseudo-color image generated from the high-energy and low-energy data.

[0004] Therefore, how to measure and evaluate the differences in X-ray data output by different security inspection machines, and then guide the improvement of production processes to ensure that the data output by the produced security inspection machines is as consistent as possible, is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] This invention provides a method for evaluating the consistency of security X-ray data, which addresses the shortcomings of existing technologies that cannot measure and evaluate the differences in security X-ray data output by different security screening machines, and achieves the goal of effectively evaluating the differences in security X-ray data output by different security screening machines.

[0006] This invention provides a method for consistency assessment of security inspection X-ray data, comprising: The scanning area perpendicular to the conveyor belt running direction of the first security inspection machine and the second security inspection machine is divided into multiple areas, and the area obtained by the first security inspection machine after division corresponds one-to-one with the area obtained by the second security inspection machine. For each of the aforementioned regions, a first high- and low-energy data set and a second high- and low-energy data set are obtained for calibration objects of different materials and thicknesses within the region. The first high- and low-energy data set includes the high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes the high- and low-energy data collected by the second security inspection machine. Based on the first high- and low-energy data set and the second high- and low-energy data set of the same material and thickness within the region, a first data consistency coefficient corresponding to the thickness of the material within the region is determined. Based on the first data consistency coefficient corresponding to each thickness of the material within the region, a second data consistency coefficient for the material within the region is determined. Based on the second data consistency coefficient for each material within the region, a third data consistency coefficient corresponding to the region is determined. Based on the third data consistency coefficient corresponding to all regions, a target data consistency coefficient is determined between the first security screening machine and the second security screening machine. The target data consistency coefficient is used to evaluate the consistency between the security screening X-ray data of the first security screening machine and the second security screening machine.

[0007] According to the present invention, a method for consistency evaluation of security X-ray data is provided, wherein the scanning area perpendicular to the conveyor belt running direction of the first and second security inspection machines is divided into multiple regions, including: A first pseudo-color image is obtained after the first security inspection machine scans a lead ruler placed on the conveyor belt of the first security inspection machine, wherein the lead ruler includes scales engraved at uniform physical intervals along the length direction. Based on the scale in the first pseudo-color image, the scanning area of ​​the first security inspection machine is divided into multiple areas; Obtain a second pseudo-color image after the second security inspection machine scans the lead ruler placed on the conveyor belt of the second security inspection machine; Based on the scale in the second pseudo-color image, the scanning area of ​​the second security inspection machine is divided into multiple regions.

[0008] According to the consistency evaluation method for security inspection X-ray data provided by the present invention, the step of determining a first data consistency coefficient corresponding to the thickness of the material within the region based on a first high-energy data set and a second high-energy data set of calibration objects of the same material and thickness within the region includes: Based on a first set of high and low energy data of calibrators of the same material and thickness within the region, a first set of scattered points is determined, and based on a second set of high and low energy data of the same calibrators within the region, a second set of scattered points is determined. Each first scattered point in the first set of scattered points is composed of a pair of high-energy data and low-energy data from the first set of high and low energy data, and each second scattered point in the second set of scattered points is composed of a pair of high-energy data and low-energy data from the second set of high and low energy data. For each first scattered point in the first scattered point set, determine the distance from the first scattered point to each second scattered point in the second scattered point set; Determine the minimum distance among multiple distances corresponding to each of the first scattered points, and determine the first average of the minimum distances of all the first scattered points; For each second scattered point in the second scattered point set, determine the distance from the second scattered point to each first scattered point in the first scattered point set; Determine the minimum distance among multiple distances corresponding to each of the second scattered points, and determine the second average of the minimum distances of all the second scattered points; The first data consistency coefficient is determined based on the first average value and the second average value.

[0009] According to the consistency assessment method for security inspection X-ray data provided by the present invention, the step of determining a first scatter set based on a first set of high and low energy data of calibration objects of the same material and thickness within the region includes: From the first high- and low-energy data set, obtain the high- and low-energy data pairs corresponding to each pixel of the calibrated object within the region to obtain the first scatter point set; or... According to a preset sampling interval, the pixels of the calibration object are sampled in a grid pattern in the width and height directions of the region, and the high and low energy data pairs corresponding to each sampling point are obtained from the first high and low energy data set to obtain the first scatter set.

[0010] According to the present invention, a method for evaluating the consistency of security X-ray data, wherein determining the first data consistency coefficient based on the first average value and the second average value includes: Based on the first average value and the second average value, determine the target distance between the first scatter set and the second scatter set; Determine the first average distance from all the first scattered points to the origin, and determine the second average distance from all the second scattered points to the origin; Based on the target distance, the first average distance, and the second average distance, the cosine value of the target angle is determined, wherein the target angle is the angle formed by connecting the center of the first scatter set and the center of the second scatter set from the origin; The cosine value is determined as the first data consistency coefficient.

[0011] According to the present invention, a method for evaluating the consistency of security inspection X-ray data, wherein determining a second data consistency coefficient for the material within the region based on a first data consistency coefficient corresponding to each thickness of the material within the region includes: Based on the first data consistency coefficient corresponding to each thickness of the material within the region and the number of thicknesses corresponding to the material, the average data consistency coefficient of the material within the region is determined for all thicknesses. The average data consistency coefficient of the material at all thicknesses within the region is determined as the second data consistency coefficient.

[0012] According to the present invention, a method for evaluating the consistency of security X-ray data, wherein determining a third data consistency coefficient corresponding to the region based on a second data consistency coefficient for each material within the region includes: Based on the second data consistency coefficient for each material in the region and the weighting coefficient corresponding to each material, the weighted average data consistency coefficient for all materials in the region is determined. The weighted average data consistency coefficient corresponding to all materials in the region is determined as the third data consistency coefficient.

[0013] According to the present invention, a method for evaluating the consistency of security X-ray data includes determining a target data consistency coefficient between the first security screening machine and the second security screening machine based on a third data consistency coefficient corresponding to all regions. Determine the average value of the third data consistency coefficients for all regions, and use the average value as the target data consistency coefficient.

[0014] According to the present invention, a method for evaluating the consistency of security X-ray data is provided, the method further includes: If the difference between the target data consistency coefficient and the preset coefficient is greater than or equal to the allowable coefficient deviation value, the detector parameters or X-ray source parameters of the non-benchmark security inspection machine are adjusted and the target data consistency coefficient is recalculated until the difference between the calculated target data consistency coefficient and the preset coefficient is less than the allowable coefficient deviation value. At this point, the non-benchmark security inspection machine is deemed qualified. The non-benchmark security inspection machine is the security inspection machine with poor data quality among the first security inspection machine and the second security inspection machine.

[0015] According to the consistency assessment method for security inspection X-ray data provided by the present invention, the material of the calibration object includes at least two of iron, copper, aluminum, polyvinyl chloride (PVC) and plexiglass, and the length of the calibration object is the same as the length of the lead ruler.

