Object projection in CT X-ray images

Abstract

A method for generating a test X-ray image, the method comprising: preparing (S10) first CT data (B) of an inspection object, second CT data (BM) for a metal part of the inspection object, and third CT data (TM) for a metal part of a target object, the target object to be inserted into the first CT data (B); transforming (S20) the first, second and third CT data (B, BM, TM) from an image space (BR) to corresponding first sinogram data (SB), second sinogram data (SBBM) in a radon space (RR); and ) and third sinogram data (STM); calculating (S30, S40) artifact sinogram data (SA) in radon space (RR), the artifact sinogram data (SA) representing a metal artifact to be added to the first CT data (B) based on a target object to be inserted; inversely transforming (S50) the artifact sinogram data (SA) from radon space (RR) to image space (BR) in CT artifact data (A) for the artifact to be inserted; and inserting the CT artifact data (A) into the first CT data (B). The computer program product has a computer program having software means for executing the generation method when the computer program is run in an automatic system. A generating device (26, 26.1, 26.2) for generating a test X-ray image, wherein the generating device (26, 26.1, 26.2) comprises: an examination object database (34) for storing generated first CT data (B) of an examination object (14); a target object database (36), the target object database (36) having fourth CT data (T) of the target object and possibly third CT data (TM) of the target object; and an image projection unit (38) for projecting the fourth CT data (T) of the target object from the target object database (36) onto the first CT data (B) from the examination object database (34) to generate a test X-ray image, wherein the image projection unit (38) is configured to perform a generating method. A central control unit (18) of at least one X-ray inspection unit (10.1, 10.2, 10.n) for non-destructive inspection of inspection objects (14), in particular items of baggage or other parcels, the central control unit (18) having a generating device (26, 26.1, 26.2), wherein the control unit (18) is preferably connected to the at least one X-ray inspection unit (10.1, 10.2, 10.n) via a communication network (42). A baggage item or parcel screening system (1) having at least one X-ray inspection unit (10.1, 10.2, 10.n) and a central control unit (18).

Description

[0001] The invention generally relates to an X-ray image of a test examination object, which is generated, for example, for an X-ray examination unit at a security checkpoint at an airport, wherein an object has been inserted into the test X-ray image. The invention particularly relates to the generation of a test X-ray image based on a three-dimensional X-ray image of the examination object generated by computed tomography or a two-dimensional X-ray image derived therefrom, and in this respect, in particular to the generation of realistic metal artifacts which can be attributed to metal parts contained in the examination object and metal parts of the object to be inserted and their interaction. Background Art

[0002] Inserting a threat object into an X-ray image is called "Threat Image Projection" (TIP), as described, for example, by V. Cutler et al. in the article "Use of Threat Image Projection (TIP) to Enhance Security Performance" presented at the 43rd International Carnahan Conference on Security Technology, October 5-8, 2009 (pp. 46-51, DOI 10.1109 / CCST.2009.5335565). An X-ray image of a real baggage item into which a false threat object is inserted (as if the false threat object really existed in the real baggage item) is called a "False Threat Image" (FTI).

[0003] DE 10 2014 109 214 A1, US 6 899 540 B1 and US 5 243 693 A give examples of computer systems for training and testing operators of X-ray inspection units for baggage screening.

[0004] The known TIP function essentially involves a 2D test X-ray image of the kind usually produced by an X-ray examination unit, which works on the same principle as a line scanner, wherein the examination object is imaged line by line using X-rays and, based on the intensity values ​​of the X-rays after passing through the examination object, the associated line of the 2D X-ray image is produced.

[0005] At the same time, X-ray inspection units whose functions are based on the principle of computed tomography (CT) are also used for non-destructive inspection of inspection objects. For example, EP 1 585 995 A1 discloses a corresponding inspection system and inspection method for luggage.

[0006] CT X-ray images of examination objects containing metal parts generally show so-called metal artifacts. Threat objects such as weapons or explosive devices generally contain metal, such as metal components, or consist mainly or entirely of metal. Therefore, CTFTI must have authentic metal artifacts, which can be caused by metal objects that may already exist in the examination object and metal contained in false threat objects. Otherwise, CT FTI is identified as FTI due to unreal or implausible metal artifacts.

[0007] Imaging in CT is based in principle on radon data detected in the detector, i.e. the detected intensity and direction of the X-ray beam associated with the detector. Each X-ray beam attenuates as it passes through the object under examination and the material contained in the beam path. In computer-aided image reconstruction, the attenuation coefficient can be determined as a material property for each volume element (voxel) of the object under examination. Based on these attenuation data present for all voxels, corresponding 3D X-ray images and / or 2D cross-sectional views can be generated.

[0008] Metallic objects have very high attenuation coefficients, whereby starting from a certain metal thickness, virtually total absorption can occur, since, for example, the useful signal detected at the detector falls below the noise threshold and can no longer be distinguished. At sharp edges between a metallic object and its surroundings, the attenuation coefficient change can be considered to be particularly high values. If the combined detector elements detect X-rays at this transition, the detected intensities are averaged and the resulting intensity values ​​are assigned to the corresponding detector elements. Since the logarithm of the linear average does not correspond to the logarithmic sum, this results in the average attenuation coefficient not matching the average intensity of the detected X-rays. Image artifacts attributable to this are called partial volume effects. In addition, the spectrum of the X-ray tube is not monofrequency. The radiation absorption of the material through which the X-rays pass is energy-dependent. Low-energy X-rays are absorbed more than high-energy X-rays. This leads to an increase in the average energy of the total spectrum. This occurs in all materials, but is more intense in metals. Image artifacts attributable to this are called hardening artifacts.

[0009] The above-mentioned effects lead to inconsistencies in the radon data detected by directing radiation through the object under examination. Since known image reconstruction by means of filtered backprojection assumes ideal conditions, the radon data detected in the reconstruction are "smeared" in the image area. Therefore, the effect attributable to the metal in the object under examination exerts an effect on the entire image. In other words, metal artifacts in CT X-ray images are essentially caused by the deviation of the above-mentioned radon data from the ideal mathematical model, and the radon data is detected for the reconstruction of the voxel data. For example, metal artifacts appear as black or white lines that radiate from the metal part and extend in particular between different metal parts.

[0010] In medical applications of computed tomography, metal artifacts are known to be unwanted image distortions. For example, T. Koehler et al. proposed a method for removing metal artifacts from CT X-ray images in "A New Method for Metal Artifact Reduction in CT" published at "The International Conference in X-ray Computed Tomography" held in Salt Lake City, Utah, USA in 2011.

