Image processing apparatus, image processing method, and storage medium

Abstract

To provide an image processing device that acquires divided areas without dividing a defect of an image analysis result as much as possible to improve efficiency of inspection work of the defect.SOLUTION: An image processing apparatus includes an image analysis unit 103 that executes image analysis on an image to be inspected and acquires a defect of an image analysis result, and a division unit 104 that divides a region of the image analysis to acquire divided regions, and the division unit 104 divides the region of the image analysis on the basis of a state of the defect.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] In image-based inspections of infrastructure structures such as bridges and tunnels, a method is employed in which multiple high-resolution images are combined to generate a single composite image in order to detect defects such as cracks and exposed rebar on the structure's wall surface. In an image obtained by combining multiple high-resolution images, there is a possibility that areas containing many progressing defects may be captured. In this case, when inspectors check the progressing areas, the number of areas to be checked becomes enormous, which may make it difficult to perform the inspection efficiently. Therefore, a method has been proposed in which the inspection area is divided and displayed to the inspector to improve the efficiency of the inspection work. For example, in Patent Document 1, the inspection area is displayed in sections separated by a predetermined distance, making it easy to check and edit the defects. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6937355 Summary of the Invention [Problem to be solved by the invention]

[0004] However, when dividing an inspection area using the technology shown in Patent Document 1, the area may be divided in such a way as to separate deformations that are longer than the division size or areas where deformations are concentrated. When the area is divided in such a way that deformations are separated, the worker must check multiple divided areas to confirm the overall picture of the deformation. Furthermore, when multiple workers are working on divided areas, each worker may check and edit the separated deformations, which may lead to inefficiencies in the work, or long deformations may be mistaken for short deformations, resulting in an underestimation of deterioration.

[0005] The present disclosure has been made in consideration of the above-mentioned problems, and aims to provide an image processing device that prevents misidentification of abnormalities in the image analysis results of the inspection object and makes inspection work for abnormalities more efficient. [Means for solving the problem]

[0006] The image processing device of the present disclosure includes an image analysis unit that performs image analysis on an image of an inspection target and acquires a state of deformation as a result of the image analysis, and a division unit that divides the image analysis area to acquire divided areas. The division unit divides the image analysis area based on the state of the deformation. [Effects of the Invention]

[0007] According to the present disclosure, an image processing device is realized that prevents misidentification of abnormalities in the image analysis results of an inspection target, thereby improving the efficiency of inspection work for abnormalities. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a functional configuration diagram showing a schematic configuration of an image processing device according to a first embodiment. [Figure 2] 1 is a block diagram illustrating an example of the hardware configuration of an image processing device according to a first embodiment. [Figure 3] 3 is a schematic diagram showing an example of the relationship between a drawing and an inspection object image in the first embodiment. FIG. [Figure 4] FIG. 2 is a data configuration diagram showing an example of a deformation data table showing the data structure of deformation data in the first embodiment. [Figure 5] 3 is a flowchart showing an image processing method according to the first embodiment. [Figure 6] 10 is a flowchart showing an example of area division processing in the first embodiment. [Figure 7] 7A to 7C are schematic diagrams showing deformation data in each step of the area division processing in FIG. 6. [Figure 8] 10 is a schematic diagram showing an example of a UI screen relating to division of an inspection area in the first embodiment. FIG. [Figure 9] FIG. 2 is a schematic diagram illustrating an example of a UI screen related to display of divided areas in the first embodiment. [Figure 10] 10 is a flowchart showing an example of area division processing in the second embodiment. [Figure 11] 10 is a flowchart showing an example of area division processing in the second embodiment. [Figure 12] FIG. 11 is a schematic diagram showing an example of a UI screen relating to display of divided areas in the third embodiment. [Figure 13] 13 is a flowchart showing an example of area division processing in the fourth embodiment. [Figure 14] FIG. 13 is a schematic diagram showing an example of a UI screen relating to display of divided areas in the fourth embodiment. [Figure 15] FIG. 13 is a diagram showing an example of a deformation importance evaluation table in the fifth embodiment. [Figure 16] 13A to 13C are schematic diagrams illustrating an example of a circumscribing rectangle calculation process according to the fifth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] -Basic configuration of image processing device in various embodiments- Before specifically disclosing the embodiments, the basic configuration of the image processing device in each embodiment will be described.