[0016] The present invention also provides a consistency evaluation device for security inspection X-ray data, comprising: The segmentation module is used to divide the scanning area perpendicular to the conveyor belt running direction of the first security inspection machine and the second security inspection machine into multiple areas respectively, and the area obtained by the first security inspection machine after segmentation corresponds one-to-one with the area obtained by the second security inspection machine after segmentation. The acquisition module is used to acquire, for each of the areas, a first high-low energy data set and a second high-low energy data set of calibration objects of different materials and thicknesses in the area. The first high-low energy data set includes the high-low energy data collected by the first security inspection machine, and the second high-low energy data set includes the high-low energy data collected by the second security inspection machine. The determination module is used to determine a first data consistency coefficient for the thickness of the material in the region based on a first high- and low-energy data set and a second high- and low-energy data set of calibrators of the same material and thickness in the region. The determining module is further configured to determine a second data consistency coefficient for the material within the region based on a first data consistency coefficient corresponding to each thickness of the material within the region; The determining module is further configured to determine a third data consistency coefficient corresponding to the region based on a second data consistency coefficient for each of the materials within the region. The determining module is further configured to determine a target data consistency coefficient between the first security inspection machine and the second security inspection machine based on a third data consistency coefficient corresponding to all regions. The target data consistency coefficient is used to evaluate the consistency between the security inspection X-ray data of the first security inspection machine and the second security inspection machine.

[0017] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the consistency assessment method for security X-ray data as described above.

[0018] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the consistency evaluation method for security X-ray data as described above.

[0019] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the consistency evaluation method for security inspection X-ray data as described above.

[0020] The consistency assessment method for security X-ray data provided by this invention divides the scanning area perpendicular to the conveyor belt running direction of the first and second security inspection machines into multiple regions. The regions obtained by the first security inspection machine after division correspond one-to-one with the regions obtained by the second security inspection machine. For each region, a first high- and low-energy data set and a second high- and low-energy data set of calibration objects of different materials and thicknesses are acquired within the region. The first high- and low-energy data set includes high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes high- and low-energy data collected by the second security inspection machine. Based on the first high- and low-energy data sets of calibration objects of the same material and thickness within the region, a first data consistency coefficient corresponding to the thickness of the material within the region is determined. Based on the first data consistency coefficient corresponding to each thickness of the material within the region, a second data consistency coefficient corresponding to the material within the region is determined. Based on the second data consistency coefficient corresponding to each material within the region, a third data consistency coefficient corresponding to the region is determined. Based on the third data consistency coefficient corresponding to all regions, a target data consistency coefficient between the first and second security inspection machines is determined. The target data consistency coefficient is used to assess the consistency between the security X-ray data of the first and second security inspection machines. Because the scanning areas of the first and second security inspection machines can be precisely divided into corresponding regions, and a multi-layered progressive data consistency determination method from thickness, material, region to the security inspection machine itself is adopted, an objective, stable, and quantifiable target data consistency coefficient can be obtained. This target data consistency coefficient can reflect the overall data output similarity of the first and second security inspection machines under conditions of complete scanning field of view, multiple material properties, and different thicknesses. It can effectively measure and evaluate whether there are differences between the X-ray data output by different security inspection machines, thus providing a basis for improving the security inspection machine manufacturing process to ensure that the produced security inspection machines can output X-ray data as consistently as possible. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating the consistency evaluation method for security inspection X-ray data provided in an embodiment of the present invention.

[0023] Figure 2 Visible light image of a plumb line provided for an embodiment of the present invention.

[0024] Figure 3 The pseudo-color image of the ruler provided in the embodiment of the present invention.

[0025] Figure 4 This is a schematic diagram of region division provided for an embodiment of the present invention.

[0026] Figure 5 False-color images of iron calibration plates of different thicknesses.

[0027] Figure 6 False-color images of PVC calibration boards of different thicknesses.

[0028] Figure 7 The high and low energy scatter plots of the calibration material of region n and material m at different thicknesses provided in the embodiments of the present invention.

[0029] Figure 8 This is a schematic diagram illustrating the target distance between the first scatter point set and the second scatter point set provided in an embodiment of the present invention.

[0030] Figure 9 This is a schematic diagram of the structure of the consistency evaluation device for security inspection X-ray data provided in an embodiment of the present invention.

[0031] Figure 10 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0033] Currently, security personnel or check-in staff typically identify item categories using pseudo-color images output by the security scanner. These pseudo-color images are obtained by assigning specific colors to substances with different densities and equivalent atomic numbers; therefore, different colors in the pseudo-color image represent different item categories. For example, the correspondence between item categories, specific items, and corresponding pseudo-color colors is shown in Table 1: Table 1

[0034] Therefore, the color of pseudo-color images is particularly important for determining the category of items. However, in practical applications, the color consistency of pseudo-color images is often affected by many factors. Besides differences in coloring algorithms between different manufacturers leading to color deviations, even for the same model of security screening machine from the same manufacturer, differences in manufacturing processes and hardware components can cause discrepancies between the high-energy and low-energy data collected at the underlying level. This results in differences between the equivalent atomic number map and the pseudo-color image generated from the high-energy and low-energy data. Therefore, how to effectively measure and evaluate whether there are differences in the X-ray data output by different security screening machines, and thus guide the improvement of security screening machine manufacturing processes to ensure that the produced security screening machines output as consistent X-ray data as possible, is a pressing technical problem that needs to be solved.

[0035] In view of the above-mentioned problems, this invention proposes a consistency evaluation method for security inspection X-ray data, which can effectively measure and evaluate whether there are differences between the X-ray data output by different security inspection machines, thereby providing a basis for improving the security inspection machine manufacturing process to ensure that the produced security inspection machines can output X-ray data as consistently as possible.

[0036] The following is combined Figures 1 to 8 The consistency assessment method for security X-ray data provided in this invention is described. This invention can be applied to scenarios involving the production, quality inspection, or maintenance of security inspection machines. The security inspection machine described in this invention refers to a security X-ray machine, specifically a transmission-type baggage security inspection device within a micro-dose X-ray security inspection system.

[0037] The subject executing this method can be a first security inspection machine, a second security inspection machine, or other terminal equipment, a computer, a server, a server cluster, or a specially designed security X-ray data consistency assessment device, or a security X-ray data consistency assessment device installed in the electronic device. The security X-ray data consistency assessment device can be implemented through software, hardware, or a combination of both.

[0038] Figure 1 This is a flowchart illustrating the consistency assessment method for security X-ray data provided in an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes: Step 101: Divide the scanning area perpendicular to the conveyor belt running direction of the first security inspection machine and the second security inspection machine into multiple areas respectively. The area obtained by the first security inspection machine after division corresponds one-to-one with the area obtained by the second security inspection machine after division.

[0039] In this step, the scanning area perpendicular to the conveyor belt running direction of the first and second security scanners can be understood as the entire cross-section through which the X-rays of the first and second security scanners can penetrate and image at a certain scanning instant. This cross-section is perpendicular to the direction of baggage travel and covers the entire effective width of the conveyor belt, representing the complete data line that the detector array can acquire in one acquisition.