[0011] It is known that the intensity and orientation of the metal artifacts that appear depends on the number of metal parts, the size of the metal parts and the position of the metal parts relative to other metal parts in the object under examination. Therefore, projecting an object with metal parts into a 3DCT X-ray image requires not only adding metal artifacts caused by the threat object alone, but also adding metal artifacts that appear due to the interaction of the metal of the object with the metal already present in the luggage item. Likewise, metal artifacts should be visible in the threat object even if the threat object does not contain metal.

[0012] For the projection of objects onto CT X-ray images, N. Megherbi et al. have proposed a method for achieving this projection. The known method is described in 3D Image Processing by Najla Megherbi et al., 2012. rd The following document “Fully automatic 3D Threat Image Projection: Application to Densely Cluttered 3D Computed Tomography Baggage Images” published at the International Conference on Image Processing Theory, Tools and Applications (IPTA) (pages 153-159) and the following document “Radon Transform Based Metal Artefacts Generation in 3D Threat Image Projection,” published by N. Megherbi et al. in Proc. SPIE Optics and Photonics for Counterterrorism, Crime Fighting and Defense, volume 8901, pages 1-7, in October 2013 (DOI: 10.1117 / 12.2028506),.

[0013] In the known method, metal artifacts in the real luggage item are caused in the CT X-ray image of the luggage item by the metal of the object and its possible interaction with metal already present in the luggage item. Basically, the metal artifacts are added by processing the sinogram of the 3D CT X-ray image of the luggage item in Radon space and subsequent back-transformation to image space; the Radon transformation of a plane (slice) of the CT data from image space is here called a "sinogram". The object itself is inserted into the luggage item in image space, because the voxels in the 3D CT X-ray image of the luggage item associated with the object to be inserted are attributed to the corresponding attenuation coefficients of the associated voxels of the object to be inserted.

[0014] Figure 1 A schematic flow chart of a known image processing method proposed by Megherbi et al. is shown.

[0015] In a first step S1, for an inspection object such as a baggage item ("bag" B), first CT data B and associated second CT data BM ("bag metal" BM) representing optional metal parts ("metal" M) already contained in the baggage item are generated. In addition, third CT data TM ("threat metal" TM) representing optional metal parts of a threat object ("threat" T) such as a weapon to be inserted into the first CT data B are generated based on fourth CT data T of the threat object. The result of step S1 is the first CT data B, the second CT data TM, the third CT data BM and the fourth CT data T.

[0016] In a second step S2, a known radon transformation is used to transform planes (slices) from the first CT data B, the second CT data BM and from the third CT data TM from the image space BR to the radon space RR. The result of step S2 is first sinogram data SB of a first sinogram of the first CT data B, second sinogram data SBM of a second sinogram of the second CT data BM and third sinogram data STM of a third sinogram of the third CT data TM.

[0017] The known "radon transform" transforms the image f(x, y) in the image space BR(x, y) to the radon space RR(α, s). In the radon space, for each angle α, the line integral of the function f(x, y) along all the lines s of the xy plane is determined, wherein for each of these lines s, the radon transform Rf corresponds to the projection of the function f(x, y) to the vertical line of these lines s. The superposition (line integral) of all these projections produces a sinogram, i.e. a 2D data set. Thus, the radon space RR is therefore only two-dimensional at first. If this principle is applied to each plane of the 3D volume of the 3D CT X-ray image, then a plurality of sinograms, i.e. a sinogram per plane, are obtained accordingly. Here, the term "radon transform" is understood to be a "forward transform" from the image space BR to the radon space RR all the time. The inverse transform of the Radon transform is called "radon inverse transform", "inverse radon transform" or "filtered reverse projection" (because this implementation basically includes two steps, i.e. filtering and transforming).

[0018] In a third step S3, the second sinogram data SBM are added to the corresponding third sinogram data STM, generating corresponding fourth sinogram data SM. The result of step S3 is fourth sinogram data SM for a fourth sinogram, which represents all metal parts, i.e. metal parts already contained in the baggage item and virtual metal parts of the threat object to be added.

[0019] In a fourth step S4, the fourth sinogram data SM are used to identify and (more or less in the reverse method of the known methods for reducing metal artifacts in medical CT systems) process locations in the respectively associated first sinogram data SB, which are affected by metal parts. In this case, the fourth sinogram data SM are used as a mask to determine whether metal is present on a specific sinogram beam in the examination object with the virtual insertion of the threat object. In this context, a sinogram beam is a line into which the intensity data of a specific point in an associated (image) plane (slice) in the image space BR flows. The first sinogram data SB are modified at the locations thus identified in a manner hardened due to metal simulation. To this end, the respective values ​​of the pixels in the first sinogram of the first sinogram data SB, which pixels have been identified as having been affected by metal, are increased by bringing the respective values ​​of the pixels closer to the maximum value. These local modifications in the first sinogram data SB lead to the expected metal artifacts appearing in the reverse transformation to the image space BR. In other words, the first sinogram data SB of the examination object are modified at all locations in the first sinogram data SB which are affected by metal in the examination object (second sinogram data SBM) or by metal in the threat object (third sinogram data STM). The result of the fourth step S4 is modified first sinogram data SB* which respectively contain the anomalies necessary for a realistic and credible metal artifact, which would be contained in the first sinogram data SB of the real first CT data based on the actually contained metal parts.

[0020] The inventors have realized that it is problematic that the first sinogram data SB of the original first CT data already contain anomalies which represent metal artifacts which may have been contained in the first CT data B. In known methods, these locations in the first sinogram data SB are also modified, whereby already existing metal artifacts are unrealistically enhanced.

[0021] In a fifth step S5, the modified fourth sinusoidal data SB* are transformed back into the image space BR by means of an inverse radon transformation. The result of step S5 is modified first CT data B* of the examination object, which should include all necessary metal artifacts, i.e. artifacts of metal parts already contained in the examination object, artifacts of metal of the threat object to be inserted, and artifacts between all these metal parts.

[0022] In a sixth step S6, the fourth CT data T are inserted into the modified first CT data B*, which do not yet contain the threat object itself. To this end, the attenuation coefficients of the voxels of the modified first CT data B* are replaced by the corresponding attenuation coefficients of the associated voxels of the fourth CT data T of the threat object to be inserted. The result of step S6 is the deviated CT result image data FTI for the virtual threat image.

