Information processing device, determination method, and determination program

Abstract

To make accurate determination even for an image to be easily determined erroneously.SOLUTION: An information processing device (1) includes: a classification unit (105) that acquires an output value by inputting an inspection image to a classification model generated by learning so that a distance in a feature space among feature amounts extracted from an image group without noise is reduced; and a determination unit (102) that determines the presence or absence of a defect by applying a method for the image group without noise or an image group with noise depending on the output value.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device or the like that makes a determination based on an image. [Background technology]

[0002] The use of images to assess various evaluation items has been widely practiced. For example, Patent Document 1 listed below discloses an ultrasonic flaw detection method using a phased array TOFD (Time Of Flight Diffraction) method. In this ultrasonic flaw detection method, an ultrasonic beam is transmitted from a phased array flaw detection element and focused on a stainless steel weld, and a flaw detection image generated based on the diffracted waves is displayed. This makes it possible to detect weld defects occurring inside a stainless steel weld. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2014-48169 Summary of the Invention [Problem to be solved by the invention]

[0004] The technology of Patent Document 1 involves visually checking flaw detection images to detect welding defects, which poses a problem of high manpower and time costs required for inspection. One possible means for solving this problem is to automatically determine whether or not there are welding defects by analyzing flaw detection images with a computer, for example.

[0005] However, flaw detection images may contain noise that looks similar to the echo of a welding defect, and when automatic detection is performed, this noise may be erroneously determined to be a welding defect. Such erroneous determinations are not limited to flaw detection images, but may occur in any image that may contain an object that looks similar to the target. Furthermore, even in object detection, which detects an object in an image, it is difficult to correctly detect the object from such an image.

[0006] As described above, various automatic judgment processes using images have a problem in that the judgment accuracy decreases when the target image to be judged is an image that is likely to cause an erroneous judgment. One aspect of the present invention aims to realize an information processing device or the like that is capable of making a highly accurate judgment even for images that are likely to cause an erroneous judgment. [Means for solving the problem]

[0007] In order to solve the above problem, an information processing device according to one embodiment of the present invention includes an acquisition unit that acquires an output value obtained by inputting a target image into a classification model generated by learning so that the distance between multiple feature quantities extracted from a first group of images having common characteristics becomes small when the feature quantities are embedded in a feature space, and a judgment unit that judges predetermined judgment items regarding the target image by applying a first method for the first group of images or a second method for a second group of images consisting of images that do not belong to the first group of images depending on the output value.

[0008] In addition, in order to solve the above-mentioned problems, a judgment method according to one aspect of the present invention is a judgment method executed by an information processing device, and includes: an acquisition step of inputting a target image into a classification model generated by learning so that when multiple feature quantities extracted from a first group of images having common characteristics are embedded in a feature space, the distance between the feature quantities becomes small, and acquiring an output value obtained; and a judgment step of applying a first method for the first group of images or a second method for a second group of images consisting of images that do not belong to the first group of images, depending on the output value, to judge predetermined judgment items regarding the target image. [Effects of the Invention]

[0009] According to one aspect of the present invention, it is possible to perform highly accurate judgment even for images that are prone to erroneous judgment. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a block diagram showing an example of a configuration of a main part of an information processing device according to a first embodiment of the present invention. [Figure 2] FIG. 1 is a diagram illustrating an overview of an inspection system including the information processing device. [Figure 3] FIG. 2 is a diagram illustrating an outline of an inspection performed by the information processing device. [Figure 4] 3A and 3B are diagrams illustrating an example of the configuration of a determination unit included in the information processing device and an example of a method for determining the presence or absence of a defect by the determination unit. [Figure 5] FIG. 10 is a diagram showing an example in which feature quantities extracted from a large number of inspection images by a classification model are embedded in a feature space. [Figure 6] FIG. 10 is a diagram illustrating an example of an inspection method using the information processing device. [Figure 7] FIG. 10 is a block diagram showing an example of a configuration of a main part of an information processing device according to a second embodiment of the present invention. [Figure 8] FIG. 10 is a diagram illustrating an example of an inspection method using the information processing device. [Figure 9] FIG. 10 is a block diagram showing an example of a configuration of a main part of an information processing device according to a third embodiment of the present invention. [Figure 10] FIG. 10 is a diagram illustrating an example of an inspection method using the information processing device. [Figure 11] FIG. 10 is a block diagram showing an example of a main configuration of an information processing device according to a fourth embodiment of the present invention. [Figure 12] FIG. 10 is a diagram illustrating an example of an inspection method using the information processing device. DETAILED DESCRIPTION OF THE INVENTION

[0011] [Embodiment 1] [System Overview] An overview of an inspection system according to one embodiment of the present invention will be described with reference to Fig. 2. Fig. 2 is a diagram showing an overview of an inspection system 100. The inspection system 100 is a system that inspects an object for defects based on an image of the object, and includes an information processing device 1 and an ultrasonic flaw detector 7.

[0012] In this embodiment, an example will be described in which the inspection system 100 is used to inspect the presence or absence of defects in tube end welds of a heat exchanger. A tube end weld is a welded portion between multiple metal tubes constituting a heat exchanger and a metal tube plate that bundles the tubes. A defect in a tube end weld is a defect in which a void occurs inside the tube end weld. The tubes and tube plate may be made of a non-ferrous metal such as aluminum, or may be made of resin. The inspection system 100 can also inspect the presence or absence of defects in the welds (root welds) between pipe stubs and tubes in boiler equipment used in waste incineration facilities, for example. Of course, the inspection area is not limited to welds, and the inspection target is not limited to heat exchangers.

[0013] During inspection, as shown in Figure 2, a probe coated with couplant is inserted from the end of the pipe, and ultrasonic waves are propagated from the inner wall of the pipe toward the pipe end weld, measuring the resulting echo. If a defect that creates a void has occurred in the pipe end weld, an echo from the void will be measured, which can be used to detect the defect. Note that the above couplant and its application method may be any that can obtain an ultrasonic image. For example, the couplant may be water. If water is used as the couplant, the water may be supplied to the area around the probe using a pump.

[0014] For example, in the enlarged view of the area around the probe shown in the lower left of Figure 2, the ultrasonic wave indicated by arrow L3 propagates to a part of the pipe end weld where there is no void. Therefore, the echo of the ultrasonic wave indicated by arrow L3 is not measured. On the other hand, the ultrasonic wave indicated by arrow L2 propagates toward a part of the pipe end weld where there is a void, and the echo of the ultrasonic wave reflected by this void is measured.

[0015] In addition, ultrasonic waves are reflected from the periphery of the pipe end weld, so the echo of the ultrasonic waves propagating to the periphery is also measured. For example, the ultrasonic waves indicated by arrow L1 propagate closer to the pipe end than the pipe end weld, so they do not hit the pipe end weld but are reflected by the pipe surface on the pipe end side of the pipe end weld. Therefore, the echo from the pipe surface is measured by the ultrasonic waves indicated by arrow L1. In addition, the ultrasonic waves indicated by arrow L4 are reflected by the pipe surface on the inner side of the pipe end weld, so that echo is measured.

[0016] Since the pipe end weld exists over a 360-degree circumference of the pipe, measurements are repeatedly performed while rotating the probe by a predetermined angle (for example, 1 degree). Data indicating the measurement results from the probe is then transmitted to the ultrasonic flaw detector 7. For example, the probe may be an array probe consisting of multiple array elements. With an array probe, by arranging the array elements so that their arrangement direction coincides with the extension direction of the pipe, pipe end welds that are wide in the extension direction of the pipe can be efficiently inspected. Note that the array probe may also be a matrix array probe in which multiple array elements are arranged vertically and horizontally.

[0017] The ultrasonic flaw detector 7 uses data indicating the measurement results from the probe to generate an ultrasonic image by imaging the echoes of ultrasonic waves propagated through the pipe and the pipe-end weld. Fig. 2 shows an ultrasonic image 111, which is an example of an ultrasonic image generated by the ultrasonic flaw detector 7. Note that the information processing device 1 may be configured to generate the ultrasonic image 111. In this case, the ultrasonic flaw detector 7 transmits data indicating the measurement results from the probe to the information processing device 1.

[0018] The intensity of the measured echo is represented as a pixel value for each pixel in the ultrasonic image 111. The image area of ​​the ultrasonic image 111 can be divided into a pipe area ar1 corresponding to the pipe, a weld area ar2 corresponding to the pipe end weld, and peripheral echo areas ar3 and ar4 in which echoes from around the pipe end weld appear.

[0019] As described above, ultrasonic waves propagated from the probe in the direction indicated by arrow L1 are reflected by the pipe surface on the pipe end side of the pipe end weld. These ultrasonic waves are also reflected by the inner surface of the pipe, and these reflections occur repeatedly. Therefore, repeated echoes a1 to a4 appear in the peripheral echo region ar3 along arrow L1 in the ultrasonic image 111. Furthermore, ultrasonic waves propagated from the probe in the direction indicated by arrow L4 are also repeatedly reflected by the outer and inner surfaces of the pipe. Therefore, repeated echoes a6 to a9 appear in the peripheral echo region ar4 along arrow L4 in the ultrasonic image 111. These echoes that appear in the peripheral echo regions ar3 and ar4 are also called back surface echoes.

[0020] The ultrasonic waves propagating from the probe in the direction indicated by arrow L3 have no reflection, and therefore no echo appears in the area along arrow L3 in ultrasonic image 111. On the other hand, the ultrasonic waves propagating from the probe in the direction indicated by arrow L2 are reflected by the void, i.e., the defect, in the pipe end weld, resulting in echo a5 appearing in the area along arrow L2 in ultrasonic image 111.