[0040] The scanning area of ​​the first security inspection machine is divided into multiple areas, and the scanning area of ​​the second security inspection machine is also divided into multiple areas. The number of multiple areas divided by the first security inspection machine and the second security inspection machine is equal, and they correspond one-to-one with each other. That is, the physical position and physical width of the conveyor belt represented by any area in the first security inspection machine are exactly the same as the physical position and physical width of the conveyor belt represented by the corresponding area in the second security inspection machine.

[0041] Step 102: For each area, obtain the first high-low energy data set and the second high-low energy data set for calibration objects of different materials and thicknesses within the area. The first high-low energy data set includes the high-low energy data collected by the first security inspection machine, and the second high-low energy data set includes the high-low energy data collected by the second security inspection machine.

[0042] In this step, flat calibration objects of different materials and thicknesses can be placed on the conveyor belt of the first security inspection machine. These calibration objects can be, for example, calibration plates. When placing the calibration plates, the center point of the calibration plate along its length coincides with the center point of the conveyor belt along its width to ensure that it can be correctly mapped to each area during subsequent scanning and imaging.

[0043] By controlling the conveyor belt carrying the calibration object to pass through the scanning channel of the first security inspection machine at a constant speed and triggering the first security inspection machine to scan, the original response data of the detector during the process of the calibration object completely passing through the scanning area can be collected. By processing the original response data of the detector, the first high and low energy data set of each area can be obtained.

[0044] The first high- and low-energy data set includes multiple pairs of high-energy and low-energy data. Each high- and low-energy data pair corresponds to a measurement output of an independent pixel unit in the detector of the first security inspection machine under the conditions of this region, material, and thickness, characterizing the attenuation of X-rays after penetrating the calibration material.

[0045] In practical applications, calibration objects of the same material but different thicknesses can be placed on a conveyor belt all at once to collect high and low energy data of the calibration objects of that material at different thicknesses in each area. Similarly, by placing calibration objects of different materials on the conveyor belt each time in this manner, high and low energy data of different materials at different thicknesses in each area can be obtained. Of course, calibration objects of one material and one thickness can also be placed on the conveyor belt each time, and high and low energy data of different materials at different thicknesses in each area can be obtained through multiple collections. Alternatively, a mixed placement method can be used, in which several calibration plates of different materials and thicknesses are combined in a specific order and passed through the scanning channel at once. By performing material region identification and data separation on the collected composite image, high and low energy data of multiple materials at different thicknesses can be obtained simultaneously. The embodiments of the present invention do not limit the acquisition method of the first set of high and low energy data, as long as it is possible to collect high and low energy data of calibration objects of different materials and thicknesses in different areas.

[0046] Similarly, the same calibration object is placed on the conveyor belt of the second security inspection machine to collect the second high- and low-energy data set for each area pair.

[0047] Step 103: Based on the first high- and low-energy data set and the second high- and low-energy data set of the calibration objects of the same material and thickness within the region, determine the first data consistency coefficient corresponding to the thickness of the material within the region.

[0048] In this step, for the first set of high and low energy data collected by the first security inspection machine A, for each region, the pairs of high and low energy data from the first set of high and low energy data of the same material but different thicknesses within the corresponding region can be plotted as a scatter plot on a high-energy-low-energy distribution map. Each region, each material, and each thickness corresponds to a set of scatter points. Assuming there are N regions, M materials, and T thicknesses, there are N×M×T sets of scatter points, represented as follows: ,in, And n, m, and t are all positive integers. If it is a certain region, then there are M×T sets of scattered points. For example, region 1 is represented as... For example, N has a maximum value of 400, M has a maximum value of 30, and T has a maximum value of 16.

[0049] Similarly, for the second set of high and low energy data collected by the second security inspection machine B, a scatter plot of calibration objects of different thicknesses of the same material (a certain material) in each divided area can be obtained, represented as follows: .

[0050] Furthermore, it can be detected through the first security screening machine A. The second security inspection machine B Scattered datasets are used to evaluate the consistency of output data (collected data) from two security scanners. It should be understood that data consistency cannot be assessed simply by calculating the pairwise distances between all scattered points in the dataset; only scattered datasets with the same location, material, and material thickness are comparable. Therefore, a scattered dataset based on the same material and thickness in the same location area of ​​both security scanners A and B can be used, such as... and By calculating the distance between these two scattered point sets, and based on this distance, a first data consistency coefficient is determined for the calibration object of this material and thickness within this region. The larger the distance, the worse the consistency of the data output by the two security inspection machines for the calibration object of this material and thickness within this region. The first data consistency coefficient can be used to characterize the degree of consistency of the data distribution of the two security inspection machines under this region, material, and thickness.

[0051] Step 104: Based on the first data consistency coefficient corresponding to each thickness of the material within the region, determine the second data consistency coefficient for the material within the region.

[0052] In this step, after determining the first data consistency coefficient for a certain material at various thicknesses within a certain region, these first data consistency coefficients can be aggregated, such as by averaging or weighted averaging, to obtain the second data consistency coefficient for that material within the same region. The second data consistency coefficient characterizes the consistency of the output data from the first and second security inspection machines across the entire thickness variation range within that region for that material; in other words, it represents the average consistency performance of the two machines.

[0053] Step 105: Based on the second data consistency coefficient for each material within the region, determine the third data consistency coefficient corresponding to the region.

[0054] In this step, after determining the second data consistency coefficients for various materials within a certain region, these second data consistency coefficients can be aggregated, such as by averaging or weighted averaging, to obtain the third data consistency coefficient for that region. The third data consistency coefficient characterizes the overall consistency of the output data from the first and second security inspection machines across various calibration objects within the physical space corresponding to that region.

[0055] Step 106: Based on the third data consistency coefficient corresponding to all regions, determine the target data consistency coefficient between the first security screening machine and the second security screening machine. The target data consistency coefficient is used to evaluate the consistency between the security screening X-ray data of the first security screening machine and the second security screening machine.

[0056] In this step, after determining the third data consistency coefficients for all regions, these third data consistency coefficients can be aggregated, such as by averaging or weighted averaging. This fuses the consistency coefficients of all spatial locations across the entire scanning field of view of the security inspection machine, yielding the target data consistency coefficient between the first and second security inspection machines. The target data consistency coefficient integrates comparison information from multiple dimensions, including spatial distribution, various materials, and different thicknesses. It provides a reliable and intuitive quantitative basis for evaluating the global similarity of the output data from the first and second security inspection machines. Through the target data consistency coefficient, the consistency between the security X-ray data of the first and second security inspection machines can be directly assessed, thereby determining whether the consistency between the two machines meets the preset standard.