[0023] In the known method, the graphic quality of the resulting image is clearly altered. The X-ray image appears blurred and less rich in detail. Furthermore, in the known method, metal artifacts that were already present in the original first CT data of the luggage item appear unnaturally enhanced. Even if there is only a slight difference in image quality compared to the real X-ray image, these metal artifacts are noticeable and, even if this deviation is relatively small, the resulting image cannot be used for testing and training operators in practice. Summary of the invention

[0024] The object of the invention is to propose a generation method and a corresponding generation device for inserting a virtual target object into an X-ray image that has been generated using a CT X-ray examination unit, wherein the above-mentioned disadvantages associated with the known methods should be at least partially avoided or at least reduced.

[0025] The purpose of the generation method and the corresponding generation device is to generate a test X-ray image of the examination object, into which a virtual target object has been inserted and is planned to be found by an operator, for example for the purpose of training and / or performance review. Preferably, the test X-ray image is obtained or generated based on CT data generated by CT in the form of a 3D image or a 2D image of the examination object.

[0026] The generated test X-ray images should preferably comprise realistic and plausible metal artifacts attributable to the [sic] metal parts contained in the examination object and to the metal parts of the virtual target object to be inserted and their mutual influence.

[0027] Preferably, metal artifacts already contained in the X-ray image should not be unrealistically enhanced.

[0028] Furthermore, the influence on the image quality of the test X-ray images generated by the generation method should be as imperceptible as possible.

[0029] The above objects are achieved by the features of the independent claims. Further exemplary embodiments and advantageous modifications are defined in the respective subsequent dependent claims. The features and details described with respect to the production method according to the invention are also essentially relevant to the production device according to the invention, and vice versa. In order to avoid repetition, the disclosures of the respective aspects are therefore mutually referenced.

[0030] About At the beginning About Figure 1 Having described the known methods and their disadvantages, the present invention proposes an improved generation method and a corresponding generation device for generating test X-ray images.

[0031] The new method avoids an unrealistic intensification of metal artifacts already present in the first CT data B and a reduction in the image sharpness of the X-ray image obtained from the first CT data B. The term "image sharpness" is understood here as a measure of the local contrast, i.e. how pronounced the difference in value of the attenuation coefficients of adjacent voxels is. When the CT data are transformed from the image space into a sinusoid in the radon space by means of a radon transformation and from there back again into the X-ray image obtained from the CT data, this leads to a reduction in the local contrast (image sharpness). In other words, when compared with the original image, the resulting image looks blurred and the detail richness is compromised, thus less sharp. This is caused, for example, by the fact that the material property values ​​of adjacent voxels (e.g. attenuation coefficient and / or material density value and / or atomic number) have been brought closer in value by these two transformations. This reduces the local contrast. The reduction in local contrast is noticeable as a significant reduction in image sharpness; the image looks "washed out" or "blurred". In contrast to known methods, in which the artifacts to be added are added to the first sinogram data SB of a first sinogram of the examination object, in the new method proposed here only the actually missing artifact sinogram data (which represent the (metal) artifacts to be added) are calculated in the radon space RR and, after back-transformation into the image space BR, are inserted into the original first CT data B.

[0032] In order to achieve this aim, the inventors have proposed to generate new fourth sinogram data SM** (corresponding to the fourth sinogram) which essentially corresponds to the logical subtraction of the third sinogram data STM (corresponding to the third sinogram) minus the corresponding second sinogram data SBM (corresponding to the second sinogram). As a result, the new fourth sinogram data SM** contain only those points for which the corresponding point in the associated third sinogram is not equal to zero (STM<>0) and the corresponding point in the associated second sinogram is equal to zero (SBM==0). By means of the new fourth sinogram data SM**, points in the first sinogram data SB (corresponding to the first sinogram) are identified in order to be able to change / process the intensity values ​​of these points, whereby the effect of metal virtually inserted into the examination object is simulated in the first sinogram data SB. As a result, this avoids a situation in which artifact data already contained in the first sinogram data SB and representing metal artifacts already present in the first X-ray image B are unrealistically enhanced when the method continues.

[0033] In the new method, instead of modified first sinogram data SB*, only new artifact sinogram data SA are generated, which represent only the metal artifacts that are actually added and that are present due to the metal of the target object and the interaction with any metal present. This is achieved in that the new fourth sinogram data SM** are used more or less as a mask to identify those pixels in the corresponding first sinogram data SB that belong to the corresponding artifact sinogram data SA. All other pixels in the first sinogram data SB do not flow to the artifact sinogram data SA to be calculated. The intensity values ​​in the pixels thus identified that are present in the first sinogram data SB are processed according to an empirically determined function, whereby an additional hardening of the metal of the target object by the virtually added is simulated for an X-ray passing through the examination object.

[0034] The artifact sinogram data SA calculated in the radon space RR for the metal artifacts to be added are thus transformed into the image space BR and can then be inserted into the original (i.e. unchanged) first CT data B. In other words, using the new method, not all first sinogram data SB* with artifacts that have been added by processing in the radon space are transformed back into the image space BR. The original first CT data are exclusively processed in the image space by adding the CT artifact data A of the target object and the fourth CT data T. Thus, the image clarity of the 3D and / or 2D X-ray image obtained from the modified first CT data B remains unchanged relative to the X-ray image obtained from the original unchanged first CT data B.

[0035] A specific modification of the method involves inserting fourth CT data T (which represent the virtual target object) and CT artifact data A (which represent the metal artifacts to be added) into the first CT data B of the examination object.

[0036] Preferably, the fourth CT data T of the target object is generated in advance. For this purpose, the target object is fixed by a low-density support material (e.g., polystyrene) and its fourth CT data T is detected by a CT X-ray inspection unit. To this end, the support material can be placed under the target object or can evenly and completely wrap the target object. The obtained fourth CT data T of the target object can be saved in a database for subsequent use.

[0037] In this modification, the fourth CT data T of the target object to be inserted and / or the authentic CT artifact data A calculated for this purpose are directly used without much labor to release the target object, and are inserted into the first CT data B of the examination object. In this case, it is not that the material property values ​​of the voxels in the first CT data B are replaced by the corresponding material property values ​​in the fourth CT data T of the target object or in the CT artifact data A to be added, but rather, a "mixing" or "weighted combination" of these two material property values ​​is performed in the transition region according to the corresponding material property values ​​of the voxels to be added. For example, 20% of one value and 80% of the other value can flow into the new value; in practice, other weightings are also possible.