[0021] As will be described in detail below, the information processing device 1 analyzes such ultrasonic images 111 to inspect whether or not there is a defect in the pipe end weld. The information processing device 1 may also determine the type of defect. For example, if the information processing device 1 determines that there is a defect, it may determine whether the defect corresponds to one of the following known defects in pipe end welds: poor first layer penetration, poor fusion between weld passes, undercut, or blowhole.

[0022] As described above, the inspection system 100 includes an ultrasonic flaw detector 7 that generates an ultrasonic image 111 of the pipe end weld, and an information processing device 1 that analyzes the ultrasonic image 111 to inspect the pipe end weld for defects. As will be described in detail below, the information processing device 1 inputs an inspection image generated from the ultrasonic image 111 into a classification model generated by training so that when multiple feature quantities extracted from a group of noise-free images are embedded in a feature space, the distance between the feature quantities becomes small. The information processing device 1 then obtains an output value obtained by inputting the inspection image generated from the ultrasonic image 111 into the classification model. Depending on the output value, the information processing device 1 applies either a first method for images that do not contain noise or a second method for images that contain noise to determine the presence or absence of a defect. This makes it possible to accurately determine the presence or absence of a defect even if the ultrasonic image 111 contains noise whose appearance is confusingly similar to an echo at a defect site.

[0023] [Configuration of information processing device] The configuration of the information processing device 1 will be described with reference to Fig. 1. Fig. 1 is a block diagram showing an example of the configuration of the main parts of the information processing device 1. As shown in Fig. 1, the information processing device 1 includes a control unit 10 that controls each part of the information processing device 1, and a storage unit 11 that stores various data used by the information processing device 1. The information processing device 1 also includes an input unit 12 that accepts input operations to the information processing device 1, and an output unit 13 that allows the information processing device 1 to output data.

[0024] The control unit 10 includes an examination image generation unit 101, a determination unit 102A, a determination unit 102B, a determination unit 102C, a reliability determination unit 103, a comprehensive determination unit (determination unit) 104, and a classification unit (acquisition unit) 105. The storage unit 11 stores ultrasound images 111 and examination result data 112. In the following, when there is no need to distinguish between the determination units 102A, 102B, and 102C, they will be simply referred to as the determination unit 102.

[0025] The inspection image generating unit 101 generates an inspection image for determining whether or not an object to be inspected has a defect by cutting out an inspection target area from the ultrasound image 111. The method for generating the inspection image will be described later.

[0026] The judgment unit 102 judges predetermined judgment items from the target image together with the comprehensive judgment unit (judgment unit) 104. In this embodiment, an example will be described in which the inspection image generated by the inspection image generation unit 101 is the target image, and the presence or absence of a welding defect in the pipe end weld of the heat exchanger shown in the inspection image is the predetermined judgment item. Hereinafter, the welding defect may be abbreviated to simply "defect."

[0027] The definition of a "defect" to be judged may be determined in advance depending on the purpose of the inspection, etc. For example, in a quality inspection of the tube end welds of a manufactured heat exchanger, the presence of an echo in the inspection image due to a void inside the tube end weld or an unacceptable depression on the surface of the tube end weld may be deemed to be a "defect." Such depressions are caused by, for example, burn-through. The presence or absence of a defect can also be rephrased as the presence or absence of a part (an abnormal part) that differs from a normal product. In addition, in the field of non-destructive testing, an abnormal part detected using an ultrasonic waveform or ultrasonic image is generally called a "flaw." Such a "flaw" is also included in the category of the above-mentioned "defect." The above-mentioned "defect" also includes chipping, cracks, etc.

[0028] Determination units 102A, 102B, and 102C all determine the presence or absence of defects from the inspection image generated by inspection image generation unit 101, but as will be described below, each unit uses a different method for making the determination.

[0029] The determination unit 102A determines the presence or absence of a defect based on an output value obtained by inputting an inspection image into a trained model generated by machine learning. More specifically, the determination unit 102A determines the presence or absence of a defect using a generated image generated by inputting the inspection image into a generative model, which is a trained model generated by machine learning. Furthermore, the determination unit 102B identifies an inspection target portion in the inspection image by analyzing each pixel value of the inspection image, and determines the presence or absence of a defect based on the pixel values ​​of the identified inspection target portion.

[0030] Similarly to the determination unit 102A, the determination unit 102C also determines the presence or absence of a defect based on an output value obtained by inputting an inspection image into a trained model generated by machine learning. More specifically, the determination unit 102C determines the presence or absence of a defect based on an output value obtained by inputting an inspection image into a determination model that has been trained by machine learning so that the inspection image is input and the determination model outputs the presence or absence of a defect. Details of the determinations made by the determination units 102A to 102C and the various models used will be described later.

[0031] The reliability determination unit 103 determines the reliability, which is an index showing the likelihood, of each determination result of the determination units 102A to 102C. Specifically, the reliability determination unit 103 determines the reliability of the determination unit 102A when making a determination on the test image from an output value obtained by inputting the test image used when the determination unit 102A derived the determination result into a reliability prediction model for the determination unit 102A.

[0032] The reliability prediction model for the determination unit 102A can be generated by learning using training data in which the correct or incorrect result of the determination by the determination unit 102A based on the test image is associated with the test image as correct answer data. The test image may be generated from an ultrasound image 111 in which the presence or absence of a defect is known.

[0033] When the inspection image 111A is input to the reliability prediction model generated in this manner, a value between 0 and 1 indicating the probability that the judgment result when the judgment unit 102A performs a judgment using the inspection image 111A will be correct is output. Therefore, the reliability judgment unit 103 can use the output value of the reliability prediction model as the reliability of the judgment result of the judgment unit 102A. Similarly, a reliability prediction model for the judgment unit 102B and a reliability prediction model for the judgment unit 102C can be generated. Then, the reliability judgment unit 103 judges the reliability of the judgment result of the judgment unit 102B using the reliability prediction model for the judgment unit 102B, and judges the reliability of the judgment result of the judgment unit 102C using the reliability prediction model for the judgment unit 102C.

[0034] The overall judgment unit 104 judges the presence or absence of a defect using the judgment results of the judgment units 102A to 102C and the reliability judged by the reliability judgment unit 103. This makes it possible to obtain a judgment result that appropriately takes into account the judgment results of the judgment units 102A to 102C with a reliability according to the inspection image. The judgment method by the overall judgment unit 104 will be described in detail later.

[0035] The classification unit 105 classifies the inspection image using a predetermined classification model. Details will be explained based on FIG. 5 , but the classification model is a model generated by learning so that when multiple feature amounts extracted from a first group of images having a common feature are embedded in a feature space, the distance between the feature amounts becomes small. The common feature is that the feature does not contain noise. The classification unit 105 obtains an output value obtained by inputting the inspection image into this classification model.

[0036] Then, depending on the output value obtained by the classification unit 105, the judgment unit 102 applies a first method for the first group of images or a second method for the second group of images consisting of images that do not belong to the first group of images to judge whether or not there is a defect.

[0037] Specifically, the output value acquired by the classification unit 105 indicates whether the test image is a noisy image or a noiseless image. If the output value indicates a noiseless image, a first technique for noiseless test images is applied. On the other hand, if the output value indicates a noisy image, a second technique for noisy test images is applied.

[0038] Specifically, the first method is a method in which the overall determination unit 104 determines the presence or absence of a defect using the determination results of the determination units 102A to 102C and their reliability determined by the reliability determination unit 103. On the other hand, the second method is a method in which the determination unit 102B determines the presence or absence of a defect.

[0039] As described above, the ultrasonic image 111 is an image obtained by imaging the echo of ultrasonic waves propagated through the object to be inspected, and is generated by the ultrasonic flaw detector 7.

[0040] The inspection result data 112 is data indicating the results of the defect inspection by the information processing device 1. The inspection result data 112 records the presence or absence of a defect in the ultrasound image 111 stored in the storage unit 11. Furthermore, when the type of defect is determined, the determination result of the type of defect may be recorded as the inspection result data 112.

[0041] As described above, the information processing device 1 includes a classification unit 105 that acquires an output value obtained by inputting an inspection image into a classification model generated by learning so that the distance between multiple feature amounts extracted from a group of noise-free images (a first group of images having common features) becomes small when the feature amounts are embedded in a feature space, and a judgment unit 102 that judges the presence or absence of a defect (a predetermined judgment item related to the inspection image) by applying a first method for the first group of images or a second method for a group of noise-containing images (a second group of images consisting of images not belonging to the first group of images) according to the output value.

[0042] The classification model is generated by learning to minimize the distance between features when they are embedded in a feature space. Therefore, even if the test image contains noise that is likely to cause erroneous judgment, inputting the test image into the classification model will produce an output value indicating whether the test image's features are close to the features of the first group of noise-free images.

[0043] In other words, if the feature values ​​of the test image are close to the feature values ​​of the first group of noise-free images, the test image is likely to be noise-free. On the other hand, if the feature values ​​of the test image deviate from the feature values ​​of the first group of noise-free images, the test image is likely to contain noise. It is generally difficult to collect sufficient training data for amorphous noise due to its diverse shapes, making it difficult to determine whether noise is present using a trained model generated by machine learning. However, using the output values ​​described above, it is possible to determine whether the test image contains noise.

[0044] According to the above configuration, the determination item is determined by applying either the first method for images without noise or the second method for images with noise, depending on the output value. This makes it possible to apply an appropriate method according to the characteristics of the inspection image, and also makes it possible to perform highly accurate determinations even for inspection images that are prone to erroneous determinations.