[0010] The image processing device disclosed herein includes an image analysis unit that performs image analysis on an image of an object to be inspected and acquires the deformation results of the image analysis, and a division unit that divides the image analysis area to acquire divided areas. The division unit divides the image analysis area based on the deformation status of the image analysis results. The deformation status includes the type of deformation (crack, exposed rebar, etc.), shape (linear, arc-shaped, etc.), size, distribution of the deformation, degree of density, etc. The image analysis area is divided into individual divided areas by the division unit according to a predetermined division standard (e.g., the number of divisions or the division size) in response to the deformation status, so as to minimize the separation of deformations between adjacent divided areas, and the image analysis area is then used for deformation inspection. If the image analysis area is simply divided without considering the deformation status, many deformations may be separated depending on the deformation status, hindering the inspection work for the deformation. In the present disclosure, by taking into consideration the state of deformation in the image analysis area and appropriately dividing the image analysis area based on that state, it is possible to prevent misidentification of deformation, improve the efficiency of inspection work for deformation, and realize rapid and accurate inspection of deformation.

[0011] Specifically, the dividing means acquires candidate dividing line areas that are candidates for dividing the image analysis area, sets dividing lines in all or part of the candidate dividing line areas, and divides the image analysis area along the dividing lines. Considering the need to avoid dividing the deformation as much as possible, it is preferable for the dividing means to acquire a deformation surrounding area that surrounds the deformation in the image analysis results, and acquire a candidate dividing line area that does not overlap with the deformation surrounding area. This eliminates the need to check multiple divided areas associated with dividing the deformation, thereby improving the efficiency of the deformation inspection work. This configuration will be described in detail in embodiment 1.

[0012] Depending on the condition of the deformation in the image analysis results, it may be impossible to avoid dividing the deformation when dividing the image analysis area. Taking such a situation into consideration, the dividing means acquires a dividing line candidate area corresponding to the condition of the deformation so as to minimize the number of overlaps with the deformation surrounding area surrounding the deformation. By dividing the area in this manner, the division of the deformation is minimized, making it easier to confirm and edit the deformation, and improving the efficiency of the deformation inspection work. This configuration will be described in detail in the second embodiment. When setting a dividing line in a dividing line candidate area to obtain the desired divided area, a priority may be assigned to the selection of the dividing line candidate area in which the dividing line is set. For example, the dividing means sets a low priority to a dividing line candidate area that includes overlapping portions of deformation surrounding areas where the deformation surrounding areas are densely packed. Setting a dividing line in a dividing line candidate area that includes overlapping portions of deformation surrounding areas makes it more likely that two or more deformations will be divided between the divided areas divided by the dividing line. Therefore, from the viewpoint of minimizing the fragmentation of deformation as much as possible, it is desirable to give a low priority to the setting of such dividing lines so as to avoid the setting of such dividing lines as much as possible.

[0013] If dividing the image analysis area into sections unavoidably results in the division of a deformation, the division means divides the image so that adjacent divided areas overlap. In this case, the division means assigns the deformation included in the overlapping area to one of the divided areas. Specifically, the overlapping area in one divided area includes the entire first deformation and a portion of the second deformation, while the overlapping area in the other divided area includes the entire second deformation and a portion of the first deformation. Therefore, the first deformation can be inspected in one divided area, and the second deformation can be inspected in the other divided area. This eliminates the need to check multiple divided areas due to the division of the deformation, thereby improving the efficiency of the deformation inspection work. This configuration will be described in detail in embodiment 3.

[0014] When examining the image analysis results, it is possible that a portion of a large deformation compared to surrounding deformations is present within the image analysis area. Taking such a situation into consideration, the division means divides the image analysis area while ignoring large deformations, and assigns large deformations to multiple divided areas. Specifically, a deformation surrounding area that surrounds the image analysis results, excluding large deformations, is obtained, and a division line candidate area that does not overlap with the deformation surrounding area is obtained. Treating large deformations as an exception in this way makes it easier to confirm and edit the deformation, and improves the efficiency of the deformation inspection work. This configuration will be described in detail in embodiment 4.

[0015] Furthermore, in the present disclosure, the dividing means assigns importance to the deformations in the image analysis results according to their circumstances, and divides the image analysis area based on the importance. By dividing the image using the importance assigned to the deformation as an index, such as dividing the image analysis area so that deformations with high importance are not separated, the efficiency of the deformation inspection work can be improved. This configuration will be described in detail in embodiment 5.

[0016] -Specific Description of Various Embodiments- Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the drawings. Note that the following embodiments do not limit the invention according to the claims. Although multiple features are described in the various embodiments, not all of these multiple features are necessarily essential, and multiple features may be combined arbitrarily. Furthermore, in the drawings, the same reference numerals are used to designate the same or similar components, and redundant explanations will be omitted.

