Automatic detection of tampered metal objects in X-ray images

Abstract

This disclosure relates to a method for detecting a tampered metal article (FC), wherein a non-metallic material is concealed within the metal article (FZ) and the metal article (FZ) has been tampered with, such that the tampered metal article (FC) produces a two-dimensional X-ray image in two-dimensional X-ray data of an inspection object (O1, O2, O3) containing the metal article (FC) equivalent to that of a corresponding untampered metal article. The method includes the steps of: (S10) determining a region containing the metal article (FC) in the X-ray data; (S20) providing, based on the X-ray data, a region along the path through the metal article (FZ)... (S30) Evaluate whether the attenuation curve (D(r)) or intensity curve (I(r)) of the detected X-ray radiation on the line (L) of the metal object (FC) exhibits characteristic anomalies (P1, P2) in a predetermined area (e.g., the edge (R1, R2) or edge region (RB1, RB2) of the metal object (FZ); (S40) Trigger an alarm function when the attenuation curve (D(r)) or intensity curve (I(r)) within the predetermined range exhibits the characteristic anomalies (P1, P2).

Description

[0001] This disclosure generally relates to the non-destructive inspection of objects and the detection of potentially hazardous materials contained therein. In particular, this disclosure relates to the detection of articles that have been tampered with to conceal hazardous substances, such as batteries or battery cells, which have been modified to conceal hazardous substances (e.g., explosives or drugs) to prevent detection. Background Technology

[0002] The following introductory description is intended only to better understand this disclosure and should not in any way be construed as an admission of prior art unless expressly stated otherwise.

[0003] It is now known to detect inherently dangerous organic explosives hidden within otherwise harmless metallic objects in transmission imaging (e.g., C. Endt et al., in "Die" in the Süddeutsche Zeitung, May 17, 2017). The task of "ist entscheidend" ("Size Materials") is, if not impossible, extremely difficult. That is to say, dangerous "explosives" in X-ray images used for optical orientation analysis are currently almost impossible, or no longer possible, to distinguish from their environment in a similar mimicry manner. Specifically, this means that explosives may be hidden within the battery cells or accumulators of mobile electronic devices, which themselves are not dangerous.

[0004] In the following text, we will only discuss battery cells (i.e., rechargeable battery cells), without excluding batteries as primary cells. Mobile devices (such as laptops) currently primarily use lithium-ion batteries (here referred to simply as lithium-ion batteries). To conceal explosives within one or more battery cells of a battery pack, a portion or all of the interior of one or more battery cells is replaced with the explosive. This ensures that such a battery pack with tampered or incorrect battery cells can still provide sufficient power to the mobile device for functional testing.

[0005] X-ray-based inspections of carry-on baggage at airports typically only produce two-dimensional X-ray images of the baggage to be inspected. When inspecting baggage designated for use in the aircraft cargo hold, the current practice is to use computed tomography (CT) methods, which reliably detect explosives. Therefore, the inspection systems at security checkpoints used for carry-on baggage inspections (which display two-dimensional X-ray images to operators) should be improved to detect concealed explosives.

[0006] US20160084984A1 discloses a system and method for locating lithium batteries in an object under inspection. This is sufficient if the sole purpose is to detect the presence of lithium batteries or batteries that resemble them. However, if the sole concern is finding tampered batteries, this is not a solution. Implementing such known systems and methods would require a more thorough inspection of every mobile device with a battery cell to rule out explosives hidden within, or generally to rule out such mobile devices carried on aircraft. No single method is suitable for practical use.

[0007] US20110206240A1 also relates to the detection of potentially threatening items that may be hidden inside objects such as portable electronic devices, wherein the object to be inspected is processed by computed tomography (CT) imaging, and 3D CT data of the object is obtained. During the analysis of the CT data, items in a piece of luggage that require more detailed inspection (e.g., a laptop computer) are first identified as objects to be inspected. Then, the object is further divided into multiple partitions for further inspection by generating a one-dimensional self-projection of the CT data. The feature vectors of these partitions and the CT image data are then used to generate layout feature vectors. One or more layout feature vectors are then compared with training data from a class of imaged items containing and not containing threats to determine whether the imaged item contains a threat.

[0008] US4539648A discloses an imaging X-ray inspection system for detecting agricultural contraband contained in an object (e.g., a piece of luggage or parcel). Optionally, imaging of items with circular cross-sections is improved to better distinguish these items from those with rectangular cross-sections. Based on the assumption that the item is surrounded by a material having a different density absorption coefficient product than the item, a gradient image of the spatially resolved intensity of the X-ray radiation transmitted by the object is proposed to remove the edges of items with rectangular cross-sections in the X-ray image. Summary of the Invention

[0009] It is desirable to have an X-ray inspection system and an X-ray inspection method that improves, or at least makes fully possible, the detection of tampered, substantially metallic articles such as battery cells (particularly lithium battery cells) that conceal hazardous organic substances, such as explosives.

[0010] The features and details defined in relation to the X-ray examination method according to this disclosure are also applicable to the X-ray examination equipment according to this disclosure, and vice versa. Therefore, the disclosures of each aspect are interrelated.

[0011] In order to find a technical solution based on this disclosure, several technical considerations must be taken into account.

