Image processing apparatus, image processing method, and program

Abstract

To reduce the burden of inputting a crack width measured in the image of an object to be inspected.SOLUTION: An image processing apparatus is configured to present a user interface including an image of an inspection object having crack deformation and a plurality of crack model shapes each having a different width that can be superimposed on the image and can be moved to positions corresponding to the cracks shown in the image according to a user input; to acquire the user input for inputting the crack shape of a crack shown in the image included in the user interface; and to determine a crack width candidate for the crack according to the user input based on the crack shape according to the user input and the width of the crack model shape superimposed and displayed at a position corresponding to the crack shape according to the user input.SELECTED DRAWING: Figure 7

Description

[Technical Field]

[0001] The present disclosure relates to an image processing device, an image processing method, and a program, and in particular to a technique for detecting an abnormality from a captured image of an inspection object. [Background technology]

[0002] A method using captured images of an inspection target, such as the wall surface of a concrete structure, to inspect for abnormalities is known. For example, a computer device can automatically detect abnormalities such as cracks from the captured images and determine the attributes of the abnormalities, such as crack width, using image analysis with machine learning technology. In this case, partial image quality degradation due to blurring or blurring during shooting may result in erroneous or undetected abnormalities in part of the captured image. In this case, a user may review the captured images and detection results and add or edit abnormalities. In particular, when adding a crack, a user may input the crack width determined by looking at the captured image. Even without image analysis, a user can visually detect cracks in the captured images and input the crack shape and width using a user interface provided by the computer device.

[0003] Patent Documents 1 and 2 disclose techniques that make it easy for users to measure crack widths on captured images. Patent Document 1 proposes that when a scale image with a line drawing for measuring crack widths is superimposed on an image of a structure, the position and orientation of the scale image are set according to the position and direction of the crack in the image of the structure. Patent Document 2 also proposes superimposing an object indicating the scale of the size of the object in real space on the image of the structure.

[0004] On the other hand, various techniques have been proposed for automatically detecting deformations through image processing. Patent Document 3 describes a method for automatically detecting cracks and their widths from images. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 2017 / 122641 [Patent Document 2] International Publication No. 2017 / 168737 [Patent Document 3] Japanese Patent Application Laid-Open No. 2013-195074 Summary of the Invention [Problem to be solved by the invention]

[0006] However, even when the techniques described in Patent Documents 1 and 2 are used, the user still has to input the measured crack width, which is a heavy workload.

[0007] The present disclosure aims to reduce the burden of the operation of inputting the crack width measured in an image of an object to be inspected. [Means for solving the problem]

[0008] An image analysis device according to an embodiment includes: A presentation means for presenting a user interface including an image of an inspection object having crack deformation and model shapes of a plurality of cracks each having a different width that can be superimposed on the image and can be moved to positions corresponding to the cracks shown in the image according to user input; an input acquisition means for acquiring a user input for inputting a shape of a crack shown in the image included in the user interface; A determination means for determining a crack width candidate for the crack according to the user input based on the shape of the crack according to the user input and the width of the model shape of the crack superimposed and displayed at a position corresponding to the shape of the crack according to the user input; Equipped with. [Effects of the Invention]

[0009] This reduces the burden of inputting the crack width measured in the image of the object to be inspected. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing an example of the hardware configuration of an image analysis apparatus according to an embodiment. [Figure 2] FIG. 1 is a block diagram showing an example of the functional configuration of an image analysis apparatus according to an embodiment. [Figure 3] FIG. 4 is a diagram illustrating a part of a captured image. [Figure 4] FIG. 10 is a diagram illustrating an example of an editing screen. [Figure 5] FIG. 1 is a diagram illustrating a crack scale. [Figure 6] 10A and 10B are diagrams illustrating examples of changes in actual size indicated by a crack scale according to display magnification. [Figure 7] A diagram showing an example of the use of crack scale. [Figure 8] 10A to 10C are diagrams illustrating a method for adding cracks. [Figure 9] 1 is a flowchart of an image analysis method according to an embodiment. [Figure 10] FIG. 10 is a diagram illustrating intermediate data. [Figure 11] 1 is a flowchart of an image analysis method according to an embodiment. [Figure 12] FIG. 10 is a diagram illustrating an example of an editing screen when two crack width candidates are obtained. [Figure 13] FIG. 10 is a diagram illustrating an example of an editing screen when multiple crack width candidate determination methods are used. [Figure 14] FIG. 10 is a diagram illustrating an example of an editing screen when a plurality of default values ​​are used. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. Note that the following embodiments do not limit the scope of the claims. Although multiple features are described in the embodiments, not all of these multiple features are necessarily essential to the embodiments, and multiple features may be combined arbitrarily. Furthermore, in the accompanying drawings, the same reference numerals are used to designate the same or similar components, and redundant explanations will be omitted.

[0012] An image analysis device according to one embodiment can assist in determining crack widths in an image of an inspection target. In the embodiment described below, a user uses the image analysis device to input the shape of a crack the user has discovered in a captured image. In one embodiment, the image analysis device performs image analysis to detect crack deformations in the captured image of the inspection target. In this case, the user can input the shape of a crack not detected by image analysis to the image analysis device. At this time, the image analysis device records the width of the discovered crack. In this embodiment, the burden on the user of inputting new crack widths added in this way can be reduced.

[0013] <Determining crack width candidates based on crack model> In the embodiment described below, a crack model shape that can be superimposed on a captured image is used to measure crack widths. In this embodiment, a crack scale including multiple crack model shapes is used. Then, crack width candidates to be added are determined according to the superimposed position of such crack model shapes.

[0014] In this specification, crack deformation refers to cracks that occur on the surface of an object to be inspected. The object to be inspected is, for example, a concrete structure such as a highway, a bridge, a tunnel, or a dam. A crack is a linear damage that has a starting point, an end point, a length, and a width. Cracks can occur on the surface of an object to be inspected, for example, a concrete surface, due to damage, aging, earthquake impact, or other factors.

[0015] (Hardware configuration) An image analysis device according to an embodiment can be realized using a computer device. First, an example of the hardware configuration of the image analysis device will be described with reference to FIG. 1. FIG. 1 is a block diagram showing an example of the hardware configuration of a computer device that operates as the image analysis device 100. Note that the processing of the image analysis device 100 can be realized by a single computer device. Furthermore, the functions of the image analysis device 100 may be distributed among multiple computer devices. In this way, the image analysis device 100 may be configured by, for example, multiple information processing devices connected via a network. For example, the functions of the image analysis device 100 may be provided as a cloud service.

