Image processing device, image processing method, and image processing program

Abstract

An acquisition unit 36 acquires an input image capturing a pattern of a predetermined target. A first detection unit 38A uses pattern information for a portion of the target to detect, from the input image acquired by the acquisition unit 36, a portion of the pattern of the target in the input image. A second detection unit 38B detects the pattern of the target by using the detection result of the portion of the pattern of the target and target information for identifying the target.

Description

Image processing device, image processing method, and image processing program

[0001] The present disclosure relates to an image processing device, an image processing method, and an image processing program.

[0002] JP 2024-28042 A discloses an information processing device that performs processing to correct defective pixels in a radiographic image obtained by radiographing a physical phantom on which a test pattern is formed, the information processing device including a processor, wherein the processor inputs an image based on the radiographic image into a first machine-learned model to detect an area including the test pattern, determines the edge direction of the test pattern within the area, treats defective pixels within the area as pixels to be corrected, and corrects the pixels to be corrected using normal pixels within the area that are present in the edge direction from the pixels to be corrected, other than the defective pixels.

[0003] Japanese Patent Application Laid-Open No. 2004-150908 discloses a non-destructive inspection method that eliminates cumbersome manual operations by capturing an image of an inspection object and an image quality confirmation pattern and automating the image quality confirmation process.

[0004] Japanese Patent Publication No. 2009-136421 discloses an X-ray diagnostic apparatus having a search unit that searches a search area for a pattern that is the same as the pattern of pixel values ​​in a reference area, and determines the position of the pattern that has been found.

[0005] When taking X-ray images for medical treatment or non-destructive testing, a chart is photographed for proper diagnosis or testing, and the image is analyzed to evaluate the captured image for its resolution, contrast, etc. Furthermore, the captured image is corrected and the imaging device is adjusted based on the evaluation results of the captured image.

[0006] When evaluating X-ray images based on a chart, unclear portions of the chart are difficult to identify because their position, size, and angle cannot be determined. For example, when using a Duplex-type IQI chart, thin line pairs among multiple line pairs are sometimes unclear and difficult to identify. When using a Hole-type IQI chart, small holes among multiple holes are sometimes unclear and difficult to identify. When using a Wire-type IQI chart, thin wires among multiple wires are sometimes unclear and difficult to identify. In either case, unclear and unidentifiable line pairs, holes, or wires are difficult to evaluate. IQI stands for Image Quality Indicator, and is a chart used to evaluate images during radiography for nondestructive testing. There are several types of IQIs: Duplex-type IQIs are also called multi-line image quality meters, Hole-type IQIs are called hole-type penetrometers, and Wire-type IQIs are called wire-type penetrometers.

[0007] Furthermore, because evaluating a chart in a captured image is time-consuming, it is preferable to be able to automatically detect each part of the chart to be evaluated. However, when the chart in the captured image has individual differences or a distorted shape, it has been difficult to automatically and accurately detect each part of the chart. This is not limited to when the captured image is an X-ray image, and it is not limited to when the detection target is a chart. When the target includes unclear parts, distorted parts, or parts with individual differences, it has been difficult to accurately detect those parts.

[0008] The present disclosure has been made in consideration of the above facts, and aims to provide an image processing device, an image processing method, and an image processing program that can accurately detect the pattern of an object even if the object includes unclear and / or distorted parts or parts with individual differences.

[0009] In order to achieve the above object, an image processing device according to a first aspect of the present disclosure includes a processor, which detects a portion of a target pattern in an input image showing a predetermined target pattern using pattern information of the portion of the target, and performs processing to detect the target pattern using the partial detection result and target information for identifying the target.

[0010] An image processing device according to a second aspect of the present disclosure is the image processing device according to the first aspect, wherein the target is a chart for evaluating an image, and the input image is an X-ray transmission image.

[0011] An image processing device according to a third aspect of the present disclosure is the image processing device according to the first aspect, in which the processor accepts a range in which to detect a portion of a target pattern in an input image, and detects a portion of the target pattern from the accepted range.

[0012] An image processing device according to a fourth aspect of the present disclosure is the image processing device according to the first aspect, wherein the processor detects a target pattern using a detection result of a portion of the pattern and the relative positional relationship of the pattern as target information.

[0013] An image processing device according to a fifth aspect of the present disclosure is an image processing device according to the first aspect, wherein the pattern information is information including at least one of clear parts of the target pattern, parts with little individual variation, and parts with little distortion, information about parts of the target pattern that do not include characters, or information including machine learning parameters.

[0014] In an image processing device according to a sixth aspect of the present disclosure, in the image processing device according to the first aspect, the target information includes information including at least one of unclear parts of the target pattern, parts with individual differences, and parts with large distortion, or information of parts that are different from the pattern information.

[0015] An image processing device according to a seventh aspect of the present disclosure is the image processing device according to the first aspect, wherein the processor detects the target pattern by detecting at least one of the presence or absence, position, number, type, angle, scale, and shape of the target pattern.

[0016] An image processing device according to an eighth aspect of the present disclosure is an image processing device according to the first aspect, wherein the processor further performs processing to display at least one of the presence or absence, number, type, position, angle, scale, and shape of the detected object.

[0017] In the image processing device according to a ninth aspect of the present disclosure, the processor further performs processing to evaluate the input image.

[0018] An image processing device according to a tenth aspect of the present disclosure is the image processing device according to the second aspect, wherein the chart is a multi-linear image quality meter.

[0019] An image processing device according to an eleventh aspect of the present disclosure is the image processing device according to the tenth aspect, wherein the pattern information is information about a pattern including at least one of the first to sixth line pairs of the multi-linear image quality meter.

[0020] An image processing device according to a twelfth aspect of the present disclosure is the image processing device according to the tenth aspect, wherein the target information includes information of at least one line pair of the multi-linear image quality meter.

[0021] An image processing device according to a thirteenth aspect of the present disclosure is the image processing device according to the twelfth aspect, wherein the target information includes information on line pairs that are not included in the pattern information.

[0022] An image processing device according to a fourteenth aspect of the present disclosure is the image processing device according to the twelfth aspect, wherein the processor detects at least one line pair from the third onward as the detection of the target pattern.

[0023] An image processing device according to a fifteenth aspect of the present disclosure is the image processing device according to the twelfth aspect, wherein the processor detects at least one line pair from the fourteenth onward as the detection of the target pattern.

[0024] An image processing device according to a sixteenth aspect of the present disclosure is the image processing device according to the twelfth aspect, wherein the processor calculates at least a one-dimensional density profile as the pattern detection.

[0025] An image processing device according to a seventeenth aspect of the present disclosure is the image processing device according to the tenth aspect, wherein the processor further performs processing to display at least one detected line pair.

[0026] An image processing device according to an eighteenth aspect of the present disclosure is the image processing device according to the tenth aspect, wherein the processor further performs processing to evaluate contrast of at least one line pair.