[0012] Typically, a lithium-ion battery cell has a cylindrical shape with a thin sheath of metal foil and is internally constructed of known functional materials (e.g., aluminum, lithium oxide metal, highly porous separators, carbon, and copper). If the interior of such a lithium-ion battery cell is tampered with due to the complete or partial filling of the cell with organic explosives, the tampered battery will produce an X-ray image different from that of an actual lithium-ion battery cell. Because of its predominantly metallic composition, an actual lithium-ion battery attenuates X-ray radiation more readily than an organic material (such as an explosive). For a tampered cell, the X-ray image will therefore differ significantly from that expected of a lithium-ion battery cell due to the presence of organic explosives, and will thus be very noticeable to the operator in the X-ray image.

[0013] To compensate for the lack of attenuation properties in a tampered cell, the tampered cell can be manufactured with a much thicker metal sheath, making the resulting transmission-based X-ray image of the tampered cell look similar to that of a regular lithium-ion battery cell. If the metal sheath is designed to have a uniform thickness, i.e., a sleeve with a uniform thickness, then in an X-ray inspection system with multiple transmission directions (a so-called multi-view device), the resulting X-ray image will look equally realistic in all viewing directions. This is problematic.

[0014] Such a tampered unit is referred to here as a "pseudo-unit". A pseudo-unit is a tampered battery unit or battery cell in which foreign material has been hidden within the unit for its actual function, and the unit has been altered so that it looks as real as possible in an X-ray image.

[0015] The inventors recognized that characteristic features of untampered cells, especially anomalies of tampered cells, could be detected in X-ray images of the attenuation or intensity curves of the detected X-ray radiation passing through dummy cells. Particularly suitable are the attenuation or intensity curves of the detected X-ray radiation (which will be analyzed and which crosses the cell as much as possible, i.e., extends perpendicular to its longitudinal axis).

[0016] For example, it has been found that the metallic sheath or sleeve of a dummy unit in an X-ray image causes a characteristic peak in the attenuation curve or, correspondingly, at the edge of the dummy unit in the intensity curve passing through it. This is because the metallic sheath or sleeve is radiated tangentially rather than perpendicularly by X-ray radiation at the edge of the unit. As a result, the X-ray radiation at the edge of the unit, especially in the edge region defined by the metallic sheath or sleeve, is attenuated more strongly than in the region of the dummy unit between the edge regions defined by the metallic sheath or sleeve, because it is radiated perpendicularly through the metallic sheath or sleeve, and therefore through less metal than in the edge region, and thus the X-ray radiation is effectively attenuated less. The thicker the metallic sheath (i.e., the metallic sheath or sleeve), the more pronounced this effect. This is precisely the case for dummy units, as the lack of attenuation caused by organic hazardous materials (such as explosives) is compensated for by the greater amount of metal in the unit sheath.

[0017] Finally, during the development of the above scheme, it became clear that a major objective was also to find a sufficiently robust method to detect spurious cells. In real X-ray images, various interferences arise; that is, the cells to be analyzed are not ideally isolated, but rather, for example, stacked differently, laptop components and cables may cover the cells, and the metal sleeves of spurious and real cells may have different sizes. Here, idealized examples are used to explain the principle presented here, upon which the developed detection method is based. In this respect, the X-ray images and scenes shown may be misleading in themselves, because in each case the analyzed cell is "exposed."

[0018] It should be noted that the principles of this invention are not limited to detecting counterfeit units containing explosives. Rather, the principle can be applied to the detection of any tampered metallic article (in which organic material is concealed). The organic material may also be, for example, a pharmaceutical product. Therefore, in the following text, the term "(one or more) counterfeit units" should not be construed as limiting, but should generally be used as a substitute for any kind of counterfeit article tampered with according to the same principle.

[0019] The core idea of ​​this disclosure is to apply the above findings to the detection of metallic pseudo-elements during X-ray inspection, particularly in the evaluation process of detecting pseudo-elements in transmission images (e.g., two-dimensional X-ray images) of the inspected object.

[0020] The first aspect of this disclosure relates to a method for detecting tampered metallic articles, wherein non-metallic material is hidden in two-dimensional (2D) transmission data, particularly 2D X-ray data, of the object being examined, which contains the article. For example, the metallic article can be tampered with such that the tampered metallic article produces a transmission image, such as a 2D X-ray image, equivalent to that of a corresponding untampered metallic article. The 2D transmission data is data acquired by irradiating the object being examined in one spatial dimension along a second spatial dimension and sensing radiation not absorbed by the object. In the case of 2D X-ray data, these are attenuation or intensity values ​​of X-ray radiation always sensed along the transmission direction of one (or more) lines. Of course, the 2D transmission data can also be derived from three-dimensional (3D) transmission data acquired for the spatial elements of the object being examined and used in the process presented herein.

[0021] The method includes the following steps: determining a region containing a metal object in the X-ray data; providing an attenuation curve or intensity curve from the X-ray data of detected X-ray radiation (which has passed through the object being inspected) along a line passing through the metal object; evaluating whether the attenuation curve or intensity curve exhibits abnormal characteristics in a predetermined area, such as at the edge or edge region of the metal object; and triggering an alarm function if the attenuation curve or intensity curve exhibits abnormal characteristics in the predetermined area.