[0057] The consistency evaluation method for security X-ray data provided in this invention divides the scanning area perpendicular to the conveyor belt running direction of the first and second security inspection machines into multiple regions. The regions obtained by the first security inspection machine after division correspond one-to-one with the regions obtained by the second security inspection machine. For each region, a first high- and low-energy data set and a second high- and low-energy data set of calibration objects of different materials and thicknesses are acquired within the region. The first high- and low-energy data set includes high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes high- and low-energy data collected by the second security inspection machine. Based on the first high- and low-energy data sets of calibration objects of the same material and thickness within the region, a first data consistency coefficient corresponding to the thickness of the material within the region is determined. Based on the first data consistency coefficient corresponding to each thickness of the material within the region, a second data consistency coefficient corresponding to the material within the region is determined. Based on the second data consistency coefficient corresponding to each material within the region, a third data consistency coefficient corresponding to the region is determined. Based on the third data consistency coefficient corresponding to all regions, a target data consistency coefficient between the first and second security inspection machines is determined. The target data consistency coefficient is used to evaluate the consistency between the security X-ray data of the first and second security inspection machines. Because the scanning areas of the first and second security inspection machines can be precisely divided into corresponding regions, and a multi-layered progressive data consistency determination method from thickness, material, region to the security inspection machine itself is adopted, an objective, stable, and quantifiable target data consistency coefficient can be obtained. This target data consistency coefficient can reflect the overall data output similarity of the first and second security inspection machines under conditions of complete scanning field of view, multiple material properties, and different thicknesses. It can effectively measure and evaluate whether there are differences between the X-ray data output by different security inspection machines, thus providing a basis for improving the security inspection machine manufacturing process to ensure that the produced security inspection machines can output X-ray data as consistently as possible.

[0058] For example, based on the above embodiments, when dividing the scanning area perpendicular to the conveyor belt running direction of the first and second security inspection machines into multiple areas, it can be done in the following way: A first pseudo-color image is obtained after the first security inspection machine scans a lead ruler placed on the conveyor belt of the first security inspection machine. The lead ruler includes scales marked at uniform physical intervals along its length. Based on the scales in the first pseudo-color image, the scanning area of ​​the first security inspection machine is divided into multiple areas.

[0059] A second pseudo-color image is obtained after the second security inspection machine scans the lead ruler placed on the conveyor belt of the second security inspection machine. Based on the scale in the second pseudo-color image, the scanning area of ​​the second security inspection machine is divided into multiple areas.

[0060] Specifically, both the first and second security screening machines use a structure consisting of an X-ray source and an L-shaped detector. The L-shaped detector comprises a vertical array and a horizontal array, which results in the X-ray image data collected by the security screening machine being non-uniform. Therefore, in this embodiment, a lead ruler can be designed. The base of the lead ruler is a triangular prism made of plexiglass, with small lead sheets inlaid on one side of the prism according to uniform graduations (e.g., 1 cm).

[0061] Figure 2 Visible light images of a plumb line provided in an embodiment of the present invention. Figure 3 The pseudo-color image of the ruler provided in the embodiments of the present invention, such as Figure 2 and Figure 3 As shown, the lead ruler includes scales marked at uniform physical intervals along the length direction (the intervals are equidistant). However, due to the structure of the security inspection machine, the scales of the captured pseudo-color image are non-uniform (the intervals gradually narrow from left to right). Therefore, a lead ruler is needed to divide the area, and each divided area represents the same actual area width.

[0062] Figure 4 This is a schematic diagram of region division provided in an embodiment of the present invention, such as... Figure 4 As shown, the aforementioned lead ruler is placed on the conveyor belts of the first and second security inspection machines, respectively. When placing the lead ruler, its length center coincides with the width center of the conveyor belt. Taking the first security inspection machine as an example, during the conveyor belt's operation, the lead ruler can be scanned by the first security inspection machine to obtain a first pseudo-color image. Furthermore, the scanning area of ​​the first security inspection machine can be divided into multiple regions according to the scale of the lead ruler in the first pseudo-color image, ensuring that each region has the same actual width. For example... Figure 2 As shown, the number of intervals of the plumb line is 40, and the number of regions after division is also 40, that is, N=40.

[0063] Similarly, the scanning area of ​​the second security inspection machine can be divided into multiple regions based on the scale in the second pseudo-color image obtained from the scan.

[0064] In this embodiment, the scanning areas of the first security inspection machine and the second security inspection machine are divided based on the scale of the lead ruler. This ensures that the actual physical width represented by each corresponding area divided by the two security inspection machines is exactly the same and the spatial position is strictly aligned. This provides a basis for subsequent data comparison and fundamentally overcomes the interference of image distortion caused by differences in hardware structure on consistency assessment. This allows the final assessment result to truly reflect the differences in the data itself.

[0065] For example, when determining the first data consistency coefficient corresponding to the thickness of a material within a region, based on a first high-energy data set and a second high-energy data set of calibrators of the same material and thickness within the region, it can be done in the following way: Based on a first set of high and low energy data for calibration objects of the same material and thickness within a region, a first set of scattered points is determined. Based on a second set of high and low energy data for the same calibration objects within the region, a second set of scattered points is determined. Each first scattered point in the first set consists of a pair of high-energy and low-energy data from the first set of high and low energy data. Each second scattered point in the second set consists of a pair of high-energy and low-energy data from the second set of high and low energy data. For each first scattered point in the first set, the distance from the first scattered point to each second scattered point in the second set is determined. The minimum distance among multiple distances corresponding to each first scattered point is determined, and a first average of the minimum distances of all first scattered points is determined. For each second scattered point in the second set, the distance from the second scattered point to each first scattered point in the first set is determined. The minimum distance among multiple distances corresponding to each second scattered point is determined, and a second average of the minimum distances of all second scattered points is determined. Based on the first and second averages, a first data consistency coefficient is determined.

[0066] Specifically, taking the first security inspection machine as an example, flat calibration objects of different materials and thicknesses can be placed on the conveyor belt of the first security inspection machine. The calibration object can be, for example, a calibration plate. When placing the calibration plate, the center of the calibration plate in the length direction coincides with the center of the conveyor belt in the width direction. The length of the calibration plate is the same as the length of the lead ruler. Calibration plates of different thicknesses are laid flat along the running direction of the conveyor belt. After passing through the first security inspection machine, high and low energy data and pseudo-color image data can be collected.

[0067] For example, the material of the calibrator includes at least two of the following: iron, copper, aluminum, polyvinyl chloride (PVC), and plexiglass. Figure 5 False-color images of calibration plates of different thicknesses. Figure 6False-color images of PVC calibration boards of different thicknesses, such as Figure 5 and Figure 6 As shown, the darker the color in a pseudo-color image, the thicker it is.

[0068] For each material in each region, the high and low energy data within the region can be scatter-mapped onto a high-energy-low-energy distribution map. Figure 7 This is a scatter plot of high and low energy distribution of the calibration material of region n and material m at different thicknesses, provided in an embodiment of the present invention. Figure 7 As shown, the circular scatter plot represents the high and low energy distribution of different thicknesses of the first security scanner A, and the cross scatter plot represents the high and low energy distribution of different thicknesses of the second security scanner B. The shades of color correspond to... Figure 4 The material thickness corresponds to the thickness. For example, assuming the thickness is 8, we can obtain 8 sets of scattered points for security inspection machines A and B.