[0027] An image processing device according to a nineteenth aspect of the present disclosure is the image processing device according to the second aspect, wherein the chart is a perforated penetrameter.

[0028] An image processing device according to a twentieth aspect of the present disclosure is the image processing device according to the nineteenth aspect, wherein the pattern information is information about a pattern that includes at least characters.

[0029] An image processing device according to a 21st aspect of the present disclosure is the image processing device according to the 19th aspect, wherein the target information includes at least one of information on at least one hole and information on an outer frame.

[0030] An image processing device according to a twenty-second aspect of the present disclosure is the image processing device according to the twenty-first aspect, wherein the processor detects at least one hole in a perforated penetrometer as the detection of the target pattern.

[0031] An image processing device according to a twenty-third aspect of the present disclosure is the image processing device according to the nineteenth aspect, wherein the processor further performs processing to display the at least one detected hole.

[0032] An image processing device according to a 24th aspect of the present disclosure is the image processing device according to the 19th aspect, wherein the processor further performs processing to evaluate the contrast of the at least one detected hole.

[0033] An image processing device according to a 25th aspect of the present disclosure is the image processing device according to the first aspect, wherein at least one of the pattern information and the target information is registered by a user.

[0034] An image processing method according to a 26th aspect of the present disclosure includes detecting a portion of a target pattern in an input image showing a predetermined target pattern using pattern information of the portion of the target, and detecting the target pattern using the partial detection result and target information for identifying the target.

[0035] An image processing program according to a 27th aspect of the present disclosure is a program for causing a computer to execute processing including detecting a portion of a target pattern in an input image showing a predetermined target pattern using pattern information of the portion of the target, and detecting the target pattern using the partial detection result and target information for identifying the target.

[0036] According to the present disclosure, it is possible to provide an image processing device, an image processing method, and an image processing program that can accurately detect patterns even when there are unclear and / or distorted parts of the object, or parts that vary from one individual to another.

[0037] 1 is a block diagram showing an example of the overall configuration of a radiographic image capturing system according to the present embodiment. FIG. 2 is a block diagram showing an example of the configuration of a main part of a control device. FIG. 3 is a functional block diagram showing the functional configuration of a control unit in the control device of the radiographic image capturing system according to the present embodiment. FIG. 4 is a flowchart showing an example of the flow of processing performed by the control unit in the control device of the radiographic image capturing system according to the present embodiment. FIG. 5 is a diagram for explaining an example of pattern information of a part of a chart. FIG. 6 is a diagram showing an example of an input image in which a chart is reflected. FIG. 7 is a diagram showing an example of the result of detecting a chart. FIG. 8 is a diagram showing an example of the overall shape of a Duplex type IQI. FIG. 9 is a diagram for explaining an example of target information of the chart when the chart is a Duplex type IQI. FIG. 10 is a diagram showing an example of a display of evaluation results. FIG. 11 is a diagram showing an example of a chart of a Hole type IQI. FIG. 12 is a diagram showing an example of a chart 50 of a Wire type IQI.

[0038] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings, but the present invention is not limited to the embodiment.

[0039] In this embodiment, a radiographic image capturing system including a control device as an example of an image processing device will be described as an example. Fig. 1 is a block diagram showing an example of the overall configuration of a radiographic image capturing system according to this embodiment.

[0040] As shown in FIG. 1, the radiation image capturing system 10 includes a radiation irradiation device 12, a radiation image capturing device 16, and a control device 18.

[0041] The radiation irradiator 12 according to this embodiment includes a radiation source 14 that irradiates a subject W, which is an example of an imaging target, with radiation R, such as X-rays (X-rays). The method of instructing the radiation irradiator 12 to irradiate radiation R is not particularly limited. For example, if the radiation irradiator 12 includes an irradiation button or the like, a user such as a radiologist may issue an instruction to irradiate radiation R using the irradiation button, thereby irradiating radiation R from the radiation irradiator 12. Alternatively, for example, a user such as a radiologist may issue an instruction to irradiate radiation R by operating the control device 18, thereby irradiating radiation R from the radiation irradiator 12.

[0042] When the radiation irradiating device 12 receives an instruction to irradiate radiation R, it irradiates radiation R from the radiation source 14 in accordance with irradiation conditions such as set tube voltage, tube current, and irradiation period. Note that, hereinafter, the dose of radiation R will be referred to as the "radiation dose."

[0043] The radiographic imaging device 16 according to this embodiment includes a radiation detector 20 that detects radiation R that is irradiated from the radiation irradiation device 12 and passes through the subject W. The radiographic imaging device 16 uses the radiation detector 20 to capture a radiographic image of the subject W.

[0044] 2, the control device 18 includes a control unit 22, a display unit 32, an operation unit 34, and a communication I / F (interface) unit 35. FIG. 2 is a block diagram showing an example of the configuration of the main parts of the control device 18.

[0045] The main part of the control unit 22 is configured as a general computer, and controls the overall operation of the control device 18 .

[0046] The control unit includes a CPU 24, a ROM 26, and a RAM 28. The ROM 26 pre-stores various programs 30 executed by the CPU 24, such as an image capture processing program and / or a display processing program executed when controlling image capture. The RAM 28 temporarily stores various data.

[0047] The display unit 32 displays a screen for operating the radiation irradiating device 12 and / or the radiation image capturing device 16, and / or captured radiation images, etc. The display unit 32 is, for example, a direct-view electronic display.

[0048] The operation unit 34 is used by the user to input instructions and / or various information related to photography, etc. The operation unit 34 is not particularly limited, and examples thereof include various switches, a touch panel, a touch pen, and a mouse.

[0049] The communication I / F unit 35 communicates captured images such as radiographic images and various information between the radiation irradiating device 12 and the radiographic image capturing device 16 via wireless communication or wired communication.

[0050] In radiography for medical and / or non-destructive testing, a chart is photographed for proper diagnosis and / or testing, and the photographed image is analyzed to determine resolution, contrast, etc., and evaluated. Based on the evaluation results, the photographed image is corrected and / or the photographing device is adjusted.

[0051] However, when evaluating an X-ray image based on the chart, it is difficult to identify the position, size, angle, etc. of unclear parts of the chart, making it difficult to evaluate the X-ray image.

[0052] For example, in the case of a Duplex type IQI, the thin line pair, in the case of a Hole type IQI, the small hole, and in the case of a Wire type IQI, the thin wire were unclear and could not be identified, making it difficult to evaluate contrast.

[0053] Furthermore, since evaluating a chart in a photographed image is time-consuming, it is preferable to automate the process. Even when the evaluation is performed manually, the time and effort can be reduced by simply automatically detecting each part of the chart to be evaluated. However, when the chart in the photographed image has individual differences and / or shape distortions, it is difficult to automatically detect each part to be evaluated.

[0054] Therefore, in this embodiment, the control device 18 detects a portion of the target pattern in an input image that shows a predetermined target pattern using pattern information of the portion of the target, and then performs a process of detecting the target pattern using the partial detection result and target information for identifying the target.