[0022] Determining that the region containing the metallic article in the X-ray data may include: determining that the region contains metal based on whether the region in the X-ray data substantially causes minimal attenuation of the X-ray radiation and / or based on what atomic number or effective atomic number (Z value or Z effective) is assigned to the material in the X-ray data.

[0023] Providing the attenuation curve or the intensity curve along the line passing through the metal article from the X-ray data may include: determining the longitudinal direction of the metal article; and providing the attenuation curve or the intensity curve along the line passing through the metal article. The line may be aligned such that it is orthogonal to the determined longitudinal direction.

[0024] Assessing whether the attenuation curve or the corresponding intensity curve exhibits abnormal characteristics in the predetermined region may include at least one of the steps (a)-(d). Clearly, the intensity curve is substantially complementary to the attenuation curve; therefore, taking this into consideration, all steps used for the attenuation curve can be applied accordingly to the intensity curve.

[0025] Step (a): Determine whether the value of the attenuation increases along the attenuation curve from the edge of the metal article to the center of the metal article.

[0026] Step (b): Determine whether the attenuation curve has a step (jump) in the edge or peripheral area of ​​the metal article.

[0027] Step (c): Determine whether the attenuation curve has maximum attenuation at the edge of the metal article, and based on the maximum attenuation, the attenuation curve decreases toward the center of the metal article and / or extends at a low level compared to the edge.

[0028] Step (d): Determine the first derivative of the attenuation curve based on the location, and determine whether the derivative exhibits a peak value in each of the edge regions of the metal article.

[0029] In conjunction with the above measures, and for the entire document at hand, it should be noted that “edge area” here is understood to mean the area extending from the edge of the metal article toward the center of the metal article, wherein the edge area can be defined by approximately 10% of the distance from the edge to the center.

[0030] The attenuation curve or corresponding intensity curve of the contour signal can be analyzed using various methods, such as machine learning, to identify or detect characteristic anomalies, where empirically-based "human" knowledge is stored in the system. In principle, the system learns patterns and regularities in the data through presented examples of pseudo-units to be identified, where knowledge about the examples is generalized after the learning phase. Ultimately, the system can not only identify the presented examples but also evaluate unknown (i.e., new) data. The principles of machine learning are known to experts in the art and therefore do not need to be explained in detail here.

[0031] The alarm triggering function has at least one of the following steps (i)-(iii)

[0032] Step (i): Overlay the attenuation curve or the intensity curve onto the X-ray image of the object being examined along the line presented by the attenuation curve or the intensity curve.

[0033] Step (ii): Trigger a manual inspection of the object to be inspected.

[0034] Step (iii): At the inspection system implementing the method, trigger a visual and / or auditory alarm.

[0035] The metal article may be an article with a metal sleeve or metal cover. Furthermore, the tampered metal article may be a tampered battery cell or battery unit (pseudo-cell), such as a tampered lithium battery unit.

[0036] A second aspect of this disclosure relates to a processing apparatus for evaluating transmission images, the processing apparatus including a computer unit. The computer unit is adapted to perform the method according to a first aspect of this disclosure.

[0037] The processing device includes an output unit configured to perform an alarm function.

[0038] The computer unit has a communication interface through which it can network with one or more inspection devices for non-destructive testing of the object under inspection to communicate data and receive X-ray images from the one or more inspection devices for evaluation (e.g., visual) via data communication.

[0039] The third aspect of this disclosure relates to an inspection apparatus arranged for transmitting an object to be inspected through the inspection apparatus, performing an imaging inspection method for non-destructive inspection of the object to be inspected, and providing a transmission image of the object to a processing apparatus according to the second aspect of this disclosure, wherein the inspection apparatus is connected to the processing apparatus for data communication.

[0040] A fourth aspect of this disclosure relates to an inspection system having at least one processing device according to a second aspect of this disclosure, the processing device being spatially independently connected to at least one inspection device according to a third aspect of this disclosure for data communication, wherein a transmission image of an object being inspected at the at least one inspection device is transmitted to the processing device for visual evaluation.

[0041] The fifth aspect of this disclosure relates to a computer program product comprising a computer program having software means which, when executed on a computer, particularly on a processing device according to the second aspect of this disclosure, is used to perform a method according to the first aspect of this disclosure.

[0042] The sixth aspect of this disclosure relates to a data carrier comprising a computer program product according to the fifth aspect of this disclosure.

[0043] The seventh aspect of this disclosure relates to a data stream having electronically readable control signals that can interact with a programmable computer in such a way that, when the computer executes the electronically readable control signals, the computer performs the method according to the first aspect of this disclosure. Example

[0044] Other advantages, features, and details of this disclosure will become apparent from the following description, in which examples of this disclosure are described in detail with reference to the accompanying drawings. Features mentioned in the claims and specification may be essential to this disclosure individually or in any combination. Similarly, the foregoing features and the features further detailed herein may each be used individually or in any desired combination with several other features. Components or parts having similar or identical functions are partially denoted by the same reference numerals. The terms “left,” “right,” “top,” and “bottom” as used in the description of the design examples refer to figures aligned with generally readable reference numerals or symbols. The illustrated and described embodiments should not be construed as exhaustive, but rather as exemplary features for explaining this disclosure. The detailed description is intended to inform those skilled in the art; therefore, known structures and processes are not shown or explained in detail in the description so as not to complicate the understanding of this description.