[0069] It should be noted that, Figure 7 The example given uses only 6 scattered points in each set, but it can include other numbers of scattered points. Additionally, Figure 7 This only shows a scatter plot of the same material at different thicknesses within a single region. When there are N regions and M materials, N×M such scatter plots will be determined. Figure 7 The set of scattered points shown.

[0070] for Figure 7 One of the scattered point sets in the data, such as Figure 7 The red circle indicates the first scatter set of points for the first security inspection machine A in the same area, with the same material and the same thickness. The second scatter set of the second security inspection machine B In this set, each first scatter point in the first scatter point set consists of a pair of high-energy data and low-energy data from the first high-low-energy data set, and each second scatter point in the second scatter point set consists of a pair of high-energy data and low-energy data from the second high-low-energy data set.

[0071] For example, when determining the first scatter set based on the first high and low energy data set of calibration objects of the same material and thickness within a region, it can be done by obtaining the high and low energy data pairs corresponding to each pixel of the calibration object within the region from the first high and low energy data set; or, according to a preset sampling interval, the pixels of the calibration object are sampled in a grid in the width and height directions of the region, and the high and low energy data pairs corresponding to each sampling point are obtained from the first high and low energy data set to obtain the first scatter set.

[0072] Specifically, the first set of scattered points can be determined by full-point sampling or by grid sampling.

[0073] For the full-point acquisition method, a pair of high-energy and low-energy data corresponding to each pixel in each region of the calibration object image can be directly extracted from the first high- and low-energy data set. Each such pair of values ​​constitutes a scatter point in the scatter plot, and the set of all scatter points forms the first scatter set. By traversing all high- and low-energy data pairs of all pixels in the region, this method can accurately reflect the overall response distribution of the detector in that region, thereby improving the accuracy of subsequent data evaluation.

[0074] For grid sampling, uniform grid sampling can be performed within the calibration image area of ​​the region, according to a pre-set sampling interval (e.g., every few pixels), in both the width (horizontal) and height (vertical) directions. From the first high- and low-energy data set, the high- and low-energy data pairs corresponding to the pixels at the grid intersections (sampling points) are extracted to form a first scattered point set. This method significantly reduces the number of data points in the first scattered point set, thereby greatly reducing the complexity of subsequent distance calculations and improving processing speed and computational efficiency.

[0075] The determination method for the second set of scattered points is similar to that for the first set of scattered points, and will not be repeated here.

[0076] After determining the first set of scattered points Second scatter set Next, it is necessary to calculate the distance of this set of scattered points, i.e. and The goal is to quantify the “overall difference” between two sets of scattered points (rather than the distance between individual pairs of points).

[0077] Suppose the first set of scattered points There are I first scattered points and two scattered point sets. There are J second scattered points in total, among which, R is the set of all real numbers. Let represent the set of all ordered pairs of real numbers (x, y).

[0078] First Scatter Set Second scatter set The target distance between them is ChamferDistance. The core idea is to sum the minimum distances in both directions, that is, the average of the minimum distances from all points in one set to all points in another set. Specifically, it can be determined by the following formula (1): (1) The first part of formula (1) represents the first scatter set. The first average of the minimum distances of all first scattered points in the second scattered point set is given by formula (2), where the minimum distance of each first scattered point is the minimum of all distances from that first scattered point to each second scattered point in the second scattered point set. The second average of the minimum distances of all second scattered points in the first scattered point set is given by the first average of the minimum distances of each second scattered point to each first scattered point in the first scattered point set.

[0079] In addition, a single point pair can be determined according to formula (2). and Euclidean distance between : (2) in, and Representing the i-th first scatter point High-energy data and low-energy data, and These represent the j-th second scatter point. High-energy data and low-energy data.

[0080] After determining the first average and the second average, the first scatter set can be determined based on the first average and the second average. Second scatter set The distance between them is also known as the first data consistency coefficient. The smaller the first data consistency coefficient, the closer the two scatter sets are in spatial distribution.

[0081] In this embodiment, the first data consistency coefficient is determined by a bidirectional minimum distance summation method. This method not only considers the proximity of each first scatter point in the first scatter point set to the second scatter point set (first average value), but also symmetrically evaluates the proximity of each second scatter point in the second scatter point set to the first scatter point set (second average value). By combining the two, this bidirectional evaluation method can effectively overcome the evaluation bias that may be caused by single-point matching or unidirectional measurement, thereby obtaining a more fair and accurate first data consistency coefficient.

[0082] For example, based on the above embodiments, when determining the first data consistency coefficient based on the first average value and the second average value, it can be done in the following way: Based on the first average and the second average, the target distance between the first scatter set and the second scatter set is determined. The first average distance from all first scatter points to the origin is determined, and the second average distance from all second scatter points to the origin is determined. Based on the target distance, the first average distance, and the second average distance, the cosine value of the target angle is determined. The target angle is the angle formed by connecting the origin to the center of the first scatter set and the center of the second scatter set. The cosine value is determined as the first data consistency coefficient.

[0083] Specifically, based on formula (1), the target distance between the first scatter set and the second scatter set is determined according to the first average value and the second average value. Subsequently, the target distance is related to the absolute value of the high and low energy data; that is, the larger the high and low energy data, the larger the value of this distance will be. Figure 8 This is a schematic diagram illustrating the target distance between the first scatter set and the second scatter set provided in an embodiment of the present invention, as shown below. Figure 8 As shown, the target distances in the blue and red circles are the same, but the degree of data deviation is different. Therefore, it is necessary to normalize the target distances.

[0084] Among them, it can be calculated Figure 8 The included angle AOB ( The cosine value of the angle is used to represent the consistency of the data. When the angle is 0, If it equals 1, then and The data is completely consistent. The specific calculation process is as follows: The first average distance from all the first scattered points to the origin can be calculated using formula (3). And the second average distance from all the second scattered points to the origin is calculated using formula (4). : (3) (4) Furthermore, based on the target distance First average distance Second average distance The target included angle is calculated. The cosine value, and use this cosine value as... and First data consistency coefficient of two scatter sets .

[0085] In this embodiment, by determining the target distance, the first average distance, and the second average distance, and by determining the cosine value of the target angle based on the target distance, the first average distance, and the second average distance, the target distance is normalized, so that the determined first data consistency coefficient is independent of the absolute value of the original high and low energy data, providing a unified measurement benchmark for subsequent data consistency evaluation.

[0086] For example, based on the above embodiments, when determining the second data consistency coefficient for a material within a region based on the first data consistency coefficient corresponding to each thickness of the material within the region, it can be based on the first data consistency coefficient corresponding to each thickness of the material within the region and the number of thicknesses corresponding to the material to determine the average data consistency coefficient of the material within the region under all thicknesses, and the average data consistency coefficient of the material within the region under all thicknesses is determined as the second data consistency coefficient.

[0087] Specifically, following the methods described in the above embodiments, a first data consistency coefficient can be determined for the first security inspection machine and the second security inspection machine under a certain thickness of a certain material in a certain area. Furthermore, the second data consistency coefficient of the two security inspection machines under this material in this area can be determined based on the following formula (5): (5) Where T represents the thickness quantity. This represents the second data consistency coefficient corresponding to material m in region n.