[0017] <Embodiment 1> Hereinafter, a first embodiment of the present disclosure will be described.

[0018] [Configuration of image processing device] An example of the configuration of an image processing apparatus according to this embodiment will be described with reference to the functional configuration diagram of FIG. The image processing device 101 includes a structure image management unit 102, an image analysis unit 103, an area division unit 104, a divided area display unit 105, a divided area editing unit 106, a structure image storage unit 107, and an analysis result storage unit 108.

[0019] The structure image management unit 102 has functions for saving, deleting, listing, and viewing structure images of the inspection target. The image analysis unit 103 performs image analysis using a learning model created by, for example, machine learning or deep learning of AI (artificial intelligence) to detect abnormalities in the images of the inspection target. The area division unit 104 divides the image-analyzed abnormality data into areas based on the analysis results by the image analysis unit 103. The divided area display unit 105 presents the user with an area division screen (described later in FIG. 8) and a work screen (described later in FIG. 9) for checking and editing the divided areas. The divided area editing unit 106 has a function for the user to edit abnormalities in the divided areas on the work screen (described later in FIG. 9). The structure image storage unit 107 stores structure images of the inspection target. The analysis result storage unit 108 stores the results of image analysis and editing.

[0020] Next, an example of the hardware configuration of the image processing device 101 according to this embodiment will be described with reference to the block diagram of FIG. The image processing device 101 includes at least a CPU 201 , a RAM 202 , a ROM 203 , a network interface 204 , an external storage device 205 , a display device 206 , and an input device 207 .

[0021] The CPU 201 controls the operation of each component of the image processing device 101 and is responsible for executing various processes (described later) performed by the image processing device 101. The RAM 202 is a memory for temporarily storing data and control information and serves as a work area used by the CPU 201 when executing various processes. The ROM 203 stores fixed operating parameters and operating programs of the image processing device 101. The network interface 204 provides a function for transmitting and receiving data to and from external devices. The external storage device 205 is a device for storing data and has an interface for receiving I / O commands for reading and writing data. The external storage device 205 may be a hard disk drive (HDD), a solid-state drive (SSD), an optical disk drive, a semiconductor storage device, or other storage device. The external storage device 205 stores computer programs and data for causing the CPU 201 to execute various processes (described later) performed by the image processing device 101. The display device 206 is, for example, an LCD (Liquid Crystal Display) and displays information required by the user. The input device 207 is, for example, a keyboard, a mouse, a touch panel, etc., and receives necessary input from the user.

[0022] In explaining this embodiment, the relationship between the image and the deformation data, and the structural information relating to the structure of the inspection target will be explained using the schematic diagram of FIG. In image inspection, it is preferable to manage images of structure walls in association with drawings. FIG. 3(a) shows an example of an infrastructure structure, in which an image 311 of a bridge wall is pasted onto a drawing 300. The drawing 300 has drawing coordinates 301 with point 302 as the origin. The position of the image on the drawing is defined by the coordinates of the vertex at the upper left of the image. For example, the coordinates of image 311 are the position of vertex 312 (X312, Y312). The image, along with its coordinate information, is stored in the RAM 202 or the external storage device 205, which are storage units. In this embodiment, images used in image inspection of infrastructure structures are large in size because they are captured at high resolution (e.g., 1 mm per pixel) so that minute cracks and the like can be identified. For example, image 311 in FIG. 3(a) is an image of a 20 m x 10 m bridge deck. If the image resolution is 1.0 mm per pixel (1.0 mm / pixel), the image size of the image 311 is 20,000 pixels by 10,000 pixels.

[0023] The high-resolution image 311 contains numerous (e.g., 1,000 or more) cracks, exposed rebar, and other defects. However, because it is difficult to display all of the defects on paper, only a portion of the defects are displayed on the paper. In the following explanation, even in diagrams displaying wide-area images and defect data, only a portion of the defects are displayed. Defect data is information resulting from the automatic detection of defects, such as cracks on concrete wall surfaces, through image analysis processing. Defect data also includes information on defects obtained through image analysis processing that has been manually supplemented by a human, as appropriate, or where erroneous detections have been corrected. In explaining this embodiment, it is assumed that the input image and defect data are managed in association with drawing coordinates.

[0024] Figure 3(b) shows the state in which deformation data 321 corresponding to image 311 has been pasted onto drawing 300 in the same position as image 311. The deformation data 321 contains a large number of deformations (e.g., over 1,000), including deformations that are not displayed on the paper. The position on the drawing of each deformation data in the deformation data 321 is defined by the coordinates of the pixels that make up the deformation data.