[0055] Specifically, the CPU 24 executes a program 30 stored in the ROM 26 of the control unit 22 to provide the functions shown in Fig. 3. Fig. 3 is a functional block diagram showing the functional configuration of the control unit 22 in the control device 18 of the radiation image capturing system 10 according to this embodiment.

[0056] That is, the control unit 22 has the functions of an acquisition unit 36, a detection unit 38, and an evaluation unit 40, as shown in FIG.

[0057] The acquisition unit 36 ​​acquires an input image showing a pattern of a predetermined target. In this embodiment, a chart for evaluating an image is used as an example of the predetermined target, and a radiographic image is used as an example of the input image. That is, the input image is acquired by inputting the radiographic image showing the chart from the radiographic imaging device 16 to the control device 18 as an input image.

[0058] The detection unit 38 includes a first detection unit 38A and a second detection unit 38B, and detects a target pattern in an input image using predetermined pattern information of a portion of the target and target information. The first detection unit 38A detects a portion of the target pattern in the input image acquired by the acquisition unit 36 ​​using the pattern information of the portion of the target. The second detection unit 38B detects the target pattern using the detection result of the portion of the target pattern and target information for identifying the target. In the above configuration, the pattern information of the portion of the target refers to information including at least one of a clear portion of the target, a portion with little individual variation, and a portion with little influence of distortion, or information about a portion of the target pattern that does not contain characters. This configuration enables pattern detection even if the target pattern in the input image contains unclear portions, distortion, or individual variations. Note that an example of the detection result of the detection unit 38 is the result of detecting at least one of the presence, absence, number, type, position, angle, scale, and shape of the target pattern. The target information may also include information about a portion different from the pattern information of the portion of the target. The target information may also include information on at least one of unclear portions of the target pattern, portions with individual differences, and portions significantly affected by distortion. If the target chart is a duplex type IQI, the target information may also include information on line pairs that are not included in some of the pattern information of the IQI.

[0059] The evaluation unit 40 evaluates the input image using the detection result of the target pattern. The evaluation unit 40 may evaluate the input image using the input image and evaluation information necessary for evaluating the chart. The evaluation information includes information on a portion different from the target's partial pattern information, or information obtained from the target's partial pattern information and / or target information. An example of the evaluation information may be information on each portion that is the subject of the target evaluation. Furthermore, if the target is a Duplex type IQI, the evaluation information may include information on line pairs that are not included in the target's partial pattern information.

[0060] 4 is a flowchart showing an example of the flow of processing performed by the control unit 22 in the control device 18 of the radiographic image capturing system 10 according to this embodiment. The processing in FIG. 4 starts, for example, when an instruction to evaluate a radiographic image is given.

[0061] In step 100, the CPU 24 inputs an image and proceeds to step 102. That is, the acquisition unit 36 ​​acquires an input image by inputting an input image showing a chart for evaluating the image from the radiographic image capturing device to the control device 18.

[0062] In step 102, the CPU 24 detects a part of the object in the input image using the pattern information of the part of the object, and then proceeds to step 103. That is, the first detection unit 38A detects a part of the pattern of the chart in the input image acquired by the acquisition unit 36 ​​using the pattern information of the part of the chart.

[0063] In step 103, the CPU 24 detects a target pattern using the input image and the target information, and then proceeds to step 104. That is, the second detection unit 38B detects the chart pattern using the detection result of a portion of the chart pattern and the target information for identifying the chart. For example, the second detection unit 38B detects the target pattern using the detection result of a portion of the pattern and the relative positional relationship of the pattern as the target information.

[0064] In step 104, the CPU 24 evaluates the input image and proceeds to step 106. That is, the evaluation unit 40 evaluates the input image using the input image, the detection result of the target chart, and the evaluation information of the target chart. Note that the evaluation unit 40 may evaluate the input image using the detection result of the target chart. Alternatively, the evaluation unit 40 may identify relative position information of at least some of the patterns obtained from the evaluation information based on the state of the chart identified from the detected chart, and evaluate the input image based on the identification result.

[0065] In step 106, the CPU 24 displays the evaluation result of the input image on the display unit 32, and then the series of processes ends.

[0066] With the above configuration, the present invention first reliably detects an object in an input image using information on a portion of the pattern where the object is clear, is less affected by distortion, and / or has little individual variation, and then accurately detects the pattern including the portion of the object where there is unclearness, distortion, and / or individual variation based on the detection result and the object information for identifying the object. When the object is a chart, accurately detecting the chart pattern makes it possible to accurately identify and evaluate the position, size, angle, etc. of the portion of the chart where there is unclearness, distortion, and individual variation.

[0067] Hereinafter, a specific example will be described in which the input image is an X-ray transmission image and the target is a chart and an IQI. However, the present disclosure is also effective for other charts and images that have similar issues other than IQI. Furthermore, even if the target is not a chart, the present disclosure is effective in cases where the target contains unclear, distorted, and / or individual-varying parts, but accurate detection of the pattern is required. In other words, the input image in this disclosure is not limited to an X-ray transmission image and the target is not limited to a chart.

[0068] (Specific Example 1) A specific example of the pattern information of a portion of a chart that is used by the first detection unit 38A when detecting a chart in an input image using the pattern information of a portion of the chart in step 102 will be described. An example of the pattern information of a portion of a chart 50 is shown in FIG. 5 by a solid white frame 60.

[0069] In the case of Duplex type IQI, the characters may vary from one character to another. The 14th and 15th line pairs may or may not be present. In an X-ray transmission image, the shape of the chart 50 may be distorted depending on the positional relationship between the radiation source 14 and the chart 50 and / or the inclination of the chart 50 relative to the radiation detector 20. In other words, the spacing between each line pair on the chart 50 may be distorted. Depending on the input image, thin line pairs may be unclear.

[0070] 5, only the first to third line pairs, which are thick and clear with no individual differences and are less affected by distortion, are used as part of the pattern information of the chart 50, so that a chart including individual differences, distortion, and unclear portions (for example, at least one line pair from the third onward and / or at least one line pair from the fourteenth onward) can be detected. Note that a pattern including at least one of the first to sixth line pairs, which are thick and clear with little effect of distortion, may also be used as part of the pattern information of the chart.