[0088] In this embodiment, by using the first data consistency coefficient corresponding to each thickness of the material within the region and the number of thicknesses corresponding to the material, the average data consistency coefficient for all thicknesses is determined, and then the second data consistency coefficient is determined. This avoids dependence on the test results of a single thickness and effectively integrates the consistency coefficients of the material under different thicknesses, making the obtained second data consistency coefficient more stable and reliable.

[0089] For example, based on the above embodiments, when determining the third data consistency coefficient of a region based on the second data consistency coefficient for each material within the region, the weighted average data consistency coefficient of all materials within the region can be determined based on the second data consistency coefficient for each material within the region and the weighted coefficient of each material, and the weighted average data consistency coefficient of all materials within the region can be determined as the third data consistency coefficient.

[0090] Specifically, if the material of the calibrated object includes five types: iron, copper, aluminum, PVC, and plexiglass, assuming m=1 represents iron, m=2 represents copper, m=3 represents aluminum, m=4 represents PVC, and m=5 represents plexiglass, different weighting coefficients can be assigned to different materials. In practical applications, the significance of metallic materials such as iron and copper to public safety and their frequency of appearance in luggage may differ from those of non-metallic materials such as PVC and plexiglass. By assigning higher weights to key materials (such as common metals or materials simulating typical hazardous materials), the consistency performance of these key material data can be more prominently reflected in the overall consistency coefficient.

[0091] Furthermore, the third data consistency coefficient of the first and second security inspection machines in region n can be determined by weighted averaging based on the following formula (6). : (6) Where M represents the number of material types of the calibrated object.

[0092] In this embodiment, different weighting coefficients are assigned to different materials, and the third data consistency coefficient of each region is determined by weighted averaging. This makes the third data consistency coefficient of each region more closely match the degree of attention paid to different materials in the security inspection scenario, and allows for targeted evaluation of the consistency of the output data of the security inspection machine.

[0093] For example, based on the above embodiments, when determining the target data consistency coefficient between the first security inspection machine and the second security inspection machine based on the third data consistency coefficient corresponding to all regions, the average value of the third data consistency coefficient corresponding to all regions can be determined, and the average value can be determined as the target data consistency coefficient.

[0094] Specifically, the target data consistency coefficient r can be determined according to the following formula (7): (7) Where N represents the number of regions.

[0095] By determining the average of the third data consistency coefficients for all regions and setting this average as the target data consistency coefficient, which integrates the data performance of location, material, and different thicknesses, the consistency of data output from the two security inspection machines across the entire scanning field of view can be comprehensively and evenly characterized.

[0096] For example, based on the above embodiments, if the difference between the target data consistency coefficient and the preset coefficient is greater than or equal to the allowable coefficient deviation value, the detector parameters or X-ray source parameters of the non-baseline security inspection machine are adjusted and the target data consistency coefficient is recalculated until the difference between the calculated target data consistency coefficient and the preset coefficient is less than the allowable coefficient deviation value, at which point the non-baseline security inspection machine is deemed qualified; the non-baseline security inspection machine is the security inspection machine with poorer data quality among the first and second security inspection machines.

[0097] Specifically, the security inspection machine with better data quality from the first and second security inspection machines is designated as the benchmark machine, while the machine with poorer data quality is designated as a non-benchmark machine. Once the target data consistency coefficient *r* is determined, it is compared with a preset coefficient. If the difference is greater than or equal to the allowable deviation value, it indicates a significant difference between the data output of the non-benchmark machine and the benchmark machine, failing to meet the preset consistency standard. In this case, the detector parameters, radiation source parameters, or other hardware structure parameters of the non-benchmark machine need to be adjusted.

[0098] Subsequently, following the methods described in the aforementioned embodiments, the target data consistency coefficient between the baseline security inspection machine and the non-baseline security inspection machine after parameter adjustment is recalculated, and the recalculated target data consistency coefficient is compared with the preset coefficient. By repeating the above process, until the difference between the calculated target data consistency coefficient and the preset coefficient is less than the allowable coefficient deviation value, the non-baseline security inspection machine is determined to be qualified, thereby improving the quality of the security inspection machine.

[0099] It should be noted that when evaluating other security inspection machines, the target data consistency coefficient between other security inspection machines and the aforementioned benchmark security inspection machine can be determined. This benchmark security inspection machine with better performance can be used as a unified standard to evaluate and calibrate all other security inspection machines on the production line or in the system in turn, so as to achieve the consistency of output data of all security inspection machines in the whole batch or the whole system.

[0100] The consistency evaluation device for security X-ray data provided by the present invention is described below. The consistency evaluation device for security X-ray data described below and the consistency evaluation method for security X-ray data described above can be referred to in correspondence.

[0101] Figure 9 This is a schematic diagram of the structure of the consistency evaluation device for security X-ray data provided in an embodiment of the present invention, as shown below. Figure 9 As shown, the consistency evaluation device 900 for security X-ray data includes: The dividing module 11 is used to divide the scanning area perpendicular to the conveyor belt running direction of the first security inspection machine and the second security inspection machine into multiple areas respectively, and the area obtained by the first security inspection machine after division corresponds one-to-one with the area obtained by the second security inspection machine after division. The acquisition module 12 is used to acquire, for each of the regions, a first high-low energy data set and a second high-low energy data set of calibration objects of different materials and thicknesses in the region. The first high-low energy data set includes the high-low energy data collected by the first security inspection machine, and the second high-low energy data set includes the high-low energy data collected by the second security inspection machine. The determining module 13 is used to determine a first data consistency coefficient for the thickness of the material in the region based on a first high- and low-energy data set and a second high- and low-energy data set of calibrators of the same material and thickness in the region. The determining module 13 is further configured to determine a second data consistency coefficient for the material within the region based on a first data consistency coefficient corresponding to each thickness of the material within the region. The determining module 13 is further configured to determine a third data consistency coefficient corresponding to the region based on a second data consistency coefficient for each of the materials within the region. The determining module 13 is further configured to determine a target data consistency coefficient between the first security inspection machine and the second security inspection machine based on a third data consistency coefficient corresponding to all regions. The target data consistency coefficient is used to evaluate the consistency between the security inspection X-ray data of the first security inspection machine and the second security inspection machine.

[0102] In one example embodiment, the partitioning module 11 is specifically used for: A first pseudo-color image is obtained after the first security inspection machine scans a lead ruler placed on the conveyor belt of the first security inspection machine, wherein the lead ruler includes scales engraved at uniform physical intervals along the length direction. Based on the scale in the first pseudo-color image, the scanning area of ​​the first security inspection machine is divided into multiple areas; Obtain a second pseudo-color image after the second security inspection machine scans the lead ruler placed on the conveyor belt of the second security inspection machine; Based on the scale in the second pseudo-color image, the scanning area of ​​the second security inspection machine is divided into multiple regions.

