Apparatus and method for removing artifact in x-ray image

Abstract

An apparatus and a method for removing an artifact in an X-ray image are disclosed. According to an embodiment of the present application, an apparatus for detecting low-density foreign substances in food by using dark-field X-ray imaging based on a single imaging grid comprises: an acquisition unit for acquiring a sample image, including a food sample and a grid, and a reference image, including only the grid without including the food sample, the sample image and the reference sample being captured using grid-based X-ray imaging; a restoration unit for restoring an absorption image and a dark-field image by using the sample image and the reference image; and a fusion unit for generating a fusion image by fusing the absorption image and the dark-field image.

Description

Device and method for removing artifacts from X-ray images

[0001] The present invention relates to a device and method for removing artifacts from X-ray images.

[0002] The technology of obtaining X-ray phase contrast images and dark-field images by detecting and imaging the refraction and small-angle scattering of X-rays can detect information that was not accessible in conventional X-ray absorption images, and is currently being used in various fields such as medicine, industry, and material science.

[0003] In this regard, in the medical and biomedical fields, a grating-based dark-field X-ray imaging (DFXI) method has been proposed that can be used to obtain microstructural information related to small-angle X-ray scattering (SAXS) of samples, and to complement the challenges of high-precision gratings, mechanical phase stepping, and precise alignment of grating-based dark-field X-ray imaging and large-area gratings, single-shot grid-based dark-field X-ray imaging (SG-DFXI) that requires only a single shot and one X-ray grid has been developed.

[0004] Meanwhile, X-ray imaging technology can also be used to detect foreign bodies (FBs) in food, and demand is growing as an inspection tool to avoid food contamination. However, the detection of foreign bodies in food using X-ray inspection is directly related to the density of the food and the foreign body. Low-density foreign bodies, among these, are difficult to detect and detect using conventional X-ray inspection.

[0005] Therefore, there is a need to improve the detection accuracy of the presence and location of low-density foreign substances in food by applying a single-shot grid-based dark-field X-ray imaging technique to detect foreign substances in food.

[0006] Meanwhile, when applying a single-shot grid-based dark-field X-ray imaging technique, a spectral overlap artifact may occur in which the harmonic frequency spectrum and the spectrum of the sample overlap, and in such a case, the acquired image may be damaged.

[0007] To prevent image damage caused by spectral overlap artifacts, methods such as increasing the magnification of an object or reducing the bandwidth of a window function have been used in the past, but these methods can cause problems such as a decrease in image sensitivity and field of view (FOV) and a decrease in image resolution.

[0008] The technology underlying this application is disclosed in Korean Patent Publication No. 10-2127711.

[0009] The present invention is intended to solve the problems of the prior art described above, and to provide a device and method for detecting low-density foreign substances in food using single-shot grid-based dark-field X-ray photography, which can detect and clearly display low-density foreign substances in food by obtaining dark-field images and fusion images through single-shot grid-based dark-field X-ray photography with low system setup complexity and required dose using only a single shot and one X-ray grid.

[0010] The present invention is intended to solve the problems of the prior art described above, and to provide a device and method for detecting low-density foreign substances in food using dark-field X-ray photography based on a single-shot grid, which can remove spectral overlap artifacts without loss of image sensitivity and reduction of resolution by moving a sample at a constant speed corresponding to a condition in which spectral overlap artifacts do not occur, thereby generating a motion blurring effect.

[0011] The present invention is intended to solve the problems of the prior art described above, and to provide a device and method for detecting low-density foreign substances in food using dark-field X-ray photography based on a single-shot grid, which can improve resolution and remove spectral overlap artifacts by applying a window function only to an area that does not include overlapping artifacts using a partial Fourier technique. However, the technical problems to be achieved by the embodiments of the present invention are not limited to the technical problems described above, and other technical problems may exist.

[0012] As a technical means for achieving the above-described technical task, a device for detecting low-density foreign substances in food using single-shot grid-based dark-field X-ray imaging according to one embodiment of the present invention includes an acquisition unit that acquires a sample image captured by grid-based X-ray imaging to include a food sample and a grid and a reference image that includes only the grid but not the food sample, a restoration unit that restores an absorption image and a dark-field image using the sample image and the reference image, and a fusion unit that generates a fusion image by fusing the absorption image and the dark-field image, wherein the grid may be arranged at a position such that at least one low-density foreign substance included in the food sample is detected at a level that can be identified in the dark-field image.

[0013] According to one embodiment of the present invention, the grid may be disposed between the food sample and the detector, and may be disposed at a position adjacent to the food sample relative to the detector.

[0014] According to one embodiment of the present invention, the food sample, the grid, and the detector may be arranged in a straight line.

[0015] According to one embodiment of the present invention, the restoration unit may include a region identification unit that identifies a region surrounding a 0th harmonic peak and a region surrounding a 1st harmonic peak by applying a Fourier transform to the sample image and the reference image, a region separation unit that separates the region surrounding the 0th harmonic peak and the region surrounding the 1st harmonic peak using a band-pass filter, a harmonic image acquisition unit that obtains a plurality of harmonic images by applying an inverse Fourier transform to the separated region surrounding the 0th harmonic peak and the region surrounding the 1st harmonic peak, and an image restoration unit that restores the absorption image and the dark-field image using the plurality of harmonic images.

[0016] According to one embodiment of the present invention, the fused image may clearly reflect the location and shape of the at least one low-density foreign substance compared to the dark field image.

[0017] According to one embodiment of the present invention, the restoration unit may be such that the food sample is linearly moved by a moving device at a preset speed for a preset shooting time.

[0018] According to one embodiment of the present invention, an artifact in an X-ray image detected through the detector may be removed due to a motion blurring phenomenon caused by movement of the sample.

[0019] According to one embodiment of the present invention, the region separation unit may remove the artifact by applying an extended window function using a partial Fourier technique to an absence region other than an existence region of the artifact in an X-ray image detected by a detector so that the existence region is not reflected in the dark field image.

[0020] As a technical means for achieving the above-described technical task, a system for detecting low-density foreign matter in food using dark-field X-ray photography based on a single-shot grid according to one embodiment of the present invention may include an X-ray imaging device that irradiates X-rays to a food sample and detects the X-rays transmitted through a grid, and a food low-density foreign matter detection device that detects at least one low-density foreign matter contained in the food sample based on an X-ray image of the detected X-rays.

[0021] According to one embodiment of the present invention, the food low-density detection device may obtain a sample image captured to include a food sample and a grid as the X-ray image and a reference image including only the grid but not the food sample, restore an absorption image and a dark-field image using the sample image and the reference image, and fuse the absorption image and the dark-field image to generate a fused image, wherein the grid may be positioned such that at least one low-density foreign substance included in the food sample is detected at a level that can be identified in the dark-field image.

[0022] As a technical means for achieving the above-described technical task, a method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray imaging according to one embodiment of the present invention includes the steps of: acquiring a sample image captured to include a food sample and a grid through grid-based X-ray imaging and a reference image including only the grid but not the food sample; restoring an absorption image and a dark-field image using the sample image and the reference image; and generating a fusion image by fusing the absorption image and the dark-field image, wherein the grid may be arranged at a position such that at least one low-density foreign matter included in the food sample is detected at a level that can be identified in the dark-field image.

[0023] According to one embodiment of the present invention, the restoring step may include a step of applying a Fourier transform to the sample image and the reference image to identify a surrounding area of ​​a 0th harmonic peak and a surrounding area of ​​a 1st harmonic peak, a step of separating the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak using a bandpass filter, a step of applying an inverse Fourier transform to the separated surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak to obtain a plurality of harmonic images, and a step of restoring the absorption image and the dark field image using the plurality of harmonic images.