[0045] Figure 1 A simplified block diagram of an inspection system with a processing device for evaluating transmission images of an object under inspection is shown.

[0046] Figure 2 An inspection system is shown, which has several (N) at one inspection point. Figure 1 The inspection equipment, and the spatially separated evaluation points have several (M) evaluation devices.

[0047] Figure 3 The attenuation curves of X-ray radiation are shown as a comparison of the lines passing through a real lithium-ion battery cell (left) and through a tampered lithium-ion battery (so-called a fake cell (right)) to illustrate and explain the principle presented here for detecting fake cells.

[0048] Figure 4 Another example of a degradation curve is shown along a line passing through an untampered lithium-ion battery cell.

[0049] Figure 5 Another example of a decay curve is shown on a pseudo-unit installed in a laptop computer.

[0050] Figure 6A The diagram, from top to bottom, shows a simplified cross-section of an unaltered lithium-ion battery cell, a portion of a 2D X-ray image of the lithium-ion battery cell, an intensity curve of the detected X-ray radiation along a line passing through the X-ray image, and a curve derived from the intensity curve using a position variable.

[0051] Figure 6B The diagram, from top to bottom, shows a simplified cross-section of two adjacent, tampered lithium-ion battery cells (i.e., pseudo-cells), a portion of a 2D X-ray image of the pseudo-cells, an intensity curve along a line passing through the X-ray image, and a curve derived from the intensity curve by the position variable.

[0052] Figure 7 A flowchart of a method for detecting pseudo-units in an object under inspection, according to this disclosure, is shown.

[0053] Figure 1 A simplified block diagram of an inspection system 300 with an evaluation device 100 is shown, which is used to evaluate the transmission images B1 of inspected objects O1, O2, and O3. Figure 3 B2 Figure 4 B3 Figure 5 B4 Figure 6A ) and B5 ( Figure 6B To simplify, Figure 1 The objects to be inspected, O1, O2, and O3, are shown only symbolically.

[0054] The evaluation device 100 can essentially be an operator's workstation and includes a processing unit 110, an input unit 120, and an output unit 130.

[0055] The output unit 130 has been configured to display transmission images of the inspected objects O1, O2, and O3 to the operator, such as transmission image B1. Figure 3 B2 Figure 4 ) and B3 ( Figure 5 This is to check whether the contents of the inspected objects O1, O2, and O3 contain a target item. The target item can be a dangerous item, such as a weapon, explosive, hazardous liquid, and / or contraband and / or drugs. In the context of this disclosure, it is particularly important to detect hazardous substances (especially explosives) hidden within otherwise harmless metallic articles. As described elsewhere, harmless metallic articles that can be altered or counterfeited in such a way (i.e., altered or counterfeited articles (counterfeit articles)) appear similar to or nearly identical to genuine (i.e., unaltered) articles in two-dimensional X-ray images. This will correspondingly affect the detection rate. To improve or facilitate the detection of such counterfeit articles, appropriate measures are described herein using altered or counterfeited battery cells or accumulators as an example.

[0056] A key part of the task is finding a robust enough approach because various interferences can occur in real images, such as the units being stacked differently, laptop components and cables overlapping, and sleeves and units having different sizes and types; in this respect, the image / drawing itself can be misleading because the units there are placed "freely".

[0057] Figure 1The input unit 120 is essentially configured to input the operator's input to control the evaluation device 100. Furthermore, the operator can input the results of the visual evaluation in response to the displayed transmission images B1, B2, B3, such as whether the inspected objects O1, O2, O3 are "safe," or whether the contents of the inspected objects O1, O2, O3 must be further examined, especially in cases where manual examination is necessary. A more detailed investigation can also be automatically triggered by the processing unit 110 via an alarm function, for example, to examine objects with traces of explosives in a more detailed investigation (explosive trace detection, ETD). This measure is time-consuming and expensive, and should therefore only be necessary when there is a sufficient reason to do so.

[0058] The transmission image B1 of the inspected objects O1, O2, and O3 ( Figure 3 A), B2 Figure 4 A) and B3( Figure 5 A) By inspection equipment 200 ( Figure 1 The X-ray examination equipment is generated using imaging techniques. In this embodiment, the examination device 200 is an X-ray examination device known from DE10149254A1. This X-ray examination device may have one or more radiation planes.

[0059] The inspection device 200 is essentially designed to transport the objects O1, O2, and O3 through the inspection device 200 via a conveyor device T, such as a conveyor belt. Inside the inspection device 200, the objects O1, O2, and O3 are non-destructively irradiated with X-rays in a known manner. As a result, the inspection device 200 generates two-dimensional X-ray data of the objects O1, O2, and O3, thereby generating 100 X-ray images on an evaluation device and displaying them to the operator on the display unit 130.

[0060] exist Figure 1In an embodiment of the inspection system 300, the evaluation device 100 is spatially separated from the inspection device 200 and connected to the inspection device 200 via a network connection 310 for data communication. For this purpose, the computer unit 110 is equipped with a first communication interface 112, and the inspection device 200 is equipped with a second communication interface 212. The computer unit 110 is connected to one of the inspection devices shown through this second communication interface, and also to other inspection systems 200 via the network connection 310 for data communication, for non-destructive testing of the objects being inspected. Through the network connection 310, the acquired transmission data of the objects O1, O2, and O3 inspected at the inspection device 200 is transmitted to the evaluation device 100 for evaluation. In addition to multiple inspection devices 200, the inspection system 300 may also have multiple evaluation devices 100, thereby optimally distributing the workload of evaluating the transmission data across the existing evaluation devices 100 (such a system in...). Figure 2 (as shown in the diagram). Of course, the evaluation device 100 can always be installed near the inspection device 200, or as an integral part of the inspection device 200.

