Conical X-ray beam projection compensation method and industrial detection system
By adjusting the sampling frequency and pixel spacing in the detector component, the problem of geometric artifacts at the edge of the object in X-ray detection is solved, and the correct display of the image edge and the detection accuracy are improved.
Patent Information
- Application Number
- CN202311633776.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-30
- Publication Date
- 2025-05-30
AI Technical Summary
In X-ray detection, geometric artifacts at the edge of an object cause image blurring, affecting measurement accuracy and algorithm recognition accuracy. It is difficult to effectively identify and correct all artifacts in existing methods.
By setting the detector assembly in the detector assembly that has a sampling frequency of the detector relatively close to the projection edge is greater than that of the detector relatively close to the projection center, the pixel spacing is adjusted to increase the sampling frequency of the edge and reduce the impact of geometric artifacts.
It effectively reduces the impact of geometric artifacts on the edge of the object on the image, ensures the correct display of the shape and geometry of the edge of the image, and improves detection accuracy.
Smart Images
Figure CN120064332A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of detection, and more particularly to the field of X-ray detection. Background Art
[0002] Any place where the attenuation coefficient shown in the X-ray image is inconsistent with the actual situation can be regarded as an artifact. There are many sources of artifacts and many forms of manifestation. Common artifacts can be classified into three categories according to the phenomena they exhibit: motion artifacts, geometric artifacts, and beam hardening artifacts. The formation principles of these artifacts are different, the corresponding solutions are different, and their harmful degrees are also different in different situations.
[0003] For example, a box is a very common type among many food packaging types. When using X-rays to detect food packaged in a box, due to the projection effect of the side wall in the conical beam field of view when the box passes through the detector, a projection of the box side wall will be generated on the detector plane. This situation will cause the side wall to appear blurred in the image, as Figure 2B-2C shown. Blurring is considered an "artifact" formed during the imaging process and belongs to a type of geometric artifact. This will introduce errors in measurement, affect the operator's observation, and also affect the recognition accuracy of foreign objects by the algorithm. And whether using an LDA (line detector array) detector or a TDI (time delayed integration) detector, this kind of "artifact" will occur. The existing methods generally identify the differences between them and the real detail features by analyzing the properties such as the morphology and position of the artifacts, and then infer the physical principle of the generation of the "artifacts" and make relevant corrections. However, there are also some artifacts that are difficult to distinguish, and it is impossible to obtain the significant properties for discriminating the artifacts, and they are prone to confusion and difficult to correct. Summary of the Invention
[0004] An object of the present invention is to provide an industrial detection system that can solve the problem of geometric artifacts at the edges of objects.
[0005] The industrial detection system for achieving the above object includes an X-ray source and a detector assembly. The X-ray source is used to emit an X-ray beam; the detector assembly includes a plurality of detectors for receiving the projection of the X-ray beam; wherein, the sampling frequency of the detectors relatively closer to the projection edge is set to be greater than the sampling frequency of the detectors relatively closer to the projection center.
[0006] In one or more embodiments, the pixel pitch of the detectors relatively closer to the projection edge is set to be less than the pixel pitch of the detectors relatively closer to the projection center.
[0007] In one or more embodiments, the detector assembly includes detectors distributed in rows and / or columns.
[0008] In one or more embodiments, the sizes of the detectors are set to be equal or unequal.
[0009] In one or more embodiments, the length of the detector relatively closer to the edge of the conical projection is set to be greater than the length of the detector relatively closer to the center of the conical projection.
[0010] In one or more embodiments, the detector assembly is a combination of one or more detectors with a pixel pitch of 0.1 mm, 0.2 mm, 0.4 mm, 0.6 mm, or 0.8 mm.
[0011] In one or more embodiments, the detectors are arranged along a curved surface, and the distances from each detector to the X-ray source are equal.
[0012] In one or more embodiments, each detector is replaceably arranged.
[0013] In one or more embodiments, the system further includes a conveyor belt disposed between the X-ray source and the detector assembly.
[0014] Another object of the present invention is to provide a method for compensating a conical X-ray beam projection. The conical beam is provided by an X-ray source. The method includes the following steps of arranging a detector assembly for receiving the X-ray projection according to the following distribution rule: making the sampling frequency of the detector relatively closer to the projection edge in the detector assembly greater than the sampling frequency of the detector relatively closer to the projection center.