[0024] According to one embodiment of the present invention, the fused image may clearly reflect the location and shape of the at least one low-density foreign substance compared to the dark field image.

[0025] According to one embodiment of the present invention, the food sample may be linearly moved by a moving device at a preset speed for a preset shooting time.

[0026] According to one embodiment of the present invention, an artifact in an X-ray image detected through a detector may be removed due to a motion blurring phenomenon caused by movement of the sample.

[0027] According to one embodiment of the present invention, the separating step may be to remove the artifact by applying an extended window function using a partial Fourier technique to an absence area other than an existence area of ​​the artifact in an X-ray image detected by a detector so that the existence area is not reflected in the dark field image.

[0028] The above-described problem-solving methods are merely exemplary and should not be construed as limiting the present invention. In addition to the exemplary embodiments described above, additional embodiments may be included in the drawings and detailed description of the invention.

[0029] According to the aforementioned means for solving the problem of the present invention, dark-field images and fusion images can be obtained through dark-field X-ray photography based on a single-shot grid with low system setup complexity and low required dose using only a single shot and one X-ray grid, thereby detecting and clearly displaying low-density foreign substances in food.

[0030] According to the solution to the problem of the present invention described above, by moving a sample at a constant speed corresponding to a condition in which a spectral overlap artifact does not occur and thereby generating a motion blurring effect, a spectral overlap artifact can be removed without loss of image sensitivity or reduction in resolution.

[0031] According to the solution to the problem of the present invention described above, by applying a window function only to a region that does not contain overlapping artifacts using a partial Fourier technique, the resolution can be improved and spectral overlapping artifacts can be removed.

[0032] However, the effects that can be obtained from this center are not limited to the effects described above, and other effects may exist.

[0033] FIG. 1 is a schematic diagram of a food low-density foreign matter detection system using single-shot grid-based dark-field X-ray imaging according to one embodiment of the present invention.

[0034] Figure 2 is a schematic block diagram of an X-ray imaging device according to one embodiment of the present invention.

[0035] Figure 3 is a schematic diagram of an X-ray imaging device according to one embodiment of the present invention.

[0036] Figure 4 is a plan view of a schematic configuration of an X-ray imaging device according to one embodiment of the present invention.

[0037] FIG. 5 is a schematic block diagram of a food low-density foreign matter detection device (X-ray image processing device) using a single-shot grid-based dark-field X-ray imaging according to one embodiment of the present invention.

[0038] Figure 6 is a schematic diagram illustrating a process for obtaining absorption images and dark-field images for a sample.

[0039] FIG. 7 is a diagram illustrating an extended window function according to one embodiment of the present invention for generating spectral overlap artifacts and removing spectral overlap artifacts.

[0040] Figure 8 is a diagram showing an example of an experimental setting based on the autocorrelation length for detection of low-density foreign substances in food.

[0041] Figure 9 is a diagram showing an example of an experiment using absorption images and dark-field images according to autocorrelation length for detection of low-density foreign substances in food.

[0042] Figure 10 is a diagram showing the evaluation results for an experimental example of absorption images and dark-field images by autocorrelation length for detection of low-density foreign substances in food.

[0043] Figure 11 is a diagram showing an example of an experiment of absorption images, dark-field images, and fusion images for samples containing low-density foreign substances such as wooden chopsticks and Styrofoam.

[0044] Figure 12 is a diagram showing the evaluation results for an experimental example of absorption images, dark-field images, and fusion images for a sample including a wooden chopstick, which is a low-density foreign substance.

[0045] Figure 13 is a diagram showing the evaluation results for an experimental example of absorption images, dark-field images, and fusion images for a sample containing low-density foreign matter, Styrofoam.

[0046] Figure 14 is a diagram showing an experimental example of absorption images and dark-field images for samples containing low-density foreign substances such as mechanical pencil lead and plastic bullets.

[0047] Figure 15 is a diagram showing an experimental example of spectral overlap artifact removal using an extended window function.

[0048] Figure 16 is a diagram showing another experimental example for removing spectral overlap artifacts using an extended window function.

[0049] Figure 17 is a diagram showing the evaluation results for another experimental example of spectral overlap artifact removal using an extended window function.

[0050] FIG. 18 is a flowchart illustrating a method for removing artifacts in an X-ray image using motion blurring through an X-ray imaging device according to one embodiment of the present invention.

[0051] Figure 19 is a flowchart of a method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography.

[0052] FIG. 20 is a flowchart illustrating a method for restoring absorption images and dark-field images using a food low-density foreign substance detection device (X-ray image processing device) using a single-shot grid-based dark-field X-ray imaging, and a method for removing artifacts in X-ray images using an extended window function.

[0053] Below, with reference to the attached drawings, embodiments of the present invention are described in detail to facilitate easy implementation by those skilled in the art. However, the present invention can be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, irrelevant parts have been omitted for clarity, and similar reference numerals have been used throughout the specification to indicate similar elements.

[0054] Throughout this specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected," but also the case where it is "electrically connected" or "indirectly connected" with another element in between.

[0055] Throughout this specification, when it is said that a member is located “on,” “above,” “upper,” “lower,” “lower” or “lower” another member, this includes not only cases where the member is in contact with the other member, but also cases where another member exists between the two members.

[0056] Throughout this specification, whenever a part is said to "include" a component, this means that it may include other components, but not to the exclusion of other components, unless otherwise specifically stated.

[0057] The present invention relates to a device and method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray imaging. Specifically, the present invention relates to a system for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray imaging, which can detect low-density foreign matter in food using single-shot grid-based dark-field X-ray imaging and remove artifacts in X-ray images using at least one of an extended window function and motion blurring.

[0058] FIG. 1 is a schematic diagram of a food low-density foreign matter detection system using single-shot grid-based dark-field X-ray imaging according to one embodiment of the present invention.

[0059] Referring to FIG. 1, a food low-density foreign matter detection system (10) using a single-shot grid-based dark-field X-ray imaging may include an X-ray imaging device (100) and a food low-density foreign matter detection device (200) using a single-shot grid-based dark-field X-ray imaging.

[0060] In this regard, throughout the specification of the present application, the food low-density foreign matter detection device (200) using a single-shot grid-based dark-field X-ray imaging according to one embodiment of the present application may refer to an X-ray image processing device (200) that performs image processing on an X-ray image captured or detected by an X-ray imaging device (100). Accordingly, hereinafter, the 'food low-density foreign matter detection device (200) using a single-shot grid-based dark-field X-ray imaging' may be described interchangeably as the 'X-ray image processing device (200)', and for the convenience of explanation, may be specifically referred to as the 'image processing device (200)'.

[0061] In addition, the 'food low-density foreign matter detection system (10) using single-shot grid-based dark-field X-ray photography' may be used interchangeably with the 'X-ray image processing system (10)' or the 'X-ray imaging system (10)', such as the 'X-ray image processing device (200)', and in the present invention, for the convenience of explanation, it will be specifically referred to as the 'detection system (10)' below.

[0062] In addition, the detection system (10) is for single-shot grid-based dark-field imaging, and may be a single-shot grid-based dark-field X-ray imaging (SG-DFXI) system, but is not limited thereto, and may be a spatial harmonic imaging (SHI) method or various other X-ray imaging systems.

[0063] Referring to FIG. 1, in a detection system (10), an X-ray imaging device (100) and an image processing device (200) can communicate with each other through a network. Specifically, the X-ray imaging device (100) detects X-rays irradiated toward a sample (S) from an X-ray source (110), and the image processing device (200) can receive an X-ray image (X-ray) captured (detected) by the X-ray imaging device (100) through a network and perform at least one image processing to detect a foreign substance in the sample (S).