[0045] [Inspection overview] An overview of inspection by the information processing device 1 will be described with reference to Fig. 3. Fig. 3 is a diagram showing an overview of inspection by the information processing device 1. Fig. 3 shows processing after an ultrasonic image 111 generated by the ultrasonic flaw detector 7 is stored in the storage unit 11 of the information processing device 1.

[0046] First, the test image generation unit 101 extracts an area to be inspected from the ultrasound image 111 to generate the test image 111A. An extraction model constructed by machine learning may be used to extract the area to be inspected. The extraction model may be constructed using any learning model suitable for extracting an area from an image. For example, the test image generation unit 101 may construct the extraction model using YOLO (You Only Look Once), which has excellent extraction accuracy and processing speed.

[0047] The inspection target area is an area sandwiched between two peripheral echo areas ar3 and ar4, in which echoes from the peripheral portion of the inspection target area in the inspection object repeatedly appear. As shown in FIG. 2, predetermined echoes (echoes a1 to a4 and a6 to a9) are repeatedly observed in the peripheral portion of the inspection target area in the ultrasound image 111, due to the shape of the peripheral portion, etc. Therefore, the area corresponding to the inspection target area in the ultrasound image 111 can be identified from the positions of the peripheral echo areas ar3 and ar4 in which such echoes repeatedly appear. Note that the appearance of predetermined echoes in the peripheral portion of the inspection target area is not limited to the ultrasound image 111 of the pipe end weld. Therefore, the configuration of extracting the area surrounded by the peripheral echo areas as the inspection target area can be applied to inspections other than those of the pipe end weld.

[0048] Next, the classification unit 105 classifies the inspection image 111A. Then, for the inspection image 111A classified as having noise by the classification unit 105, the presence or absence of a defect is determined by the second method as described above. Specifically, as shown in FIG. 3, for the inspection image 111A classified as having noise, the determination unit 102B determines the presence or absence of a defect by numerical analysis. Then, this result is added to the inspection result data 112. The determination unit 102B may also cause the output unit 13 to output the determination result.

[0049] On the other hand, for inspection image 111A classified as having no noise by classification unit 105, the presence or absence of a defect is determined by the first method. Specifically, first, determination units 102A, 102B, and 102C determine the presence or absence of a defect based on inspection image 111A. Details of the determination will be described later.

[0050] Next, reliability determination unit 103 determines the reliability of each determination result of determination unit 102A, determination unit 102B, and determination unit 102C. Specifically, the reliability of the determination result of determination unit 102A is determined from an output value obtained by inputting inspection image 111A into a reliability prediction model for determination unit 102A. Similarly, the reliability of the determination result of determination unit 102B is determined from an output value obtained by inputting inspection image 111A into a reliability prediction model for determination unit 102B. Furthermore, the reliability of the determination result of determination unit 102C is determined from an output value obtained by inputting inspection image 111A into a reliability prediction model for determination unit 102C.

[0051] Then, the overall judgment unit 104 makes an overall judgment on the presence or absence of defects using the judgment results of the judgment units 102A, 102B, and 102C and the reliability of those judgment results judged by the reliability judgment unit 103, and outputs the result of the overall judgment. This result is added to the inspection result data 112. The overall judgment unit 104 may also cause the output unit 13 to output the result of the overall judgment.

[0052] In the overall judgment, the judgment result of the judgment unit 102 may be expressed as a numerical value, and the reliability judged by the reliability judgment unit 103 may be used as a weight. For example, the judgment units 102A, 102B, and 102C may output a judgment result of "1" when they judge that there is a defect, and output a judgment result of "-1" when they judge that there is no defect. Furthermore, the reliability judgment unit 103 may output a reliability in a numerical range from 0 to 1 (the closer to 1 the higher the reliability).

[0053] In this case, the overall judgment unit 104 may calculate a total value by adding up the value obtained by multiplying the numerical values ​​"1" or "-1" output by the judgment units 102A, 102B, and 102C by the reliability output by the reliability judgment unit 103. Then, the overall judgment unit 104 may judge the presence or absence of a defect based on whether the calculated total value is greater than a predetermined threshold value.

[0054] For example, suppose the threshold value is set to "0," which is an intermediate value between "1," which indicates the presence of a defect, and "-1," which indicates the absence of a defect. Then, suppose the output values ​​of the determination unit 102A, the determination unit 102B, and the determination unit 102C are "1," "-1," and "1," respectively, and the corresponding reliability levels are "0.87," "0.51," and "0.95," respectively.

[0055] In this case, the overall judgment unit 104 performs the calculation 1×0.87+(−1)×0.51+1×0.95. The result of this calculation is 1.31, which is greater than the threshold value of “0,” so the overall judgment by the overall judgment unit 104 is that there is a defect.

[0056] [Determination by Determination Unit 102A] As described above, the determination unit 102A determines the presence or absence of defects using a generated image generated by inputting an inspection image into a generative model. This generative model is constructed by machine learning using images of defect-free inspection objects as training data, so as to generate a new image having similar features to the input image. Note that the "features" mentioned above are any information obtained from an image, and include, for example, the distribution and variance of pixel values ​​in an image.

[0057] The generative model is constructed by machine learning using images of defect-free inspection objects as training data. Therefore, when an image of a defect-free inspection object is input to the generative model as an inspection image, there is a high possibility that a new image with similar characteristics to the inspection image will be output as the generated image.

[0058] On the other hand, if an image of an object to be inspected containing a defect is input into this generation model as an inspection image, no matter what position, shape, and size of the defect is shown in the inspection image, the generated image is likely to have different characteristics from the inspection image.

[0059] In this way, there is a difference between a generated image generated from an inspection image that shows a defect and a generated image generated from an inspection image that does not show a defect, in that the target image input into the generative model is either not correctly restored or correctly restored.

[0060] Therefore, according to the information processing device 1 that performs a comprehensive judgment taking into consideration the judgment results of the judgment unit 102A that judges the presence or absence of defects using the generated image generated by the above-mentioned generative model, it becomes possible to accurately judge the presence or absence of defects whose position, size, shape, etc. are indefinite.

[0061] Details of the determination by the determination unit 102A will be explained below with reference to Fig. 4. Fig. 4 is a diagram showing an example of the configuration of the determination unit 102A and an example of a method for determining the presence or absence of a defect by the determination unit 102A. As shown in Fig. 4, the determination unit 102A includes an inspection image acquisition unit 1021, a restored image generation unit 1022, and a defect presence / absence determination unit 1023.

[0062] Inspection image acquisition unit 1021 acquires an inspection image. Since information processing device 1 includes inspection image generation unit 101 as described above, inspection image acquisition unit 1021 acquires the inspection image generated by inspection image generation unit 101. Note that the inspection image may be generated by another device. In this case, inspection image acquisition unit 1021 acquires the inspection image generated by the other device.

[0063] The restored image generation unit 1022 inputs the inspection image acquired by the inspection image acquisition unit 1021 into a generative model, thereby generating a new image having similar characteristics to the input inspection image. Hereinafter, the image generated by the restored image generation unit 1022 will be referred to as a restored image. As will be described in detail later, the generative model used to generate the restored image is also called an autoencoder, and is constructed by machine learning using images of defect-free inspection objects as training data. Note that the generative model may be an improved or modified version of an autoencoder. For example, a variational autoencoder or the like may be applied as the generative model.

[0064] The defect presence / absence determination unit 1023 determines whether or not the object to be inspected has a defect using the restored image generated by the restored image generation unit 1022. Specifically, the defect presence / absence determination unit 1023 determines that the object to be inspected has a defect when the variance of the difference value for each pixel between the inspection image and the restored image exceeds a predetermined threshold.

[0065] In the method for determining the presence or absence of a defect by determination unit 102A having the above configuration, first, inspection image acquisition unit 1021 acquires inspection image 111A. Then, inspection image acquisition unit 1021 sends the acquired inspection image 111A to restored image generation unit 1022. Inspection image 111A is generated by inspection image generation unit 101 from ultrasound image 111 as described above.

[0066] Next, restored image generation unit 1022 inputs inspection image 111A into a generative model and generates restored image 111B based on the output value. Then, inspection image acquisition unit 1021 removes the peripheral echo region from inspection image 111A to generate removed image 111C, and removes the peripheral echo region from restored image 111B to generate removed image (restored) 111D. Note that the position and size of the peripheral echo region appearing in inspection image 111A will be roughly constant if the inspection object is the same. Therefore, inspection image acquisition unit 1021 may remove a predetermined range in inspection image 111A as the peripheral echo region. Alternatively, inspection image acquisition unit 1021 may analyze inspection image 111A to detect the peripheral echo region and remove the peripheral echo region based on the detection result.

[0067] By removing the peripheral echo region in the above manner, the defect presence / absence determining unit 1023 determines the presence / absence of a defect based on the remaining image region excluding the peripheral echo region from the image region of the restored image 111B. This makes it possible to determine the presence / absence of a defect without being affected by echoes from the peripheral portion, thereby improving the accuracy of determining the presence / absence of a defect.

[0068] Next, defect presence / absence determination unit 1023 determines the presence or absence of a defect. Specifically, defect presence / absence determination unit 1023 first calculates the difference between removed image 111C and removed image (restored) 111D on a pixel-by-pixel basis. Next, defect presence / absence determination unit 1023 calculates the variance of the calculated difference. Then, defect presence / absence determination unit 1023 determines the presence or absence of a defect based on whether the calculated variance value exceeds a predetermined threshold.

[0069] Here, the difference value calculated for a pixel containing an echo due to a defect is larger than the difference values ​​calculated for other pixels, and therefore the variance of the difference values ​​calculated for removed image 111C and removed image (restored) 111D based on inspection image 111A containing an echo due to a defect becomes larger.