[0015] In one or more embodiments, the pixel pitch of the detector relatively closer to the projection edge in the detector assembly is made smaller than the pixel pitch of the detector relatively closer to the projection center.
[0016] The above industrial detection system and method start from the aspect of hardware compensation. By increasing the pixel pitch of the detectors at the edge, the sampling frequency at the edge is adjusted, and the influence of geometric artifacts at the object edge on the image is reduced by using the change in the sampling frequency of the detectors themselves, ensuring the correct display of the shape and geometry of the image; and detector modules of different sizes can be conveniently purchased and assembled, with strong operability. Description of the Drawings
[0017] The above and other features, properties, and advantages of the present invention will become more apparent from the following description in conjunction with the drawings and embodiments, where:
[0018] Figure 1A is the object to be detected on the center line of the conveyor belt;
[0019] Figure 1B is Figure 1AThe front developed image of the object to be detected;
[0020] Figure 1C is Figure 1A The back developed image of the object to be detected;
[0021] Figure 2A is the object to be detected deviated from the center line of the conveyor belt;
[0022] Figure 2B is Figure 2A The front developed image of the object to be detected;
[0023] Figure 2C is Figure 2A The back developed image of the object to be detected;
[0024] Figure 3 is the schematic diagram of the geometric model sampled by the detector assembly;
[0025] Figure 4 is the schematic diagram of the geometric model sampled by the linear detector assembly;
[0026] Figure 5 is the schematic diagram of the geometric model sampled by the curved surface detector assembly;
[0027] Figure 6A-6C is the schematic diagram of different distribution structures of the detectors within the detector assembly. Detailed implementation manners
[0028] The present invention will be further described below in conjunction with specific embodiments and the accompanying drawings. More details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention is obviously capable of being implemented in many other ways different from this description. Those skilled in the art can make similar generalizations and deductions according to the actual application situation without departing from the connotation of the present invention. Therefore, the protection scope of the present invention should not be limited by the content of this specific embodiment.
[0029] It should be noted that these and subsequent other drawings are only examples, and they are not drawn under the condition of equal proportion, and should not be used to limit the actual protection scope required by the present invention.
[0030] Figures 1A to 2C A comparison diagram showing the relative positional relationship between the object to be detected and the center line of the conveyor belt can be seen that the more deviated from the center line, the more easily the side wall of the object to be detected is blurred. In some other applications that require measurement using images, this kind of blurring is difficult to determine the edge and will bring errors to the measurement.
[0031] The above-mentioned blurring phenomenon belongs to a kind of artifact. Most X-ray images contain artifacts. However, in applications, most artifacts can be distinguished from actual defects in the detected object by the characteristics they exhibit. Nevertheless, a large number of artifacts are difficult to distinguish, and it is impossible to determine properties such as the morphology and location of the artifacts.
[0032] Generally speaking, "artifacts" are usually related to the following two factors: one is that the shape of the light field generated by X-rays is conical, as Figure 3 shown; the other is that existing detectors all use an equal pixel pitch sampling method, making the sampling frequency fixed from the middle to both sides.
[0033] The conical X-ray beam projection compensation method described in the present invention is used to solve the above-mentioned "artifacts". According to the Nyquist sampling theorem, if a system uniformly samples an analog signal at a frequency at least twice higher than the highest frequency of the signal, then the original analog signal can be completely recovered from the discrete values generated by the sampling, which can be expressed as D represents the pixel pitch, that is, pixel pitch; f represents the Nyquist sampling frequency of the imaging system.
[0034] The above formula shows that reducing the pixel size D of the detector is beneficial to increasing the sampling frequency f.
[0035] For example, if the pixel size D of the detector module at the center is 0.4 mm, and extending from the geometric center of the detector to both the left and right sides, the length of the detector module can be gradually reduced to 0.2 mm and 0.1 mm. In this case, the sampling frequency will increase to 2 times and 4 times the original, thereby reducing the influence of artifacts at the edge of the box.