[0064] At this time, the 'detection' of the X-ray image performed in the X-ray imaging device (100) means the detection of X-rays and refers to the process of obtaining a raw image to which no post-processing has been applied, and the 'detection' of a foreign substance performed in the image processing device (200) may mean the process of clearly indicating at least one foreign substance included in the X-ray image through at least one image processing on the X-ray image. In addition, the X-ray image here refers to an image that can be obtained through X-ray detection and may be interpreted as a concept that includes everything from a raw image to which image processing has not been applied to an absorption image and a dark-field image to which image processing has been applied.

[0065] In addition, the network for communication of the X-ray imaging device (100) and the image processing device (200) may include, but is not limited to, a 3GPP (3rd Generation Partnership Project) network, an LTE (Long Term Evolution) network, a 5G network, a WIMAX (World Interoperability for Microwave Access) network, a wired / wireless Internet, a LAN (Local Area Network), a Wireless LAN (Wireless Local Area Network), a WAN (Wide Area Network), a PAN (Personal Area Network), a Bluetooth network, a Wifi network, an NFC (Near Field Communication) network, a satellite broadcasting network, an analog broadcasting network, a DMB (Digital Multimedia Broadcasting) network, etc.

[0066] According to one embodiment of the present invention, the X-ray imaging device (100) can irradiate an X-ray to a sample (S) and detect the X-ray that has passed through the grid (120). At this time, the sample (S) may mean a food sample, and although the terms sample (S) and food sample (S) may be used interchangeably herein, the sample (S) is not necessarily limited to the food sample (S).

[0067] Figure 2 is a schematic block diagram of an X-ray imaging device according to one embodiment of the present invention.

[0068] Referring to FIG. 2, the X-ray imaging device (100) may include an X-ray source (110), a grid (120), a detector (130), and a mover (140).

[0069] Figure 3 is a schematic diagram of an X-ray imaging device according to one embodiment of the present invention. In addition, Figure 4 is a plan view of a schematic diagram of an X-ray imaging device according to one embodiment of the present invention.

[0070] Referring to FIGS. 3 and 4, the X-ray source (110) may irradiate X-rays toward the sample (S). At this time, the X-ray source (11) may be a microfocus X-ray tube having a focal size of 20 μm, operated at 30 kVp, and having an X-ray beam energy evaluated to be about 10.7 keV, and an X-ray wavelength λ may correspond to 0.116, but is not limited thereto.

[0071] Referring to FIGS. 3 and 4, the grid (120) may be arranged in a straight line with the X-ray source (110) and the sample (S), and may transmit X-rays irradiated on the sample (S). At this time, the grid (120) functions as a frequency modulator for Fourier components, and thus the phase information of the sample (S) may be reconstructed through a Fourier processing technique. This reconstructing process will be described in detail later with reference to FIGS. 5 to 7.

[0072] In addition, the grid (120) may be a linear grid, but is not limited thereto. In addition, the grid (120) may be arranged to be inclined at a predetermined angle with respect to a straight line (center horizontal line in FIGS. 3 and 4) on which the X-ray source (110), the sample (S), and the grid (120) are arranged. This will be described in detail later. In addition, the grid (120) may be, for example, a focal linear X-ray grid having a strip density in the range of 6.99 to 13.16 lines / mm, but is not limited thereto.

[0073] Also, referring to FIGS. 3 and 4, the detector (130) may detect X-rays irradiated to the sample (S) transmitted from the grid (120). The detector (130) may be arranged in a straight line with the X-ray source (110), the sample (S), and the grid (120), and may be, for example, a flat panel detector, but is not limited thereto.

[0074] Referring to FIGS. 3 and 4, the grid (120) may be disposed between the sample (S) and the detector (130), but may be disposed at a position adjacent to the sample (S) with respect to the detector (130). This may be to ensure that at least one low-density foreign substance contained in the sample (S) is accurately detected in an X-ray image corresponding to the X-ray detected by the detector (130).

[0075] In other words, the grid (120) is configured to detect at least one low-density foreign substance contained in the sample (S) in an X-ray image (dark field image (I DF ) etc.) may be placed in a position to be detected at an identifiable level.

[0076] In this regard, the position of the grid (120) and the sample (S) affects the autocorrelation length (ξ) which affects the sensitivity to the small-angle X-ray scattering (SAXS) signal. G , Correlation length) should have as large a value as possible, so that the sample (S) and the grid (120) should be placed sufficiently close together.

[0077] Specifically, in the detection system (10), the autocorrelation length (ξ G ) is on the order of hundreds of nanometers, and the SAXS signal is, when the particle or structure size of the sample (S) is hundreds of nanometers or larger, the autocorrelation length (ξ G ) may be measured accurately if it is set to the largest possible value. The autocorrelation length (ξ) for this G) may be defined by the following [Formula 1].

[0078] [Formula 1]

[0079]

[0080] Here, λ is the X-ray wavelength, p is the period of the grid projected on the surface of the detector (130), and d SS is the distance between the X-ray source (110) and the sample (S), d SG is the distance between the X-ray source (110) and the grid (120), d SD is the distance between the X-ray source (110) and the detector (130).

[0081] At this time, p can be expressed by the grid period p00 as in [Equation 2] below.

[0082] [Formula 2]

[0083]

[0084] Considering the above [Equation 1] and [Equation 2], d SD The larger is, the smaller is p0, the larger is the distance (|d) between the sample (S) and the grid (120). SS - d SG The smaller |) is, and d SD = d SD / 2 The autocorrelation length (ξ) is G ) can be seen to have a large value. Experimental examples related to this will be described in detail later with reference to FIGS. 8 to 10.

[0085] Next, the design criteria for the grid (120) according to one embodiment of the present invention will be described in more detail. The lead thickness of the grid (120) is determined by a dark field image (I DF ) may be thick enough to shield all X-ray energy corresponding to the tube voltage so that visibility is at its maximum value to increase the contrast to noise ratio (CNR), and d SG corresponds to the focal length of the grid (in this case, dSG = d SD / 2, so d SD = 2 x focal length), but is not limited thereto.

[0086] According to one embodiment of the present invention, the mover (140) may linearly move the sample (S) at a preset speed. Specifically, the mover (140) may move the sample (S) in the x-axis direction in FIGS. 3 and 4 to generate a motion blurring phenomenon in order to remove artifacts that may be included in an X-ray image detected by the detector (130). The mover (140) is a component for moving the sample (S), and may be, for example, a conveyor belt, but is not limited thereto. In addition, a description related to the control of movement of the sample (S) through the mover (140) will be described in detail later as an artifact removal method according to one embodiment of the present invention.

[0087] According to one embodiment of the present invention, the image processing device (200) performs image processing on an X-ray image for an X-ray detected through the X-ray imaging device (100), and can detect at least one low-density foreign substance included in a sample (S) based on the X-ray image.

[0088] FIG. 5 is a schematic block diagram of a food low-density foreign matter detection device (X-ray image processing device) using a single-shot grid-based dark-field X-ray imaging according to one embodiment of the present invention.

[0089] Referring to FIG. 5, the X-ray image processing device (200) may include an acquisition unit (210), a restoration unit (220), and a fusion unit (230).

[0090] According to one embodiment of the present invention, the acquisition unit (210) captures a sample image (I) including a sample (S) and a grid (120) captured through grid-based X-ray imaging by an X-ray imaging device (100). s) and a reference image (I) containing only the grid (120) and not the sample (S). r ) can be obtained. At this time, a sample image (I s ) and reference images (I r ) may be a raw image acquired through an X-ray imaging device (100) and may be acquired with the same image settings.