[0025] FIG. 4 shows an example of a deformation data table 401 representing the data structure of deformation data. The deformation data table 401 is composed of a deformation ID, deformation type, coordinates, line width, maximum width, and line length. The deformation ID indicates the ID of the detected deformation, and the deformation type indicates the type, such as crack or exposed rebar. The coordinates represent multiple coordinate information constituting the deformation data. The line width is an attribute value representing the width of the deformation at that coordinate when the deformation type is a crack. The maximum width represents the maximum numerical value of the line width when the deformation type is a crack, and the line length represents the total length of the crack. The area represents the area when the deformation type indicates an area, such as exposed rebar. For example, crack Ca001 is represented by n consecutive pixels from (Xca001_1, Yca001_1) to (Xca001_n, Yca001_n). In this manner, in this embodiment, the deformation data is represented by pixels.

[0026] Deformation data may be expressed as vector data such as polylines or curves consisting of multiple points. When deformation data is expressed as vector data, the data volume is reduced and the expression is simpler. An example of deformation data for a deformation other than a crack is when the deformation type is exposed rebar and the deformation ID is Ta001. When a deformation such as exposed rebar is expressed using coordinate information, it becomes a deformation with an area surrounded by a polyline. Note that the information contained in the deformation data is not limited to the information shown in the deformation data table 401, and other attribute information may also be held.

[0027] Other attribute information may include structural information related to the structure of the inspection target, such as the structure type, basic structure, various dimensions of the structure, component information, and year of completion. Furthermore, repair history information may include maintenance-related information such as the repair year, repair locations, and repair method. In this embodiment, structural information related to specific positions of the structure, such as component information and repair information, is stored along with position information on the drawings. That is, the positions of each component and repair locations on the drawings are stored as part of the structural information. Therefore, the correspondence between the structural information and images and deformation data can be determined via the drawings. The structural information, together with images and deformation data, can be stored and retrieved in the RAM 202 or the external storage device 205, which are storage units. Note that the information included in the structural information is not limited to the above information and may include other information. Furthermore, information specific to each type of structure may be stored.

[0028] [Image processing method] The image processing method according to this embodiment will be described below with reference to a flowchart shown in Fig. 5.

[0029] First, in S501, the structure image management unit 102 acquires the structure image of the inspection target stored in the structure image storage unit 107. Next, in S502, the image analysis unit 103 performs an analysis process on the acquired structure image and calculates deformation data corresponding to the structure image. The calculated deformation data is stored in the deformation data table 401.

[0030] Next, in S503, the region dividing unit 104 divides the inspection region (region for image analysis) based on the calculated deformation data and the state of the deformation. The state of the deformation includes the type of deformation (crack, exposed rebar, etc.), shape (linear, arc, etc.), size, distribution of the deformation, degree of density, etc. In this embodiment, the entire structure image is the inspection region (region for image analysis). Details of the region dividing process will be described later using the flowchart in FIG. 6.

[0031] Next, in S504, the divided area display unit 105 presents to the user the divided areas into which the inspection area (area for image analysis) is divided and the structure images corresponding to the divided areas. Next, in S505, the divided area editing unit 106 presents a screen for checking and editing the divided areas to the user, and the user checks and edits the state of the deformation in the divided areas from the screen. Next, in S506, the divided area editing unit 106 saves the results of the user's anomaly confirmation and editing of the divided areas, and at this time, reflects the changes to the divided areas in the original anomaly detection results.

[0032] Fig. 6 is a flowchart detailing the area division process S503 in this embodiment. Fig. 7 is a schematic diagram showing deformation data in each step of the area division process S503 in Fig. 6.

[0033] 7(a) shows an example of deformation data 701 obtained by image analysis of an image of a structure to be inspected. First, in S601, the area dividing unit 104 calculates a deformation surrounding area surrounding each deformation in the deformation data 701, in this case, a rectangular circumscribing rectangle surrounding the deformation.

[0034] Next, in S602, the area division unit 104 maps the calculated circumscribing rectangle to the deformation data. Figure 7(b) shows an example of deformation data 702 in which a circumscribing rectangle 711 is mapped to each deformation. S601 and S602 are executed in a loop the number of times equal to the number of deformations. In this way, by creating a circumscribing rectangle 711 corresponding to each deformation, the state of the deformation (position, etc.) when dividing the area can be numerically grasped, and the desired dividing line candidate area can be obtained.