[0016] The image analyzing device 100 includes a control unit 101, a nonvolatile memory 102, a work memory 103, a storage device 104, an input device 105, an output device 106, a network interface 107, and a system bus 108.

[0017] The control unit 101 controls the entire image analysis device 100. The control unit 101 may include an arithmetic processor such as a CPU or an MPU. The non-volatile memory 102 is a ROM that stores programs or parameters executed by the processor of the control unit 101. These programs are programs for executing the processes according to each embodiment. The work memory 103 is a RAM that temporarily stores programs or data supplied from an external device or a storage device 104. The storage device 104 is an internal device such as a hard disk or a memory card built into the image analysis device 100, or an external device such as a hard disk or a memory card detachably connected to the image analysis device 100. The storage device 104 may be a memory card or a hard disk formed of a semiconductor memory or a magnetic disk. The storage device 104 may also be an optical disc such as a DVD or Blu-ray (registered trademark) disc. In this case, the storage device 104 may include a disk drive that reads or writes data from or to the optical disc.

[0018] The input device 105 is an operating member such as a mouse, keyboard, or touch panel that accepts user operations. The input device 105 can output the accepted operation instructions to the control unit 101. The output device 106 is a display device such as a monitor or a display device composed of an LCD or an organic EL display. The output device 106 can display data stored in the image analyzing device 100 or data supplied from an external device. The network interface 107 communicably connects the image analyzing device 100 to a network such as the Internet or a LAN (Local Area Network). The system bus 108 connects each component of the image analyzing device 100 so that data can be exchanged. The system bus 108 can include an address bus, a data bus, and a control bus.

[0019] The nonvolatile memory 102 stores an OS (operating system), which is basic software executed by the control unit 101. The nonvolatile memory 102 also stores applications that cooperate with the OS to realize applied functions. In this embodiment, the nonvolatile memory 102 stores an application that realizes image analysis processing by the image analyzing device 100 to detect abnormalities in captured images of an inspection target, which will be described later.

[0020] In this way, the processor of the control unit 101 executes a program stored in a memory such as the non-volatile memory 102, the work memory 103, or the storage device 104, thereby realizing the functions of each unit shown in FIG. 2 (described later). The processing of the image analyzing device 100 according to this embodiment is realized by loading software provided by an application. The application has software for utilizing the basic functions of the OS installed in the image analyzing device 100. Note that the OS of the image analyzing device 100 may have software for realizing the processing according to this embodiment.

[0021] (Functional configuration) Next, functional blocks of an image analysis device according to an embodiment will be described with reference to Fig. 2. Fig. 2 is a functional block diagram of an image analysis device 100 according to an embodiment. As described above, the functions of the image analysis device 100 shown in Fig. 2 can be realized by a computer, but some or all of the functions of the image analysis device 100 may also be realized by dedicated hardware.

[0022] The image analysis device 100 has a folder management unit 211, a folder setting storage unit 212, an image management unit 213, an image storage unit 214, an image analysis unit 215, an analysis result management unit 216, an analysis result storage unit 217, an analysis result editing unit 218, and an editing result storage unit 219.

[0023] The folder management unit 211 has a function of managing folders that store images. For example, the folder management unit 211 can create, set, delete, or list folders. In this embodiment, a folder is a unit of image management. For example, an image can be saved in association with a specific folder. Registering an image in a folder means associating the image with the folder.

[0024] The folder management unit 211 provides the user with folder-related functions. For example, when creating a folder, the user can input folder settings. The folder settings can include the folder name and actual size information. The folder settings can also include runtime notes. This information can be input when creating a folder. For example, the user can input this information on a folder creation screen that is displayed when creating a folder. The folder settings can also include information such as creation date and access date and time for folder management.

[0025] Actual size information is information indicating the size in real space of an object captured in an image. Actual size information can indicate the resolution of the object captured in the image. In this embodiment, the actual size information is an image actual size value indicating the actual size of the object captured in the image per pixel. Specifically, the image actual size ratio can be a conversion value indicating the actual size value per pixel (e.g., mm). The image actual size ratio is expressed as the ratio between pixels and the actual size value (mm / pixel). Such actual size information is also called image actual size conversion, resolution, pixel actual size value, image actual size value, etc.

[0026] Images often do not include actual size information. Therefore, the user can manually input the actual size information. The actual size information of an image can also be estimated from the size of a deformation whose actual size is known. Furthermore, a distance measuring device may be used to obtain the actual size information at the same time as capturing the image. In this embodiment, when creating a folder, the user manually inputs the actual size information of the images to be registered in this folder.

[0027] The folder setting storage unit 212 can store the settings of the folders described above.

[0028] The image management unit 213 has a function for managing images. For example, the image management unit 213 can save images, register images in folders, delete images, display a list of images, view images, change file names, etc. The image management unit 213 may have a function for changing the actual size ratio of each image.

[0029] The image management unit 213 can manage images by registering one or more images in a folder. For example, when saving an image, the user can specify one folder to register the image in. The image management unit 213 can apply the image actual size ratio setting set for each folder to the image registered in the folder specified by the user.

[0030] The image storage unit 214 can store image data and image-related settings such as image size ratio or file name.

[0031] The image analysis unit 215 detects cracks based on the image of the inspection target. The image analysis unit 215 can detect crack deformation in the image through image analysis processing. The image analysis unit 215 can use a learning model created by machine learning to detect deformation from the image of the inspection target. For machine learning, deep learning used in the field of AI (artificial intelligence) can be used. The image analysis unit 215 can also estimate attributes of the detected cracks. The image analysis unit 215 can use a learning model created by machine learning to estimate the attributes of the cracks. The specific method is not particularly limited, and the method described in Patent Document 3 can be used, for example. However, the method of detecting cracks and the method of estimating crack attributes are not particularly limited. In this embodiment, the image analysis unit 215 estimates the width of the crack (hereinafter, crack width) as an attribute of the crack. The image analysis unit 215 can also estimate the length of the crack as an attribute of the crack. In this way, the image analysis unit 215 can perform image analysis processing on the images registered in the folder.

[0032] The analysis result storage unit 217 stores the results of the image analysis performed by the image analysis unit 215 .

[0033] The analysis result management unit 216 has a function for viewing and acquiring the analysis results stored in the analysis result storage unit 217. The analysis result management unit 216 can present a user interface including an image of an inspection target having crack deformation. For example, the analysis result management unit 216 can present the user with an editing screen shown in FIG. 4. As shown in FIG. 4, this user interface can include multiple crack model shapes, each having a different specific width, that can be superimposed on the image of the inspection target. The crack model shapes can be moved to positions corresponding to the cracks shown in the image according to user input.