[0071] To detect the chart 50, for example, an image shown in FIG. 6 is input and a pattern similar to the pattern shown in the solid white frame 60 in FIG. 5 is searched for using a pattern matching technique. The search results are shown in FIG. 7. The color of each pixel in FIG. 7 indicates the degree of similarity between the pattern of a rectangular area in the input image with that pixel as its upper left corner and the pattern shown in the solid white frame 60 in FIG. 5; the whiter the pixel color, the greater the similarity. It can be seen that the degree of similarity is greater at the position of the chart 50 in FIG. 6. The chart 50 in the input image can be detected by extracting only those areas where the degree of similarity is equal to or greater than a predetermined threshold. There are various known methods for detecting patterns in the input image that are similar to a portion of the chart 50 pattern. A common method is to derive the degree of similarity by comparing the pattern of the chart 50 with a pattern in the input image on a pixel-by-pixel basis, and the normalized correlation coefficient is often used to measure this similarity. Other methods, such as the normalized squared difference, are also used. Another method involves extracting feature points, such as corners and / or corner points, from the pattern of the chart 50, and similarly extracting feature points in the input image to evaluate the similarity based on the positional relationship of the feature points. There is also a method of extracting and vectorizing the contour features of the pattern of the chart 50, and similarly vectorizing the contour features of the input image to evaluate the similarity of the vectors (called a contour search and / or a geometric shape search, etc.). Any of the known methods may be used for detection. The chart 50 may be detected using any of these known methods, or another method may be used for detection of the chart 50.

[0072] Furthermore, the angle of the chart 50 in the input image is unknown and varies. The distance of the chart 50 from the radiation detector 20 also varies, meaning that the scale (size, reduction) of the chart 50 is unknown. Therefore, a portion of the pattern of the chart 50 in FIG. 5 is rotated from 0 to 360 degrees, and the scale is also changed to search for similar locations. Here, the input image and the pattern of the chart 50 may be reduced in size to shorten the search processing time. Alternatively, the angle and scale may be first searched roughly to extract candidate locations with high similarity, and then the angle and scale at which each candidate has the highest similarity may be searched for and detected in detail. Alternatively, the input image and the pattern of the chart 50 may be reduced to multiple sizes at different reduction rates, and candidates with high similarity may be extracted using an image with the highest reduction rate (smallest image size), and then the candidates may be sequentially narrowed down and detected using images with smaller reduction rates (larger image sizes). By searching while varying the angle and scale of the chart 50, it is possible to detect the chart 50 in the input image and simultaneously identify the angle and scale of the chart 50.

[0073] When extracting a location where the similarity is greater than or equal to a predetermined threshold, multiple locations may be extracted. If there are multiple charts 50 in the input image, multiple locations are extracted. For example, if there are two charts 50, two charts 50 are extracted. Here, whether there is one chart 50 or multiple charts 50, each chart 50 location will have multiple pixels with similarity greater than or equal to the predetermined threshold. Therefore, the charts 50 are arranged in descending order of similarity using the angle and scale detected for each pixel. If the overlap with an already arranged chart 50 is greater than or equal to a predetermined value (e.g., if the area when all the charts 50 overlap is 1.0, the ratio of the overlapping area is 0.2 and / or 0.4 or greater), that pixel is removed from the candidates. In this way, only the pixel with the highest similarity for each chart 50 remains as a candidate for detection. When searching for a location similar to the chart 50, the relationship between the partial pattern of the chart 50 and the overall shape of the chart 50 is known. For example, the overall shape of the Duplex type IQI shown in Fig. 5 is shown by a solid white frame 54 in Fig. 8. The relationship between the partial pattern of the chart 50 (Duplex type IQI) shown in Fig. 5 and the overall shape of the chart 50 shown in Fig. 8 is known. Therefore, the overall shape of the chart 50 can be arranged using the angle and scale detected for each pixel in the input image.

[0074] Next, a specific example of target information used by the second detection unit 38B in step 103 when detecting a pattern on the chart 50 using the input image, the detection results, and the target information on the chart 50 will be described. Fig. 9 shows an example of target information on the chart 50 when the chart 50 is a Duplex type IQI.

[0075] Before describing the detection of target information and patterns on the chart 50, we will first provide an overview of the evaluation of the chart 50 as a duplex-type IQI. First, a one-dimensional density profile of each line pair is derived. An average density profile of 21 or more lines, including the center line of the IQI, is derived. Next, the contrast of at least one line pair is evaluated from the density profile. For example, the contrast of each line pair is evaluated from the density profile. Specifically, the dip values ​​for the two peaks in the profile of each line pair are derived, starting with the thickest line pair, and the first line pair with a dip value of less than 20% is identified. Finally, an iSRb value (meaning an interpolated SRb value) with a dip value of 20% is derived by interpolation from the SRb values ​​of nearby line pairs, including the line pair with a dip value of less than 20%. The SRb value represents the basic spatial resolution and is defined for each line pair.

[0076] In the above procedure, information on the position of each line pair is required as target information on chart 50 to identify the position of each line pair from the profile and derive the dip value. In FIG. 9, the position of each line pair is indicated by a downward-pointing triangle (▼) above the IQI. Furthermore, information on the wire diameter and distance between the two lines may be used as target information on chart 50 to identify the position of each line pair. Furthermore, information on the SRb value of each line pair is required as evaluation information to derive the iSRb value.

[0077] In order to evaluate the input image in step 104, in step 103, a line is first determined in a direction perpendicular to each line pair based on the position, angle, and scale detected in the input image as a pattern similar to a part of the pattern on the chart 50, and a range including 21 or more lines is set centered on that line. The start point of the set range is set to include the thickest first line pair, and the end point is set to include the thinnest 15th line pair.

[0078] It is unknown whether the chart 50 in the input image has 13 pairs or 15 pairs, but if it has 13 pairs, there is no problem even if the end point is set longer than necessary.

[0079] In Figure 9, the center line and the range of 21 or more lines including the center line are indicated by a white frame 56. Next, the set range is extracted to derive an average density profile, and the position of each line pair is tentatively identified in that density profile. The relationship between the partial pattern of the chart 50 and the position of each line pair given as the target information of the chart 50 is known, so they can be identified. Here, as described above, the chart 50 in the input image may be distorted. In other words, the position of each line pair in the chart 50 in the input image may deviate from the position of each line pair in the target information of the chart 50. Therefore, the tentatively identified positions of each line pair in the density profile may deviate from their true positions.

[0080] Therefore, after tentatively specifying the position of each line pair, the positions of the two peaks and dips of the line pair are searched for near the specified positions to detect the true positions. The peak positions can be detected by searching for the positions of two downward-convex peaks where the density is minimum near the tentative line pair positions. In addition, the position where the density between the two peaks is maximum can be detected as the dip position.

[0081] When detecting the two peaks and dips, information on the diameter and distance of the two wires may be used in addition to the position of each wire pair. In this way, the true position of each wire pair can be detected. Furthermore, since two peaks and the dip between them can be detected simultaneously, the dip value can be derived and the first wire pair with a dip value of less than 20% can be identified. Note that in the procedure of repeatedly detecting the true position and deriving the dip value starting from the thickest wire pair, the first wire pair with a dip value of less than 20% has a number less than 13, so there is no need to search for the true positions of the 13th and subsequent wire pairs.