[0061] Figure 2 An inspection system 300 is shown, which has a plurality (n) of points located at an inspection point 250. Figure 1 The system includes inspection devices 200.1, 200.2, ..., 200.n and several (m) evaluation devices 100.1, ..., 100.m located at evaluation points 150, which are spatially separated from the inspection system. The inspection devices 200.1, 200.2, ..., 200.n at the inspection points 250 are connected to the evaluation devices 100.1, ..., 100.m at the evaluation points 150 via a data network 320 for electronic data exchange in a known manner. Alternatively, the inspection points 250 may be networked with one or more evaluation points 150.

[0062] For example, checkpoint 250 could be a checkpoint at the entrance to an airport security area. At this checkpoint, the hand luggage of air passengers is inspected in a conventional manner using inspection equipment 200.1, 200.2, ..., 200.n at checkpoint 250. In this example, the X-ray data generated by one of the inspection equipment 200.1, 200.2, ..., 200.n is, for example, X-ray data of luggage being inspected by passengers. For evaluation, such as regarding potential security risks or other items of interest, the X-ray data is transmitted via network 320 to one of the evaluation equipment 100.1, ..., 100.m, where the X-ray data is used to generate X-ray images B1-B3 and made visually inspected by an operator.

[0063] The operator faces particular challenges when inspecting electronic mobile devices with battery or accumulator units. The operator must determine whether the mobile device's battery or accumulator unit may have been tampered with, thus requiring additional time-consuming and costly inspections. Avoiding unnecessary additional inspections at inspection point 250 can be avoided by more effectively evaluating X-ray data at evaluation point 100, thereby improving the overall inspection process. Furthermore, additional costs can be avoided.

[0064] Figure 3 A first embodiment of the proposed improvement for evaluation is shown, and in particular a first embodiment of the improvement for detecting possible counterfeit items in an inspection object.

[0065] Figure 3 The upper part (a) shows the first X-ray image B1 of the tray W, which is the first object of inspection O1. There are two battery packs AP1 and AP2 in the tray W, each of which consists of 9 cells.

[0066] The first battery pack AP1 consists of an arrangement of nine lithium battery cells LZ, wherein three battery cells are connected in series and three cells with series-connected lithium battery cells are connected in parallel.

[0067] The second battery pack AP2 is essentially the same as the first battery pack AP1 in terms of lithium battery cell connections, but one of its three series-connected battery cells consists of three tampered lithium battery cells, referred to here as the pseudo-cell FZ. This means that the second battery pack AP2 has only 2 / 3 the capacity of the first battery pack AP1, but it is essentially functional. Therefore, the second battery pack AP2 cannot be identified by functional testing of a mobile device containing only this battery pack AP2.

[0068] As discussed elsewhere, and Figure 3 As shown, the actual lithium battery cell LZ has a cylindrical shape with a thin metal foil sheath and is internally composed of known functional materials, such as an aluminum electrode coated with lithium oxide and a copper electrode coated with carbon. The lithium oxide layer and the carbon layer are arranged in an electrolyte and separated from each other by highly porous separators.

[0069] If the interior of a tampered pseudo-cell FZ is completely or partially filled with organic explosives, it will produce a clearly identifiable different X-ray image compared to a genuine lithium-ion battery cell LZ. Due to its predominantly metallic composition, the attenuation of X-ray radiation in a genuine lithium-ion battery cell LZ is higher than that in a pseudo-cell filled with organic materials (such as explosives). Therefore, in the case of pseudo-cell FZ, due to the organic explosive content, the X-ray image will be clearly distinguishable from that of a genuine lithium-ion battery cell LZ, making it easily identifiable to the operator; it is quite evident in X-ray image B1.

[0070] To compensate for the missing attenuation characteristics of the pseudo cell FZ, a thicker metal sheath can be designed, with the material thickness adjusted in such a way that the transmission-based X-ray image of the pseudo cell FZ looks like an X-ray image of a normal, untampered lithium battery cell LZ.

[0071] If the metal sheath is designed to have a uniform thickness, i.e., in the form of a sheath or sleeve with a uniform thickness, then in a multi-view X-ray inspection system, the resulting X-ray images in all view directions acquired during the process can also appear equally realistic. This is problematic because the FZ dummy unit may still remain undetected, and explosives could be smuggled into a safe area using such a dummy unit.

[0072] The inventors have recognized that pseudo-units FZ in X-ray images exhibit anomalous characteristics in the attenuation curve along a line passing through the pseudo-unit FZ, making it possible to automatically detect the possible presence of pseudo-units FZ. Such anomalies can be detected particularly well in the attenuation curve along a line L that is substantially transverse to or orthogonal to the longitudinal direction LR of (one or more) pseudo-units FZ.