[0035] Next, in S603, the region dividing unit 104 calculates dividing line candidate regions for calculating dividing lines. The dividing line candidate regions are rectangular regions that serve as dividing point candidates and extend vertically and / or horizontally so as not to overlap with the circumscribing rectangles 711 of each deformation. The dividing line candidate regions contain one or more circumscribing rectangles 711. Figure 7(c) shows an example of deformation data 703 in which dividing line candidate regions 712 are arranged. In this way, by first setting the dividing line candidate regions 712 so that they do not overlap with each circumscribing rectangle 711, it is possible to reliably divide the region in accordance with the user's requirements, provided that the requirement that no divided deformations exist within each divided region is met.

[0036] Next, in S604, the region dividing unit 104 sets division lines in all or part of the dividing line candidate regions to create divided regions. FIG. 7(d) shows an example of the deformation data 704 in which divided regions 714 have been created. Here, a division line 713, shown by a dashed line in the figure, is set in part of the dividing line candidate region 712, and multiple (six in the illustrated example) divided regions 714 are created. The number of divisions of the deformation data 704 (the number of divided regions 714 divided by the division lines 713) may be determined automatically by the region dividing unit 104 or may be specified by the user. In this way, by appropriately selecting and setting division lines 713 for the already created dividing line candidate regions 712 and creating divided regions 714, the desired region division that meets the division conditions (number of divisions, division size, etc.) can be achieved.

[0037] For example, if six users are assigned to check for abnormalities, it is desirable to set the number of divisions to six or a multiple of six. The number of divisions may also be determined based on the size of the division area (division size). For example, one method of division is to divide the data so that when the division area is superimposed on the corresponding structure image, the division area is 1000 pixels x 1000 pixels or less. Note that if the vertical and horizontal widths of the division area are too small, the efficiency of the inspection work will decrease. To prevent this, it is desirable to divide the abnormality data into areas by setting division lines so that the vertical and horizontal widths of the division area are equal to or greater than a certain value.

[0038] FIG. 8 is a schematic diagram showing an example of a UI screen relating to division of an inspection area (area for image analysis). The divided area display unit 105 presents the user with an area division screen 801. The area division screen 801 includes an area image 802, an area division item 803, and an area division button 804. The area image 802 displays the area of ​​the deformation data and the division lines (shown by dashed lines) that divide the area. This allows the user to easily confirm how the area is divided from the area image 802. The user can also move the division lines to adjust the divided area as needed. The area image 802 may be displayed superimposed on a structure image corresponding to the area. The area division item 803 is an example of a parameter specification for dividing the area. The area image 802 displays the division lines for area division calculated by the area division unit 104 according to the parameters specified in the area division item 803. The area division button 804 is a button that executes the area division. When the user presses the area division button 804, the division lines are determined and the area is divided.

[0039] FIG. 9 is a schematic diagram showing an example of a UI screen related to display of the divided areas. The divided area display unit 105 presents the user with a work screen 901 for checking and editing divided areas. The work screen 901 includes a divided area image 902, an entire area image 903, a divided area list 904, a deformation edit button 905, and a work completion button 906. The divided area image 902 displays an enlarged image corresponding to the divided area. The divided area image 902 may be displayed superimposed on a structure image corresponding to the area of ​​deformation data. The entire area image 903 displays the entire area of ​​deformation data to be inspected, highlighting an image 903a corresponding to the divided area enlarged and displayed in the divided area image 902 within the entire area of ​​deformation data. Note that the display may be changed by highlighting, for example, graying out, the position corresponding to the divided area set as work completed depending on the work status of the divided area. This allows the user to easily select divided areas for which work is incomplete.

[0040] The divided area list 904 shows a list of divided areas. The deformation edit button 905 provides a function for adding, deleting, and correcting deformations in divided areas. When the user presses the work completion button 906, a work completion flag is set for the target divided area, and if an edit such as adding a deformation has been made to the divided area, the edit results are reflected in the deformation data. In this way, the image processing device of this embodiment has an added function for the user to edit deformation data as appropriate, resulting in a device configuration that is easy to use and flexibly responds to the actual circumstances of deformation inspection work.

[0041] As described above, according to this embodiment, the positions of the dividing lines that divide the areas of the deformation data are calculated based on the deformation data, and the areas are divided so that the deformation is not divided. By being able to perform the deformation confirmation and editing work for each divided area so that the deformation is not divided, misidentification of the deformation is suppressed, and the efficiency of the user's inspection work is greatly improved.