[0091] According to one embodiment of the present invention, the restoration unit (220) is a sample image (I s ) and reference images (I r ) using absorption imaging (I ABS ) and dark field images (I DF ) can be restored.

[0092] Referring to FIG. 5, the restoration unit (220) may include a region identification unit (221), a region separation unit (222), a harmonic image acquisition unit (223), and an image restoration unit (224).

[0093] Figure 6 is a schematic diagram illustrating a process for obtaining absorption images and dark-field images for a sample.

[0094] Referring to Fig. 6, the area identification unit (221) is a sample image (I s ) and reference images (I r ) by applying Fourier Transform to the 0th harmonic peak (0 th harmonic, Primary) and 1st harmonic peak (1 st The surrounding area for each harmonic can be identified. In addition, the area separation unit (222) can identify the 0th harmonic peak (0 th The surrounding area of ​​the first harmonic peak (1) and the first harmonic peak (1) st The surrounding area of ​​harmonic can be separated using a bandpass filter.

[0095] In addition, the harmonic image acquisition unit (223) separates the 0th harmonic peak (0 thThe surrounding area of ​​the first harmonic peak (1) and the first harmonic peak (1) st By applying the inverse Fourier transform to the surrounding area of ​​the harmonic, multiple harmonic images can be obtained. At this time, the multiple harmonic images are a sample image (I s ) for the 0th harmonic image (H s 0), sample image (I s ) for the first harmonic image (H s 1), reference video (I r ) for the 0th harmonic image (H r 0) and reference image (I r ) for the first harmonic image (H r 1) may be included.

[0096] Referring to FIG. 6, the image restoration unit (224) uses multiple harmonic images to create an absorption image (I ABS ) and dark field images (I DF ) can be restored. Specifically, the absorption image (I ABS ) and dark field images (I DF ) can be calculated by [Equation 3] and [Equation 4] below using multiple harmonic images.

[0097] [Formula 3]

[0098]

[0099] [Formula 4]

[0100]

[0101] According to one embodiment of the present invention, the fusion unit (230) is an absorption image (I ABS ) and dark field images (I DF ) can be fused to create a fused image. At this time, the fused image is a dark field image (I DF ) may clearly reflect the location and shape of at least one low-density foreign body.

[0102] Specifically, absorption images (I ABS) is a dark field image (I DF ) clearly shows the shape of the contrast sample (S), while the dark field image (I DF ) is an absorption image (I ABS ) can clearly display low-density foreign substances. Therefore, the fusion part (230) can clearly display low-density foreign substances. ABS ) and dark field images (I DF ) by generating a fused image, thereby displaying low-density foreign substances in the form of a sample (S), making it easy to confirm where and in what form the low-density foreign substances are contained in the sample (S).

[0103] Below, two methods for removing spectral overlap artifacts according to one embodiment of the present invention are described in detail.

[0104] In this regard, when acquiring an X-ray image, the Fourier spectrum (original sample spectrum) of the sample (S) is copied to the harmonic maximum of the grid spectrum by the X-ray irradiated to the sample (S), so that multiple harmonic images can be acquired for the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak, as disclosed in the description of the restoration unit (220). However, if the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak partially overlap, a spectrum overlap artifact may occur.

[0105] FIG. 7 is a diagram illustrating an extended window function according to one embodiment of the present invention for generating spectral overlap artifacts and removing spectral overlap artifacts.

[0106] Fig. 7 (a) is a drawing for explaining a state in which a spectral overlap artifact (or artifact) occurs due to a partial overlap of the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak. Referring to Fig. 7 (a), when separating the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak through the region separation unit (222) of the restoration unit (220) in a state in which an artifact occurs, a band pass filter, i.e., a band window function can be used, but an image (absorption image (I)) restored by the overlapping area (area in which an artifact exists) ABS ) and dark field images (I DF )) may be damaged.

[0107] FIG. 7 (b) is a drawing of a conventional technique for preventing the occurrence of such artifacts. Referring to FIG. 7 (b), in the past, in order to prevent image damage caused by spectral overlap artifacts, when capturing an X-ray image, a method was used in which the focal spot size was increased or the object magnification (i.e., magnification) for the sample (S) was increased, thereby blurring the spectrum around the first harmonic peak, i.e., the first harmonic peak, thereby setting a band-limited value for the window function to prevent the occurrence of spectral overlap artifacts. However, this method has a problem in that the field-of-view (FOV) may be reduced and the image sensitivity may be lowered due to the magnification applied to the sample (S).

[0108] A detection system (10) according to one embodiment of the present invention is intended to overcome the limitations of the above-described conventional technology and may remove artifacts without lowering image sensitivity.

[0109] Specifically, the detection system (10) may perform artifact removal using an extended window function (hereinafter, referred to as a first artifact removal method) and artifact removal using motion blurring (hereinafter, referred to as a second artifact removal method). In this regard, the detection system (10) may selectively apply at least one of the first artifact removal method and the second artifact removal method to remove spectral overlapping artifacts for an X-ray image.

[0110] Below, the artifact removal method using the extended window function (first artifact removal method) will be described in detail.

[0111] (C) of FIG. 7 is a drawing for explaining a first artifact removal method. The first artifact removal method may be a method of removing artifacts by using an extended window function based on a partial Fourier technique when separating the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak in the region separation unit (222).

[0112] In this regard, the partial Fourier technique is a concept devised from the technology of magnetic resonance imaging, and is a technique that can restore an image for the entire space by utilizing the symmetry of the Fourier spectrum even if only a spectrum for a part of the entire space is secured, and the region separation unit (222) according to one embodiment of the present invention can remove artifacts by restoring an image using only the spectrum for an area where artifacts do not exist by utilizing this partial Fourier technique.

[0113] In other words, the area separation unit (222) applies an extended window function using the partial Fourier technique to the absence area outside the presence area of ​​the artifact to create an image (e.g., a dark field image (I DF )) can be used to remove artifacts by making the area of ​​existence not reflected.

[0114] Specifically, referring to (c) of FIG. 7, the region separation unit (222) may apply a window function to the surrounding region of the first harmonic peak to exclude the presence region where the spectral overlap artifact occurs, thereby obtaining only the absence region where no artifact exists, and apply a partial Fourier technique to the obtained absence region so that no artifact is included in the restored image. Here, the partial Fourier technique used in the present invention may be, for example, a zero-filling technique.

[0115] Regarding the zero-filling technique, as an example, when the spectrum has symmetry based on the first harmonic frequency peak, the component F1 of the first harmonic frequency spectrum is in the low frequency region (F) as shown in [Equation 5] below. low ) and high frequency region (F high ) can be expressed separately.

[0116] [Formula 5]

[0117]

[0118] Low-frequency function F of the above [Equation 5] low and high-frequency function F low When the inverse Fourier transform is applied to each, the first harmonic frequency image can be produced as the sum of them. The first harmonic image (H1) can be expressed by the following [Equation 6].

[0119] [Formula 6]

[0120]

[0121] At this time, if the area separation unit (222) obtains only half of the area for the high-frequency area due to the presence of an artifact, it is expressed as in [Formula 7].

[0122] [Formula 7]

[0123]

[0124] As in [Equation 7], half of the region for the high-frequency region appears in the form combined with the step function θ(u).

[0125] In [Formula 7], if the high frequency range is taken as double and organized, it is expressed as [Formula 8].

[0126] [Formula 8]

[0127]

[0128] In [Equation 8], H1(x, y) is the first harmonic image we want to see. To obtain the first harmonic image, H1(x, y), we take the real function H*1(x, y) to return the real part of a complex number, which is ultimately expressed as [Equation 9].