[0038] For example, in one embodiment, by means of a suitable predetermined threshold in the CT artifact data A of the metal artifact to be added and / or in the fourth CT data T, a distinction is made between a first voxel whose material property value is below the predetermined threshold, a second voxel whose material property value is within a predetermined range above the predetermined threshold, and a third voxel whose material property value is above the predetermined range. The first voxel is interpreted as not belonging to the target object or not being an artifact, for example, as being part of the background. In the fourth CT data T, for example, this is support material. In other words, the first voxel is ignored. When inserted into the first CT data B, the second voxel whose material property value is within the predetermined range is "blended" (weighted combination) with the material property value of the corresponding voxel in the first CT data B, respectively. As for the third voxel, when inserted into the CT data B of the examination object, the material property value of the locally corresponding voxel is replaced by the material property value of the voxel belonging to the third voxel.

[0039] The above classification also enables the release of metal parts in the inspection object or metal of the target object to be inserted, which is suitable for the above-described method for generating artifacts. In other words, by means of a predetermined threshold, the new method is able to automatically detect metal parts of the target object and metal parts of the inspection object or artifacts to be added, and is able to distinguish them from non-metal parts.

[0040] A first aspect of the invention relates to a method for producing a test X-ray image. In principle, the method comprises:

[0041] preparing first computed tomography (CT) data B representing an inspection object, second CT data BM representing a metal portion contained in the inspection object, and third CT data TM representing a metal portion of a target object to be inserted into the first CT data B;

[0042] transforming the first, second and third CT data B, BM, TM from the image space into corresponding first sinusoidal data SB, second sinusoidal data SBM and third sinusoidal data TM in the radon space;

[0043] calculating artifact sinogram data SA in the radon space, the artifact sinogram data (SA) representing metal artifacts to be added to the first CT data B based on a target object to be inserted;

[0044] inversely transforming the artifact sinusoidal data SA from the radon space to image space in CT artifact data A representing the artifact to be inserted; and

[0045] The CT artifact data A is inserted into the first CT data B.

[0046] The examination subject and the target object are preferably scanned using a CT X-ray examination unit.

[0047] Preferably, the generating method further comprises: inserting fourth CT data T representing the target object into the first CT data B to obtain CT test image data FTI.

[0048] Preferably, the generating method further comprises: obtaining a CT X-ray image and / or a 2D X-ray image of the object under examination from the CT test image data FTI. The 3D CT X-ray image and / or the 2D X-ray image of the object under examination can be used as a test X-ray image, for example for training and / or testing an operator of an X-ray examination unit.

[0049] It should be noted that the generation method can basically be implemented to generate a specific 2D X-ray image with a virtual target object inserted. To this end, it is sufficient in principle to perform the method only for the affected image planes (i.e. specific slices) in the first CT data B and the second, third and fourth CT data BM, TM and T respectively.

[0050] One possible representation of the associated CT data is a three-dimensional (3D) computed tomography (CT) X-ray image or a two-dimensional (2D) X-ray image.

[0051] CT data can be obtained by CT of a scanned object (e.g., an inspection object and a target object to be inserted). During the CT scanning process, the first radon data of the object is detected. The radon data (or sinogram data) can be transformed by filtering the CT data back-projected to the object. In other words, the CT data is essentially a 3D data set, in which each voxel of the object is associated with a material property of the material at the voxel position.

[0052] The sinogram in Radon space is essentially a possible representation of radon data, which includes intensity data detected during the object is scanned by CT. In principle, in CT, the object is penetrated by X-rays generated by an X-ray point source in a plane (slice) from multiple projection directions. In this regard, for each projection direction in the plane, an associated intensity distribution is obtained on the relative detector behind the object, which corresponds to the perspective projection of the object. If the X-ray source and the detector rotate around the object, this correspondingly produces a corresponding associated intensity distribution of multiple projection directions. If the intensity data detected in this way is plotted in the form of a Cartesian coordinate system in the projection direction and the detector width, this produces a sinogram as a possible graphical representation of the detected radon data (sinogram data).

[0053] For each pixel (p) of the associated SA, a modified intensity value is calculated according to the following rule:

[0054] (i) SA(p) = (qμ - q SB(p)) If STM(p) ≠ 0 and SBM(p) = 0 and

[0055] (ii) SA(p) = 0 Others,

[0056] Wherein q is a predetermined value between 0 and 1, wherein μ is the maximum value in the associated second sinogram data. The factor q can be a value between 0.02 and 0.06. Preferably, the factor q is a value between 0.03 and 0.04. Particularly preferably, the factor q is equal to 0.04.

[0057] Inserting the CT artifact data A and / or the fourth CT data T into the first CT data B can include: mixing or weighted combining the material property values ​​associated with specific voxels in the first CT data B according to the respective material property values ​​associated with the corresponding voxels of the CT data A, T to be added, wherein the corresponding spatial coordinates are associated with the corresponding voxels and the specific voxels.

[0058] Preferably, inserting the CT data A, T into the first CT data B comprises:

[0059] classifying, by a predetermined threshold in the CT data (T, A) to be added, a first voxel having a material property value below the predetermined threshold, a second voxel having a material property value in a predetermined range above the predetermined threshold, and a third voxel having a material property value above the predetermined threshold;

[0060] Ignore the CT data of the first voxel;

[0061] blending a material property value associated with a particular voxel of the first CT data with a material property value of a corresponding second voxel of the CT data to be interpolated; and

[0062] replacing a material property value associated with a particular voxel of the first CT data with a material property value associated with a corresponding voxel of the CT data to be added;

[0063] In the mixing step and the replacing step, the corresponding spatial coordinates are associated with the corresponding voxels and the specific voxels.

[0064] Preferably, the CT data B comprises: a value for a material property; and coordinates for determining a volume element (voxel) (position) associated with at least one value for the material property. The material property associated with the respective voxel is preferably at least one of: an absorption coefficient, a material density value, and an atomic number. The coordinates for identifying the voxel are preferably three-dimensional (3D) coordinates that identify the position of the voxel in space.

[0065] A second aspect of the invention relates to a computer program product with a computer program having software means for performing the method according to the first aspect of the invention when the computer program is run in an automatic system, such as a computer system.

[0066] A third aspect of the present invention relates to a generating device for generating a test X-ray image, the generating device comprising: an examination object database for storing generated first CT data B of the examination object; a target object database having fourth CT data T of the target object and possibly third CT data TM of the target object; and an image projection unit for projecting the target object from the target object database onto the X-ray image from the examination object database to generate the test X-ray image, wherein the image projection unit is configured to perform the method according to the first aspect of the present invention. The generating device preferably has an automatic system, such as a computer system, for performing the method according to the first aspect.