[0042] <Embodiment 2> Next, a second embodiment of the present disclosure will be described. In the first embodiment, a configuration was shown in which a dividing line was calculated to divide the area so that the deformation was not divided. However, there are cases where it is difficult to divide the area in a way that does not divide the deformation, such as when there are many deformations in the inspection area (image analysis area). In the second embodiment, an example of area division processing in which the division of the deformation is allowed to a certain extent is shown. Note that the functional configuration and hardware configuration of the image processing device are the same as those of the first embodiment, and therefore description thereof will be omitted.

[0043] In the image processing method according to this embodiment, steps S501 to S506 are executed in sequence, similarly to Fig. 5 shown in embodiment 1. In this embodiment, the area division process S503 is executed similarly to Fig. 6 shown in embodiment 1. Fig. 10 is a schematic diagram showing deformation data in each step of the area division process S503 of this embodiment.

[0044] 10(a) shows an example of deformation data 1001 obtained by image analysis of an image of a structure to be inspected. In this embodiment, the entire structure image is the inspection area (area for image analysis). First, in S601, the area division unit 104 calculates a deformation surrounding area surrounding each deformation in the deformation data 1001, in this case a rectangular circumscribing rectangle surrounding the deformation.

[0045] Next, in S602, the area dividing unit 104 maps the calculated circumscribing rectangle to the deformation data. Figure 10(b) shows an example of deformation data 1002 in which a circumscribing rectangle 1011 is mapped to each deformation. By creating a circumscribing rectangle 1011 corresponding to each deformation in this way, the state of the deformation (position, etc.) when dividing the area can be numerically grasped, and the desired dividing line candidate area can be obtained.

[0046] Next, in S603, the region dividing unit 104 calculates dividing line candidate regions for calculating dividing lines. FIG. 10(c) shows an example of the deformation data 1003 in which dividing line candidate regions 1012 are arranged. In the first embodiment, dividing line candidate regions that do not overlap with the circumscribing rectangle of each deformation are calculated. However, it is not possible to create dividing line candidate regions that do not overlap with the circumscribing rectangle 1011 of the deformation using the deformation data 1002 to which the circumscribing rectangle 1011 is mapped. Therefore, in this embodiment, overlap with the circumscribing rectangle 1011 is minimized to reduce the number of overlaps with the circumscribing rectangle 1011. For example, as shown in the figure, a dividing line candidate region 1012 with an overlapping number of 1 with the circumscribing rectangle 1011 is created. Creating the dividing line candidate region 1012 so that the number of overlaps of the dividing line candidate region 1012 with the circumscribing rectangle 1011 is reduced contributes to minimizing the complexity of the deformation inspection work.

[0047] Note that when it is possible to create a dividing line candidate area that does not overlap with the circumscribing rectangle 1011 of the deformation, even if the creation position is not appropriate for setting the dividing line, it is possible to create a dividing line candidate area whose overlapping number with the circumscribing rectangle 1011 of the deformation is 1. Furthermore, when it is not possible to create a dividing line candidate area whose overlapping number with the circumscribing rectangle 1011 of the deformation is 1, it is also possible to calculate dividing line candidate areas by successively increasing the overlapping number with the circumscribing rectangle 1011 of the deformation to 2, 3, ...

[0048] Next, in S604, the region dividing unit 104 sets dividing lines in all or part of the dividing line candidate region to create divided regions. Figure 10(d) shows an example of the deformation data 1004 from which divided regions 1014 have been created. Here, a dividing line 1013, shown as a dashed line in the figure, is set in part of a dividing line candidate region 1012 whose overlapping count with the deformation circumscribing rectangle 1011 is 1, and multiple divided regions 1014 (two in the illustrated example) are created.

[0049] Here, in S604, when setting the dividing line 1013 in the dividing line candidate region 1012, the region dividing unit 104 may assign a priority to the selection of the dividing line candidate region 1012 in which to set the dividing line 1013. For example, the region dividing unit 104 sets a low priority to the dividing line candidate region 1012 that includes an overlapping portion of the circumscribing rectangles 1011 where the deformation surrounding region 1012 is dense. If the dividing line 1013 is set in the dividing line candidate region 1012 that includes an overlapping portion of the circumscribing rectangles 1011, two or more deformation divisions are likely to occur between the divided regions 1014 divided by the dividing line 1013. In this embodiment, from the viewpoint of minimizing deformation divisions as much as possible, it is desirable to set a low priority to the setting of such a dividing line 1013 in order to avoid it as much as possible.

[0050] As described above, according to this embodiment, when it is difficult to divide the area without dividing the anomaly, the area is divided to minimize division while allowing some division of the anomaly. This prevents misidentification of the anomaly and makes it easier for the user to confirm and edit the anomaly.