[0129] [Formula 9]

[0130]

[0131] The process of [Formula 9] above can be applied to the first harmonic images for each of the sample image and the reference image, and a dark field image can be obtained through [Formula 4] above.

[0132] Meanwhile, referring to (c) of FIG. 7, in order to prevent a decrease in the resolution of the image, the region separation unit (222) may acquire the entire region where no artifacts exist, and may use a smooth filter (e.g., Hann filter) for each truncation portion to prevent an artifact (truncation artifact) caused by the image being truncated.

[0133] Accordingly, the detection system (10) according to one embodiment of the present invention can improve resolution and remove spectral overlap artifacts by applying a window function only to an area that does not include spectral overlap artifacts using the partial Fourier technique.

[0134] At this time, considering the results of the experimental example described through Fig. 17, the absence area that does not include spectral overlap artifacts to which the extended window function is applied in the detection system (10) may be an area of ​​more than half of the entire area, but is not limited thereto.

[0135] According to one embodiment of the present invention, the detection system (10) can also remove spectral overlap artifacts for X-ray images by utilizing the motion blurring phenomenon caused by movement of the sample (S) during X-ray photography.

[0136] Below, we will describe in detail the method of removing artifacts using motion blurring (second artifact removal method).

[0137] In this regard, spectral overlap artifacts may occur when the condition of [Equation 10] below is not satisfied.

[0138] [Formula 10]

[0139]

[0140] Here, M ob is the object magnification (i.e., magnification) for the sample (S), S0 is the size of the X-ray beam irradiated from the X-ray source (110), and p is the grid period.

[0141] That is, the spectral overlap artifact may not occur if the above [Equation 10] is satisfied, and as described above, increasing the magnification for the sample (S) can prevent the occurrence of the spectral overlap artifact by satisfying the above [Equation 10]. However, the increase in magnification may cause the dark field image (I DF ) may cause problems such as reduced sensitivity and deterioration of image quality.

[0142] Accordingly, in addition to the artifact removal method using the extended window function described above, the present invention proposes an artifact removal method using the motion blurring phenomenon.

[0143] In this regard, motion blurring is an effect caused by the movement of a moving object or a camera in a photograph or video, and in X-ray photography, it may mean an effect in which the Fourier spectrum of an object is restricted to the direction in which the object moves by the object's speed, magnification, and exposure time when the object moves at a constant speed during X-ray photography.

[0144] Referring again to FIGS. 3 and 4, when the sample (S) is moved in a direction perpendicular to the line direction of the grid (120) (x-axis direction) for a shooting time (T) at a constant speed (v) by the mover (140), the displacement of the image projected onto the detector (130) due to the magnification of the sample (S) is M ob It can be expressed as vT.

[0145] At this time, h, which is the motion spread function for the sample (S), motion (x) can be expressed as in [Formula 11] below.

[0146] [Formula 11]

[0147]

[0148] Here, x is the direction of movement of the sample (S). Also, h motion The image of the sample (S) projected onto the detector (130) by (x) can be expressed as in [Equation 12] below.

[0149] [Formula 12]

[0150]

[0151] Here, g(x) is the projected image, f(x) is the transfer function of the sample (S), and h system (x) can be an impulse function of a system including geometrical ambiguity. In addition, by applying the Fourier transform to [Equation 12], the following [Equation 13] can be obtained.

[0152] [Formula 13]

[0153]

[0154] Referring to the above [Formula 11] to [Formula 13], h motion (x) has width M ob vT is a rectangular function, and accordingly H motion (u) has a cutoff frequency of 1 / M ob vT may be a sinc function. According to one embodiment of the present invention, such H motion Considering the cutoff frequency of (u), a condition in which no spectral overlap artifact occurs can be derived (see [Equation 14] below).

[0155] [Formula 14]

[0156]

[0157] According to one embodiment of the present invention, if the above [Formula 14] is satisfied, the occurrence of spectral overlap artifacts can be prevented even if the condition of the above [Formula 10] is not satisfied. That is, the detection system (10) according to one embodiment of the present invention is M, which is an enlargement of the sample (S). ob Spectral overlap artifacts can be removed by specifying the moving speed v of the sample (S) and the shooting time T without increasing .

[0158] In other words, the moving device (140) may move the sample (S) at a preset speed for a preset shooting time, and the X-ray imaging device (100) may prevent and remove the occurrence of spectral overlap artifacts by utilizing the motion blurring phenomenon caused by the movement of the sample (S).

[0159] Meanwhile, when detecting X-rays, an interference artifact called a moire fringe may occur due to the interaction between the grid (120) and the detector (130) having periodic stripes, and to prevent this, the grid (120) must be tilted at a predetermined angle other than vertical with respect to the bottom surface (a plane that is horizontal to the straight line on which the X-ray source (110), the sample (S), and the grid (120) are arranged).

[0160] At this time, the u-axis coordinate of the first harmonic peak becomes cosθ / p, not 1 / p, and in this case, even if the moving speed v of the sample (S) and the shooting time T are set to satisfy the above [Equation 14], spectral overlap may occur.

[0161] In order to prevent such spectral overlap, it may be required to tilt the direction of movement of the sample (S) by taking into account the inclination of the grid (120), but in reality, manufacturing and applying a device that tilts and moves the sample (S) may cause problems such as a complicated process and reduced price competitiveness.

[0162] Accordingly, the detection system (10) according to one embodiment of the present invention can be applied by modifying the above [Formula 14] as shown in the following [Formula 15], taking into account that the bandwidth of the band-pass filter becomes cosθ / p when the grid (120) is arranged to be inclined at a predetermined angle.

[0163] [Formula 15]

[0164]

[0165] In other words, the detection system (10) according to one embodiment of the present invention can remove artifacts from an X-ray image detected by the detector (130) by utilizing a motion blurring phenomenon caused by movement of a sample (S), and specifically, when the grid (120) is arranged vertically with respect to the floor surface, the movement speed v of the sample (S) and the shooting time T are set to satisfy the above [Formula 14] to prevent the occurrence of artifacts, and when the grid (120) is arranged at a predetermined angle other than vertical with respect to the floor surface, the movement speed v of the sample (S) and the shooting time T are set to satisfy the above [Formula 15] to prevent the occurrence of artifacts.

[0166] Meanwhile, in relation to the above [Formula 15], M ob As the difference between vT and 2psecθ becomes larger, some of the high-frequency components of the X-ray image may be removed, causing blurring in the X-ray image, but M ob It was found that when vT and 2psecθ are close and the difference is small, artifacts can be removed and blurring of X-ray images can also be minimized.

[0167] Therefore, according to one embodiment of the present invention, the detection system (10) minimizes the blurring of the X-ray image and removes artifacts so that the moving speed v of the sample (S) and the shooting time T satisfy the above [Formula 15], but M ob It may be set to a level where the difference between vT and 2psecθ is minimized.

[0168] According to the above, the detection system (10) can remove spectral overlap artifacts without loss of image sensitivity and reduction of resolution by moving the sample (S) at a constant speed corresponding to a condition in which spectral overlap artifacts do not occur, thereby generating a motion blurring effect.

[0169] Hereinafter, the performance of the X-ray imaging device (100) and the image processing device (200) will be demonstrated based on an experimental example of detection of low-density foreign substances in food using the X-ray imaging device (100) and the image processing device (200) according to one embodiment of the present invention.

[0170] In this regard, in this study, we conducted an experiment on the detection of low-density foreign substances in food, and compared the contrast-to-noise ratio (CNR) and relative contrast gain (Relative Contrast Gain).

[0171] We present the results of a quantitative evaluation of the experimental results by considering the gain, RCG.