[0067] The generating device can further comprise (sic) an updating unit for the examination object data. The updating unit can be configured to update the examination object database because first CT data B older than a predetermined time interval and / or already used a certain number of times (e.g. once) to generate the CT test image data FTI are deleted from the examination object database or are prohibited from further use. The image generating unit can further be configured to store the test X-ray image in the test X-ray image database. The updating unit and / or the image projection unit can be configured to update the CT test image data FTI in the database because CT test image data FTI older than a second time interval and / or already used a certain number of times (e.g. once) for training or testing an operator are deleted or are prohibited from further use.

[0068] The generating device can be part of the X-ray examination unit or can be communicatively and / or operatively connected to the X-ray examination unit to enable "on-site" training and / or testing of operators. Preferably, the CT test image data of the examination object is generated using the X-ray examination unit at the corresponding screening point.

[0069] A fourth aspect of the invention relates to a central control unit of at least one X-ray inspection unit for non-destructive inspection of objects, in particular items of luggage or other packages, the central control unit having a generating device according to the third aspect, wherein the control unit is preferably connected to the at least one X-ray inspection unit via a communication network.

[0070] The central control unit preferably has at least one display unit for displaying an X-ray image of the current examination object for visual inspection by an operator or for displaying a test X-ray image. Preferably, the central control unit is positioned relative to the at least one X-ray examination unit such that the examination object currently being screened using the at least one X-ray examination unit is not visible to the operator.

[0071] A fifth aspect of the invention relates to an X-ray inspection unit for non-destructive inspection of inspection objects, in particular items of luggage or other packages, having or being operatively and communicatively connected to a central control unit according to the fourth aspect.

[0072] Based on the X-ray image that the X-ray inspection unit has generated, the operator can find the target object among the inspection objects. For example, the X-ray inspection unit can be a CT X-ray inspection unit of the type used, for example, at a security inspection point (e.g. at the entrance of a security inspection area or in a baggage handling system, such as an airport) for manually or automatically performing non-destructive inspections of luggage items and / or cargo loaded onto an aircraft. The X-ray inspection unit can also be used at other inspection points, such as at the entrance of a relevant security area or building, at a border inspection station, etc., for inspecting objects, such as handbags carried by a person or mailing parcels, such as letters, parcels and small parcels.

[0073] The purpose of the inspection is to find a certain target object. The target object can be an object or substance with potential hazards, such as weapons, explosives, chemicals, etc., i.e., a dangerous object. The target object can also be an object or substance that does not initially bring direct danger. The target object can, for example, be a data carrier (such as a DVD or CD-ROM), smuggled goods, money, drugs, etc. The target object can also be an object that is basically harmless but is classified as a dangerous object for some reason, such as a lithium-ion battery. The target object can be food that is prohibited from import. Basically, in order to train operators, any object can be defined as a target object. Therefore, the target object is basically understood as an object or substance that the operator wants to detect in the X-ray image of the inspection object. For these reasons, the corresponding CT image data of the target object can be generated and stored in the target object database.

[0074] Using the method proposed here, real CT test image data FTI can be generated and, during operation, can be used for testing operators of X-ray examination units or for training operators. Due to the above-described novel generation of CT test image data FTI, these data contain realistic and authentic metal artifacts, just as they would appear in a real CT X-ray image of an examination object actually containing the target object. Furthermore, the CT FTI generated using the proposed method does not contain any changes in image sharpness and for this reason it cannot be identified as FTI.

[0075] A sixth aspect of the invention relates to a baggage or parcel screening system having at least one X-ray inspection unit and a central control unit according to the fourth aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Other advantages, features and details of the invention are obtained from the following description of the exemplary embodiments described in detail with reference to the accompanying drawings. Each of the features mentioned in the claims and / or the description may be necessary in itself or in any combination. Likewise, the above-mentioned features and the features explained in more detail below can be used in any combination of one or more. Some parts or components of similar or identical functions are provided with the same reference numerals. The terms "left", "right", "above", "below" used in the description of the exemplary embodiments refer to the drawings with generally clear descriptions of the drawings and the generally clear orientation of the reference numerals. The exemplary embodiments shown and described are not to be understood as exclusive, but are schematic in nature for explanatory purposes. The detailed description is used to provide information to those skilled in the art; for this reason, known circuits, structures and methods are not described or explained in detail in the description.

[0077] Figure 1 A schematic flow chart showing a known method for inserting a virtual threat object into a CT X-ray image;

[0078] Figure 2 A schematic flow chart showing an exemplary embodiment of a method for inserting a virtual target object into a CT X-ray image according to the present invention;

[0079] Figure 3 An automated baggage or parcel screening system schematically depicted in block diagram form is shown;

[0080] Figure 4 A schematic embodiment of a device for generating a 3D CT test X-ray image is shown;

[0081] Figure 5A shows an original 3D CT X-ray image of a baggage item as an example of an inspection object with some metal artifacts caused by metal parts in the baggage item;

[0082] Figure 5B Shows Figure 5A A 3D CT X-ray image, a threat object serving as a target object and the metal artifacts generated according to the present invention are inserted into the image. DETAILED DESCRIPTION

[0083] Figure 2 A schematic flow chart showing an exemplary embodiment of the novel method proposed herein for inserting a virtual target object into a CT X-ray image is shown.

[0084] In a first step S10, a first CT X-ray image of an object under examination, i.e., first CT data B of the object under examination (e.g., a luggage item) is generated. Based on the CT data B of the object under examination, an associated second CT X-ray image (i.e., second CT data BM) with optional metal parts, which may have been contained in the luggage item, is generated. In addition, a third CT X-ray image (i.e., third CT data TM) with optional metal parts of the target object is generated, which is to be inserted into the first CT X-ray image. The third CT X-ray image (i.e., third CT data TM) is generated based on a fourth CT X-ray image (i.e., fourth CT data T) of the target object.

[0085] Therefore, the result of the first step S10 is the first CT data B, the second CT data BM, the third CT data TM, and the fourth CT data T.

[0086] The fourth and / or third CT data TM, T of the target object can be generated in advance and stored in the target object database for further use (eg, referring to Figure 4 34 in the figure).

[0087] In the second step S20, the first CT data B, the second CT data BM and the third CT data TM are transformed from the image space BR to the radon space RR by using a known radon transformation, thereby obtaining corresponding first sinogram data SB of the first CT data B, second sinogram data SBM of the second CT data BM and third sinogram data STM of the third CT data TM.