[0070] On the other hand, the variance of the difference values ​​is relatively small for removed image 111C and removed image (restored) 111D based on inspection image 111A that does not capture echoes caused by defects. This is because, when echoes caused by defects are not captured, there may be some locations where pixel values ​​are somewhat large due to the influence of noise, etc., but there is little possibility that there will be some locations where pixel values ​​are extremely large.

[0071] Such a large variance of the difference values ​​is a characteristic phenomenon when the object being inspected has a defect. Therefore, if the defect presence / absence determining unit 1023 is configured to determine that a defect exists when the variance of the difference values ​​exceeds a predetermined threshold, the presence or absence of a defect can be appropriately determined.

[0072] The timing for removing the peripheral echo region is not limited to the above example. For example, a difference image between the inspection image 111A and the restored image 111B may be generated, and the peripheral echo region may be removed from this difference image.

[0073] [Determination by Determination Unit 102B] As described above, the judgment unit 102B identifies the inspection target area in the inspection image by analyzing each pixel value of the inspection image, which is an image of the object to be inspected, and judges whether or not there is a defect based on the pixel values ​​of the identified inspection target area.

[0074] In conventional image-based inspections, inspectors visually identify the area to be inspected in the image and check that the identified area does not contain any defects such as scratches or voids that are not present in the design. There is a demand for automation of such visual inspections from the perspective of labor saving and stabilizing accuracy.

[0075] The determination unit 102B identifies an inspection target portion by analyzing each pixel value of the image, and determines whether or not there is a defect based on the pixel values ​​of the identified inspection target portion. This makes it possible to automate the above-described visual inspection. For inspection images classified as noise-free, the information processing device 1 makes a determination by comprehensively considering the determination result of the determination unit 102B and the determination results of the other determination units 102, thereby making it possible to accurately determine whether or not there is a defect. Furthermore, for inspection images classified as noisy, the information processing device 1 analyzes the pixel values, making it possible to accurately determine whether or not there is a defect without mistaking noise for a defect.

[0076] The processing (numerical analysis) executed by the determination unit 102B will be described in more detail below. First, the determination unit 102B identifies, as the inspection target area, an area sandwiched between two peripheral echo areas (peripheral echo areas ar3 and ar4 in the example of FIG. 2) in which echoes from the peripheral portions of the inspection target area repeatedly appear. Then, the determination unit 102B determines whether or not there is a defect depending on whether or not the identified inspection target area includes an area (also called a defect area) consisting of pixel values ​​equal to or greater than a threshold value.

[0077] When detecting the peripheral echo region and the defect region, the determination unit 102B may first binarize the inspection image 111A using a predetermined threshold to generate a binarized image. Then, the determination unit 102B detects the peripheral echo region from the binarized image. For example, the inspection image 111A shown in FIG. 3 contains echoes a1, a2, a6, and a7. If the determination unit 102B binarizes the inspection image 111A using a threshold that can distinguish these echoes from noise components, the determination unit 102B can detect these echoes from the binarized image. Then, the determination unit 102B can detect the edges of the detected echoes and identify the region surrounded by these edges as the inspection target area.

[0078] More specifically, the determination unit 102B identifies the right end of echo a1 or a2 as the left end of the region to be examined, and the left end of echo a6 or a7 as the right end of the region to be examined. These ends are the boundaries between the region to be examined and the peripheral echo regions ar3 and ar4. Similarly, the determination unit 102B identifies the upper end of echo a1 or a6 as the upper end of the region to be examined, and the lower end of echo a2 or a7 as the lower end of the region to be examined.

[0079] As shown in the ultrasound image 111 of FIG. 2, an echo caused by a defect may appear above echo a1 or a6, so the determination unit 102B may set the upper end of the inspection target area above the position of the upper end of echo a1 or a6.

[0080] Furthermore, the determination unit 102B can analyze the inspection target portion identified in the binarized image to determine whether or not an echo caused by a defect is captured. For example, when a continuous area consisting of a predetermined number of pixels or more exists in the inspection target portion, the determination unit 102B may determine that an echo caused by a defect is captured at the position where the continuous area exists.

[0081] The above numerical analysis is merely an example, and the content of the numerical analysis is not limited to the above example. For example, if there is a significant difference in the variance of pixel values ​​in the inspection target area between the presence and absence of a defect, the determination unit 102B may determine the presence or absence of a defect based on the value of the variance.

[0082] Furthermore, for example, the determination unit 102B may determine the presence or absence of a defect by numerical analysis based on the results of simulation by an ultrasonic beam simulator. The ultrasonic beam simulator outputs the height of the reflected echo when detecting an artificial flaw set at any position on the test piece. Therefore, the determination unit 102B can determine the presence or absence and position of a defect by comparing the height of the reflected echo corresponding to the artificial flaw at various positions output by the ultrasonic beam simulator with the reflected echo in the inspection image.

[0083] [Determination by Determination Unit 102C] As described above, the determination unit 102C determines the presence or absence of a defect based on the output value obtained by inputting the inspection image into the determination model. This determination model is constructed by performing machine learning using, for example, training data generated using the ultrasound image 111 of an inspection object having a defect and training data generated using the ultrasound image 111 of an inspection object without a defect.

[0084] The determination model can be constructed using any learning model suitable for image classification. For example, the determination model may be constructed using a convolutional neural network or the like, which has excellent image classification accuracy.

[0085] [About the classification model] The classification model used by the classification unit 105 to classify the inspection images will be described with reference to Fig. 5. Fig. 5 is a diagram showing an example in which feature amounts extracted from a large number of inspection images by the classification model are embedded in a feature space.

[0086] This classification model was generated by training so that when feature quantities extracted from a group of noise-free images (first group of images) of the images of the object to be inspected are embedded in a feature space, the distance between the feature quantities becomes small. More specifically, this classification model was generated by training so that the distance between feature quantities extracted from a group of noise-free images with defects becomes small, and so that the distance between feature quantities extracted from a group of noise-free images with defects becomes small, as well as the distance between feature quantities extracted from a group of noise-free images with no defects. In other words, this classification model classifies inspection images into two classes: no noise with defects and no noise with no defects.

[0087] The feature space shown in FIG. 5 is a two-dimensional feature space with x on the horizontal axis and y on the vertical axis. FIG. 5 also illustrates some of the inspection images from which feature quantities were extracted (inspection images 111A1 to 111A5). Of the inspection images shown in FIG. 5, inspection images 111A1 and 111A2 are noise-free, defect-free images that do not capture noise or defects. On the other hand, inspection images 111A3 and 111A4 are noisy images that capture noise in areas AR1 and AR2. Furthermore, inspection image 111A5 is a noise-free, defect-containing image that does not capture noise but captures a defect echo a10.

[0088] As shown in the figure, when the features extracted from each inspection image are embedded in a feature space using the classification model generated by the learning described above, the features of inspection images belonging to the same class are plotted in positions close to each other.

[0089] Specifically, the feature values ​​of noise-free, defect-free inspection images such as inspection images 111A1 and 111A2 are generally within a circle C1 with a radius of r1 and a center at point P1, while the feature values ​​of noise-free, defect-containing inspection images such as inspection image 111A5 are generally within a circle C2 with a radius of r2 and a center at point P2.

[0090] On the other hand, the feature quantities of noisy inspection images such as inspection images 111A3 and 111A4 are plotted at positions away from both circle C1 and circle C2. This shows that by using a model that classifies inspection images into two classes, no noise and defects and no noise and no defects, it is possible to distinguish between noisy inspection images and noise-free inspection images.

[0091] For example, the classification unit 105 may classify an inspection image as having no defects if the feature values ​​obtained by inputting the inspection image into the classification model are plotted within circle C1. The classification unit 105 may classify an inspection image as having defects if the feature values ​​obtained by inputting the inspection image into the classification model are plotted within circle C2. The classification unit 105 may classify an inspection image as having noise if the feature values ​​obtained by inputting the inspection image into the classification model are plotted at a position outside either circle C1 or circle C2.

[0092] The radius r1 of the circle C1 and the radius r2 of the circle C2 may be the same or different. For example, the radius r1 and the radius r2 may each be set to an appropriate value. Alternatively, the radius may be set to the distance from the center of the feature plot of the training data to the furthest plot. Alternatively, the radius may be set to the value obtained by doubling the standard deviation (σ) of the feature plot of the training data.

[0093] Furthermore, the plot positions of the feature amounts of the inspection image may be expressed by numerical values ​​ranging from 0 to 1. For example, the position of point P1 may be (0,0), the position of point P2 may be (0,1), and each plot of the feature amount of the inspection image may be projected onto a line L connecting point P1 and point P2.

[0094] In this case, if the feature values ​​are plotted in the range from point p11 to point p12 on the line L, the inspection image can be determined to be noise-free and defect-free. Note that point p11 is the intersection point between the circle C1 and the line L1 that is closer to the circle C2. Also, point p12 is the intersection point between the circle C1 and the line L1 that is farther from the circle C2.

[0095] Similarly, when feature quantities are plotted in the range from point p21 to point p22 on the line L, the inspection image can be determined to be noise-free and defective. Note that point p21 is the intersection point between circle C2 and line L1 that is closer to circle C1. Also, point p22 is the intersection point between circle C2 and line L1 that is farther from circle C1.

[0096] If the feature amount is plotted in the range from point p11 to point p21, the inspection image can be determined to contain noise.