[0036] Therefore, the detector assembly for receiving the X-ray projection is arranged according to the following distribution rule: the pixel pitch of the detector 40' relatively close to the projection edge in the detector assembly is smaller than the pixel pitch of the detector 40" relatively close to the projection center, so that the pixel arrangement density near the projection edge increases, and thus the sampling frequency increases.
[0037] In this way, by changing the sampling frequency, the deformation at the edge is minimized as much as possible, and the correct shape of the image edge is maintained, thereby slowing down or solving the artifacts generated at the projection edge and improving the overall quality of the image.
[0038] It should be noted that increasing the density of pixel arrangement near the edge of the projection is different from increasing the number of pixel lattices. The present application does not distinguish the edge of a smaller image by increasing the resolution, but rather enables the image shape to be correctly and completely displayed geometrically to avoid deformation of the projected object shape. For example, a circular test block will be deformed into an ellipse under the influence of the above-mentioned "artifacts", resulting in a change in shape. The above-mentioned method can effectively avoid deformation of the image to ensure the correct display of the image shape and geometry.
[0039] The term relatively close to the "projection center" refers to being relatively close to the center of the conical area projected by the conical light beam, such as Figure 3 The position of the detector 40" in the middle; "projection edge" refers to the position of the projected edge, such as Figure 3 The position of the detector 40'.
[0040] This arrangement of pixel modules can be easily realized by splicing detector sampling modules, such as Figure 6A-6C shown.
[0041] If Figure 6A As an example, the four rows of TDI (Tim Delay Integration) shown in FIG. 1 are used as examples, O represents a detector with a pixel pitch of 0.04 mm, P represents a detector with a pixel pitch of 0.02 mm, and Q represents a detector with a pixel pitch of 0.01 mm. The detector assembly includes detectors 40 arranged in rows and / or columns. The row arrangement refers to the arrangement of detectors along the Figure 6A-6C Distribution in the horizontal direction, column distribution refers to the distribution along Figure 6A-6C The longitudinal distribution. Figure 6A-6B There are 4 rows and 8 columns of detectors. Figure 6C The detectors include 4 rows and 6 columns.
[0042] It is understandable that the detector assembly may also only include detectors distributed in a single row or a single column.
[0043] In some embodiments, the distances between the detectors 40 in the detector assembly are set to be equal.
[0044] The size of the detector 40 can be set to be equal, such as Figure 6A-6B As shown, it can also be set to be unequal, such as Figure 6C As shown, the length of the detector 40' relatively close to the edge of the cone projection is made greater than the length of the detector 40" relatively close to the center of the cone projection.
[0045] The above-mentioned change in pixel arrangement density can achieve artifact reduction on the linear detector, such as Figure 4 As shown; artifact reduction can also be achieved on curved detector layouts, such as Figure 5as shown
[0046] The topological shape of the curve may include, but is not limited to, a circular surface. Taking the topological shape of a circular surface as an example, each detector module is isocentric, that is, the distance from each point on the detector module to the light source is equal. According to the inverse square law, the attenuation of X-rays from the light source to any point on the module is also the same.
[0047] Taking a curved surface as an example, the intensity of X-rays is lower towards the edge of the detector. Using an isocentric design method, the distance from the light source to any point on the detector is made equal. In this case, the attenuation of X-rays to each point is also the same. By making the pixel pitch of the detectors near the edge smaller than that of the detectors relatively closer to the projection center, the sampling frequency at the edge of the detector can be increased to a certain extent, minimizing the deformation at the edge and maintaining the correct shape display of the image edge.
[0048] Thus, as a hardware compensation method, the above method can improve the resolution at the edge of the conical projection, thereby solving the problems of blurred projection and artifacts at the object edge, being applicable to different types of artifacts, without the need to first determine the physical principle of the artifacts and make relevant corrections according to the analysis of the morphology and occurrence position of specific artifacts, which is simple and reliable.
[0049] Combined with the introduction of the above method, an industrial inspection system can also be understood. The system includes an X-ray source 20 and a detector assembly. The X-ray source is used to emit a conical X-ray beam; the detector assembly includes a plurality of detectors 40 distributed in rows and / or columns, which are used to receive the projection of the X-ray beam.
[0050] Among them, the pixel pitch of the detectors relatively closer to the projection edge is smaller than that of the detectors relatively closer to the projection center, so that the sampling frequency of the detectors relatively closer to the projection edge is greater than that of the detectors relatively closer to the projection center.