[0073] The metal sheath of the pseudo-unit FZ causes characteristic peaks P1 and P2 in the attenuation curve of the X-ray image B1 along the line L passing through the pseudo-unit FZ, appearing at the left edge R1 and right edge R2 of the pseudo-unit FZ. These peaks P1 and P2 are due to the metal sheath or sleeve being radiated tangentially rather than perpendicularly at the unit edges R1 and R2. As a result, the X-ray radiation passes through more metal at the unit edges R1 and R2, and therefore attenuates much more than in the pseudo-unit region between the unit edges R1 and R2. In the region between the unit edges R1 and R2, the metal sleeve or sheath is penetrated substantially perpendicularly, and therefore less metal is actually irradiated. Consequently, the X-ray radiation attenuates less. The thicker the metal sleeve or sheath of the pseudo-unit FZ, the more pronounced this effect and the resulting detectable anomaly become.

[0074] Therefore, by searching for the aforementioned anomalies, the pseudo-unit FZ can be identified. These anomalies are difficult for the human eye to detect, especially due to screen resolution, and thus have remained undetected until now.

[0075] In any case, the inventors have recognized that this anomaly occurs particularly in pseudo-cells FZ compared to untampered lithium-ion battery cells, because the lack of degradation, especially the lack of degradation of the organic hazardous materials (e.g., explosives) hidden within them, must be compensated for by more metal in the battery casing.

[0076] exist Figure 3 In the lower part (b), the above findings are shown by the attenuation curve D(r) along the line L with the position variable r. For optimal analysis, the attenuation curve D(r) can be examined transversely to the longitudinal direction LR of the cells of battery packs AP1 and AP2. However, it should be noted that the principle described here is also valid in other cases, i.e., anomalies indicating tampering can also be detected if the curve is examined at an angle (i.e., tilted) to the longitudinal direction LR.

[0077] In any case, Figure 3 In the lower part (b), it can be seen that for the left battery pack AP1, the attenuation curve D(r) increases from the edge of the battery cell toward the center M of the battery pack AP1 as expected, and decreases from the center M toward the edge of the battery cell.

[0078] for Figure 3 In the lower part (b), on the right side of the second battery assembly AK2, the pseudo-cell FZ shows that the decay curve D(r) at the cell edges R1 and R2 of the pseudo-cell FZ has characteristic peaks P1 and P2 in the corresponding local maxima Dmax1 and Dmax2, i.e., obvious step (jump). Then the decay curve D(r) decreases significantly towards the center M of the battery assembly AP2 and the pseudo-cell FZ, or extends noticeably at a significantly lower level. This is anomalous compared to the decay curve D(r) of the actual lithium battery cell LZ, and is described here as a possible characteristic anomaly for detecting the pseudo-cell FZ.

[0079] In the initial consideration of constructing a typical pseudo-cell, one approach is to attempt to extract the metallic portion from X-ray data in the region of the identified battery cell. However, even in simple scenarios, this proves extremely difficult due to interference from X-ray signals attenuated by organic materials. The solution proposed and discussed here precisely utilizes the metallic portion of the pseudo-cell to detect anomalies caused by it (anomaly detection), which also works in challenging scenarios. A particularly difficult scenario to detect is battery cells within built-in electronic devices (e.g., laptop computers), which are placed in a package along with many other objects.

[0080] Figure 4 and 5 Each shows another example of the principle presented here, therefore Figure 4 and 5 I will not explain in detail, but only in essence.

[0081] exist Figure 4 The upper part (a), the second X-ray image B2 shows the tray W as the second object of inspection O2. In the tray W, there is a third battery pack AP3, which has six actual (i.e., tamper-proof) lithium battery cells LZ arranged side by side.

[0082] exist Figure 4 In the lower part (b), along Figure 4 The attenuation curve D(r) of the X-ray radiation of line L in the upper part (a) is shown by the third battery pack AP3 with position variable r. The expected attenuation curve D(r) passing through the six lithium battery cells LZ is clearly visible. The attenuation increases from the edge of the battery cell to the center M of the battery cell (where the X-ray radiation must pass through most of the metal battery cell material), and decreases again from there to the edge.

[0083] exist Figure 5 In the upper part (a), there is a third X-ray image B3 of a box BO, which is the third inspection object O3. The box BO contains a laptop computer LT, which is an example of a mobile device. The laptop computer LT contains a fourth battery pack AP4 with nine battery cells arranged side by side. Three of the battery cells are tampered with to hide organic material inside the cells. These cells are pseudo-cells FZ.

[0084] exist Figure 5 In the lower part (b), it is shown again along the passage in Figure 5 The attenuation curve D(r) of X-ray radiation from the fourth battery pack AP4 in the upper part (a) is shown by line L with position variable r, wherein line L crosses the longitudinal pseudo-cell FZ. Here, the expected attenuation curve D(r) on the pseudo-cell FZ is also clearly visible. The attenuation has a peak at the corresponding cell edge R, decreases from the corresponding cell edge R to the corresponding cell center M of the corresponding pseudo-cell FZ, and increases again from the corresponding center M of the pseudo-cell FZ to the edge R.

[0085] Based on the above Figure 3-5 The findings explained in the text indicate that, with the help of detectable anomalies, pseudo-cells FZ can be reliably and automatically detected.