[0082] Furthermore, when detecting the two peaks and dips of each line pair, the angle of the line perpendicular to each line pair may be varied within a predetermined range based on the angle determined in step 102 described above, and a density profile may be derived, and the angle at which the contrast between the two peaks and the dip is maximized may be searched for and identified. Because the chart 50 in the input image may be distorted, the angle of each line pair may deviate from the angle determined in step 102 described above. Therefore, by searching while varying the angle of the line relative to each line pair as described above, the two peaks and dip of each line pair can be accurately detected. The angle of the line perpendicular to each line pair may be searched for and identified individually, or may be identified for multiple line pairs collectively. When identifying multiple line pairs, a density profile may be derived by determining one line perpendicular to the multiple line pairs, and the angle at which the contrast between the two peaks and the dip of the thinnest line pair is maximized may be identified.

[0083] In this way, the pattern of the chart 50 (the true position of each line pair) can be accurately detected using the input image, the detection result (the position, angle, and scale of a portion of the pattern of the chart 50 in the input image), and the target information of the chart 50 (the position of each line pair). Even if the chart 50 in the input image is distorted, or even if an unclear thin line pair is present, it can be accurately detected by detailed search in the vicinity of the tentative position.

[0084] In step 104, the evaluation unit 40 accurately derives the dip value of each line pair based on the true positions of the two peaks and dips of nearby line pairs, including line pairs with a dip value of less than 20% detected by the second detection unit 38B in step 103, and interpolates the SRb values ​​corresponding to each line pair to derive the iSRb value with a dip value of 20%.

[0085] Then, in step 106, the CPU 24 displays the evaluation results including the iSRb value on the display unit 32.

[0086] When searching for and identifying the angle of each line pair (the angle of the line perpendicular to each line pair), if the angle of each line pair differs by a large amount by more than a predetermined threshold value depending on the line pair, the chart 50 in the input image is deemed to be distorted and unreliable for the purpose of evaluating the input image (deriving the iSRb value), and evaluation (deriving the iSRb value) may not be performed. In this case, a message indicating that evaluation will not be performed is displayed in step 106.

[0087] FIG. 10 shows an example of the display of the evaluation results. FIG. 10 shows an example of the display of the evaluation results when the chart 50 is a Duplex type IQI. In the upper diagram of FIG. 10, the center line of the chart 50 in the input image is indicated by a white line, and the range of 21 lines or more including the center line is indicated by a dotted white frame 68. In addition, the start and end points of the lines and line pairs with dip values ​​of less than 20% are indicated by letters and / or white lines. In addition, the largest number and / or iSRb value with a dip value of 20% or more is displayed around the periphery of the chart 50.

[0088] The largest number and / or iSRb value with a dip value of 20% or more may be displayed on the graph in the lower diagram of Figure 10. The lower diagram of Figure 10 shows the profile of each line pair between the start point and the end point, as well as the number and dip value of each line pair. It can be seen that the dip value of the 10th line pair is less than 20%. The lower and upper limits of the density profile of the largest numbered line pair with a dip value of 20% or more and the dip position are also shown with white lines.

[0089] (Specific Example 2) In specific example 2, another example of pattern information of a portion of the target will be described.

[0090] When it is not necessary to consider individual differences in the characters on the chart 50, it is preferable to detect the chart 50 in the input image using a partial pattern of the chart 50, including the characters.

[0091] For example, even if there is no individual difference in the characters on the chart 50 and / or there is individual difference, if it is acceptable to detect each type of chart 50 using a partial pattern of each type of chart 50 that has individual differences, it is preferable to detect the characters as well.

[0092] (Specific Example 3) In specific example 3, another example of target information will be described.

[0093] As another example of target information, a pattern including a portion that is different from part of the pattern information of the chart 50 may be used as target information of the chart 50 .

[0094] For example, the pattern indicated by the dotted white frame 62 in the middle of Fig. 5 and / or the pattern indicated by the dotted white frame 64 on the right side may be used as target information for the chart 50. Distortion can be detected using the pattern indicated by the dotted white frame 62 in the middle of Fig. 5, and the presence or absence of the 14th and 15th line pairs can be detected using the pattern indicated by the dotted white frame 64 on the right side of Fig. 5.

[0095] First, as a result of detection using the pattern of the solid white frame 60 in FIG. 5 (pattern information of a portion of the chart 50), the position, angle, and scale of the chart 50 in the input image are identified.

[0096] Next, the aspect ratio of the pattern of the dotted white frame 62 in the middle of Figure 5 is changed within a predetermined range, and the similarity with the pattern of the chart 50 at the specified position, angle, and scale in the input image is calculated, and the aspect ratio that gives the greatest similarity is identified.

[0097] Finally, the similarity between the pattern of the dotted white frame 64 on the right side of Figure 5 and the pattern of the chart 50 in the input image with the specified position, angle, scale, and aspect ratio is calculated, and if the similarity is equal to or greater than a predetermined threshold, it is determined that the 14th and 15th line pairs exist, and if it is less than the threshold, it is determined that they do not exist.

[0098] Since the positional relationship between the pattern of the solid-line white frame 60 in FIG. 5 and the patterns of the dotted-line white frame 62 in the middle and the dotted-line white frame 64 on the right in FIG. 5 is known, it is possible to detect the aspect ratio and / or the presence or absence of the 14th and 15th line pairs after detecting the position, angle, and scale of the chart 50 in the input image as described above.

[0099] When searching using the patterns of the dotted white frame 62 in the middle of Fig. 5 and the dotted white frame 64 on the right, it is also possible to change the amount of positional deviation in addition to the aspect ratio and search by considering the positional deviation from the pattern of the solid white frame 60 in Fig. 5. By searching in this way, it is possible to detect distortion including not only the aspect ratio but also the positional deviation.

[0100] Similarly, the search may be performed by changing the amount of angular misalignment, assuming that there is also an angular misalignment with the pattern of the solid white frame 60 in Fig. 5. Specifically, when searching using the patterns of the dotted white frames 62 and 64 in the center and right of Fig. 5, the angle of the pattern may be changed within a predetermined range centered on the angle identified by the pattern of the solid white frame 60, and the similarity with the pattern of the chart 50 in the input image may be calculated, and the angle at which the similarity is greatest may be identified. By performing a search in this manner, it is possible to detect distortion, including angular misalignment due to the positions of the line pairs in the pattern of the chart 50 in the input image.

[0101] Furthermore, when searching for the pattern of the dotted white frame 62 in the middle of Fig. 5, not only the aspect ratio but also the scale may be specified again. Specifically, the aspect ratio and scale of the pattern of the dotted white frame 62 in the middle of Fig. 5 may be changed within a predetermined range to calculate the similarity with the pattern of the chart 50 in the input image, and the aspect ratio and scale that maximize the similarity may be specified. By searching for both the aspect ratio and the scale, it is possible to accurately specify both the aspect ratio and the scale.

[0102] Since the position of the chart 50 in the input image is at least specified by the pattern of the solid-line white frame 60 in Fig. 5, it is also possible to search for and specify the chart 50 in the input image by changing not only the aspect ratio of the chart 50 in the input image but also multiple parameters related to the angle, scale, and shape using the pattern of the dotted-line white frame 62 in the middle of Fig. 5 as described above. Also, a pattern including the pattern of the solid-line white frame 60 may be used as the pattern of the dotted-line white frame 62 in the middle of Fig. 5. In this way, the pattern of the chart 50 (aspect ratio, positional deviation, angle deviation, scale, and the presence or absence of the 14th and 15th line pairs) can be detected using the input image, the detection results, and target information of the chart 50 (a pattern including a portion different from the pattern information of the chart 50).