[0034] The analysis result editing unit 218 edits the results of image analysis by the image analysis unit 215. The analysis result editing unit 218 can edit the crack deformation detection results stored in the analysis result storage unit 217 or the edited result storage unit 219. More specifically, the analysis result editing unit 218 can add or change the detection results in accordance with the received user operation.

[0035] The analysis result editing unit 218 may include an input acquisition unit 221 and a width determination unit 222. The input acquisition unit 221 may acquire user input for inputting the shape of a crack shown in an image. The input acquisition unit 221 may add crack deformations according to such user input to the results of image analysis. In this embodiment, the editing screen presented to the user by the analysis result management unit 216 has components such as buttons for editing, as shown in FIG. 4. The user may perform input on such an editing screen to add crack deformations. In this embodiment, the shape of a crack is represented by a line. For example, the shape of a crack is represented by a line segment or a broken line. A broken line can be represented by multiple line segments. Furthermore, a broken line can be represented by a combination of a starting point, one or more vertices, and an ending point.

[0036] The width determination unit 222 can also determine crack width candidates for cracks based on user input. In this embodiment, the width determination unit 222 can determine crack width candidates based on the shape of the crack based on user input. The width determination unit 222 can also determine crack width candidates based on the width of a model shape of the crack superimposed and displayed at a position corresponding to the shape of the crack based on user input. The editing screen shown in FIG. 4 can include a crack scale including the model shape of the crack, as shown in FIG. 5. In such an embodiment, the width determination unit 222 can determine crack width candidates taking into account the position of the crack scale. A specific method for determining candidates will be described later.

[0037] The editing result storage unit 219 stores the results of image analysis after editing. The editing result storage unit 219 may also store image analysis settings. The analysis result management unit 216 may have a function for viewing and obtaining the analysis results after editing stored in the editing result storage unit 219.

[0038] (Structural Deformation Detection Workflow) Before describing the operation of an image analysis device according to an embodiment, an example of a workflow for detecting structural abnormalities through image analysis will be described. In this example, in this embodiment, image analysis using a learning model is performed on images of the wall surface of a concrete structure captured by a camera to detect abnormalities.

[0039] When field workers photograph structures, it is often difficult to obtain a single image that includes the entire inspection area and has a sufficient image scale ratio to recognize abnormalities. Therefore, in many cases, workers will zoom in on a portion of the inspection area so that it is visible, and then repeatedly photograph this portion while moving the photographing range. After that, the obtained images are processed by enlarging, reducing, rotating, projecting, color adjusting, or noise removal, and then stitched together to obtain a single combined image.

[0040] Such combined images are generated for each part of the structure. The worker can repeat the above process according to the number of components in the structure's drawing. For example, if the inspection target is a bridge pier with a rectangular cross section, the above process is repeated for each of the four sides to prepare a set of four images for "X Bridge Pier 1." Depending on the inspection target, a standard image scale ratio (e.g., 0.5 mm / pixel for bridges and 2.0 mm / pixel for tunnels) may be set. In this case, images are taken to satisfy this scale ratio. On the other hand, for key inspection targets, higher-resolution images may be taken. In this case, a composite image can be generated to match the resolution and position of the structural drawing. The images are then ready for image analysis.

[0041] Unlike the method of visually inspecting images and manually recording defects, defects detected by image analysis, which will be described later, may contain false positives or missed detections. Therefore, visual confirmation and correction are performed using an image analysis device or an external server. Furthermore, the image analysis device can create an inspection report based on these detection results. For example, the image analysis device can create an inspection report that includes drawings or captured images on which markers indicating detected cracks are superimposed. The image analysis device can also record the length and width of the detected cracks in the inspection report.

[0042] (User Interface) Next, the editing screen used in this embodiment will be described with reference to Fig. 3 and Fig. 4. Fig. 3 is a diagram showing an example of a portion of a captured image of a wall surface of a concrete structure to be inspected. Cracks 311 to 315 exist on a wall surface 300 of the concrete structure.

[0043] 4 is a diagram illustrating an example of an editing screen. The editing screen 400 includes an information display field 410, a detection result display field 420, an editing operation field 430, a legend display field 440, and an attribute display field 450.

[0044] The information display field 410 includes display information 411. The display information 411 may include information about the captured image being displayed, such as the image file name or the image actual size ratio. The display information 411 may also include information about the folder in which the captured image is registered, such as the folder name or runtime memo.

[0045] The detection result display field 420 includes the captured image. The detection result display field 420 also includes information showing the detection results of crack deformation. In FIG. 4, detection results 422 and 423 showing cracks are displayed superimposed on the image of the wall surface 300 of the concrete structure being inspected. The detection results 422 and 423 show the shape of the detected crack. As shown in FIG. 4, the detection results 422 and 423 may be broken lines that follow the crack. Attributes such as length and width (thickness) are determined for each crack. These lengths and widths represent the actual dimensions of the object being inspected. Each crack is displayed in a different display format (e.g., different colors or line types) depending on its length or width so that it can be distinguished.

[0046] The editing operation field 430 displays components related to editing. In the example of FIG. 4, a save editing result button 431, a cancel editing button 432, an undo button 433, a redo button 434, and a scale on / off button 435 are displayed. The save editing result button 431 is a button for saving the results of the editing operation performed by the user using the mouse cursor 423. The cancel editing button 432 is a button for discarding the results of the editing operation performed by the user without saving them. When the cancel editing button 432 is pressed, the detection result returns to the state before the start of editing. The undo button 433 is a button for canceling the previous editing operation. The redo button 434 is a button for redoing the editing operation canceled by the undo button. The scale on / off button 435 is a button for switching whether or not to display a crack scale, which will be described later with reference to FIG. 5.

[0047] The legend display field 440 displays the correspondence between crack attributes and display formats. The legend display 441 shown in FIG. 4 shows display formats corresponding to each of the four types of crack width classifications. Note that the legend display shown in FIG. 4 is only an example. The number of crack width classifications and display formats are not limited to the example in FIG. 4. Furthermore, the classification and display format may be set according to the operation when creating a folder. Furthermore, the legend display 441 may show legends for deformations other than cracks.

[0048] Actual size information such as the length and width of a crack can be obtained by the image analysis unit 215 as described above. For example, the actual size information can be calculated based on the number of pixels in the image that are determined to represent a crack using a learning model and the image actual size ratio. In addition, by comparing the actual size information of the crack with the drawing data, the coordinates of the crack on the image shown in the analysis results can be converted to coordinates in the coordinate system of the drawing. Based on these converted coordinates, the detection results can be viewed or edited in the image analysis device or an external server.