[0086] Figure 6AThe simplified cross-section Q1 of the unaltered lithium battery cell LZ, the cross-section of the 2D X-ray image B4 of the lithium battery cell LZ, the intensity curve I(r) along the line L passing through the X-ray image B4, and the curve obtained by differentiating the intensity curve I(r) with respect to the position variable (r) are shown from top to bottom.

[0087] Figure 6A The untampered lithium-ion battery cell LZ has a thin metal sheath H1, while the interior is essentially uniformly filled with functional material of the lithium-ion battery cell LZ with a high metal content up to the center M. This causes the intensity value of the intensity curve I(r) to decrease towards the center M of the untampered lithium-ion battery cell LZ, starting from the right edge RR or the left edge RL. Since the attenuation of X-ray radiation is essentially determined by the thickness of the material to be penetrated, and the cross-section Q1 is circular, it becomes thicker towards the center, producing a typical "feature" in the detection of the first derivative of the intensity curve I(r) of X-ray radiation. The first derivative of the intensity curve I(r), dI(r) / dr, exhibits peaks at the left deformation point WL and the right deformation point WR of the intensity curve I(r), which are not located at the edge or peripheral region of the lithium-ion battery cell LZ. Therefore, based on the findings here, even the untampered lithium-ion battery cell LZ can be definitively detected or confirmed.

[0088] and Figure 6B compared to, Figure 6B The simplified cross-section Q2 of two tampered lithium battery cells (i.e., pseudo cells FZ) placed adjacent to each other is shown from top to bottom. Below is a cross-section of the 2D X-ray image B5 of these pseudo cells FZ, below that is the intensity curve I(r) along the line L passing through the X-ray image B5, and below that is the curve of the derivative of the intensity curve I(r) with respect to the position variable (r), dI(r) / dr.

[0089] and Figure 6A Compared to the thin metal protective layer H1, Figure 6B The tampered pseudo-unit FZ has a thicker metal sheath H2, while the interior is substantially uniformly filled up to the center M, where organic material (e.g., explosives) is concealed within the pseudo-unit FZ. The thickness of the metal sheath H2 defines the left edge region RB1 and the right edge region RB2.

[0090] As described elsewhere, the edge region is defined as the region extending from the edges RL, RR of the cell FZ toward the center M of the cell FZ, and its thickness is at most about 10% of the distance from the edges RL, RR to the center M; this definition applies to all examples.

[0091] In the case of pseudo-units, and Figure 6ACompared to the untampered lithium-ion battery cell LZ, the intensity I(r) in the right edge region RB1 and the left edge region RB2 is significantly reduced. Due to the low attenuation characteristics of organic materials, the intensity of the detected X-ray radiation is almost uniform towards the center M of the dummy cell FZ, as the total (effective) thickness of the metal sleeve H2 through which the X-ray radiation passes is essentially unchanged. This results in a significantly different signal in the first derivative of the intensity I(r) for the dummy cell FZ. Particularly important are the clearly identifiable peaks in the edge regions RB1 and RB2, which are located at the edges of the cell and therefore differ from the actual values ​​of the dummy cell FZ. Figure 6A Compared to the peak values ​​at the deformation points WL and WR, the identifiable peak value is clearly distinguishable, and the deformation point is significantly closer to the center of the untampered lithium battery cell LZ.

[0092] Figure 7 This illustrates how to detect the tampered metal object (e.g., in two-dimensional X-ray data of an inspection object O1 or O3 containing a metal article FZ) Figure 3 A flowchart of the process for the tampered metal article (FZ) of A and 5A, wherein non-metallic substances (e.g., explosives) are hidden within the tampered metal article. To this end, the method comprises the following steps.

[0093] In step S10, the region containing the metal article FZ in the X-ray data is determined. Then, in step S20, an attenuation curve I(r) across the metal article FZ is provided from the X-ray data (see...). Figure 3 A-5B). Then, in step S30, the attenuation curve I(r) is evaluated to determine whether it exhibits characteristic anomalies P1 and P2 at the edges R1 and R2 of the metal article FZ (see A-5B). Figure 3 (B, 4B, 5B). Finally, in step S40, if the attenuation curve I(r) exhibits the characteristic anomalies P1 and P2 at the edges R1 and R2 of the metal object FZ, an alarm function is triggered.

[0094] The alarm function may have at least one of the following steps:

[0095] (i) The attenuation curve I(r) and the line along which the attenuation curve I(r) is presented are superimposed on the X-ray images B1, B2, B3 of the objects O1, O2, O3 to enable the operator to make a better visual assessment.

[0096] (ii) By using an automatic control device to automatically move the objects to be inspected from the third-party entrance and transfer them to a subsequent inspection point, manual inspection of the objects O1, O2, and O3 is triggered.

[0097] (iii) At the inspection equipment 200 where the process is applied, a visual and / or auditory alarm is triggered. This also ensures the attention of the operator and other safety personnel.

[0098] Finally, it should be noted that the principles of this disclosure are not limited to detecting counterfeit units FZ containing explosives. Rather, the principles can be applied to the detection of any tampered metallic articles in which organic material is concealed. This can also involve, for example, drugs. Therefore, in the following text we will refer to counterfeit items rather than counterfeit units in a restrictive manner.

[0099] The core idea of ​​this disclosure is to apply the above findings to the detection of metallic pseudo-units in X-ray inspection methods, particularly to evaluation methods for detecting pseudo-units in transmission images (e.g., two-dimensional X-ray images) of the inspected object.