[0097] The plot values ​​outside point P1 on line L (on the opposite side from the direction in which circle C2 exists) may be regarded as 0, and the plot values ​​outside point P2 on line L (on the opposite side from the direction in which circle C1 exists) may be regarded as 1. The plot values ​​inside circle C1 may also be regarded as 0, and the plot values ​​inside circle C2 may also be regarded as 1. In this case, an inspection image with a plot value of 0 is classified as having no defects, an inspection image with a plot value of 1 is classified as having defects, and an inspection image with a plot value other than 0 or 1 is classified as having noise.

[0098] Furthermore, even when using a classification model trained on only noise-free and defect-free inspection images, or only noise-free and defect-containing inspection images, it is possible to similarly classify inspection images into noise-containing and noise-free.

[0099] As described above, the classification unit 105 can classify inspection images into noise-containing and noise-free images by using the output values ​​of a classification model generated by learning so that when multiple feature quantities extracted from a group of noise-free images are embedded in a feature space, the distance between the feature quantities becomes small. The classification model may be designed to output an output value indicating the classification result (e.g., a confidence level for each class), or may be designed to output feature quantities. The confidence level is a numerical value between 0 and 1 indicating the likelihood of the classification result.

[0100] The classification model described above can be generated by, for example, deep metric learning. Deep metric learning is a learning method in which features embedded in a feature space are trained so that the distance Sn between features of data of the same class is small and the distance Sp between features of data of different classes is large. During training, the distance between features may be expressed as Euclidean distance or an angle.

[0101] The inventors of the present invention also attempted to classify test images with noise and without noise using a classification model based on a convolutional neural network, but classification using this classification model proved difficult. Therefore, to distinguish between test images with noise and those without noise, it is important to use a classification model generated by training so that the distance between feature values ​​becomes small when the feature values ​​are embedded in a feature space.

[0102] [Inspection process flow] The flow of processing (determination method) in the inspection will be described with reference to Fig. 6. Fig. 6 is a diagram showing an example of an inspection method using information processing device 1. At the start of the processing in Fig. 6, it is assumed that an ultrasonic image 111 for detecting flaws in the pipe end weld and its peripheral portion, which was generated by the method described with reference to Fig. 2, is stored in storage unit 11, and that inspection image generation unit 101 has already generated an inspection image from that ultrasonic image 111.

[0103] In S11, classification unit 105 acquires the inspection image generated by inspection image generation unit 101. Subsequently, in S12 (acquisition step), classification unit 105 inputs the inspection image acquired in S11 into the classification model described above and acquires an output value of the classification model. Then, in S13, classification unit 105 determines whether the inspection image acquired in S11 is an inspection image with noise or an inspection image without noise, based on the output value acquired in S12.

[0104] If it is determined in S13 that the inspection image has noise (YES in S13), the process proceeds to S17. Then, in S17 (determination step), the presence or absence of defects in the inspection image is determined by the determination unit 102B using a second technique for noisy inspection images, i.e., numerically analyzing pixel values ​​of the inspection image, and the determination result is recorded in the inspection result data 112.

[0105] On the other hand, if it is determined in S13 that the inspection image is noise-free (NO in S13), the process proceeds to S 14. Then, in S14 to S16 (determination steps), the presence or absence of defects in the inspection image is determined by the determination units 102A, 102C, etc., which use a first technique for noise-free inspection images, i.e., a trained model, to determine the presence or absence of defects.

[0106] Specifically, in S14, the determination units 102A, 102B, and 102C each determine whether or not there is a defect. Then, in the following S15, the reliability determination unit 103 determines the reliability of the determination results of the determination units 102A, 102B, and 102C. Note that the process of S15 may be performed before S14 or may be performed in parallel with S14.

[0107] Then, in S16, the overall judgment unit 104 judges the presence or absence of a defect using each judgment result of S14 and the reliability judged in S15. Specifically, the overall judgment unit 104 judges the presence or absence of a defect using a numerical value obtained by weighting the numerical values ​​indicating each judgment result of the judgment units 102A to 102C according to their reliability and adding them up. The overall judgment unit 104 also adds this judgment result to the inspection result data 112.

[0108] For example, the determination results of the determination units 102A to 102C can be expressed as a numerical value of −1 (no defect) or 1 (defect present). In this case, if the reliability is calculated as a numerical value between 0 and 1, the reliability value may be used as a weight to multiply the determination result.

[0109] As a specific example, suppose that the judgment result from judgment unit 102A is that there is a defect, the judgment result from judgment unit 102B is that there is no defect, and the judgment result from judgment unit 102C is that there is a defect. Also, suppose that the reliability of the judgment results from judgment units 102A to 102C is 0.87, 0.51, and 0.95, respectively. In this case, comprehensive judgment unit 104 performs the calculation 1×0.87+(−1)×0.51+1×0.95 and obtains the resulting numerical value of 1.31.

[0110] The overall judgment unit 104 may then compare this numerical value with a predetermined threshold value, and if the calculated numerical value is greater than the threshold value, judge that there is a defect. If no defect is represented by "-1" and the presence of a defect by "1", the threshold value may be set to "0", which is the intermediate value between these numerical values. In this case, since 1.31>0, the final judgment result by the overall judgment unit 104 is that there is a defect.

[0111] As described above, the determination method according to this embodiment is a determination method executed by the information processing device 1, and includes an acquisition step (S12) of inputting an inspection image into a classification model generated by learning so that when multiple feature quantities extracted from a group of noise-free images (a first group of images having common features) are embedded in a feature space, the distance between the feature quantities becomes small, and acquiring an output value obtained by inputting the inspection image into the classification model, and a determination step (S14 to S16 when the first method is applied, S17 when the second method is applied) of applying a first method for noise-free inspection images or a second method for noisy inspection images (a second group of images not belonging to the first group of images) according to the output value to determine the presence or absence of a defect (a predetermined determination item related to the inspection image). Thus, it is possible to perform highly accurate determination even for images that are prone to erroneous determination.

[0112] Furthermore, a noise-free inspection image is an image for which judgment based on an output value obtained by inputting the inspection image into a trained model generated by machine learning, as executed by the judgment units 102A and 102C, is effective. For this reason, as in the example of Figure 6 above, it is preferable that the first technique includes at least a process for making a judgment using a trained model, and the second technique includes at least a process for performing the above-mentioned numerical analysis.

[0113] For noise-free inspection images, since they do not contain parts that are similar in appearance to defects in the object being inspected, judgment using a trained model generated by machine learning is effective. Therefore, for noise-free inspection images, highly accurate judgment results can be expected by including a process for making judgments using a trained model generated by machine learning.

[0114] Because noise is irregular and has a similar appearance to defects in the object being inspected, judgment using a trained model generated by machine learning may not be effective for inspection images with noise. However, even for such inspection images, numerical analysis may be able to make a valid judgment.

[0115] Therefore, with the above configuration, in which the presence or absence of defects is determined for inspection images with noise using the second method including numerical analysis, a valid determination result can be expected even when the inspection image is one for which determination using a trained model is not effective. In other words, with the above configuration, it is possible to make a valid determination whether or not the inspection image is one for which determination using a trained model is effective.

[0116] The first technique may include at least one determination process using a trained model generated by machine learning. For example, the first technique may include only one of the determination processes by the determination units 102A and 102C. The second technique may include, in addition to the determination process by the determination unit 102B, determination processes by other techniques such as the determination units 102A and 102C. In this case, however, it is desirable to weight the determination result by the determination unit 102B more heavily than the determination results by the other techniques.

[0117] [Embodiment 2] Other embodiments of the present invention will be described below. For ease of explanation, the same reference numerals will be used to designate components having the same functions as those described in the above embodiments, and their descriptions will not be repeated. This also applies to the third and subsequent embodiments.

[0118] [Device configuration] The configuration of an information processing device 1A according to this embodiment will be described with reference to Fig. 7. Fig. 7 is a block diagram showing an example of the configuration of the main parts of the information processing device 1A. The information processing device 1A differs from the information processing device 1 shown in Fig. 1 in that it does not include a classification unit 105 and includes a determination unit 102X and a determination method determination unit (acquisition unit) 106.

[0119] The determination unit 102X determines the presence or absence of a defect using the classification model described in embodiment 1. More specifically, the determination unit 102X determines the presence or absence of a defect based on an output value obtained by inputting an inspection image into the classification model.

[0120] For example, the determination unit 102X may use a classification model generated by learning so as to reduce the distance between feature amounts extracted from a group of images without noise and with defects, and so as to reduce the distance between feature amounts extracted from a group of images without noise and with defects, as in the example of Fig. 5. The determination unit 102X can determine whether an inspection image is a noise-free, defect-free inspection image or a noise-free, defect-containing inspection image, based on the output value obtained by inputting the inspection image into this classification model.

[0121] The determination method determination unit 106 acquires the output value of the classification model used by the determination unit 102X for the above determination. If the output value indicates that the inspection image is either "no noise and no defect" or "no noise and defect," the determination method determination unit 106 determines that the inspection image is noise-free and determines to apply the first method for noise-free inspection images. On the other hand, if the output value indicates that the inspection image is neither "no noise and no defect" nor "no noise and defect," the determination method determination unit 106 determines that the inspection image is noisy and determines to apply the second method for noisy inspection images.

[0122] [Processing flow] The flow of the process (determination method) executed by the information processing device 1A will be described with reference to Fig. 8. Fig. 8 is a diagram showing an example of an inspection method using the information processing device 1A. At the start of the process of Fig. 8, it is assumed that an ultrasound image 111 is stored in the storage unit 11 and that the inspection image generation unit 101 has already generated an inspection image from the ultrasound image 111.