[0049] The attribute display field 450 displays attribute information of the deformation. FIG. 4 shows, as an example, a screen when a crack 421 is selected using the mouse cursor 423. In this case, the crack width display 451 indicates the crack width of the crack 421 detected by image analysis. The crack width display 451 displays a numerical value indicating that the crack width value is 3.00 mm. In this example, only the crack width is displayed in the attribute display field 450. However, other attributes such as length may also be displayed in the attribute display field 450. Furthermore, the crack width may be input into the crack width display 451. The user can edit the crack width value by inputting a new crack width.

[0050] (crack scale) Next, the crack scale and how to use the crack scale during editing will be described with reference to FIGS.

[0051] FIG. 5 shows an example of a crack scale. The crack scale 500 includes multiple crack model shapes (hereinafter simply referred to as models) having specific widths. The multiple crack models are line segments, each with a different width. In FIG. 5, the crack scale 500 includes a crack width guide 501 that indicates such crack models. In this example, the crack width guide 501 indicates crack models having widths of 1 to 5 pixels. Each model is a line segment having a width of 1 to 5 pixels. As shown in FIG. 5, the crack scale 500 may also indicate the actual width of each model (e.g., in mm). The crack scale 500 may also include a crack length guide 502. The crack length guide 502 indicates a scale that allows the length of a crack on the image to be measured and the actual size corresponding to each scale.

[0052] The actual size displayed on the crack scale 500 can be determined based on the image actual size ratio set for the folder or image and the display magnification of the captured image displayed on the editing screen 400. FIG. 6 shows an example of the actual size shown on the crack scale according to the display magnification of the captured image. In FIG. 6(A), the image size and display size of the captured image are the same size (number of pixels). Also, the image actual size ratio set for the image is 0.5 mm / pixel. In this case, one pixel represents 0.50 mm. The actual size is displayed on the crack scale 500 so that one pixel corresponds to 0.50 mm.

[0053] On the other hand, the display magnification of the captured image in FIG. 6(B) is twice that of FIG. 6(B). In this case, the actual size corresponding to one pixel is half that of FIG. 6(A), i.e., 0.25 mm. The actual size displayed on the crack scale 500 is such that one pixel corresponds to 0.25 mm. In this way, the actual size displayed on the crack scale can be dynamically changed as the displayed captured image is enlarged or reduced. On the other hand, the size of the crack scale does not need to be changed as the displayed captured image is enlarged or reduced. For example, the width of the crack model or scale included in the crack scale does not need to be changed as the displayed captured image is enlarged or reduced.

[0054] The crack scale 500 can be displayed by operating the scale ON / OFF button 435 on the editing screen 400. The display position and display angle of the crack model or crack scale can be changed according to user input. That is, the user can translate and rotate the crack scale 500. In the inspection of concrete structures, the width of the widest point of a continuous crack is typically measured as the crack width. Therefore, the user can align the crack width guide 501 with the widest part of the crack 313 that was not detected as a crack by image analysis. For example, the crack scale 500 can be moved and rotated according to user operation using the mouse cursor 423, etc. This configuration allows the width of a crack at any position in the captured image displayed in the detection result display field 420 to be measured. In the example of FIG. 7, the user visually measures the width of the crack 313 to be 2.50 mm using the crack width guide 501.

[0055] It is not necessary to use a crack scale that includes multiple crack models. For example, multiple crack models with specific widths may be independently movable and rotatable according to user input.

[0056] (Adding cracks) Next, with reference to FIGS. 8 and 9, a method for adding a new crack and inputting a crack width in accordance with user input will be described. As described above, the image analysis unit 215 can detect crack deformations through image analysis processing. Then, the analysis result management unit 216 can present the detected cracks on a user interface. On the other hand, the image may show crack deformations that were not detected by the image analysis unit 215 or not presented by the analysis result management unit 216. In this way, the user can add crack deformations that were not presented by the analysis result management unit 216 to the detection results. In one embodiment, the user input is for inputting the shape of a crack deformation that was not presented by the analysis result management unit 216. For example, the user can perform user input to input the shape of a crack deformation.

[0057] FIG. 8 shows an example of adding a new crack when editing the detection results. The user can add a crack at a desired position in the detection result display field 420 using the mouse cursor 423 or the like. The user starts adding a crack by clicking the start point of the crack using the mouse cursor 423. The user can add a vertex of the crack by moving the mouse cursor 423 to the next position and clicking. The user can also input the end point of the crack by double-clicking, for example. In this way, the addition of one crack is completed.

[0058] Next, the crack width added by the user is determined. The user may input a numerical value indicating the crack width measured visually using the crack scale 500. Meanwhile, in this embodiment, the width determination unit 222 determines a crack width candidate. By using the candidate determined in this manner, the burden of inputting information indicating the crack width added by the user can be reduced.

[0059] 9 is a flowchart of a method performed by the image analyzing device 100 according to this embodiment. According to this method, the crack width of a new crack 703 added by a user is determined. Before this process begins, the analysis result management unit 216 presents a user interface including an image of an inspection target having a crack deformation.

[0060] In S901, the width determination unit 222 acquires information specifying the shape of the crack in accordance with the user input received by the input acquisition unit 221. In this example, the crack is represented by a polygonal line including multiple line segments.

[0061] In S902, the width determination unit 222 acquires information indicating the position of the crack model superimposed on the captured image. The width determination unit 222 can also acquire information indicating the orientation of the crack model. In this example, the control unit 101 acquires information indicating the position and orientation of the crack scale 500 displayed in the detection result display field 420 by the analysis result management unit 216. The arrangement of each crack model is fixed on the crack scale 500. Therefore, the width determination unit 222 can determine the position and orientation of each crack model based on the position and orientation of the crack scale 500.

[0062] In S903, the width determination unit 222 determines a crack model corresponding to the crack according to the user input, according to the information acquired in S901 and S902. Specifically, the width determination unit 222 can determine that a crack model superimposed and displayed at a position corresponding to the crack according to the user input is the crack model corresponding to the crack according to the user input. When one of the multiple line segments indicating the crack and a specific model are displayed at approximately the same position, the width determination unit 222 can determine that the specific model is superimposed and displayed at a position corresponding to the crack. When one of the multiple line segments indicating the crack and the specific model intersect, the width determination unit 222 can determine that they are displayed at approximately the same position. Furthermore, when the distance between one of the multiple line segments indicating the crack and the specific model is within a threshold, the width determination unit 222 can determine that they are displayed at approximately the same position.