[0172] [Formula 16]

[0173]

[0174] The CNR metric can be defined as in [Equation 16] above. Here, μ is the mean value, σ 2 may be the variance of the foreign matter and food matrix (M) in the image. In this case, the areas where the foreign matter and food matrix are applied in the image are different from each other, and accordingly, the variance may be weighted by a factor provided by the ratio of the number of pixels of the foreign matter or food matrix to the total number.

[0175] Also, absorption images (I ABS ) for dark field imaging (I DF ) is the RCG metric of the absorption image (I ABS ) and dark field images (I DF ) was defined as in [Formula 17] below to directly compare the CNR characteristics between them.

[0176] [Formula 17]

[0177]

[0178] Here, CNR DF and CNR ABS are dark field images (I DF ) and absorption image (IABS ) may represent the CNR for.

[0179] Figure 8 is a diagram illustrating an example of an experimental setup based on autocorrelation length for the detection of low-density foreign matter in food. Additionally, Figure 9 is a diagram illustrating an example of an experimental implementation of absorption images and dark-field images based on autocorrelation length for the detection of low-density foreign matter in food.

[0180] Specifically, FIGS. 8 and 9 show absorption images (I) according to the autocorrelation length. ABS ) and dark field images (I DF ) may represent an example of an experiment to compare CNR characteristics.

[0181] Referring to FIGS. 8 and 9, the detection system (10) uses a cookie sample containing wooden chopsticks as a low-density foreign substance to obtain an absorption image (I ABS ) and dark field images (I DF ) was experimented to generate absorption images (I) from Setup 1 to Setup 5 according to the autocorrelation length. ABS ) and dark field images (I DF ) was obtained.

[0182] At this time, the absorption image (I ABS ), no significant change was observed according to the autocorrelation length for Setup 1 to Setup 5, and Fig. 9 shows the absorption image (I) according to Setup 1. ABS ) was displayed only (Fig. 9 (a)).

[0183] Meanwhile, (b) to (f) of FIG. 9 are dark field images (I) according to Setup 1 to Setup 5, respectively. DF ), referring to Fig. 9, as described above, the larger the autocorrelation length, the dark field image (I DF ) can clearly detect the low-density foreign matter, wooden chopsticks. This is confirmed by the dark-field image (I DF) may be due to the influence of the autocorrelation length when formed in contrast.

[0184] Figure 10 is a diagram showing the evaluation results for an experimental example of absorption imaging and dark-field imaging according to autocorrelation length for the detection of low-density foreign substances in food. Specifically, Figure 10 shows the calculated values ​​for CNR (Figure 10 (a)) and RCG (Figure 10 (b)) for the experimental examples for Figures 8 and 9.

[0185] Referring to Fig. 10 (a), CNR DF The values ​​were found to increase as the autocorrelation length increased, while the CNR ABS The values ​​did not show any significant difference even with changes in the autocorrelation length. This is because the autocorrelation length is a dark-field image (I DF ) affects the CNR characteristics of the absorption image (I ABS ) may prove that it does not have a significant effect. On the other hand, when the autocorrelation length is 59 nm (Setup 5), CNR DF The value is CNR ABS It was found to be smaller than the value, and this is because the SAXS signal becomes weaker and is not measured as the autocorrelation length becomes smaller. Considering this, in the detection system (10) according to one embodiment of the present invention, increasing the autocorrelation length is a dark field image (I) for detecting low-density foreign substances in the sample (S). DF ) is a major factor in improving the performance. In addition, referring to Fig. 10 (b), the RCG value was found to increase as the autocorrelation length increases.

[0186] Figure 11 is a diagram showing an example of an experiment of absorption images, dark-field images, and fusion images for samples containing low-density foreign substances such as wooden chopsticks and Styrofoam.

[0187] Specifically, FIG. 11 shows absorption images (I) for cookie samples, ketchup samples, and meat samples containing wooden chopsticks and Styrofoam as foreign substances. ABS ), dark field imaging (I DF ) and fusion images. Fig. 11 (a) shows an image of a cookie sample, Fig. 11 (b) shows an image of a ketchup sample, and Fig. 11 (c) shows an image of a meat sample, and from the top, the absorption image (I ABS ), dark field imaging (I DF ) and fusion images.

[0188] At this time, the image of Fig. 11 may be acquired based on Setup 1, which has the largest autocorrelation length among Setup 1 to Setup 5.

[0189] Referring to Figure 11, for all samples, including cookie samples, ketchup samples, and meat samples, absorption images (I ABS ) no foreign substances were clearly detected, but dark field images (I DF ) can clearly show the presence of foreign substances. However, in the dark field image (I DF ) does not clearly include the shape of the sample (S), so the dark field image (I DF ) alone, it is difficult to determine where the foreign substances are contained within the sample (S).

[0190] Accordingly, the detection system (10) according to one embodiment of the present invention, absorption image (I ABS ) and dark field images (I DF ) can be generated as a fused image, and referring to FIG. 11, it can be confirmed that the shape of the sample (S) and the shape and location of the foreign substance of the sample (S) are clearly displayed in the fused image.

[0191] Figure 12 is a diagram showing the evaluation results for an experimental example of absorption images, dark-field images, and fusion images for a sample containing a low-density foreign substance, such as wooden chopsticks. Furthermore, Figure 13 is a diagram showing the evaluation results for an experimental example of absorption images, dark-field images, and fusion images for a sample containing a low-density foreign substance, such as Styrofoam.

[0192] Specifically, FIGS. 12 and 13 show the calculated values ​​for CNR (FIG. 12(a) and FIG. 13(a)) and RCG (FIG. 12(b) and FIG. 13(b)) for the experimental example for FIG. 11.

[0193] Referring to (a) of Fig. 12 and (a) of Fig. 13, for the sample containing wooden chopsticks as a foreign substance and the sample containing Styrofoam, CNR DF The value is CNR ABS It was shown to show high performance for the value, and referring to Fig. 12 (b) and Fig. 13 (b), the RCG value was shown to be in the range of 2.8 to 10.5, which is a dark field image (I DF ) detects low-density foreign substances such as wooden chopsticks and Styrofoam, using absorption imaging (I ABS ) can be confirmed to be effective.

[0194] Fig. 14 is a diagram showing an example of an experiment of absorption images and dark-field images for samples including low-density foreign substances such as mechanical pencil lead and plastic bullets. Specifically, Fig. 14 shows the results of an additional experiment performed on samples and foreign substances other than the samples and foreign substances used in the experiments described with reference to Figs. 9 to 13, and absorption images (I) for a banana sample including a sharp mechanical pencil lead as a foreign substance (see (a) of Fig. 14) and a chicken breast sample including a soft plastic bullet as a foreign substance (see (b) of Fig. 14) ABS ) and dark field images (I DF )am.

[0195] Referring to Fig. 14, for other samples including foreign substances other than the samples and foreign substances shown through Figs. 9 to 13, dark field images (I DF ) It can be confirmed that foreign substances were clearly detected.

[0196] Hereinafter, the performance of the X-ray imaging device (100) and the image processing device (200) will be demonstrated based on an experimental example of artifact removal using the X-ray imaging device (100) and the image processing device (200) according to one embodiment of the present invention.

[0197] Fig. 15 is a diagram illustrating an experimental example of spectral overlap artifact removal using an extended window function. In addition, Fig. 16 is a diagram illustrating another experimental example of spectral overlap artifact removal using an extended window function.

[0198] Specifically, (a) of FIG. 15 and (a) of FIG. 16 are dark-field images obtained through a conventional technique, and (b) of FIG. 15 and (b) of FIG. 16 are dark-field images reflecting an extended window function according to one embodiment of the present invention.