[0103] In one example embodiment, the determining module 13 is specifically used for: Based on a first set of high and low energy data of calibrators of the same material and thickness within the region, a first set of scattered points is determined, and based on a second set of high and low energy data of the same calibrators within the region, a second set of scattered points is determined. Each first scattered point in the first set of scattered points is composed of a pair of high-energy data and low-energy data from the first set of high and low energy data, and each second scattered point in the second set of scattered points is composed of a pair of high-energy data and low-energy data from the second set of high and low energy data. For each first scattered point in the first scattered point set, determine the distance from the first scattered point to each second scattered point in the second scattered point set; Determine the minimum distance among multiple distances corresponding to each of the first scattered points, and determine the first average of the minimum distances of all the first scattered points; For each second scattered point in the second scattered point set, determine the distance from the second scattered point to each first scattered point in the first scattered point set; Determine the minimum distance among multiple distances corresponding to each of the second scattered points, and determine the second average of the minimum distances of all the second scattered points; The first data consistency coefficient is determined based on the first average value and the second average value.

[0104] In one example embodiment, the determining module 13 is specifically used for: From the first high- and low-energy data set, obtain the high- and low-energy data pairs corresponding to each pixel of the calibrated object within the region to obtain the first scatter point set; or... According to a preset sampling interval, the pixels of the calibration object are sampled in a grid pattern in the width and height directions of the region, and the high and low energy data pairs corresponding to each sampling point are obtained from the first high and low energy data set to obtain the first scatter set.

[0105] In one example embodiment, the determining module 13 is specifically used for: Based on the first average value and the second average value, determine the target distance between the first scatter set and the second scatter set; Determine the first average distance from all the first scattered points to the origin, and determine the second average distance from all the second scattered points to the origin; Based on the target distance, the first average distance, and the second average distance, the cosine value of the target angle is determined, wherein the target angle is the angle formed by connecting the center of the first scatter set and the center of the second scatter set from the origin; The cosine value is determined as the first data consistency coefficient.

[0106] In one example embodiment, the determining module 13 is specifically used for: Based on the first data consistency coefficient corresponding to each thickness of the material within the region and the number of thicknesses corresponding to the material, the average data consistency coefficient of the material within the region is determined for all thicknesses. The average data consistency coefficient of the material at all thicknesses within the region is determined as the second data consistency coefficient.

[0107] In one example embodiment, the determining module 13 is specifically used for: Based on the second data consistency coefficient for each material in the region and the weighting coefficient corresponding to each material, the weighted average data consistency coefficient for all materials in the region is determined. The weighted average data consistency coefficient corresponding to all materials in the region is determined as the third data consistency coefficient.

[0108] In one example embodiment, the determining module 13 is specifically used for: Determine the average value of the third data consistency coefficients for all regions, and use the average value as the target data consistency coefficient.

[0109] In one example embodiment, the device further includes an adjustment module, wherein: The adjustment module is used to adjust the detector parameters or X-ray source parameters of the non-benchmark security inspection machine and recalculate the target data consistency coefficient when the difference between the target data consistency coefficient and the preset coefficient is greater than or equal to the allowable coefficient deviation value. The non-benchmark security inspection machine is then deemed qualified when the difference between the calculated target data consistency coefficient and the preset coefficient is less than the allowable coefficient deviation value. The non-benchmark security inspection machine is the security inspection machine with poor data quality among the first security inspection machine and the second security inspection machine.

[0110] In one example embodiment, the material of the caliper includes at least two of iron, copper, aluminum, polyvinyl chloride (PVC), and plexiglass, and the length of the caliper is the same as the length of the ruler.

[0111] The apparatus of this embodiment can be used in any of the methods in the side embodiment of the consistency assessment method for security X-ray data. Its specific implementation process and technical effects are similar to those in the side embodiment of the consistency assessment method for security X-ray data. For details, please refer to the detailed description in the side embodiment of the consistency assessment method for security X-ray data, which will not be repeated here.

[0112] Figure 10 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as... Figure 10As shown, the electronic device may include: a processor 1010, a communication interface 1020, a memory 1030, and a communication bus 1040, wherein the processor 1010, the communication interface 1020, and the memory 1030 communicate with each other through the communication bus 1040. The processor 1010 can call logical instructions in the memory 1030 to execute a consistency evaluation method for security X-ray data. The method includes: dividing the scanning area perpendicular to the conveyor belt running direction of the first security inspection machine and the second security inspection machine into multiple regions, wherein the regions obtained by the first security inspection machine after division correspond one-to-one with the regions obtained by the second security inspection machine; for each region, acquiring a first high- and low-energy data set and a second high- and low-energy data set of calibration objects of different materials and thicknesses within the region, wherein the first high- and low-energy data set includes high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes high- and low-energy data collected by the second security inspection machine; and based on the same material and thickness within the region... The system uses a first set of high- and low-energy data and a second set of high- and low-energy data for a calibration object to determine a first data consistency coefficient for the thickness of the material within the specified region. Based on the first data consistency coefficients for each thickness of the material within the specified region, a second data consistency coefficient for the material within the specified region is determined. Based on the second data consistency coefficients for each material within the specified region, a third data consistency coefficient for the specified region is determined. Based on the third data consistency coefficients for all regions, a target data consistency coefficient is determined between the first security inspection machine and the second security inspection machine. The target data consistency coefficient is used to evaluate the consistency between the security inspection X-ray data of the first security inspection machine and the second security inspection machine.

[0113] Furthermore, the logical instructions in the aforementioned memory 1030 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the consistency evaluation method for security X-ray data provided by the above methods. The method includes: dividing a scanning area perpendicular to the conveyor belt running direction of a first security inspection machine and a second security inspection machine into multiple regions, wherein the regions obtained by the first security inspection machine after division correspond one-to-one with the regions obtained by the second security inspection machine; for each region, acquiring a first high- and low-energy data set and a second high- and low-energy data set of calibration objects of different materials and thicknesses within the region, wherein the first high- and low-energy data set includes high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes... The method includes: collecting high- and low-energy data from the second security inspection machine; determining a first data consistency coefficient for the thickness of the material within the region based on a first set of high- and low-energy data and a second set of high- and low-energy data for a calibration object of the same material and thickness within the region; determining a second data consistency coefficient for the material within the region based on the first data consistency coefficient for each thickness within the material within the region; determining a third data consistency coefficient for the region based on the second data consistency coefficient for each material within the region; and determining a target data consistency coefficient between the first security inspection machine and the second security inspection machine based on the third data consistency coefficients for all regions. The target data consistency coefficient is used to evaluate the consistency between the security X-ray data of the first security inspection machine and the second security inspection machine.