[0063] Furthermore, the width determination unit 222 can determine a crack model extending in a direction corresponding to the crack according to the user input as the crack model corresponding to the crack according to the user input. For example, the width determination unit 222 can determine a crack model that is superimposed and displayed at a position corresponding to the crack according to the user input and that extends in a direction corresponding to the crack according to the user input as the crack model corresponding to the crack according to the user input. When one of the multiple line segments indicating the crack and a specific model are displayed at approximately the same position and extend in approximately the same direction, the width determination unit 222 can determine that this model is the crack model corresponding to the crack according to the user input. For example, when the angle between one of the multiple line segments indicating the crack and the specific model is within a threshold, the width determination unit 222 can determine that they extend in approximately the same direction.

[0064] As a specific example, the width determination unit 222 can determine whether the position of the added crack 703 includes the width display position of the crack scale 500. The width display position refers to the position of the crack model (line segment) indicated by the crack width guide 501. Specifically, the width determination unit 222 can determine that the position of the crack 703 includes a specific width display position when any of the line segments constituting the crack 703 intersects with a specific crack model at a predetermined relative angle or less. This predetermined relative angle may be given in advance or may be set by the user. The determination of the intersection between the line segment and the model and the calculation of the relative angle can be performed according to a general geometric method.

[0065] The determination method is not limited to the above method. For example, instead of intersection determination, the distance between the line segment indicating the crack and the model can be calculated. Then, the combination of the line segment indicating the crack and the model that gives the smallest distance smaller than a predetermined value can be used. This method also makes it possible to determine the crack model corresponding to the crack according to the user input.

[0066] The width determination unit 222 determines whether or not a crack model corresponding to the crack according to the user input exists according to these methods. If the width determination unit 222 determines that such a model exists, the process proceeds to S904. On the other hand, if the width determination unit 222 determines that such a model does not exist, the process proceeds to S905.

[0067] In S904, the width determination unit 222 determines a crack width candidate for the crack according to the user input based on the width of the crack model corresponding to the crack according to the user input determined in S903. The width determination unit 222 can use the width of the crack model corresponding to the crack according to the user input as the crack width candidate for the crack according to the user input. In the example of FIG. 8, the width of the crack model corresponding to the crack according to the user input is 2.50 mm. Therefore, the width determination unit 222 determines the crack width of 2.50 mm as the crack width candidate for the crack according to the user input.

[0068] In S903, multiple crack models corresponding to the crack according to the user input may be determined. In this case, the width determination unit 222 can select, from the multiple models, a model that is displayed closer to the crack according to the user input. As another method, the width determination unit 222 can select, from the multiple models, a model that is displayed closer to the crack according to the user input. As another method, the width determination unit 222 can select one model from the multiple models so that the intersection angle between the crack according to the user input and the model is smaller. Then, the width determination unit 222 can determine a crack width candidate for the crack according to the user input based on the width of the selected model.

[0069] In S905, the width determination unit 222 sets an arbitrary crack width as a crack width candidate for the crack according to the user input. This crack width candidate is used as a default value for the crack width. For example, the width determination unit 222 can set the crack width candidate to a predetermined value, such as 0.01 mm, which is the minimum crack width. This predetermined value is not particularly limited. The predetermined value may be given in advance or may be set by the user.

[0070] In S906, the width determination unit 222 reflects the crack width candidate determined in S904 or S905 in the crack width display 451. In this way, the crack width for the crack according to the user input can be automatically input. That is, the crack width candidate can be used as the crack width for the crack according to the user input. On the other hand, the user may further modify the crack width for the crack according to the user input by modifying the numerical value of the crack width display 451.

[0071] According to the above embodiment, when a crack deformation is added by a user operation, the crack width indicated by the crack model superimposed on the crack deformation for measurement can be used as the added crack width. Therefore, when a new crack deformation is added, the burden of inputting the crack width for each crack can be reduced.

[0072] <Determining crack width candidates based on intermediate data> The method for determining crack width candidates is not limited to the method of referring to the position of the crack scale. Below, an example of reflecting crack width candidates using intermediate data obtained in image analysis processing will be described with reference to Figures 9, 10, and 11. The same reference numerals are used for components already described, and their description will be omitted.

[0073] (Intermediate data) In the image analysis process, crack detection results may be determined to be inappropriate depending on various analysis parameters and a preset reliability threshold. Such inappropriate detection results are excluded from the final detection results. Meanwhile, the detection results before the exclusion process, i.e., information on crack candidates, can be saved as intermediate data. Such intermediate data includes location information on crack and other potential defects, along with attribute information such as the reliability of the detection. In other words, the image analysis unit 215 can detect the location of crack candidates and estimate the width of the crack candidates based on images of the inspection target having crack and other defects through image analysis. The image analysis unit 215 can then save such intermediate data, including information on crack candidates, in the analysis result storage unit 217.

[0074] FIG. 10 shows an example of intermediate data. FIG. 10 shows crack candidates corresponding to IDs 1 to 6. For example, if the reliability threshold is set to 0.5, the detection result for ID 6 will be excluded from the final analysis result. In this case, the detection result for ID 6 will not be displayed in the detection result display field 420. Such a reliability threshold may be determined in advance or may be set by the user. Note that the method for determining inappropriate detection results is not limited to the method using reliability. For example, cracks of a certain length or less may be excluded in order to remove small cracks, etc.

[0075] In this way, the image analysis unit 215 can select some of the detected crack candidates as crack deformation detection results. At this time, the analysis result management unit 216 can present the crack candidates selected from the crack candidates as crack deformation detection results on the user interface. On the other hand, as already described, the image may show crack deformations that were not detected by the image analysis unit 215 or not presented by the analysis result management unit 216. The user can add crack deformations that were not presented by the analysis result management unit 216 to the detection results in this way. Here, the crack deformations added by the user may correspond to crack candidates detected by the image analysis unit 215. In this embodiment, the width determination unit 222 determines crack width candidates for the cracks according to the user input based on the crack shape according to the user input and the widths estimated for the crack candidates detected at positions corresponding to the crack shape according to the user input.

[0076] In this embodiment, the editing operation of the detection results can be performed as described above with reference to Fig. 8. However, it is not necessary to display the crack scale 500 as in Fig. 8.

[0077] 11 is a flowchart of a method performed by the image analyzing device 100 according to this embodiment. According to this method, the crack width of a new crack 703 added by a user is determined. Before this process begins, the analysis result management unit 216 presents a user interface including an image of an inspection target having a crack deformation. S1101 is performed in the same manner as S901.