[0199] Referring to FIG. 15, it can be seen that the dark-field image reflecting the extended window function according to one embodiment of the present invention has improved spatial resolution compared to the dark-field image acquired through the prior art. In addition, when the high-resolution portions were extracted for (a) and (b), it can be seen that the corresponding portions appear more clearly in the dark-field image reflecting the extended window function compared to the dark-field image acquired through the prior art.

[0200] Referring to FIG. 16, it can be seen that the spectral overlap artifact clearly appears in the form of black / white stripes in the dark field image obtained through the conventional technique shown in (a) of FIG. 16, whereas the spectral overlap artifact is cleanly removed in the dark field image reflecting the extended window function according to one embodiment of the present invention shown in (b) of FIG. 16.

[0201] Fig. 17 is a diagram showing the evaluation results for another experimental example for removing spectral overlap artifacts using an extended window function. Specifically, Fig. 17 is a diagram for evaluating CNR according to partial Fourier domain amounts (PF), where (a) of Fig. 17 shows dark-field images according to different partial Fourier domain amounts (PF), and (b) of Fig. 17 shows dark-field images according to different partial Fourier domain amounts (PF).

[0202] At this time, the partial Fourier domain amount (PF) means the size of the first frequency function, and in (a) of Fig. 17, a partial Fourier domain amount (PF) of 0 means that sampling was not performed for the first frequency, and a partial Fourier domain amount of 0.5 may mean that sampling was performed to obtain half the domain for the first frequency.

[0203] In this regard, referring to Fig. 17, it can be confirmed that the CNR value is higher when the partial Fourier domain amount (PF) is 0.5 or more and closer to 0.5 (PF = 0.55). That is, it can be meant that the stronger the dark field signal with artifacts removed can be obtained when the partial Fourier domain amount (PF) is 0.5 or more and closer to 0.5.

[0204] On the other hand, when the partial Fourier domain amount (PF) is 0.5, the CNR is actually reduced. This is believed to be due to noise or multiple artifacts in the experimental data. Consequently, when more than half of the area in the first frequency domain is acquired, artifacts are removed, and a darkfield image with improved resolution can be obtained.

[0205] As described above, the detection system (10) according to one embodiment of the present invention can detect and clearly display low-density foreign substances in food by obtaining dark-field images and fusion images through single-shot grid-based dark-field X-ray photography with low system setup complexity and required dose using only a single shot and one X-ray grid.

[0206] Below, we will briefly review the operating flow of the present invention based on the detailed description above.

[0207] FIG. 18 is a flowchart illustrating a method for removing artifacts in an X-ray image using motion blurring through an X-ray imaging device according to one embodiment of the present invention.

[0208] The method for removing artifacts in an X-ray image using motion blurring through an X-ray imaging device illustrated in FIG. 18 can be performed by the X-ray imaging device (100) described above. Therefore, even if the content is omitted below, the content described for the X-ray imaging device (100) can be equally applied to the description of the method for removing artifacts in an X-ray image using motion blurring through an X-ray imaging device.

[0209] Referring to FIG. 18, in step S11, the X-ray source (110), the sample (S), and the grid (120) may be arranged in a straight line. At this time, the grid (120) may be arranged between the sample (S) and the detector (130), but may be arranged at a position adjacent to the sample (S) with respect to the detector (130).

[0210] Next, in step S12, X-rays can be irradiated from an X-ray source (110) toward the sample (S).

[0211] Next, in step S13, the sample (S) may be linearly moved at a preset speed by the mover (140). At this time, in step S13, the sample (S) may be moved at a preset speed for a preset shooting time.

[0212] Specifically, when the grid (120) is arranged vertically based on a straight line where the X-ray source (110), the sample (S) and the grid (120) are arranged, the preset speed and the preset shooting time are It may be satisfying.

[0213] Here, M ob may represent the sample magnification, v may represent the preset speed, T may represent the preset shooting time, and p may represent the grid period.

[0214] In addition, when the grid (120) is arranged at a predetermined angle with respect to the straight line where the X-ray source (110), sample (S) and grid (120) are arranged, the preset speed and preset shooting time are It may be satisfying.

[0215] Here, M ob is the sample magnification, v is the preset speed, T is the preset shooting time, p is the grid period, and the predetermined angle may be an angle other than vertical.

[0216] Next, in step S14, X-rays transmitted from the grid (120) to the sample (S) can be detected through the detector (130).

[0217] Here, artifacts in the X-ray image detected through the detector (130) can be removed due to the motion blurring phenomenon caused by the movement of the sample (S) by the mover (140).

[0218] Figure 19 is a flowchart of a method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography.

[0219] The method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography, as illustrated in FIG. 19, can be performed by the device (200) for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography, as described above. Therefore, even if the content is omitted below, the content described for the device (200) for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography can be equally applied to the description for the method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography.

[0220] Referring to FIG. 19, in step S21, the acquisition unit (210) captures a sample image (I) including a sample (S) and a grid (120) that is captured through grid (120)-based X-ray photography. s ) and a reference image (I) containing only the grid (120) and not including food samples. r ) can be obtained.

[0221] At this time, the grid (120) is included in the sample (S) at least one low-density foreign substance in the dark field image (I DF ) is positioned so as to be detected at an identifiable level, and may be positioned between the sample (S) and the detector (130), but may be positioned adjacent to the sample (S) with respect to the detector (130).

[0222] Next, in step S22, the restoration unit (220) samples the image (I s ) and reference images (I r ) using absorption imaging (I ABS ) and dark field images (I DF ) can be restored.

[0223] Next, in step S23, the fusion part (230) is an absorption image (I ABS ) and dark field images (I DF ) can be fused to create a fused image. At this time, the fused image is a dark field image (I DF ) may clearly reflect the location and shape of at least one low-density foreign body.

[0224] FIG. 20 is a flowchart illustrating a method for restoring absorption images and dark-field images using a food low-density foreign substance detection device (X-ray image processing device) using a single-shot grid-based dark-field X-ray imaging, and a method for removing artifacts in X-ray images using an extended window function.

[0225] Specifically, FIG. 20 is a flowchart of a method for restoring an absorption image and a dark-field image included in a method for detecting low-density foreign substances in food using the single-shot grid-based dark-field X-ray imaging described above, and a method for removing artifacts in an X-ray image using an extended window function is included in the method for restoring an absorption image and a dark-field image, and a description of a method for removing artifacts in an X-ray image using an extended window function may be included in a description of a method for restoring an absorption image and a dark-field image.

[0226] Therefore, the method for removing artifacts in an X-ray image using the extended window function illustrated in FIG. 20 can be performed by the food low-density foreign substance detection device (200) using the single-shot grid-based dark-field X-ray imaging described above. Therefore, even if the content is omitted below, the content described for the food low-density foreign substance detection device (200) using the single-shot grid-based dark-field X-ray imaging can be equally applied to the description of the method for removing artifacts in an X-ray image using the extended window function.

[0227] Referring to FIG. 20, in step S201, the area identification unit (221) of the restoration unit (220) identifies a sample image (Is ) and reference images (I r ) by applying the Fourier transform, the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak can be identified.

[0228] Next, in step S202, the area separation unit (222) of the restoration unit (220) can separate the area surrounding the 0th harmonic peak and the area surrounding the 1st harmonic peak using a band-pass filter.

[0229] At this time, in step S202, the area separation unit (222) of the restoration unit (220) applies the partial Fourier technique to the absence area outside the presence area of ​​the artifact to generate a dark field image (I DF ) may be used to remove artifacts by not reflecting the area of ​​presence of the artifacts. In this case, the partial Fourier technique may be a zero-filling technique.

[0230] Next, in step S203, the harmonic image acquisition unit (225) of the restoration unit (220) can acquire multiple harmonic images by applying inverse Fourier transform to the surrounding area of ​​the separated 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak.