[0123] In S21, all of the determination units 102, i.e., determination units 102A, 102B, 102C, and 102X, acquire the inspection image generated by the inspection image generation unit 101. Then, in S22, all of the determination units 102 that acquired the inspection image in S21 each use the inspection image to determine whether or not there is a defect.

[0124] In S23 (acquisition step), the determination method determination unit 106 acquires the output value obtained by the determination unit 102X in S22 when the test image is input to the classification model. Then, based on the acquired output value, the determination method determination unit 106 determines whether the test image acquired in S21 is a test image with noise or a test image without noise.

[0125] If it is determined in S23 that the inspection image contains noise (YES in S23), the determination method determination unit 106 instructs the determination unit 102B to perform a determination, and the process proceeds to S26. Then, in S26 (determination step), the presence or absence of defects in the inspection image is determined by the determination unit 102B using a second method for noisy inspection images, i.e., numerically analyzing the pixel values ​​of the inspection image, and the determination result is added to the inspection result data 112.

[0126] Since the judgment by the judgment unit 102B has already been made in S22, if the judgment in S23 is YES, instead of performing the processing in S26, the judgment result of the judgment unit 102 in S22 may be added to the test result data 112 as the final judgment result.

[0127] If it is determined in S23 that the inspection image is noise-free (NO in S23), the determination method determination unit 106 instructs the reliability determination unit 103 and the overall determination unit 104 to make a determination, and the process proceeds to S24. Then, in S24 to S25 (determination step), the presence or absence of defects in the inspection image is determined using a first method for noise-free inspection images, i.e., a method that combines the determination results from the multiple methods in S22 to make a final determination.

[0128] Specifically, in S24, the reliability determination unit 103 determines the reliability of each of the determination results of the determination units 102A, 102B, 102C, and 102X. The method for determining the reliability of the determination results of the determination units 102A, 102B, and 102C is as described in embodiment 1. The reliability of the determination result of the determination unit 102X may be determined using a reliability prediction model generated for the determination unit 102X in the same manner as the reliability prediction model for the determination unit 102A described in embodiment 1.

[0129] Then, in S25, the overall judgment unit 104 judges the presence or absence of a defect using each judgment result of S22 and the reliability judged in S24. Then, the overall judgment unit 104 adds this judgment result to the inspection result data 112.

[0130] A noise-free inspection image is an image for which judgment based on the output value obtained by inputting the inspection image into a trained model generated by machine learning, as executed by the judgment units 102A and 102C, is effective. Therefore, when the first method is a method for making a final judgment by integrating the results of judgments on the presence or absence of defects by multiple methods, it is preferable that the method includes a method for making a judgment using a trained model generated by machine learning. It is also preferable that the method includes a method for making a judgment based on the output value of a classification model. In this case, it is preferable that the second method is a method for making a judgment by numerically analyzing the pixel values ​​of the inspection image.

[0131] According to the above configuration, the first method, which is a judgment method for noise-free inspection images, is a method of making a final judgment by using multiple judgment methods and then integrating the judgment results, and the multiple judgment methods include a method of making a judgment using a trained model by judgment units 102A and 102C. For noise-free inspection images, judgment using a trained model is effective, thereby enabling highly accurate judgment. Furthermore, this judgment also takes into account the judgment result of judgment unit 102X, which judges the judgment items based on the output value of the classification model, so further improvement in judgment accuracy can be expected.

[0132] Furthermore, according to the above configuration, the second method, which is a determination method for noisy inspection images, is a method in which the determination unit 102B performs a determination by numerically analyzing pixel values ​​of the inspection image. For noisy inspection images, a determination using a trained model may not be effective, but even in such cases, a valid determination may be made using numerical analysis.

[0133] Therefore, with the above configuration, it is possible to make a valid judgment whether or not the inspection image is one for which judgment using a trained model is effective.

[0134] [Embodiment 3] [Device configuration] The configuration of an information processing device 1B according to this embodiment will be described with reference to Fig. 9. Fig. 9 is a block diagram showing an example of the configuration of the main parts of the information processing device 1B. The information processing device 1B includes an inspection image generation unit 101, a determination unit 102B, a determination unit 102Y, and a determination method determination unit (acquisition unit) 106.

[0135] The determination unit 102Y determines the presence or absence of a defect using a classification model, similar to the determination unit 102X of embodiment 2. More specifically, the determination unit 102Y determines the presence or absence of a defect based on an output value obtained by inputting an inspection image into the classification model.

[0136] For example, a classification model may be used that is generated by learning so as to reduce the distance between feature amounts extracted from a group of images without noise and with defects, and so as to reduce the distance between feature amounts extracted from a group of images without noise and with defects, as in the example of Fig. 5. The determination unit 102Y can determine whether an inspection image is a noise-free, defect-free inspection image or a noise-free, defect-containing inspection image, based on the output value obtained by inputting the inspection image into this classification model.

[0137] [Processing flow] The flow of the process (determination method) executed by information processing device 1B will be described with reference to Fig. 10. Fig. 10 is a diagram showing an example of an inspection method using information processing device 1B. At the start of the process of Fig. 10, it is assumed that an ultrasound image 111 is stored in storage unit 11 and that inspection image generation unit 101 has already generated an inspection image from the ultrasound image 111.

[0138] In S31, determination unit 102Y acquires the inspection image generated by inspection image generation unit 101. Then, in S32 (determination step), determination unit 102Y determines the presence or absence of a defect using the inspection image acquired in S31.

[0139] In S33 (acquisition step), the judgment method determination unit 106 acquires the output value obtained by the judgment unit 102X in S32 when inputting the inspection image into the classification model, and determines whether the inspection image acquired in S31 is an inspection image with noise or an inspection image without noise based on the output value.

[0140] If it is determined in S33 that the inspection image contains noise (YES in S33), the determination method determination unit 106 instructs the determination unit 102B to perform determination, and the process proceeds to S35. Then, in S35 (determination step), the presence or absence of defects in the inspection image is determined by the determination unit 102B using a second method for noisy inspection images, i.e., numerically analyzing the pixel values ​​of the inspection image, and the determination result is added to the inspection result data 112.

[0141] On the other hand, if the determination method determination unit 106 determines in S33 that the inspection image is noise-free (NO in S33), the determination method determination unit 106 proceeds to the process of S34. Then, in S34, the determination method determination unit 106 adds the determination result of S32 to the inspection result data 112 as the final determination result.

[0142] In this embodiment, as in the first and second embodiments, the presence or absence of defects in the object to be inspected, i.e., whether or not there is an abnormal portion, is determined. When a defective portion is called an abnormal portion, noise has a similar appearance to the abnormal portion, so images included in the image group with noise can be said to be images that include pseudo-abnormal portions that have a similar appearance to the abnormal portion. Furthermore, inspection images included in the image group without noise can be said to be images that do not include pseudo-abnormal portions.

[0143] The output value of the classification model used by the determination unit 102Y indicates whether the inspection image belongs to a group of images with noise, an image that belongs to a group of images without noise and includes an abnormal portion, or an image that belongs to a group of images without noise and does not include an abnormal portion. In this case, as in the above example, the first method may include a process of determining whether or not the inspection object has an abnormal portion based on the output value of the classification model. The second method may include a process of determining whether or not the inspection object has an abnormal portion by numerically analyzing the pixel values ​​of the inspection image.

[0144] According to the above configuration, the first method is applied to inspection images included in the noise-free image group, and the presence or absence of an abnormality is determined based on the output value of the classification model. As described above, this classification model is generated by learning so as to reduce the distance between feature quantities extracted from the noise-free and defect-containing image group, and so as to reduce the distance between feature quantities extracted from the noise-free and defect-free image group. Therefore, by performing a determination using this classification model, it is possible to accurately determine whether an inspection image is a noise-free and defect-containing image or a noise-free and defect-free image.

[0145] However, even if the above classification model is used, it may be difficult to distinguish between pseudo-abnormal regions and abnormal regions. Therefore, with the above configuration, for an inspection image for which the output value of the classification model indicates that it belongs to a noisy image (second image group), that is, an inspection image containing pseudo-abnormal regions, the determination unit 102B determines whether or not an abnormal region exists in the inspection target by performing numerical analysis on the pixel values ​​of the inspection image. This makes it possible to accurately determine the presence or absence of an abnormal region even for an inspection image containing pseudo-abnormal regions that are difficult to distinguish from abnormal regions.

[0146] Therefore, with the above configuration, it is possible to perform appropriate judgments on both inspection images that include pseudo-abnormal regions and inspection images that do not include pseudo-abnormal regions. Of course, the first technique may include judgment processing by the judgment unit 102B and judgment processing by the judgment units 102A, 102C, etc. described in embodiment 1. Similarly, the second technique may include judgment processing by the judgment unit 102Y and judgment processing by the judgment units 102A, 102C, etc. described in embodiment 1 in addition to judgment processing by the judgment unit 102B.

[0147] In S32, the determination unit 102Y may perform the determination using a classification model that classifies the inspection images into a total of four classes: no noise and defects, no noise and no defects, no noise and defects, and no noise and no defects. Such a classification model can be generated by learning so that the distance between feature amounts extracted from a group of images with noise and defects becomes small, and so that the distance between feature amounts extracted from a group of images with noise and no defects becomes small.

[0148] In this case, in S35, both the determination result of the determination unit 102Y (noise and defect) or (noise and no defect) and the determination result of the determination unit 102B (defect or not) may be used as the final determination result. Alternatively, these determination results may be combined to determine the final determination result. The multiple determination results may be combined based on reliability, as in the first and second embodiments. In this case, however, it is desirable to weight the determination result of the determination unit 102B more heavily than the determination result of the determination unit 102Y.