[0078] In S1102, the width determination unit 222 acquires the intermediate data described above obtained in the image analysis process for the image displayed in the detection result display field 420. In this example, the image analysis unit 215 detects the position of a crack candidate based on the image of the inspection object having a crack deformation, and estimates the crack width of the crack candidate.

[0079] In S1103, the width determination unit 222 determines the crack candidate detected by the image analysis unit 215 that corresponds to the shape of the crack according to the user input, according to the information acquired in S1101 and S1102. Specifically, the width determination unit 222 can determine the crack candidate detected at the position corresponding to the shape of the crack according to the user input as the crack candidate corresponding to the crack according to the user input. In this way, the width determination unit 222 can determine whether the position of the added crack matches the position of the crack candidate indicated by the intermediate data.

[0080] This determination can be made based on the positions of the line segments that make up the added crack and the coordinates of the crack candidate. For example, if the distance between the center coordinates of the added crack and the center coordinates of a specific crack candidate is within a certain range, the width determination unit 222 can determine that the added crack corresponds to this specific crack candidate.

[0081] Furthermore, the width determination unit 222 can determine a crack candidate having a shape corresponding to a crack specified by a user input as a crack candidate corresponding to the crack specified by the user input. The width determination unit 222 can determine a crack candidate located at a position corresponding to a crack specified by a user input and having a corresponding shape as a crack candidate corresponding to the crack specified by the user input. For example, if the similarity between the shape of the added crack and the shape of a specific crack candidate indicated by the intermediate data is equal to or greater than a threshold, the width determination unit 222 can determine that the added crack corresponds to this specific crack candidate. Such similarity can be determined by treating the coordinates of the line segments or polygonal lines constituting the added crack and the coordinates of the line segments or polygonal lines constituting the crack candidate as feature points. Specifically, such similarity can be calculated using a technique such as feature point matching. Furthermore, the width determination unit 222 may further perform a determination based on the similarity in the size or position of the circumscribing rectangle between the added crack and the crack candidate. For example, an additional condition may be that the difference in size of the circumscribing rectangles falls within a certain range.

[0082] The width determination unit 222 determines whether or not a crack candidate corresponding to the crack specified by the user input exists according to these methods. If the width determination unit 222 determines that such a model exists, the process proceeds to S1104. On the other hand, if the width determination unit 222 determines that such a model does not exist, the process proceeds to S1105.

[0083] In S1104, the width determination unit 222 determines a crack width candidate for the crack according to the user input based on the width of the crack candidate corresponding to the crack according to the user input determined in S1103. The width determination unit 222 can use the width of the crack candidate corresponding to the crack according to the user input as the crack width candidate for the crack according to the user input.

[0084] In S1103, multiple crack candidates corresponding to the crack according to the user input may be determined. In this case, the width determination unit 222 can select, from the multiple crack candidates, the crack candidate detected at a position closer to the crack according to the user input. As an alternative method, the width determination unit 222 can select, from the multiple crack candidates, the crack candidate having a shape closer to the crack according to the user input. Then, the width determination unit 222 can determine a crack width candidate for the crack according to the user input based on the width of the selected crack candidate.

[0085] S1105 and S1106 are performed in the same manner as S905 and S906.

[0086] According to the above embodiment, when a crack deformation is added by a user operation, the crack width of the corresponding crack candidate obtained by the image analysis process can be used as the added crack width. Therefore, when a new crack deformation is added, the burden of inputting the crack width for each crack can be reduced.

[0087] (Suggestion of multiple crack candidates) So far, a method for determining one crack width candidate corresponding to the added crack has been described. However, two or more crack width candidates may be determined. For example, in S903 or S1103, multiple models or crack candidates corresponding to the crack according to the user input may be determined. In this case, in S904 or S1104, the width determination unit 222 can determine multiple crack width candidates. For example, the width determination unit 222 can determine multiple crack width candidates, each corresponding to one of the multiple models or crack candidates.

[0088] In such an embodiment, in S906 or S1106, the analysis result management unit 216 can present a user interface including two or more of the crack width candidates.

[0089] 12 shows an example of a user interface when two crack width candidates are obtained. The analysis result management unit 216 can present a user interface that includes two or more crack width candidates. For example, the user interface can include multiple components, such as buttons 1201 and 1202, each of which corresponds to one of the two or more crack width candidates. Such multiple buttons 1201 and 1202 can be displayed in the attribute display field 450.

[0090] By using such buttons, the user can easily input the numerical value of the width corresponding to the button. That is, the input acquisition unit 221 can acquire a user input for selecting one of multiple components as a user input indicating the crack width of the crack according to the user input. Specifically, when the user presses button 1201 or 1202, the analysis result editing unit 218 reflects the value corresponding to the pressed button in the crack width display 451. This configuration can reduce the effort required for the user to input a numerical value when changing the value of the crack width display 451.

[0091] <Modification> So far, we have described cases in which crack width candidates are determined using one method, such as a method that references a crack scale or a method that references intermediate data. However, crack width candidates may also be determined using multiple methods. For example, the width determination unit 222 can determine crack width candidates by using at least some of the multiple methods in order of priority. The width determination unit 222 may use each method in order of priority. In this case, the width determination unit 222 can determine multiple crack width candidates for an added crack using each of the multiple methods. Furthermore, the width determination unit 222 may determine one crack width candidate for an added crack using a method with a higher priority. In this case, if the width determination unit 222 is unable to determine a crack width candidate, it may determine one crack width candidate using a method with a lower priority. Such priorities may be changeable. Furthermore, it may be possible to switch between enabling and disabling each method.

[0092] Furthermore, the method for determining a crack width candidate is not limited to the method of referencing a crack scale or the method of referencing intermediate data. Other methods can be used to determine the crack width. For example, the width determination unit 222 can determine a crack width candidate according to a user input based on width information prepared in advance that matches the position and shape of the crack according to the user input. The width information that matches the position and shape of the crack according to the user input is a single piece of width information estimated based on the position and shape of the crack. When referring to a crack scale, the width information that matches the position and shape of the crack according to the user input is width information of a model corresponding to the crack according to the user input, which is prepared in advance. Such a model is determined as described above based on the position and direction of the crack and the position and direction of the model. When referring to intermediate data, the width information that matches the position and shape of the crack according to the user input is information of a crack width candidate corresponding to the crack according to the user input, which is prepared in advance. Such a crack width candidate is determined as described above based on the position and shape of the crack and the position and shape of the model.