[0103] Even if the chart 50 in the input image is distorted or if there are unclear thin line pairs, they can be detected by detecting some patterns that are thick and clear without individual differences and are little affected by distortion, and then performing a detailed search using the target information of the chart 50. After this, based on the position, angle, scale, aspect ratio, positional deviation, angular deviation, and the presence or absence of the 14th and 15th line pairs detected in the input image, a range including 21 or more lines including the center line of the chart 50 is set and extracted to derive an average density profile, identify the position of each line pair, derive a dip value, and derive the iSRb value.

[0104] When deriving the dip value, the true position of each line pair may be searched for and detected using the procedure described in Specific Example 1. Furthermore, if at least one of the detected aspect ratio, positional misalignment amount, and angle misalignment amount is greater than a predetermined threshold, the chart 50 in the input image is deemed to be distorted and unreliable for the purpose of evaluating the input image (deriving the iSRb value), and evaluation (deriving the iSRb value) may not be performed.

[0105] (Specific Example 4) In specific example 4, the display of the evaluation results of the input image in step 106 will be specifically described.

[0106] The evaluation results may be displayed by displaying at least one of the presence or absence, number, type, position, angle, scale, and shape of the detected objects. In addition, if the chart 50 is a duplex type IQI, at least one detected line pair may be displayed.

[0107] For example, the detected angle of the chart 50 may be displayed. When the chart 50 is a Duplex type IQI, it must be tilted by 2 to 5 degrees relative to the rows or columns of pixels in the image. Therefore, the color and / or font of the text may be changed depending on whether the angle is within the 2 to 5 degree range. Furthermore, when two Duplex type IQI charts 50 are used, if one chart 50 is aligned with a row (or column), the other chart must also be aligned with a column (or row). Therefore, whether the relationship between the two charts is appropriate may be displayed. Furthermore, if the distance from the radiation source 14 to the radiation detector 20 in an X-ray transmission image is known, the distance between the radiation detector 20 and the chart 50 can be derived from the scale of the chart 50. The scale of the chart 50 and / or the distance between the radiation detector 20 and the chart 50 derived from the scale may be displayed.

[0108] Furthermore, each line pair of the chart 50 may be displayed. Furthermore, if the chart 50 is a duplex type IQI, the type of IQI may be displayed as 13 pairs or 15 pairs. Furthermore, if no chart 50 is detected and / or three or more charts 50 are detected, a message indicating that the number of IQIs is outside the range of 1 or 2 may be displayed. Furthermore, if the distortion of the chart is deemed to be large in step 103 and evaluation of the input image is not performed, a message to that effect may be displayed.

[0109] Specific Example 5 In specific example 5, a case will be described in which Hole type IQI is applied as the chart 50. Fig. 11 is a diagram showing an example of the chart 50 of Hole type IQI.

[0110] When the chart 50 is a Hole-type IQI, there are many types (large individual differences), and detection using a common pattern is difficult. It is preferable to detect each type of chart 50 using partial pattern information of each type of chart 50.

[0111] For each type of chart 50, the position, angle, and scale of the chart 50 in the input image are first detected using clear characters as part of the pattern information of the chart 50. Next, at least one of the patterns of each hole and / or outer frame of the chart 50 is used as the target information of the chart 50. The position, angle, and scale are varied within a predetermined range, centered on the position, angle, and scale detected by the characters, to calculate the similarity with the pattern in the input image, and the position, angle, and scale that maximize the similarity are detected. Here, the aspect ratio may also be varied within a predetermined range. Furthermore, if the similarity calculated for the hole and outer frame patterns of the chart 50 is less than a predetermined threshold, the pattern may be excluded from detection. This eliminates false detections when detecting only characters. For hole-type IQIs, the position of the hole and outer frame relative to the characters may shift over time, so pattern detection using the above procedure is effective.

[0112] The Hole-type IQI has three holes of different sizes. The positions of the three holes are indicated by white circles in FIG. 11 . Information about the positions and sizes of these holes may be used as target information for the chart 50. The position and size of each hole in the chart 50 in the input image can be identified from the position, angle, and scale information (and also aspect ratio information) of the chart 50 detected in the input image and the target information for the chart 50. However, the true positions can be searched and detected in the vicinity of the tentative positions. The true positions of the holes can be detected by searching for the hole positions that maximize the contrast with the surrounding area at the identified size in the vicinity of the tentative hole positions. Then, by accurately detecting the true positions of each hole, the contrast of at least one detected hole (e.g., each hole) can be accurately evaluated.

[0113] Note that for charts 50 of different types but with the same hole positions and sizes, the same common information on the hole positions and sizes may be linked as the target information for each chart 50 .

[0114] The detection results may include the presence or absence, number, type, position, angle, and scale of the Hole type IQI and / or each hole. Also, at least one of the detected holes may be displayed.

[0115] (Specific Example 6) In specific example 6, a case will be described in which a wire type IQI is applied as the chart 50. Fig. 12 is a diagram showing an example of a chart 50 of a wire type IQI.

[0116] If the chart 50 is a wire type IQI, only clear characters may be used as part of the pattern information of the chart 50, or only clear wires that have little individual variation and are less affected by distortion may be used as part of the pattern information of the chart 50. A pattern that includes both clear characters and wires may be used as part of the pattern information of the chart 50. Then, information on the position and / or wire diameter of each wire may be used as the target information of the chart 50.

[0117] First, the position, angle, and scale of the chart 50 in the input image are detected using information about a portion of the pattern of the chart 50. Then, since the relationship between the pattern of that portion of the chart 50 and the position of each wire in the chart 50 is known, the tentative position of each wire in the chart 50 in the input image can be identified from the information about the position, angle, and scale detected in the input image, and the wire diameter can also be identified. Then, by searching for a position near the tentative position of each wire at the identified wire diameter where the contrast with the surrounding area is maximized, the true position of the wire can be detected. Then, by accurately detecting the true position of each wire, the contrast of each wire can be accurately evaluated.

[0118] (Specific Example 7) A modified example of detecting a target pattern will be described.

[0119] In the above-mentioned step 102, when a part of an object in the input image is detected, the detected predetermined range may be cut out, and in step 103, the pattern of the object may be detected using the cut-out image, the detection result of the part of the object, and the object information.

[0120] (Eighth Specific Example) In an eighth specific example, before detecting a part of a target pattern, an object is detected in advance in an input image quickly and roughly using a different method.

[0121] For example, a learning model for detecting a part of an object from an input image may be generated in advance by learning, and then, before detecting the part of the object in step 102, the generated learning model may be used to roughly detect the part of the object from the input image at high speed by a machine learning technique.