[0149] [Embodiment 4] [Device configuration] The configuration of an information processing device 1C according to this embodiment will be described with reference to Fig. 11. Fig. 11 is a block diagram showing an example of the configuration of the main parts of the information processing device 1C. The information processing device 1C includes an inspection image generation unit 101, judgment units 102A to 102C, a reliability judgment unit 103, an overall judgment unit 104, a weight setting unit (acquisition unit) 107, and an overall weight determination unit 108.

[0150] The weight setting unit 107 obtains an output value obtained by inputting a target image into a classification model generated by learning so that when multiple feature amounts extracted from a group of noise-free images (a first group of images having common features) are embedded in a feature space, the distance between the feature amounts becomes small.

[0151] Then, based on the acquired output value, the weight setting unit 107 sets a weight for each determination result when combining the determination results of the determination units 102A to 102C. Specifically, when applying the first method for noise-free inspection images, the weight setting unit 107 assigns a heavier weight to the determination results of the determination units 102A and 102C that use a trained model generated by machine learning than to the determination result of the determination unit 102B that uses a numerical analysis method. On the other hand, when applying the second method for noise-containing inspection images, the weight setting unit 107 assigns a heavier weight to the determination result of the determination unit 102B than to the determination results of the determination units 102A and 102C.

[0152] The specific method for determining the weight values ​​can be determined in advance. For example, suppose we use a classification model generated by training so that the distance between feature values ​​extracted from a group of images with and without noise and defects is small, as in the example in Figure 5, and so that the distance between feature values ​​extracted from a group of images with and without noise and defects is small.

[0153] In this case, the weight setting unit 107 may convert the coordinate values ​​of the plot of the feature amounts extracted from the inspection image in the feature space into weight values ​​between 0 and 1 using a predetermined formula. In this case, the procedure for calculating the weight values ​​is, for example, as follows: (1) Calculate the distance in the feature space from the plot position of the feature amounts extracted from the inspection image to point P1, which is the center point of the class without noise and without defects. (2) As in (1) above, calculate the distance from the plot position of the feature amounts extracted from the inspection image to point P2, which is the center point of the class without noise and with defects. (3) Calculate the shorter of the calculated distances into a predetermined formula to calculate the weight value.

[0154] The above formula is a function with the distance and the weight value as variables, and the shorter the distance, the greater the weight value of the determination results of the determination units 102A and 102C. Furthermore, when a distance shorter than the radius r1 or r2 is substituted into this formula, the determination results of the determination units 102A and 102C are calculated to have a weight value equal to or greater than the weight value of the determination result of the determination unit 102B. Note that the above term "equivalent" also includes the same value. On the other hand, when a distance longer than the radius r1 or r2 is substituted into this formula, the determination results of the determination units 102A and 102C are calculated to have a weight value smaller than the weight value of the determination result of the determination unit 102B.

[0155] Furthermore, the weight setting unit 107 may determine the weight values ​​using, for example, a method similar to the method used by the reliability determination unit 103 to determine the reliability. In this case, the weight setting unit 107 calculates the reliability of the output value of the classification model using the reliability prediction model for the determination unit 102X described in embodiment 2. Then, the weight setting unit 107 may set a larger weight value for the determination results of the determination units 102A and 102C as the calculated reliability becomes higher.

[0156] As a result, for an inspection image that is similar to an image with a high classification success rate by the classification model and for which the judgment results of the judgment units 102A and 102C are likely to be valid, a larger weight is assigned to the judgment results of the judgment units 102A and 102C. On the other hand, for an inspection image that is dissimilar to the above images and for which the judgment results of the judgment units 102A and 102C are likely to be invalid, a larger weight is assigned to the judgment results of the judgment unit 102B.

[0157] Furthermore, for example, if the output value of the classification model is a value indicating the degree of certainty that the test image is a noise-free image, the weight setting unit 107 may set the weight for each determination result to a predetermined value depending on whether the degree of certainty is equal to or greater than a predetermined threshold. For example, if the degree of certainty is 0.8 or greater, the weight setting unit 107 may set the weights for the determination units 102A and 102C to 0.4 and the weight for the determination unit 102B to 0.2. In this case, if the degree of certainty is less than 0.8, the weight setting unit 107 may set the weights for the determination units 102A and 102C to 0.2 and the weight for the determination unit 102B to 0.6.

[0158] The total weight determination unit 108 calculates a weight (hereinafter referred to as a total weight) to be used when combining the determination results of the determination units 102A to 102C, using the weight set by the weight setting unit 107 and the reliability determined by the reliability determination unit 103. The total weight may be any weight that reflects both the weight set by the weight setting unit 107 and the reliability determined by the reliability determination unit 103. For example, the total weight determination unit 108 may set the arithmetic mean value of the weight set by the weight setting unit 107 and the reliability determined by the reliability determination unit 103 as the total weight.

[0159] [Processing flow] The flow of the process (determination method) executed by the information processing device 1C will be described with reference to Fig. 12. Fig. 12 is a diagram showing an example of an inspection method using the information processing device 1C. At the start of the process of Fig. 12, it is assumed that an ultrasound image 111 is stored in the storage unit 11 and that the inspection image generation unit 101 has already generated an inspection image from the ultrasound image 111.

[0160] In S41, all of the determination units 102, i.e., determination units 102A, 102B, and 102C, acquire the inspection image generated by the inspection image generation unit 101. The weight setting unit 107 and the reliability determination unit 103 also acquire the inspection image. Then, in S42, all of the determination units 102 that acquired the inspection image in S41 each use the inspection image to determine whether or not there is a defect.

[0161] In S43 (acquisition step), the weight setting unit 107 inputs the inspection image acquired in S41 into the classification model and acquires its output value. Then, in S44, the weight setting unit 107 calculates a weight according to the output value acquired in S43.

[0162] Specifically, when the output value acquired in S43 indicates that the test image is a noise-free image, weight setting unit 107 assigns a heavier weight to the determination results of determination units 102A and 102C than to the determination result of determination unit 102B. On the other hand, when the output value acquired in S43 indicates that the test image is a noise-containing image, weight setting unit 107 assigns a heavier weight to the determination result of determination unit 102B than to the determination results of determination units 102A and 102B.

[0163] In S45, the reliability determination unit 103 determines the reliability of the determination results of each of the determination units 102A, 102B, and 102C. Note that the process of S45 may be performed before S42 to S44, or may be performed in parallel with any of the processes of S42 to S44.

[0164] In S46, the total weight determination unit 108 calculates total weights using the weights calculated in S44 and the reliability calculated in S45. For example, suppose that the weights of the determination units 102A to 102C are set to 0.2, 0.7, and 0.1, respectively, and the reliability is determined to be 0.3, 0.4, and 0.3, respectively. In this case, the total weight determination unit 108 may calculate the total weights of the determination units 102A to 102C to be 0.25, 0.55, and 0.2, respectively.

[0165] In S47, the overall judgment unit 104 judges the presence or absence of a defect using each judgment result of S42 and the overall weight calculated in S46. Note that the judgment using the overall weight is similar to the judgment using the reliability described in the first and second embodiments. Then, the overall judgment unit 104 adds this judgment result to the inspection result data 112.

[0166] A noise-free inspection image is an image for which judgment based on the output value obtained by inputting the inspection image into a trained model generated by machine learning, as executed by the judgment units 102A and 102C, is effective. Therefore, when the first technique is a technique for making a final judgment by integrating the results of judgments on the presence or absence of defects by multiple techniques, it is preferable that the technique includes a technique for making a judgment using a trained model generated by machine learning. The technique may also include a technique for making a judgment by numerically analyzing the pixel values ​​of the inspection image.

[0167] In this case, when the first technique for noise-free inspection images is applied, it is preferable that the weight setting unit 107 assigns weights to the judgment results of the judgment units 102A and 102C that use a trained model equal to or greater than the weight assigned to the judgment result of the judgment unit 102B that performs numerical analysis. Basically, it is sufficient that the weight setting unit 107 assigns equal weights to each judgment result, and the final judgment result is calculated based on the reliability determined by the reliability judgment unit 103. On the other hand, when the second technique for noise-containing inspection images is applied, it is preferable that the weight setting unit 107 assigns a heavier weight to the judgment result of the judgment unit 102B that performs numerical analysis than the weight assigned to the judgment result of the judgment units 102A and 102C that use a trained model.

[0168] According to the above configuration, when the first method, which is a determination method for noise-free inspection images, is applied, the weighting of the determination result of the method using the trained model generated by machine learning is made heavier than or equal to the weighting of the determination result of the method using numerical analysis. For noise-free inspection images, determination using the trained model generated by machine learning is effective, thereby enabling highly accurate determination.

[0169] Furthermore, with the above configuration, when the second method, which is a determination method for noisy inspection images, is applied, the weighting of the determination result of the numerical analysis method is made heavier than the weighting of the determination result of the method using a trained model. For noisy inspection images, determination using a trained model may not be effective, but even in such cases, numerical analysis may be able to make a valid determination. Therefore, with the above configuration, the likelihood of obtaining a valid determination result can be increased.

[0170] Therefore, with the above configuration, it is possible to make a valid judgment whether or not the inspection image is one for which judgment using a trained model is effective.

[0171] As described above, the information processing device 1C also includes a reliability determination unit 103 that determines the reliability of each determination unit 102 based on the test image. The overall determination unit 104 then makes a determination using each determination result by the determination unit 102, the reliability determined by the reliability determination unit 103, and the weight set by the weight setting unit 107. This configuration makes it possible to derive a final determination result by appropriately considering each determination result according to the test image.