[0093] In this case, the user interface presented by the analysis result management unit 216 may include information indicating the method used to determine the crack width candidate. FIG. 13 shows an example of a user interface when multiple methods are used in combination to determine the crack width candidate. FIG. 13 shows input buttons corresponding to multiple crack width candidates, each determined using one of the multiple methods. In addition, determination method displays 1301, 1302, and 1303 are shown for each crack width candidate. The determination method displays 1301, 1302, and 1303 indicate the method used to determine the crack width candidate. In the example of FIG. 13, each method is represented by an icon.

[0094] The user interface presented by the analysis result editing unit 218 may also include information indicating the reliability of the crack width candidate. For example, FIG. 13 shows a warning indicator 1304 indicating that the reliability of the crack width candidate is low. For example, a crack width candidate determined based on the detection result of a crack with low reliability contained in intermediate data requires caution when used. Therefore, to alert the user to the crack width, the warning indicator 1304 can be attached to such a crack width candidate.

[0095] In the above-described embodiment, when the width determination unit 222 determines that there is no model or crack candidate corresponding to the crack specified by the user, a predetermined default value is used as the crack width candidate. However, in such a case, the analysis result management unit 216 may present a user interface including multiple default values. Specifically, the analysis result management unit 216 may include multiple buttons for inputting the crack width in the user interface.

[0096] Crack width candidates that can be input using such buttons may be set based on legend settings. That is, multiple default values ​​may be determined according to the crack width classification for display in the user interface. FIG. 14 shows such an example. In this example, default values ​​are automatically input as crack widths. Furthermore, crack width candidates corresponding to each of the multiple crack width input buttons 1401 are set based on the legend display 441 displayed in the legend display field 440. For example, boundary values ​​of crack width classifications in the legend display can be used as crack width candidates. In FIG. 14, buttons corresponding to each such crack width candidate are displayed in the display format shown in the legend display 441. This configuration reduces the user's effort in directly inputting the crack width even when a crack width candidate cannot be determined based on the crack scale or intermediate data, etc.

[0097] In the above-described embodiment, cracks not detected by the image analysis unit 215 are added according to user input. However, it is not necessary to obtain crack detection results by the image analysis unit 215 in advance. According to the above-described embodiment, a crack width candidate can be determined for any crack added by the user. The crack width of a crack may also be corrected using the above-described method. In this case, the user can perform user input to input the shape of a crack shown in an image by specifying an already detected crack on the user interface.

[0098] (Other Examples) The present disclosure can also be realized by supplying a program that realizes one or more functions of the above-described embodiments to a system or device via a network or a storage medium, and having one or more processors in the computer of the system or device read and execute the program. It can also be realized by a circuit (e.g., ASIC) that realizes one or more functions.

[0099] The disclosure of this specification includes the following image processing device, image processing method, and program. (Item 1) A presentation means for presenting a user interface including an image of an inspection object having crack deformation and model shapes of a plurality of cracks each having a different width that can be superimposed on the image and can be moved to positions corresponding to the cracks shown in the image according to user input; an input acquisition means for acquiring a user input for inputting a shape of a crack shown in the image included in the user interface; A determination means for determining a crack width candidate for the crack according to the user input based on the shape of the crack according to the user input and the width of the model shape of the crack superimposed and displayed at a position corresponding to the shape of the crack according to the user input; An image processing device comprising: (Item 2) The image processing device described in item 1, characterized in that the determination means determines the crack width candidate based on the width of a model shape of the crack that is superimposed and displayed at a position corresponding to the crack according to the user input and extends in a direction corresponding to the crack according to the user input. (Item 3) 3. The image processing device according to any one of items 1 to 2, wherein the user interface includes a crack scale including model shapes of the plurality of cracks. (Item 4) 4. The image processing device according to item 3, wherein the display position and display angle of the crack scale can be changed according to user input. (Item 5) 5. The image processing device according to any one of items 1 to 4, further comprising an analysis means for detecting the cracks in the image. (Item 6) the presentation means presents the cracks detected by the analysis means on the user interface; 6. The image processing device according to item 5, wherein the user input is input of a shape of a crack deformation that has not been presented by the presentation means. (Item 7) An analysis means for detecting the position of a crack candidate based on an image of an inspection object having a crack deformation and estimating the width of the crack candidate; a presentation means for presenting a user interface including an image of an inspection object having a crack deformation; an input acquisition means for acquiring a user input for inputting a shape of a crack shown in the image included in the user interface; A determination means for determining a crack width candidate for the crack according to the user input based on a shape of the crack according to the user input and a width estimated for the crack candidate detected at a position corresponding to the shape of the crack according to the user input; An image processing device comprising: (Item 8) the presenting means presents, on the user interface, crack candidates selected from the crack candidates detected by the analyzing means as crack detection results; 8. The image processing device according to item 7, wherein the user input is input of a crack shape that has not been presented by the presentation means. (Item 9) The image processing device described in any one of items 7 to 8, characterized in that the determination means determines the crack width candidate based on the width of a crack candidate that is detected at a position corresponding to the crack according to the user input and has a shape corresponding to the crack according to the user input. (Item 10) The determining means determines a plurality of the crack width candidates, 10. The image processing device according to any one of items 1 to 9, wherein the presentation means presents the user interface including the plurality of crack width candidates. (Item 11) the user interface includes a plurality of parts, each of which corresponds to one of the plurality of crack width candidates; Item 11. The image processing device according to item 10, wherein the input acquisition means acquires a user input to select one of the plurality of parts as a user input indicating the width of the crack according to the user input. (Item 12) 12. The image processing device according to any one of items 1 to 11, wherein the user interface includes information indicating the reliability of the crack width candidate. (Item 13) 13. The image processing device according to any one of items 1 to 12, wherein the user interface includes information indicating a method used to determine the crack width candidate. (Item 14) 14. The image processing device according to any one of items 1 to 13, wherein the determination means determines the crack width candidate by using at least some of a plurality of methods in order of priority. (Item 15) 15. The image processing device according to any one of items 1 to 14, characterized in that, in response to the determination means determining that there is no model shape of the crack superimposed and displayed at a position corresponding to the crack according to the user input, the presentation means presents the user interface including a plurality of default values. (Item 16) Item 16. The image processing device according to item 15, wherein the plurality of default values ​​are determined according to classification of crack widths for display in the user interface. (Item 17) a presentation means for presenting a user interface including an image of an inspection object having a crack deformation; an input acquisition means for acquiring a user input for inputting the shape of the crack shown in the image included in the user interface; A determination means for determining a candidate width of the crack according to the user input based on one piece of width information, which is estimated based on the position and shape of the crack according to the user input, from among width information prepared in advance; An image processing device comprising: (Item 18) presenting a user interface including an image of an inspection object having crack deformation and model shapes of a plurality of cracks each having a different width that can be superimposed on the image and can be moved to positions corresponding to the cracks shown in the image according to user input; obtaining a user input for inputting a shape of a crack shown in the image included in the user interface; determining a crack width candidate for the crack according to the user input based on the shape of the crack according to the user input and the width of the model shape of the crack superimposed and displayed at a position corresponding to the shape of the crack according to the user input; 10. An image processing method for an image processing apparatus, comprising: (Item 19) Detecting the position of a crack candidate based on an image of an inspection object having a crack deformation and estimating the width of the crack candidate; presenting a user interface including an image of an inspection object having a crack deformation; obtaining a user input for inputting a shape of a crack shown in the image included in the user interface; determining a crack width candidate for the crack according to the user input based on a shape of the crack according to the user input and a width estimated for the crack candidate detected at a position corresponding to the shape of the crack according to the user input; 10. An image processing method for an image processing apparatus, comprising: (Item 20) 18. A program for causing a computer to function as the image processing device according to any one of items 1 to 17.