[0051] The system further includes a conveyor belt 30 arranged between the X-ray source and the detector assembly, as Figure 3 shown in FIGS. 5 to 6. The conveyor belt 30 in the drawings runs in a direction perpendicular to the paper surface to convey the object 10 to be tested.
[0052] Preferably, the detectors in the detector assembly 40 are replaceably arranged to flexibly change the sampling frequency of the detectors in different regions according to the development situation.
[0053] The detector assembly can be a combination of one or more detectors with pixel pitches of 0.1 mm, 0.2 mm, 0.4 mm, 0.6 mm, and 0.8 mm.
[0054] The sizes of the detectors are set to be equal or unequal. The detectors include length, width, and height, so as toFigure 4 and Figure 5 Taking Figure 5 as an example, the length refers to the horizontal direction, the height refers to the vertical direction, and the width refers to the direction perpendicular to the paper surface. It can be understood that the width is consistent with the running direction of the conveyor belt 30. Figure 6A-6C is a top view of the detector assembly. At this time, the length direction is the horizontal direction and the width direction is the vertical direction.
[0055] For the embodiment in which the sizes of the detectors are set to be equal, refer to Figure 6A-6B For the embodiment in which the sizes of the detectors are set to be unequal, refer to Figure 6C For example, the length of the detector relatively close to the edge of the conical projection is set to be greater than the length of the detector relatively close to the center of the conical projection.
[0056] In the above system, the change in the pixel arrangement density can reduce geometric artifacts on the linear detector, as shown in Figure 4 ; it can also reduce geometric deformation on the curved detector layout to ensure the correct shape, as shown in Figure 5 shown.
[0057] This application uses specific terms to describe the embodiments of this application. Such as "one embodiment", "an embodiment", and / or "some embodiments" mean a certain feature, structure or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that the "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more at different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this application can be combined appropriately.
[0058] Although the present invention is disclosed above with preferred embodiments, it is not used to limit the present invention. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, any modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention shall fall within the protection scope defined by the claims of the present invention.
Claims
1. An industrial inspection system, characterized in that, it includes: an X-ray source for emitting a conical X-ray beam; a detector assembly including a plurality of detectors for receiving the projection of the X-ray beam; wherein, the sampling frequency of the detectors relatively closer to the projection edge is set to be greater than the sampling frequency of the detectors relatively closer to the projection center.
2. The industrial inspection system according to claim 1, characterized in that, the pixel pitch of the detectors relatively closer to the projection edge is set to be less than the pixel pitch of the detectors relatively closer to the projection center.
3. The industrial inspection system according to claim 1, characterized in that, the detector assembly includes detectors distributed in rows and / or columns.
4. The industrial inspection system according to claim 1, characterized in that, the sizes of the respective detectors are set to be equal or unequal.
5. The industrial inspection system according to claim 4, characterized in that, the length of the detectors relatively closer to the conical projection edge is set to be greater than the length of the detectors relatively closer to the conical projection center.
6. The industrial inspection system according to claim 1, characterized in that, the detector assembly is a combination of detectors with one or more of pixel pitches of 0.1 mm, 0.2 mm, 0.4 mm, 0.6 mm, and 0.8 mm.
7. The industrial inspection system according to claim 1, characterized in that, the detectors are arranged along a curved surface, and the distances from each of the detectors to the X-ray source are equal.
8. The industrial inspection system according to claim 1, characterized in that, each of the detectors is replaceably provided.
9. The industrial inspection system according to claim 1, characterized in that, the system further includes a conveyor belt provided between the X-ray source and the detector assembly.
10. A method for compensating a conical X-ray beam projection, the conical beam being provided by an X-ray source, characterized in that, it includes the following steps: arranging a detector assembly for receiving an X-ray projection according to the following distribution rule: making the sampling frequency of the detectors relatively closer to the projection edge in the detector assembly greater than the sampling frequency of the detectors relatively closer to the projection center.
11. The method according to claim 10, characterized in that, making the pixel pitch of the detectors relatively closer to the projection edge in the detector assembly less than the pixel pitch of the detectors relatively closer to the projection center.