[0231] Next, in step S204, the image restoration unit (224) of the restoration unit (220) uses multiple harmonic images to create an absorption image (I ABS ) and dark field images (I DF ) can be restored.

[0232] In the above description, steps S11 to S14, steps S21 to S23, and steps S201 to S204 may be further divided into additional steps or combined into fewer steps, depending on the implementation example of the present invention. In addition, some steps may be omitted as needed, and the order between the steps may be changed.

[0233] According to one embodiment of the present invention, a method for detecting low-density foreign substances in food using single-shot grid-based dark-field X-ray imaging, a method for removing artifacts in X-ray images using motion blurring, and a method for removing artifacts in X-ray images using an extended window function may be implemented in the form of program commands that can be executed by various computer means and recorded on a computer-readable medium. The computer-readable medium may include program commands, data files, data structures, etc., either singly or in combination. The program commands recorded on the medium may be those specially designed and configured for the present invention or may be known and usable by those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program commands such as ROMs, RAMs, and flash memories. Examples of program instructions include not only machine language codes, such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

[0234] In addition, the method for detecting low-density foreign substances in food using the single-shot grid-based dark-field X-ray imaging described above, the method for removing artifacts in X-ray images using motion blurring, and the method for removing artifacts in X-ray images using an extended window function can also be implemented in the form of a computer program or application executed by a computer and stored in a recording medium.

[0235] The above description of the present invention is for illustrative purposes only, and those skilled in the art will readily appreciate that the present invention can be readily modified into other specific forms without altering the technical spirit or essential characteristics of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single entity may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined manner.

[0236] The scope of the present invention is indicated by the claims described below rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention.

Claims

1. In a device for detecting low-density foreign matter in food using dark-field X-ray photography based on a single-shot grid, An acquisition unit for acquiring a sample image including a food sample and a grid, and a reference image including only the grid and not the food sample, captured through grid-based X-ray imaging; A restoration unit that restores an absorption image and a dark-field image using the sample image and the reference image; and A fusion unit that fuses the above absorption image and the above dark field image to generate a fusion image, Including, but not limited to, A detection device wherein the grid is positioned so that at least one low-density foreign substance contained in the food sample is detected at a level identifiable in the dark field image.

2. In paragraph 1, A detection device, wherein the grid is positioned between the food sample and the detector, but is positioned adjacent to the food sample relative to the detector.

3. In paragraph 2, A detection device wherein the food sample, the grid, and the detector are arranged in a straight line.

4. In paragraph 1, The above restoration part, An area identification unit that applies Fourier transform to the sample image and the reference image to identify an area around the 0th harmonic peak and an area around the 1st harmonic peak; A region separation unit that separates the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak using a bandpass filter; A harmonic image acquisition unit that obtains multiple harmonic images by applying an inverse Fourier transform to the surrounding area of ​​the separated 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak; and An image restoration unit that restores the absorption image and the dark field image using the plurality of harmonic images, A detection device comprising:

5. In paragraph 1, A detection device wherein the above fused image clearly reflects the location and shape of at least one low-density foreign substance compared to the dark field image.

6. In paragraph 3, A detection device wherein the above food sample is linearly moved by a moving device at a preset speed for a preset shooting time.

7. In paragraph 6, A detection device in which an artifact is removed from an X-ray image detected through the detector due to a motion blurring phenomenon caused by movement of the sample.

8. In paragraph 4, The above area separation section is, A detection device that removes an artifact by applying an extended window function using a partial Fourier technique to an absence area outside the presence area of ​​an artifact in an X-ray image detected by a detector so that the presence area is not reflected in the dark field image.

9. In paragraph 8, An X-ray image processing device wherein the above partial Fourier technique is a zero-filling technique.

10. In paragraph 7, When the above grid is arranged vertically based on the above straight line, the preset speed and the preset shooting time satisfy the following equation 1, [Formula 1] Here, Mob represents the sample magnification, v represents the preset speed, T represents the preset shooting time, and p represents the grid period.

11. In paragraph 7, When the above grid is arranged at a predetermined angle with respect to the above straight line, the preset speed and the preset shooting time satisfy the following equation 2. [Formula 2] Here, Mob represents the sample magnification, v represents the preset speed, T represents the preset shooting time, and p represents the grid period. A detection device wherein the above-mentioned predetermined angle is an angle other than vertical.

12. In a food low-density foreign matter detection system using a single-shot grid-based dark-field X-ray imaging, An X-ray imaging device that irradiates a food sample with X-rays and detects the X-rays passing through a grid; and A food low-density foreign matter detection device that detects at least one low-density foreign matter contained in the food sample based on an X-ray image of the detected X-ray, A detection system comprising:

13. In paragraph 12, The above food low density detection device is, A detection system, wherein the X-ray image comprises a sample image including a food sample and a grid, and a reference image including only the grid but not the food sample, the sample image and the reference image are used to restore an absorption image and a dark-field image, and the absorption image and the dark-field image are fused to generate a fused image, wherein the grid is positioned at a position such that at least one low-density foreign substance included in the food sample is detected at a level that can be identified in the dark-field image.

14. A method for detecting low-density foreign matter in food using single-shot grid-based dark-field X-ray photography, A step of obtaining a sample image including a food sample and a grid, and a reference image including only the grid and not the food sample, which is captured using grid-based X-ray imaging; A step of restoring an absorption image and a dark field image using the sample image and the reference image; and A step of generating a fused image by fusing the above absorption image and the above dark field image, Including, but not limited to, A detection method, wherein the grid is positioned at a position such that at least one low-density foreign substance contained in the food sample is detected at a level that can be identified in the dark field image.

15. In paragraph 14, The above restoration steps are: A step of applying a Fourier transform to the sample image and the reference image to identify a surrounding area of ​​the 0th harmonic peak and a surrounding area of ​​the 1st harmonic peak; A step of separating the surrounding area of ​​the 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak using a band-pass filter; A step of obtaining multiple harmonic images by applying an inverse Fourier transform to the surrounding area of ​​the separated 0th harmonic peak and the surrounding area of ​​the 1st harmonic peak; and A step of restoring the absorption image and the dark field image using the plurality of harmonic images, A detection method comprising:

16. In paragraph 14, A detection method wherein the above fused image clearly reflects the location and shape of at least one low-density foreign substance compared to the dark field image.

17. In paragraph 14, A detection method wherein the above food sample is linearly moved by a moving device at a preset speed for a preset shooting time.

18. In paragraph 17, A detection method in which an artifact is removed from an X-ray image detected through a detector due to a motion blurring phenomenon caused by movement of the sample.

19. In paragraph 16, The above separation step is, A detection method for removing an artifact by applying an extended window function using a partial Fourier technique to an absence area outside the presence area of ​​an artifact in an X-ray image detected by a detector so that the presence area is not reflected in the dark field image.

20. In paragraph 19, The above partial Fourier technique is a detection method that is a zero-filling technique.

21. In paragraph 18, When the above grid is arranged vertically based on the above straight line, the preset speed and the preset shooting time satisfy the following equation 1, [Formula 1] Here, Mob represents the sample magnification, v represents the preset speed, T represents the preset shooting time, and p represents the grid period, a method for removing artifacts.

22. In paragraph 18, When the above grid is arranged at a predetermined angle with respect to the above straight line, the preset speed and the preset shooting time satisfy the following equation 2. [Formula 2] Here, Mob represents the sample magnification, v represents the preset speed, T represents the preset shooting time, and p represents the grid period. A method for removing an artificial object, wherein the above-mentioned predetermined angle is an angle other than vertical.

Images