[0088] In a third step S30, corresponding new fourth sinogram data SM** are generated based on the second sinogram data SBM and the respectively corresponding third sinogram data STM. The new fourth sinogram essentially corresponds to a combination in the sense of a logical difference to determine an exclusive point of the third sinogram data STM. The new fourth sinogram only contains pixels for which the corresponding pixel in the associated third sinogram (exclusively) is not equal to zero (i.e. STM<>0) and at the same time the corresponding pixel in the associated second sinogram is equal to zero (i.e. SBM==0). This avoids the situation where artifact data (=SBM) representing metal artifacts already present in the first 3D CT X-ray image are unrealistically reinforced when the method continues. The result of the third step S30 of the new method is new fourth sinogram data SM**, which exclusively represent the metal part of the threat object to be added.

[0089] In a fourth step S40, the new fourth sinogram data SM** are used to identify corresponding pixels of the corresponding first sinogram SB as locations affected by the metal parts to be added. The corresponding intensity data present at these identified locations are processed to generate a new artifact sinogram. Thus, the new method does not generate a modified first sinogram, but only new artifact sinogram data SA, which only represent the metal artifacts A that are actually to be added due to the metal of the target object and due to the interaction with the metal that may be present, the other intensity data of the corresponding first sinogram data SB not flowing into the associated new artifact sinogram. This is achieved because these points in the first sinogram are identified and flow into the new artifact sinogram, while all other locations of the first sinogram do not flow into the artifact sinogram, in case the new fourth sinogram data SM** are used as a mask. For all identified locations in the first sinogram data SB, the metal-wise hardening by the intensity present at these locations in the first sinogram is calculated for the X-rays passing through the object under examination.

[0090] Steps S30 and S40 are preferably performed together. For example, in order to generate an artifact sinusoid SA for each pixel p, the intensity value SA(p) can be calculated according to the following rule according to the present invention:

[0091] (i) SA(p) = (qμ - qSB(p)) if STM(p) ≠ 0 and SBM(p) = 0 and

[0092] (ii) SA(p) = 0, other

[0093] Wherein q is a predetermined value between 0 and 1, and μ=max(SBM). Rule (i) identifies and modifies those pixels of the corresponding first sinogram that can be associated with metal parts. Rule (ii) ensures that pixels from the first sinogram that are not associated with the metal to be virtually added do not flow into the new artifact sinogram. The q value can be a value greater than or equal to 0.02 and less than or equal to 0.06. Preferably, the q value is a value greater than or equal to 0.02 and less than or equal to 0.06. In a preferred embodiment, the q value is equal to 0.04.

[0094] In an advantageous manner, the method according to the invention avoids the influence of metal artifacts already contained in the first 3D CT X-ray image B and caused by metal parts already contained in the first 3D CT X-ray image B. For this purpose, in a second modification step, the method according to the invention only generates metal artifacts which are additionally due to the metal in the target object and the interaction of this metal with the already existing metal parts.

[0095] The result of the fourth step S40 is therefore new artifact sinogram data SA which respectively contain the necessary anomalies for the metal artifacts to be added and which, based on the actually contained metal parts, are also contained in the first sinogram data SB of the real CT data.

[0096] In a fifth step S50 , the artifact sinusoidal data SA calculated in the radon space RR for the artifact to be added are inversely transformed into the image space BR, whereby the CT data A of the artifact to be added are contained in the image space BR.

[0097] In a sixth step S60, the CT data A with the artifacts to be added and the fourth CT data T of the target object to be added are inserted into the original first CT data B. The result of the sixth step S60 is the CT test image data FTI of the examination object into which the virtual target object is inserted. Based on the CT test image data FTI, a corresponding 3D CT X-ray image or a corresponding 2D X-ray image of the examination object can be obtained for further use. The obtained 3D CT X-ray image or 2D X-ray image is significantly improved in terms of the credible and realistic performance of metal artifacts and in terms of image clarity compared to those obtained using known methods.

[0098] Figure 3An automatic baggage or parcel screening system 1 (short: screening system 1) of the type used, for example, in a Class 2 baggage item or parcel at an airport is shown. In that environment, baggage items or parcels, for example brought by a third party or a passenger and to be loaded into the cargo hold of an aircraft, are checked as detection objects. It should be noted that the threat objects (TIP) described here in connection with image projection are used in substantially all embodiments of the X-ray inspection unit.

[0099] The screening system 1 comprises n CT X-ray inspection units 10.1, 10.2 and 10.n known per se, each schematically illustrated by a rectangle. On conveyor belts 12.1, 12.2 and 12.n associated with the respective CT X-ray inspection unit 10.1, 10.2 and 10.n, the baggage items or parcels are each transported as inspection objects 14 through one of the CT X-ray inspection units for automatic inspection. The CT X-ray inspection units 10.1, 10.2 and 10.n operate according to known CT principles and do not need to be explained in more detail in this context.

[0100] All CT X-ray examination units 10.1, 10.2, 10.n of the screening system 1 are networked by means of a data network 42 via corresponding data connections 16.1, 16.2 and 16.n, which has at least one central control unit 18 which is spatially remote from the individual X-ray examination units 10.1, 10.2 and 10.n. The central control unit 18 itself is connected to the data network 42 via corresponding data connections 17.

[0101] The central control unit 18 is an operator workstation. A display unit 22 is provided for visual inspection of the X-ray images of the inspection object. For example, if one of the X-ray inspection units 10.1, 10.2, 10.n cannot automatically decide that the inspection object 14 is harmless, one or more X-ray images of the inspection object in question are displayed to the operator on the display unit 22. In other words, if the inspection procedure implemented in the X-ray inspection unit 10.1, 10.2, 10.n based on one or more X-ray images of the current inspection object 14 cannot establish the required degree of certainty that no target objects (e.g. dangerous goods) are contained in the inspection object 14, then it is the operator's responsibility to have to carry out a visual inspection via the display unit 22. The operator then decides, for example, whether the inspection object in question has to be opened and subjected to manual inspection.

[0102] The control unit 18 has an input device 24 (e.g. a keyboard and / or separate input keys) and control elements (e.g. a joystick, a mouse, a trackball, etc.). It is thus possible for the operator to carry out control interventions etc. in a customary and customary manner. The central control unit 18 also accommodates the hardware and software of a conventional and thus known data processing system 20 (computer system) required for the functions described here. In other words, all functions described below can be implemented essentially by means of a data processing system 20 known per se in one computing unit or distributed over a plurality of computing units. The data processing system 20 is then essentially configured, i.e. programmed in a customary manner, to carry out the described functions and methods or parts thereof.