Claims

1. A method for detecting a tampered metal article (FZ), wherein a non-metallic substance is concealed within the metal article (FZ) and the metal article (FZ) has been tampered with, such that the tampered metal article (FZ) produces a two-dimensional X-ray image in two-dimensional X-ray data of an inspection object (O1, O2, O3) containing the metal article (FZ) equivalent to that of a corresponding untampered metal article, the method comprising the steps of: - (S10) Determine the region in the X-ray data that contains a metallic object (FZ); - (S20) Provide from the X-ray data a decay curve (D(r)) or intensity curve (I(r)) of the detected X-ray radiation along a line (L) passing through the metal article (FZ); - (S30) Evaluate whether the attenuation curve (D(r)) or the intensity curve (I(r)) exhibits characteristic anomalies (P1, P2) in a predetermined region, said predetermined region being at the edge (R1, R2) or within the edge region (RB1, RB2) of the metal article (FZ); and - If the attenuation curve (D(r)) or the intensity curve (I(r)) exhibits the characteristic anomaly (P1, P2) in the predetermined area, (S40) an alarm function is triggered.

2. The method according to claim 1, wherein, (S10) Determining the region in the X-ray data that contains the metallic article (FZ) includes: - The region is determined to include a metal based on whether a region in the X-ray data causes minimal attenuation of X-ray radiation and / or based on what atomic number or effective atomic number the material in the X-ray data is assigned.

3. The method according to claim 1, wherein, (S20) Providing the attenuation curve (D(r)) or intensity curve (I(r)) from the X-ray data along the line (L) passing through the metal article (FZ) includes: - Determine the longitudinal direction (LR) of the metal article (FZ); and - Establish the attenuation curve (D(r)) or intensity curve (I(r)) for the position variable (r) along the line (L) passing through the metal article (FZ).

4. The method according to claim 3, wherein, The line (L) is configured such that it extends orthogonally to the determined longitudinal direction (LR).

5. The method according to claim 3, wherein, (S30) Evaluating whether the attenuation curve (D(r)) or the corresponding intensity curve (I(r)) in the predetermined region exhibits characteristic anomalies (P1, P2) includes at least one of the following steps: - Determine whether the value of the attenuation increases along the attenuation curve (D(r)) from the edge (R1, R2) of the metal article (FZ) toward the center (M) of the metal article (FZ); - Determine whether the attenuation curve (D(r)) has a step at the edge (R1, R2) of the metal article (FZ) or in the edge region (RB1, RB2); - Determine whether the attenuation curve (D(r)) in the edge region (RB1, RB2) of the metal article (FZ) has an attenuation maximum value (Dmax1, Dmax2), and based on the attenuation maximum value, the attenuation curve (D(r)) decreases toward the center (M) of the metal article (FZ) and / or extends at a lower level compared to the edge region (RB1, RB2); - Determine the first derivative of the attenuation curve (D(r)) based on the position variable (r), and determine whether the derivative (D(r) / dr) exhibits a peak in each of the edge regions (RB1, RB2) of the metal article (FZ).

6. The method according to claim 1, wherein, (S50) Triggering the alarm function includes at least one of the following steps: - Overlay the attenuation curve (D(r)) or the intensity curve (I(r)) along the line (L) it presents onto the X-ray images (B1, B2, B3, B4, B5) of the objects under inspection (O1, O2, O3). - Trigger a manual inspection of the inspected objects (O1, O2, O3); as well as - Trigger a light and / or sound alarm on the inspection system (200) to which the method is applied.

7. The method according to claim 1, wherein, The altered metal article (FZ) is an article with a metal sleeve (H2) or metal covering.

8. The method according to claim 1, wherein, The altered metal article (FZ) is an altered battery cell or battery unit.

9. The method according to claim 8, wherein, The altered metal article (FZ) is an altered lithium battery cell (LZ).

10. An evaluation apparatus (100) for evaluating transmission images (B1, B2, B3, B23), the evaluation apparatus (100) comprising: Processing unit (110) is adapted to perform the method according to any one of claims 1 to 9; An input unit (120) is used to input operator input to control the evaluation device; as well as The output unit (130) is used to display a transmission image to the operator.

11. The evaluation device (100) according to claim 10, wherein, The output unit (130) is adapted to perform alarm functions.

12. The evaluation device (100) according to claim 10, in, The processing unit (110) has a communication interface (112) through which it can network with one or more inspection devices (200) for non-destructive testing of the inspection objects (O1, O2, O3) to perform data communication, thereby receiving transmission images (B1, B2, B3) from the one or more inspection devices (200) for visual evaluation.

13. An inspection system (300) having at least one evaluation device according to claim 10, said evaluation device being spatially separated from and connected to at least one inspection device (200; 200.1, 200.2, ..., 200.n) for data communication, wherein, Transmission images (B1, B2, B3) of the objects (O1, O2, O3) inspected at the at least one inspection device (200; 200.1, 200.2, ..., 200.n) are transmitted to the evaluation device for visual evaluation.

14. A computer program product having a computer program, the computer program including a software device, wherein if the computer program is executed on a computer, the software device is used to perform the method according to any one of claims 1 to 9.

15. A data carrier having a computer program product according to claim 14.

Images