[0051] <Embodiment 3> Next, a third embodiment of the present disclosure will be described. In the second embodiment, when it is difficult to perform region division without dividing the deformation, the region division is performed to suppress the division while allowing a certain degree of division of the deformation. In the third embodiment, another aspect of the region division process when a certain degree of division of the deformation is allowed is illustrated. Note that the functional configuration and hardware configuration of the image processing device are the same as those of the first embodiment, and therefore description thereof will be omitted.

[0052] In the image processing method according to this embodiment, steps S501 to S506 are executed in sequence, similarly to Fig. 5 shown in embodiment 1. In this embodiment, the area division process S503 is executed similarly to Fig. 6 shown in embodiment 1. Fig. 11 is a schematic diagram showing deformation data in each step of the area division process S503 of this embodiment.

[0053] 11(a) shows an example of deformation data 1101 obtained by image analysis of an image of a structure to be inspected. In this embodiment, the entire structure image is the inspection area (area for image analysis). First, in S601, the area division unit 104 calculates a deformation surrounding area surrounding each deformation in the deformation data 1101, in this case a rectangular circumscribing rectangle surrounding the deformation.

[0054] Next, in S602, the area dividing unit 104 maps the calculated circumscribing rectangle to the deformation data. Figure 11(b) shows an example of deformation data 1102 in which a circumscribing rectangle 1112 is mapped to each deformation. By creating a circumscribing rectangle 1112 corresponding to each deformation in this way, the state of the deformation (position, etc.) when dividing the area can be numerically grasped, and the desired dividing line candidate area can be obtained.

[0055] Next, in S603, the region dividing unit 104 calculates dividing line candidate regions for calculating dividing lines. Fig. 11(c) shows an example of deformation data 1103 in which dividing line candidate regions 1113 are arranged. The region dividing unit 104 allows a minimum amount of overlap with the circumscribing rectangle 1112 so as to reduce the number of overlaps with the circumscribing rectangle 1112, and creates a dividing line candidate region 1113 whose number of overlaps with the circumscribing rectangle 1112 is 1 or 2, as shown in the figure, for example.

[0056] Next, in S604, the region dividing unit 104 sets a dividing line in all or part of the dividing line candidate region to create a divided region. In the first and second embodiments, the region was divided into two regions using a single dividing line as a boundary. In this embodiment, the region is divided into divided regions 1110 and 1111 using two dividing lines. The divided regions 1110 and 1111 partially overlap, and the deformation 1121 divided by the divided region 1110 is not divided by the divided region 1111. Conversely, the deformation 1120 divided by the divided region 1111 is not divided by the divided region 1110 and is included in the divided region. To prevent double editing of deformations when checking and editing each divided region, deformations included in the overlapping portions of the divided regions 1110 and 1111 are treated as belonging to one of the divided regions. In this case, if the entire deformation is included in a divided region without being divided, the deformation is considered to belong to that divided region. In other words, since the deformation 1120 is contained in the divided area 1110 without being divided, it is considered to belong to the divided area 1110, and since the deformation 1121 is contained in the divided area 1111 without being divided, it is considered to belong to the divided area 1111.

[0057] In the configuration of this embodiment, both the deformations 1120 and 1121 can be inspected within a single divided area without dividing them. This eliminates the need to check multiple divided areas associated with dividing the deformation, even when it is not possible to divide the area without dividing the deformation with a single dividing line, thereby improving the efficiency of the deformation inspection work.

[0058] FIG. 12 is a schematic diagram showing an example of a work screen 1201 for checking and editing the deformation of a divided area in this embodiment. The divided area display unit 105 presents the user with a work screen 1201 for checking and editing the divided area. The work screen 1201 includes a divided area image 1202, an entire area image 903, a divided area list 904, an edit defect button 905, and a complete work button 906. The divided area image 1202 displays an enlarged image corresponding to the divided area. The divided area image 1202 may be displayed superimposed on the structure image corresponding to the area. Furthermore, if the displayed divided area image 1202 includes a defect that is separated and does not belong to the divided area, the circumscribing rectangle of the defect may be highlighted. As shown in the divided area image 1202 in FIG. 12, a defect 1211 that does not belong to the divided area may be grayed out to notify the user that the defect does not need to be checked in the divided area being inspected. This allows the user to easily select defects that do not require inspection.

[0059] As described above, according to this embodiment, when it is difficult to divide the area without dividing the deformation, by allowing the divided areas to overlap partially, it is possible to divide the area without dividing the deformation within the divided area. This prevents misidentification of the deformation and makes it easier for the user to confirm and edit the deformation.