[0103] The data processing system 20 generally comprises (not shown): one or more processors, serving as central computing units, and internal and / or external memories, which are non-volatile for the software and are provided in the form of RAM (Random Access Memory) for working data. For interaction with the functional components of the device, the data processing system 20 is connected via a communication interface to, for example, a data network 42, which comprises the respective CT X-ray examination units 10.1, 10.2, 10.n. As input / output interfaces for interaction with an operator, one (or more) screens 22 serve as one (or more) display units, and a keyboard 24 and / or a mouse 25 serve as one or more input units.

[0104] It should also be noted that the functions and methods described above can be implemented completely by means of a computer program or completely on a hardware level and in any hybrid form between hardware and software. For image processing sequences, for example when projecting an image of a threat object (TIP) onto an X-ray image of the object under examination, all or some of the method steps can also be implemented by dedicated image processing hardware (e.g. a graphics processing unit (GPU)) with corresponding programming commands.

[0105] In order to generate test X-ray images for training or testing operators, a generating device 26 for CT test X-ray images is provided, which is arranged to generate CT test X-ray images in FTI form, preferably based on an examination object that has been examined recently on site, and to display the images for training and / or testing operators. Figure 2 The CT test X-ray images produced by the explained TIP method can be called pseudo-FTI.

[0106] Figure 4 An exemplary embodiment of a generation unit 26 with an updating unit 32 is shown, which is connected to the data processing system 20 of the control unit 18 via a data line 28. Via the data line 28, the updating unit 32 is sent updated [sic] during operation of the CT X-ray examination units 10.1, 10.2, 10.n of the screening system 1.

[0107] The updating unit 32 can be arranged for storing CT X-ray images of examination objects which have been examined on site in the screening system 1 to the examination object database 34. The updating unit 32 can also be arranged for updating the examination object database 34, whereby the stored X-ray images are no older than a predetermined second time interval, and / or the stored X-ray images are deleted or prohibited from further use once they have been used a predetermined number of times (e.g. once) for generating a CT test X-ray image. This has the advantage that only content which is authentic for the respective time period is contained in the examination object, which is used as the basis for the FTI; for example, a suitcase containing objects which are normally used in winter in the middle of summer would be noticeable for this reason.

[0108] The generation unit 26 is also provided with a target object database 36, in which a library of fourth CT data T (and possibly third CT data M) of target objects is stored. The target object database 36 stores image data of those target objects within the examination object which should be found by the screening system 1, in particular by an operator. The target objects may be the objects already explained above.

[0109] The image projection unit 38 is configured to project the virtual target object from the target object database 36 onto the first CT data B of the examination object database 34 to generate the CT test image data FTI. The test X-ray image generated in this way can be provided by the test X-ray image generation unit 26 to the data processing system 20 of the control unit 18 via the data line 30.

[0110] The data processing system 20 is configured to display 3D test X-ray images or 2D test X-ray images to an operator during operation in accordance with applicable legal provisions (e.g. no notification or alert required) to test the training level, detection ability and possibly the operator's attention.

[0111] With regard to system capabilities related to the projection of images of threat objects and / or contraband, in order to avoid duplication, reference is made to Regulation (EU) no. 185 / 2010 of 4 March 2010 laying down detailed measures establishing common basic standards for aviation security, in particular Articles 11.4 and 12.5. For completeness, it should be noted that TIP is also a requirement of corresponding regulations outside the European Community.

[0112] The generation unit 26 can also be provided with a test X-ray database 40, in which the test X-ray images generated by the image projection unit 38 can be stored. The image projection unit 38 can then additionally be provided for managing the test X-ray images FTI in the test X-ray image database 40, whereby test X-ray images which are older than a predetermined time interval and / or have been used a certain number of times (e.g. once) are deleted or prohibited for further use. Thus, a large number of test X-ray images in the test X-ray image database 40 can be generated in advance during operation. The test X-ray image database 40 can be updated in a similar manner as described with respect to the examination object database 34. Thereby, management measures can be implemented so that when testing an operator, test X-ray images which are already known to the operator are not used.

[0113] The above system can generate test X-ray images in the form of virtual threat images (FTI) by using actual CT data of real inspection objects that have been inspected on site. The test X-ray image essentially corresponds to an X-ray image of a real baggage item with a real threat object inserted in advance; such a test image is called a so-called combined threat image (CTI). In this way, it is possible to advantageously avoid the disadvantages of the FTI and CTI concepts; in particular, the ongoing operation of the screening system 1 will not be adversely affected. As a result, the operator is not accustomed to a specific choice of test X-ray images. Due to the continuous updating of the test X-ray image database 40, the operator is unlikely to be faced with a test X-ray image several times. If the generation of the test X-ray image only uses X-ray images from inspection objects that have been inspected on site, this ensures that baggage items or packages are used, which are typical for the screening point and are currently prevalent at the screening point (for example, the time of year). In essence, it is ensured that possible distinguishing features of the detection object that are typical for the screening point also appear in the test X-ray image. As a result, the test X-ray image cannot appear to the operator to be "suspicious" due to consciously or unconsciously perceptible deviations from the X-ray image of a specific part.

[0114] Since the new generation method for the test X-ray image in radon space RR exclusively generates artifact sinogram data SA which only represent metal artifacts A which have not yet been added to the first CT data B, the distinguishing features of the test X-ray image generated with the new generation method discussed above in connection with the known method are avoided. With the new method, in radon space, only the sinogram beams representing metal artifacts not contained in the first CT data B are modified, i.e. these metal artifacts due to the metal of the target object and the interaction of the metal of the target object with metal parts already contained in the examination object. This avoids the situation where metal artifacts which may already be present in the first CT data B are additionally enhanced. Furthermore, only the new artifact sinogram data SA are transformed back into the image space by the radon inverse transformation. The CT artifact data A which are thus present in the image space are inserted into the first CT data B of the examination object in the image space. In other words, with the new method, only the new artifact A is calculated and added to the image space, whereas in the known method, the artifact in the radon space is added to the first sinogram data SB and the entire modified first X-ray image has to be calculated due to the inverse transformation to the image area. In other words, in the novel method, it is not necessary to transform the arbitrarily modified first sinogram data SB* from the radon space back into the image space.

[0115] Using the new method, the fourth CT data of the target object and the calculated CT artifact data A are finally inserted into the original first CT data B in the image space, and new virtual CT test image data FTI** are obtained, from which 3D and / or 2D X-ray images of the inspection object can be obtained as needed.

[0116] The novel method enables a new virtual test X-ray image with undistorted already existing metal artifacts and metal artifacts to be added and an inserted target object to have the same image sharpness as the original X-ray image based on the original first CT data B.