[0100] The present disclosure is not limited to the above-described embodiments, and various modifications and variations are possible without departing from the spirit and scope of the present disclosure. Accordingly, the following claims are appended to apprise the public of the scope of the present disclosure. [Explanation of symbols]

[0101] 211: Folder management unit, 212: Folder setting storage unit, 213: Image management unit, 214: Image storage unit, 215: Image analysis unit, 216: Analysis result management unit, 217: Analysis result storage unit, 218: Analysis result editing unit, 219: Editing result storage unit, 221: Input acquisition unit, 222: Width determination unit

Claims

1. A presentation means for presenting a user interface including an image of an inspection object having crack deformation and model shapes of a plurality of cracks each having a different width that can be superimposed on the image and can be moved to positions corresponding to the cracks shown in the image according to user input; an input acquisition means for acquiring a user input for inputting a shape of a crack shown in the image included in the user interface; A determination means for determining a crack width candidate for the crack according to the user input based on the shape of the crack according to the user input and the width of the model shape of the crack superimposed and displayed at a position corresponding to the shape of the crack according to the user input; An image processing device comprising:

2. The image processing device described in claim 1, characterized in that the determination means determines the crack width candidate based on the width of a model shape of the crack that is superimposed and displayed at a position corresponding to the crack according to the user input and extends in a direction corresponding to the crack according to the user input.

3. The image processing apparatus according to claim 1 , wherein the user interface includes a crack scale including model shapes of the plurality of cracks.

4. 4. The image processing device according to claim 3, wherein the display position and display angle of the crack scale can be changed in accordance with a user input.

5. The image processing device according to claim 1 , further comprising an analysis means for detecting the cracks in the image.

6. the presentation means presents the cracks detected by the analysis means on the user interface; The image processing device according to claim 5 , wherein the user input is an input of a crack shape that has not been presented by the presentation means.

7. An analysis means for detecting the position of a crack candidate based on an image of an inspection object having a crack deformation and estimating the width of the crack candidate; a presentation means for presenting a user interface including an image of an inspection object having a crack deformation; an input acquisition means for acquiring a user input for inputting a shape of a crack shown in the image included in the user interface; A determination means for determining a crack width candidate for the crack according to the user input based on a shape of the crack according to the user input and a width estimated for the crack candidate detected at a position corresponding to the shape of the crack according to the user input; An image processing device comprising:

8. the presentation means presents, on the user interface, crack candidates selected from the crack candidates detected by the analysis means as crack detection results; The image processing device according to claim 7 , wherein the user input is an input of a crack shape that has not been presented by the presentation means.

9. The image processing device described in claim 7, characterized in that the determination means determines the crack width candidate based on the width of a crack candidate that is detected at a position corresponding to the crack according to the user input and has a shape corresponding to the crack according to the user input.

10. The determining means determines a plurality of the crack width candidates, The image processing device according to claim 1 , wherein the presenting means presents the user interface including the plurality of crack width candidates.

11. the user interface includes a plurality of parts, each of which corresponds to one of the plurality of crack width candidates; The image processing device according to claim 10 , wherein the input acquisition means acquires a user input for selecting one of the plurality of components as the user input indicating the width of the crack according to the user input.

12. The image processing device according to claim 10 , wherein the user interface includes information indicating the reliability of the crack width candidate.

13. The image processing apparatus according to claim 10 , wherein the user interface includes information indicating a method used to determine the crack width candidate.

14. The image processing device according to claim 13, wherein the determining means determines the crack width candidate by using at least some of a plurality of methods in order of priority.

15. The image processing device described in claim 10, characterized in that, in response to the determination means determining that there is no model shape of the crack superimposed and displayed at a position corresponding to the crack according to the user input, the presentation means presents the user interface including a plurality of default values.

16. The image processing device according to claim 15, wherein the plurality of default values ​​are determined according to classification of crack widths for display on the user interface.

17. a presentation means for presenting a user interface including an image of an inspection object having a crack deformation; an input acquisition means for acquiring a user input for inputting the shape of the crack shown in the image included in the user interface; A determination means for determining a candidate width of the crack according to the user input based on one piece of width information, which is estimated based on the position and shape of the crack according to the user input, from among width information prepared in advance; An image processing device comprising:

18. presenting a user interface including an image of an inspection object having crack deformation and model shapes of a plurality of cracks each having a different width that can be superimposed on the image and can be moved to positions corresponding to the cracks shown in the image according to user input; obtaining a user input for inputting a shape of a crack shown in the image included in the user interface; determining a crack width candidate for the crack according to the user input based on the shape of the crack according to the user input and the width of the model shape of the crack superimposed and displayed at a position corresponding to the shape of the crack according to the user input; 10. An image processing method for an image processing apparatus, comprising:

19. Detecting the position of a crack candidate based on an image of an inspection object having a crack deformation and estimating the width of the crack candidate; presenting a user interface including an image of an inspection object having a crack deformation; obtaining a user input for inputting a shape of a crack shown in the image included in the user interface; determining a crack width candidate for the crack according to the user input based on a shape of the crack according to the user input and a width estimated for the crack candidate detected at a position corresponding to the shape of the crack according to the user input; 10. An image processing method for an image processing apparatus, comprising:

20. A program for causing a computer to function as the image processing device according to any one of claims 1 to 17.

Images