[0172] [Defect type determination] In the above embodiments, examples of determining the presence or absence of a defect have been described. However, a configuration may be adopted in which the type of defect is determined in addition to or instead of determining the presence or absence of a defect. For example, in embodiment 1, an inspection image determined to contain a defect may be input to a type determination model for determining the type of defect, and the type of defect may be determined from the output value. The type determination model can be constructed by performing machine learning using images containing defects of known types as training data. Furthermore, instead of using the type determination model, it is also possible to determine the type by image analysis or the like. Furthermore, a configuration may be adopted in which the determination unit 102 performs determination using the type determination model.

[0173] In the second embodiment as well, the determination units 102A to 102C may be configured to perform determination using a type determination model. In this case, the classification model used by the determination unit 102X may be a model that classifies based on the presence or absence of defects and the type of defects in addition to the presence or absence of noise. In learning this model, the distance between feature amounts may be expressed by Euclidean distance or the like, or may be expressed by an angle. Similarly, in the third embodiment, the determination unit 102Y may determine the type of defect.

[0174] [Application example] In the above embodiment, an example of determining whether or not there is a defect in a pipe end weld based on an ultrasonic image 111 is described, but the determination item can be any item, and the target image used for this determination can also be any image corresponding to the determination item, and is not limited to the example of the above embodiment.

[0175] For example, the information processing device 1 can be applied to radiographic testing (RT) to determine whether or not an object being inspected has a defect (which can also be called an abnormal portion). In this case, instead of a radiograph, an image resulting from the abnormal portion is detected from image data obtained using an electronic device such as an imaging plate. In this way, the information processing devices 1, 1A, 1B, and 1C can be applied to various non-destructive inspections using various data. Furthermore, in addition to non-destructive inspections, the information processing devices 1, 1A, 1B, and 1C can also be applied to object detection from still images and moving images, and to determining the classification of detected objects.

[0176] [Modification] In the above embodiment, an example was described in which the output value obtained by inputting an inspection image into a reliability prediction model was used as the reliability, but the reliability is not limited to this example as long as it is derived based on the data used by the judgment unit 102 for judgment.

[0177] For example, if the determination unit 102B determines the presence or absence of a defect using a binarized image obtained by binarizing an inspection image, the reliability prediction model for the determination unit 102B may be a model that uses the binarized image as input data. On the other hand, in this case, if the determination unit 102C determines the presence or absence of a defect using the inspection image as is, the reliability prediction model for the determination unit 102C may be a model that uses the inspection image as input data. In this way, the input data for the reliability prediction models for each determination unit 102 does not need to be exactly the same.

[0178] Furthermore, in the first embodiment, an example in which three determination units 102 are used has been described, but the number of determination units 102 may be two, or four or more. Furthermore, in the first embodiment, the determination methods of the three determination units 102 are different, but the determination methods of the determination units 102 may be the same. For determination units 102 that use the same determination method, it is sufficient that the threshold values ​​used for the determination and the teacher data that construct the trained model used for the determination are different. The same applies to the second and fourth embodiments, and the total number of determination units 102 used may be two or more.

[0179] Furthermore, the entity that executes each process described in each of the above embodiments can be changed as appropriate. For example, all or part of S12 (classification based on the presence or absence of noise), S14 (determination by each determination unit 102), S15 (reliability determination), and S16 (overall determination) in the flowchart of Fig. 6 may be executed by another information processing device. Similarly, all or part of the processes executed by the determination units 102A to 102C may be executed by another information processing device. In these cases, the number of other information processing devices may be one or more.

[0180] In this way, the functions of the information processing device 1 can be realized in a variety of system configurations. Furthermore, when building a system including multiple information processing devices, some of the information processing devices may be located on the cloud. In other words, the functions of the information processing device 1 can also be realized using one or more information processing devices that perform information processing online. This also applies to the information processing devices 1A, 1B, and 1C.

[0181] [Software implementation example] The functions of the information processing devices 1, 1A, 1B, and 1C (hereinafter referred to as "devices") can be realized by a program for causing a computer to function as the device, and a program (determination program) for causing a computer to function as each control block of the device (particularly each part included in the control unit 10).

[0182] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., a memory) as hardware for executing the program. The control device and storage device execute the program, thereby realizing the functions described in each of the above embodiments.

[0183] The program may be non-transitory and may be recorded on one or more computer-readable recording media. The recording media may or may not be included in the device. In the latter case, the program may be supplied to the device via any wired or wireless transmission medium.

[0184] Furthermore, some or all of the functions of the control blocks can be realized by logic circuits. For example, an integrated circuit in which a logic circuit that functions as each of the control blocks is formed is also included in the scope of the present invention. In addition, the functions of the control blocks can also be realized by, for example, a quantum computer.

[0185] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0186] 1, 1A, 1B, 1C Information processing equipment 102 (102A, 102B, 102C, 102X, 102Y) Judgment section 103 Reliability determination unit 104 General Judgment Department (Judgment Department) 105 Classification section (acquisition section) 106 Judgment method determination unit (acquisition unit) 107 Weight setting unit (acquisition unit)

Claims

1. an acquisition unit that acquires an output value obtained by inputting a target image into a classification model that has been generated by learning so that, when a plurality of feature amounts extracted from a first image group having a common feature are embedded in a feature space, the distance between the feature amounts becomes small; and a determination unit that determines predetermined determination items related to the target image by applying a first method for the first image group or a second method for a second image group consisting of images not belonging to the first image group, depending on the output value; the predetermined determination item is whether or not the object shown in the target image has an abnormal part; An information processing device, wherein the images included in the first image group are images of the object that do not include a pseudo-abnormal portion that has an appearance similar to the abnormal portion.

2. The first image group is an image group for which determination based on an output value obtained by inputting the target image into a trained model generated by machine learning is valid, The first method includes at least a process of determining the determination item using the trained model, The information processing apparatus according to claim 1 , wherein the second method includes at least a process of determining the determination item by numerically analyzing pixel values ​​of the target image.

3. The first image group is an image group for which determination based on an output value obtained by inputting the target image into a trained model generated by machine learning is valid, The first method is a method in which the determination items are determined using a plurality of methods, and then the results of each determination are integrated to make a final determination; The multiple techniques include: A method for determining the determination item using the trained model; and a method for determining the determination item based on the output value of the classification model, The information processing apparatus according to claim 1 , wherein the second method is a method for determining the determination item by numerically analyzing pixel values ​​of the target image.

4. An acquisition unit that acquires an output value obtained by inputting a target image into a classification model that has been generated by learning so that when multiple feature quantities extracted from a first group of images having common features are embedded in a feature space, the distance between the feature quantities becomes small; and a determination unit that determines predetermined determination items related to the target image by applying a first method for the first image group or a second method for a second image group consisting of images not belonging to the first image group, depending on the output value; the predetermined determination item is whether or not the object shown in the target image has an abnormal part; the images included in the first image group are images of the object that do not include a pseudo-abnormal portion that has an appearance similar to the abnormal portion, the images included in the second image group are images of the object including the pseudo-abnormal site, an output value of the classification model indicating whether the target image belongs to the second image group, is an image that belongs to the first image group and includes the abnormal portion, or is an image that belongs to the first image group and does not include the abnormal portion; the first technique includes at least a process of determining whether or not the object has an abnormal portion based on the output value; The information processing device, wherein the second technique includes at least a process of determining whether or not the target object has an abnormal portion by performing a numerical analysis of pixel values ​​of the target image.

5. An acquisition unit that acquires an output value obtained by inputting a target image into a classification model that has been generated by learning so that when multiple feature quantities extracted from a first group of images having common features are embedded in a feature space, the distance between the feature quantities becomes small; and a determination unit that determines predetermined determination items related to the target image by applying a first method for the first image group or a second method for a second image group consisting of images not belonging to the first image group, depending on the output value; The first image group is an image group for which determination based on an output value obtained by inputting the target image into a trained model generated by machine learning is valid, the determination unit determines the determination item using each of a plurality of methods, and then determines the determination item by combining the respective determination results; the plurality of methods include a method of determining the determination item using the trained model and a method of determining the determination item by numerically analyzing pixel values ​​of the target image; a weight setting unit that sets a weight for each determination result when combining the determination results, The weight setting unit When the first method is applied, a weight for the determination result of the method using the trained model is set to be equal to or greater than a weight for the determination result of the method performing numerical analysis; When the second method is applied, the information processing device weights the determination result of the method that performs numerical analysis more heavily than the weighting of the determination result of the method that uses the trained model.

6. a reliability determination unit that performs a process of determining a reliability, which is an index indicating the likelihood of each of the determination results, based on the target image, for each of the plurality of methods; The information processing apparatus according to claim 5 , wherein the determination unit determines the determination item using each of the determination results, the reliability determined by the reliability determination unit, and the weight set by the weight setting unit.

7. A determination method executed by an information processing device, an acquisition step of acquiring an output value obtained by inputting a target image into a classification model generated by training the classification model so that, when a plurality of feature amounts extracted from a first group of images having a common feature are embedded in a feature space, the distance between the feature amounts becomes small; a determination step of determining predetermined determination items related to the target image by applying a first method for the first group of images or a second method for a second group of images that does not belong to the first group of images according to the output value, the predetermined determination item is whether or not the object shown in the target image has an abnormal part; A determination method, wherein the images included in the first image group are images of the object that do not contain a pseudo-abnormal portion that has an appearance similar to the abnormal portion.

8. 2. A determination program for causing a computer to function as the information processing device according to claim 1, the determination program causing the computer to function as the acquisition unit and the determination unit.

Images