[0117] Figure 5A An original 2D X-ray image 500 of a baggage item as an example of an inspection object having metal artifacts 520 caused by metal parts 510, 512, 514 included in the baggage item is shown. The metal artifacts 520 are substantially caused by the metal parts 510 and appear in the form of rays extending from the metal parts 510.

[0118] Figure 5B Shown from Figure 5AAn X-ray image in which a threat object with a metal part / component 516 has been inserted as a target object and its metal artifacts 522, 524, 526 generated according to the present invention. Due to the virtually inserted metal part 516, the metal artifact 522 caused only by the metal part 516 and the metal artifacts 524 and 526 due to the interaction of the already existing metal parts 510, 512, 514 with the virtually inserted metal part 516 must also be inserted into the X-ray image so that they are authentic. The present invention achieves all of the above.

Claims

1. A method for generating a test X-ray image, the method comprising: preparing (S10) first computed tomography (CT) data (B) representing an inspection object, second CT data (BM) representing a metal portion contained in the inspection object, and third CT data (TM) representing a metal portion of a target object to be virtually inserted into the first CT data (B); transforming (S20) the first CT data, the second CT data and the third CT data (B, BM, TM) from the image space (BR) to corresponding first sinusoidal data (SB), second sinusoidal data (SBM) and third sinusoidal data (STM) in the radon space (RR); Based on the second sinogram data (SBM) and the third sinogram data (STM), calculating fourth sinogram data (SM**) in the radon space (RR), the fourth sinogram data (SM**) corresponding to the logical subtraction of the third sinogram data (STM) minus the corresponding second sinogram data (SBM); calculating (S30, S40) in said radon space (RR) artifact sinogram data (SA) representing metal artifacts to be added to said first CT data (B) based on said target object to be inserted and also based on said fourth sinogram data (SM**); inversely transforming (S50) said artifact sinogram data (SA) from said radon space (RR) to said image space (BR) in CT artifact data (A) representing said artifact to be inserted; as well as The CT artifact data (A) is inserted into the first CT data (B).

2. The method according to claim 1, further comprising: Fourth CT data (T) representing the target object is inserted into the first CT data (B) to obtain CT test image data (FTI).

3. The method according to claim 2, further comprising: A 3D X-ray image and / or a 2D X-ray image of the examination object is acquired from the CT test image data (FTI).

4. The method according to claim 1 , wherein for each pixel p in the associated artifact sinusoidal map, the intensity value SA(p) is calculated according to the following rule: (i) SA(p) = (qμ - q SB(p)) if STM(p) ≠ 0 and SBM(p) = 0 and (ii) SA(p) = 0, other Wherein q is a predetermined value between 0 and 1, and μ is the maximum value in the associated second sinusoidal data (SBM). The production method according to claim 4 , wherein q is a value between 0.02 and 0.

06. The production method according to claim 4 , wherein q is a value between 0.03 and 0.

04. The production method according to claim 4 , wherein q is 0.

04.

8. The generating method according to any one of claims 1 to 3, wherein inserting the CT artifact data or the fourth CT data (A, T) into the first CT data (B) comprises: Material property values ​​associated with a specific voxel in the first CT data (B) are weighted combined according to corresponding material property values ​​associated with the voxel corresponding to the CT data to be added, wherein corresponding spatial coordinates are associated with the corresponding voxel and the specific voxel.

9. The generating method according to claim 8, wherein inserting the CT artifact data or the fourth CT data (A, T) into the first CT data (B) comprises: classifying, by a predetermined threshold in the fourth CT data or the CT artifact data (T, A) to be added, a first voxel having a material property value lower than the predetermined threshold, a second voxel having a material property value within a predetermined range higher than the predetermined threshold, and a third voxel having a material property value higher than the predetermined threshold; Ignoring the CT data of the first voxel; weighted combination of the material property value associated with a particular voxel of the first CT data (B) and the material property value of a corresponding second voxel of the CT data to be interpolated; as well as replacing the material property value associated with a specific voxel of the first CT data (B) with a material property value associated with a corresponding voxel of the CT data to be added; Wherein in the mixing step and the replacing step, corresponding spatial coordinates are associated with the corresponding voxel and the specific voxel.

10. The generating method according to any one of claims 1 to 3, wherein the first CT data, the second CT data, the third CT data, the fourth CT data and the CT artifact data (B, BM, TM, T, A) at least include: A value for a material property; and three-dimensional (3D) coordinates for determining a position of the voxel, the position being associated with the at least one value for the material property. The method of producing according to claim 10 , wherein the material property comprises at least one of the following: an absorption coefficient, a material density value, and an atomic number.

12. A computer program product comprising a computer program having software means for carrying out the method according to any one of claims 1 to 11 when the computer program is run in an automated system.

13. A generating device (26, 26.2) for generating a test X-ray image, wherein the generating device (26, 26.2) comprises: An examination object database (34) for storing generated first CT data (B) of an examination object (14); a target object database (36) having fourth CT data (T) of the target object and possibly third CT data (TM) of the target object; and an image projection unit (38) for projecting the fourth CT data (T) of the target object from the target object database (36) onto the first CT data (B) from the examination object database (34) to generate test X-ray image data (FTI), wherein the image projection unit (38) is configured to perform a generation method according to any one of claims 1 to 11.

14. The production device (26, 26.2) according to claim 13, wherein the production device has an automatic system for carrying out the production method.

15. A central control unit (18) of at least one X-ray inspection unit (10.1, 10.2, 10.n) for non-destructive inspection of an inspection object (14), the central control unit (18) having a generating device (26, 26.2) according to one of claims 13 or 14, wherein the control unit (18) is connected to the at least one X-ray inspection unit (10.1, 10.2, 10.n) via a communication network (42).

16. The central control unit (18) according to claim 15, comprising at least one display unit (22) for displaying an X-ray image of the current examination object for visual inspection by an operator or for displaying a test X-ray image.

17. The central control unit (18) according to claim 15, wherein the inspection object (14) is a luggage item or other package.

18. An X-ray inspection unit (10.1, 10.2, 10.n) for non-destructive inspection of an inspection object (14), the X-ray inspection unit (10.1, 10.2, 10.n) having a central control unit (18) according to any one of claims 15 to 17.

19. X-ray inspection unit (10.1, 10.2, 10.n) according to claim 18, wherein the inspection object (14) is a baggage item or other package.

20. A baggage item or parcel screening system (1) having at least one X-ray inspection unit (10.1, 10.2, 10.n) and a central control unit (18) according to any one of claims 15-17.

Images