[0122] Then, for each location detected by another method, pattern information of a portion of the target, including the surrounding area, is used to search, and the position, angle, scale, etc. of the target can be accurately detected. Note that the user may also specify the target.

[0123] Specific Example 9 In specific example 9, a part of an object is detected from an input image using a machine learning technique as a method for detecting a part of an object.

[0124] For example, a learning model may be generated by learning the pattern characteristics of a portion of the object using a machine learning technique such as a neural network, and the generated learning model may be used to detect the portion of the object in the input image.

[0125] In this embodiment, the pattern information of the part of the target includes machine learning parameters and / or a learning model that has learned the characteristics of the part of the target pattern.

[0126] (Specific Example 10) In specific example 10, a case will be described in which the user limits the detection range of a part of the object in the input image.

[0127] In the above step 102, when detecting a part of an object, it takes time to search the entire input image for detection.

[0128] Therefore, the range for detecting a part of the object in the input image may be accepted by the user through operation of the operation unit 34 or the like, and may be limited to a specified range.

[0129] This makes it possible to reduce the time required to detect a part of the object from the input image in step 102.

[0130] Specific Example 11 In specific example 11, a specific example of registering part of the pattern information and the target information of the target will be described.

[0131] The pattern information of the part of the target and the target information are registered by user settings together with the image of the target, for example, by the user operating the operation unit 34. In other words, at least one of the pattern information of the part of the target and the target information may be registered by the user.

[0132] For example, chart target information such as the position, wire diameter and / or distance of each wire pair in a Duplex type IQI, the position and size of each hole in a Hole type IQI, and the position and wire diameter of each wire in a Wire type IQI is registered by user settings.

[0133] The pattern information and object information of a part of the object are registered for each type of object, for example, for each of multiple types of Duplex type IQI, Hole type IQI, and Wire type IQI.

[0134] Note that there may be multiple pieces of partial target pattern information for the same type. For example, for the same type of Duplex type IQI, the partial target pattern information may include the patterns of the third thickest line pairs and the fifth thickest line pairs. Although the partial target pattern information and the target information are linked, the user can register only the partial target pattern information and the target information, and can also re-register them. Multiple pieces of partial target pattern information may be registered for the same target information. Also, it may be possible to link and register one common target information to target information of multiple different types of targets.

[0135] During radiography for nondestructive testing, a chart is photographed, the length of the chart is analyzed, and the actual size of the photographed image is evaluated. Specifically, a chart such as a Duplex type IQI is photographed, and the actual length per pixel of the photographed image is evaluated based on the relationship between the length of the chart (length in pixels) in the photographed image and the actual length. The actual size of the subject in the photographed image is then evaluated based on that length.

[0136] The image processing device, image processing method, and image processing program provided by the present disclosure are effective in evaluating the actual dimensions of a captured image in this manner. For example, if the chart is a duplex-type IQI chart, the present disclosure makes it possible to detect a portion of the chart in the input image using pattern information of a portion of the chart, and to accurately detect each line pair of the chart using target information of the chart. Therefore, the pixel-by-pixel distance (length) between two accurately detected line pairs can be measured, and the actual length per pixel of the input image can be evaluated based on the relationship with the actual distance between the two line pairs.

[0137] If the chart is significantly distorted when detecting each line pair of the chart, the reliability of the evaluation may be deemed low and the evaluation may not be performed.

[0138] In the above embodiment, an example was described in which a radiological image was used as the input image, but the input image is not limited to a radiological image, and may be an image other than a radiological image as long as it contains a predetermined pattern, such as a chart for evaluating the image.

[0139] Furthermore, the various processes performed by the CPU in the above embodiments by executing software (programs) may be executed by a computer equipped with various processors other than a CPU. Examples of such processors include programmable logic devices (PLDs) (such as field-programmable gate arrays (FPGAs)) whose circuit configuration can be changed after manufacture, and dedicated electrical circuits, such as application-specific integrated circuits (ASICs), which are processors with circuit configurations specifically designed to execute specific processes. The various processes may be executed by one of these processors, or by a combination of two or more processors of the same or different types (e.g., multiple FPGAs, or a combination of a CPU and an FPGA). The hardware structure of these processors is, more specifically, an electrical circuit that combines circuit elements such as semiconductor devices.

[0140] In the above embodiment, the various programs are pre-stored (installed) in the ROM 20B, but the present invention is not limited to this. The various programs may be provided in a form recorded on a recording medium such as a CD-ROM (Compact Disk Read Only Memory), a DVD-ROM (Digital Versatile Disk Read Only Memory), or a USB (Universal Serial Bus) memory. The various programs may also be downloaded from an external information processing device or the like via a network.

[0141] The program of the present disclosure can be provided as a program product. The program product includes any product for providing the program. For example, the program product includes a program provided over a network such as the Internet, and a non-transitory computer-readable recording medium such as a CD-ROM or DVD on which the program is stored.

[0142] Furthermore, the configuration, operation, etc. of the radiation image capturing system 10 described in the above embodiment are merely examples, and it goes without saying that they can be modified according to the circumstances within the scope of the present disclosure.

[0143] The following additional notes are provided regarding the above-described embodiments.

[0144] (Supplementary Note 1) An image processing device comprising: a processor that, from an input image showing a predetermined target pattern, detects a part of the target pattern in the input image using pattern information of the part of the target; and performs processing to detect the target pattern using the detection result of the part and target information for identifying the target.

[0145] (Supplementary Note 2) The image processing device according to Supplementary Note 1, wherein the target is a chart for evaluating an image, and the input image is an X-ray transmission image.

[0146] (Supplementary Note 3) The image processing device according to Supplementary Note 1 or Supplementary Note 2, wherein the processor receives a range in which to detect a part of the target pattern in the input image, and detects the part of the target pattern from the received range.

[0147] (Appendix 4)

[0148] The image processing device according to any one of Supplementary Note 1 to Supplementary Note 3, wherein the processor detects the target pattern using a detection result of the part of the pattern and a relative positional relationship of the pattern as the target information.

[0149] (Supplementary Note 5) The image processing device according to any one of Supplementary Notes 1 to 4, wherein the pattern information is information including at least one of clear parts of the target pattern, parts with little individual variation, and parts with little distortion, information of parts of the target pattern that do not include characters, or information including machine learning parameters.

[0150] (Supplementary Note 6) The image processing device according to claim 1, wherein the target information includes information including at least one of unclear parts of the target pattern, parts with individual differences, and parts with large distortion, or includes information of parts different from the pattern information.

[0151] (Supplementary Note 7) The image processing device according to any one of Supplementary Notes 1 to 6, wherein the processor detects at least one of the presence / absence, position, number, type, angle, scale, and shape of the target pattern as the detection of the target pattern.

[0152] (Supplementary Note 8) The image processing device according to any one of Supplementary Notes 1 to 7, wherein the processor further performs processing to display at least one of the presence / absence, number, type, position, angle, scale, and shape of the detected object.