[0115] Furthermore, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements a consistency evaluation method for security X-ray data provided by the methods described above. This method includes: dividing a scanning area perpendicular to the conveyor belt running direction of a first security inspection machine and a second security inspection machine into multiple regions, wherein the regions obtained by the first security inspection machine correspond one-to-one with the regions obtained by the second security inspection machine; and for each region, acquiring a first high- and low-energy data set and a second high- and low-energy data set of calibration objects of different materials and thicknesses within the region, wherein the first high- and low-energy data set includes high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes high- and low-energy data collected by the second security inspection machine. Based on a first set of high- and low-energy data and a second set of high- and low-energy data for calibration objects of the same material and thickness within the region, a first data consistency coefficient is determined for the thickness of the material within the region. Based on the first data consistency coefficient for each thickness of the material within the region, a second data consistency coefficient for the material within the region is determined. Based on the second data consistency coefficient for each material within the region, a third data consistency coefficient for the region is determined. Based on the third data consistency coefficients for all regions, a target data consistency coefficient is determined between the first security inspection machine and the second security inspection machine. The target data consistency coefficient is used to evaluate the consistency between the security inspection X-ray data of the first security inspection machine and the second security inspection machine.

[0116] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0117] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the consistency of security inspection X-ray data, characterized in that, include: The scanning area perpendicular to the conveyor belt running direction of the first security inspection machine and the second security inspection machine is divided into multiple areas, and the area obtained by the first security inspection machine after division corresponds one-to-one with the area obtained by the second security inspection machine. For each of the aforementioned regions, a first high- and low-energy data set and a second high- and low-energy data set are obtained for calibration objects of different materials and thicknesses within the region. The first high- and low-energy data set includes the high- and low-energy data collected by the first security inspection machine, and the second high- and low-energy data set includes the high- and low-energy data collected by the second security inspection machine. Based on a first set of high and low energy data of calibrators of the same material and thickness within the region, a first set of scattered points is determined, and based on a second set of high and low energy data of the same calibrators within the region, a second set of scattered points is determined. Each first scattered point in the first set of scattered points is composed of a pair of high-energy data and low-energy data from the first set of high and low energy data, and each second scattered point in the second set of scattered points is composed of a pair of high-energy data and low-energy data from the second set of high and low energy data. For each first scattered point in the first scattered point set, determine the distance from the first scattered point to each second scattered point in the second scattered point set; Determine the minimum distance among multiple distances corresponding to each of the first scattered points, and determine the first average of the minimum distances of all the first scattered points; For each second scattered point in the second scattered point set, determine the distance from the second scattered point to each first scattered point in the first scattered point set; Determine the minimum distance among multiple distances corresponding to each of the second scattered points, and determine the second average of the minimum distances of all the second scattered points; Based on the first average value and the second average value, a first data consistency coefficient is determined; Based on the first data consistency coefficient corresponding to each thickness of the material within the region, a second data consistency coefficient for the material within the region is determined. Based on the second data consistency coefficient for each material within the region, a third data consistency coefficient corresponding to the region is determined. Based on the third data consistency coefficient corresponding to all regions, a target data consistency coefficient is determined between the first security inspection machine and the second security inspection machine. The target data consistency coefficient is used to evaluate the consistency between the security inspection X-ray data of the first security inspection machine and the second security inspection machine.

2. The consistency evaluation method for security X-ray data according to claim 1, characterized in that, The scanning area, which is perpendicular to the conveyor belt running direction of the first and second security inspection machines, is divided into multiple areas, including: A first pseudo-color image is obtained after the first security inspection machine scans a lead ruler placed on the conveyor belt of the first security inspection machine, wherein the lead ruler includes scales engraved at uniform physical intervals along the length direction. Based on the scale in the first pseudo-color image, the scanning area of ​​the first security inspection machine is divided into multiple areas; Obtain a second pseudo-color image after the second security inspection machine scans the lead ruler placed on the conveyor belt of the second security inspection machine; Based on the scale in the second pseudo-color image, the scanning area of ​​the second security inspection machine is divided into multiple regions.

3. The consistency evaluation method for security inspection X-ray data according to claim 1, characterized in that, The determination of the first scatter point set based on the first high- and low-energy data set of calibration objects of the same material and thickness within the region includes: From the first high- and low-energy data set, obtain the high- and low-energy data pairs corresponding to each pixel of the calibrated object within the region to obtain the first scatter point set; or... According to a preset sampling interval, the pixels of the calibration object are sampled in a grid pattern in the width and height directions of the region, and the high and low energy data pairs corresponding to each sampling point are obtained from the first high and low energy data set to obtain the first scatter set.

4. The consistency evaluation method for security inspection X-ray data according to claim 1, characterized in that, The step of determining the first data consistency coefficient based on the first average value and the second average value includes: Based on the first average value and the second average value, determine the target distance between the first scatter set and the second scatter set; Determine the first average distance from all the first scattered points to the origin, and determine the second average distance from all the second scattered points to the origin; Based on the target distance, the first average distance, and the second average distance, the cosine value of the target angle is determined, where the target angle is the angle formed by connecting the center of the first scatter set and the center of the second scatter set from the origin of the coordinate system. The cosine value is determined as the first data consistency coefficient.

5. The consistency evaluation method for security inspection X-ray data according to any one of claims 1-4, characterized in that, The step of determining a second data consistency coefficient for the material within the region based on a first data consistency coefficient corresponding to each thickness of the material within the region includes: Based on the first data consistency coefficient corresponding to each thickness of the material within the region and the number of thicknesses corresponding to the material, the average data consistency coefficient of the material within the region is determined for all thicknesses. The average data consistency coefficient of the material at all thicknesses within the region is determined as the second data consistency coefficient.

6. The consistency evaluation method for security inspection X-ray data according to any one of claims 1-4, characterized in that, The process of determining the third data consistency coefficient for the region based on the second data consistency coefficient for each material within the region includes: Based on the second data consistency coefficient for each material in the region and the weighting coefficient corresponding to each material, the weighted average data consistency coefficient for all materials in the region is determined. The weighted average data consistency coefficient corresponding to all materials in the region is determined as the third data consistency coefficient.

7. The consistency evaluation method for security inspection X-ray data according to any one of claims 1-4, characterized in that, The step of determining the target data consistency coefficient between the first security screening machine and the second security screening machine based on the third data consistency coefficient corresponding to all regions includes: Determine the average value of the third data consistency coefficients for all regions, and use the average value as the target data consistency coefficient.

8. The consistency evaluation method for security inspection X-ray data according to any one of claims 1-4, characterized in that, The method further includes: If the difference between the target data consistency coefficient and the preset coefficient is greater than or equal to the allowable coefficient deviation value, the detector parameters or X-ray source parameters of the non-benchmark security inspection machine are adjusted and the target data consistency coefficient is recalculated until the difference between the calculated target data consistency coefficient and the preset coefficient is less than the allowable coefficient deviation value. At this point, the non-benchmark security inspection machine is deemed qualified. The non-benchmark security inspection machine is the security inspection machine with poor data quality among the first security inspection machine and the second security inspection machine.

9. The consistency evaluation method for security inspection X-ray data according to claim 2, characterized in that, The material of the calibration object includes at least two of the following: iron, copper, aluminum, polyvinyl chloride (PVC), and plexiglass. The length of the calibration object is the same as the length of the ruler.

Images