[0153] (Supplementary Note 9) The image processing device according to any one of Supplementary Notes 1 to 8, wherein the processor further performs a process of evaluating the input image.

[0154] (Supplementary Note 10) The image processing device according to Supplementary Note 2, wherein the chart is a multi-linear image quality meter.

[0155] (Supplementary Note 11) The image processing device according to Supplementary Note 10, wherein the pattern information is information on a pattern including at least one of the first to sixth line pairs of the multi-linear image quality meter.

[0156] (Supplementary Note 12) The image processing device according to Supplementary Note 10 or Supplementary Note 11, wherein the target information includes information on at least one line pair of the multi-linear image quality meter.

[0157] (Supplementary Note 13) The image processing device according to Supplementary Note 12, wherein the target information includes information on line pairs that are not included in the pattern information.

[0158] (Supplementary Note 14) The image processing device according to Supplementary Note 12 or Supplementary Note 13, wherein the processor detects at least one line pair from the third onward as the detection of the target pattern.

[0159] (Supplementary Note 15) The image processing device according to Supplementary Note 12 or Supplementary Note 13, wherein the processor detects at least one line pair from a fourteenth line pair onward as the detection of the target pattern.

[0160] (Supplementary Note 16) The image processing device according to any one of Supplementary Notes 12 to 15, wherein the processor calculates at least a one-dimensional density profile as the detection of the pattern.

[0161] (Supplementary Note 17) The image processing device according to any one of Supplementary Notes 10 to 16, wherein the processor further performs processing to display at least one detected line pair.

[0162] (Supplementary Note 18) The image processing device according to any one of Supplementary Notes 10 to 17, wherein the processor further performs processing to evaluate contrast of at least one line pair.

[0163] (Supplementary Note 19) The image processing device according to Supplementary Note 2, wherein the chart is a perforated penetrameter.

[0164] (Supplementary Note 20) The image processing device according to Supplementary Note 19, wherein the pattern information is information about a pattern including at least a character.

[0165] (Supplementary Note 21) The image processing device according to Supplementary Note 19 or Supplementary Note 20, wherein the target information includes at least one of information on at least one hole and information on an outer frame.

[0166] (Supplementary Note 22) The image processing device according to Supplementary Note 21, wherein the processor detects at least one hole in the perforated penetrometer as the detection of the target pattern.

[0167] (Supplementary Note 23) The image processing device according to Supplementary Note 19, wherein the processor further performs processing to display the at least one detected hole.

[0168] (Supplementary Note 24) The image processing device according to Supplementary Note 19, wherein the processor further performs processing to evaluate contrast of the at least one detected hole.

[0169] (Supplementary Note 25) The image processing device according to any one of Supplementary Notes 1 to 24, wherein at least one of the pattern information and the target information is registered by a user.

[0170] (Supplementary Note 26) An image processing method including: detecting a part of a pattern of a predetermined target in an input image, using pattern information of the part of the target, from the input image showing the pattern of the target; and detecting the pattern of the target using a result of the detection of the part and target information for identifying the target.

[0171] (Supplementary Note 27) An image processing program for causing a computer to execute processing including: detecting a portion of a pattern of a predetermined target in an input image showing the pattern of the target using pattern information of the portion of the target; and detecting the pattern of the target using the detection result of the portion and target information for identifying the target.

[0172] In the above, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed by connecting them with "and / or."

Claims

1. An image processing device comprising a processor that, from an input image showing a predetermined target pattern, detects a portion of the target pattern in the input image using pattern information of the portion of the target, and performs processing to detect the target pattern using the detection result of the portion and target information for identifying the target.

2. The image processing device according to claim 1, wherein the object is a chart for evaluating an image, and the input image is an X-ray transmission image.

3. The image processing device according to claim 1, wherein the processor receives a range for detecting a portion of the target pattern in the input image, and detects the portion of the target pattern from the received range.

4. The image processing device according to claim 1, wherein the processor detects the target pattern using the detection result of the part of the pattern and the relative positional relationship of the pattern as the target information.

5. The image processing device according to claim 1, wherein the pattern information is information including at least one of clear parts of the target pattern, parts with little individual variation, and parts with little distortion, information about parts of the target pattern that do not contain characters, or information including machine learning parameters.

6. The image processing device according to claim 1, wherein the target information includes information including at least one of unclear parts of the target pattern, parts with individual differences, and parts with large distortion, or information of parts different from the pattern information.

7. The image processing device according to claim 1, wherein the processor detects at least one of the presence or absence, position, number, type, angle, scale, and shape of the target pattern as the detection of the target pattern.

8. The image processing device according to claim 1, wherein the processor further performs processing to display at least one of the presence or absence, number, type, position, angle, scale, and shape of the detected object.

9. The image processing device according to claim 1, wherein the processor further performs processing to evaluate an input image.

10. The image processing device according to claim 2, wherein the chart is a multi-linear image quality chart.

11. An image processing device according to claim 10, wherein the pattern information is information on a pattern including at least one of the first to sixth line pairs of the multi-linear image quality meter.

12. The image processing device according to claim 10, wherein the target information includes information on at least one line pair of the multi-linear image quality meter.

13. The image processing device according to claim 12, wherein the target information includes information on line pairs not included in the pattern information.

14. The image processing device according to claim 12, wherein the processor detects at least one line pair from the third onward as the detection of the target pattern.

15. The image processing device according to claim 12, wherein the processor detects at least one line pair from the fourteenth onward as the detection of the target pattern.

16. The image processing device according to claim 12, wherein the processor calculates at least a one-dimensional density profile as the detection of the pattern.

17. The image processing device according to claim 10, wherein the processor further performs processing to display at least one detected line pair.

18. The image processing device according to claim 10, wherein the processor further performs processing to evaluate the contrast of at least one line pair.

19. The image processing device according to claim 2, wherein the chart is a perforated penetrometer.

20. The image processing device according to claim 19, wherein the pattern information is information about a pattern that includes at least characters.

21. The image processing device according to claim 19, wherein the object information includes at least one of information about at least one hole and information about an outer frame.

22. The image processing device of claim 21, wherein the processor detects at least one hole in the perforated penetrometer as detecting the pattern of interest.

23. The image processing device according to claim 19, wherein the processor further performs processing to display at least one detected hole.

24. The image processing device according to claim 19, wherein the processor further performs processing to evaluate the contrast of the at least one detected hole.

25. The image processing device according to claim 1, wherein at least one of the pattern information and the target information is registered by a user.

26. An image processing method comprising: detecting a part of a pattern of a predetermined target in an input image using pattern information of the part of the target; and detecting the pattern of the target using the result of the detection of the part and target information for identifying the target.

27. An image processing program for causing a computer to execute processing including: detecting a portion of a pattern of a predetermined target in an input image showing the pattern of the target using pattern information of the portion of the target; and detecting the pattern of the target using the detection result of the portion and target information for identifying the target.

Images