[0060] <Embodiment 4> Next, a fourth embodiment of the present disclosure will be described. In the second embodiment, when it is difficult to divide the area without dividing the deformation, the area division is performed to minimize division while allowing some division of the deformation. Here, consider a long deformation that crosses the area of ​​the deformation data. In the case of a long deformation, if an attempt is made to fit the deformation into one divided area without dividing it, the divided area becomes too large, reducing the efficiency of the user's inspection work. However, since long deformations are important in checking the deterioration of a structure, it is desirable to check them without dividing them. Therefore, the fourth embodiment illustrates an example of region division processing without dividing long deformations. Note that the functional configuration and hardware configuration of the image processing device are the same as those of the first embodiment, and therefore description thereof will be omitted.

[0061] In the image processing method according to this embodiment, steps S501 to S506 are executed in sequence, similarly to Fig. 5 shown in embodiment 1. In this embodiment, the area division process S503 is executed similarly to Fig. 6 shown in embodiment 1. Fig. 13 is a schematic diagram showing deformation data in each step of the area division process S503 of this embodiment.

[0062] FIG. 13(a) shows an example of deformation data 1301 obtained by image analysis of an image of a structure to be inspected. In this embodiment, the entire structure image is the inspection area (area for image analysis). First, in S601, the area division unit 104 calculates a deformation surrounding area that surrounds each deformation in the deformation data 1301, in this case a circumscribing rectangle that is a rectangle that surrounds the deformation. Here, for a long deformation 1310 that crosses the area of ​​the deformation data 1301, the processing of S601 is terminated without calculating a circumscribing rectangle.

[0063] Next, in S602, the area division unit 104 maps the calculated circumscribing rectangle to the deformation data. Figure 13(b) shows an example of deformation data 1302 in which a circumscribing rectangle 1311 is mapped to each deformation. A long deformation 1310 that crosses the area of ​​the deformation data 1301 is not mapped because its circumscribing rectangle has not been calculated.

[0064] Next, in S603, the area dividing unit 104 calculates dividing line candidate areas for calculating dividing lines. Figure 13(c) shows an example of the deformation data 1103 in which dividing line candidate areas 1312 are arranged. The dividing line candidate areas 1312 are set so as not to overlap with the circumscribing rectangles 1311 of each deformation.

[0065] Next, in S604, the area dividing unit 104 sets dividing lines in all or part of the dividing line candidate areas to create divided areas. Figure 10(d) shows an example of deformation data 1304 in which divided areas 1314 have been created. Here, an example is shown in which a dividing line 1313, shown by a dashed line in the figure, is set in part of the dividing line candidate area 1312, and multiple (six in the illustrated example) divided areas 1314 have been created. Here, it is assumed that the deformation 1310 belongs to one of the divided areas 1314. In the illustrated example, the deformation 1310 belongs to four divided areas 1314.

[0066] Figure 14 is a schematic diagram showing an example of a work screen 1401 for checking and editing the deformation of a divided area in this embodiment, where (a) shows an example of a specified divided area, and (b) shows an example of a case where the entire deformation that does not fit into the divided area is displayed. The divided area display unit 105 presents the user with a work screen 1401 for checking and editing the divided areas. The work screen 1401 includes a divided area image 1402, an entire area image 903, a divided area list 904, an edit deformation button 905, and a work completion button 906.

[0067] As shown in FIG. 14(a), the divided area display unit 105 displays an enlarged image corresponding to the divided area in the divided area image 1402. The divided area image 1402 may be displayed superimposed on the structure image corresponding to the area. If a deformation 1310 that does not fit within the divided area image 1402 belongs to the divided area 1314, an indicator may be added to the boundary portion 1310a of the deformation 1310 in the divided area 1314 to notify the user that the deformation 1310 continues outside the divided area 1314. For example, as shown in the figure, a mark 1412 may be added to the boundary portion 1310a. By visually recognizing the mark 1412, the user can easily recognize that the deformation 1310 needs to be checked in a wider divided area image.

[0068] As shown in FIG. 14(b), when checking and editing a large deformation 1310, the divided area display unit 105 simultaneously displays multiple divided areas 1314 that include the entire deformation 1310 in the divided area image 1411 so that the entire deformation 1310 is displayed. Here, the example shows a case where the deformation 1310 is large and spans the diagonal of the entire area image, so the entire area image is displayed in the divided area image 1411. This allows the user to appropriately check and edit the long (large) deformation 1410 that spans multiple divided areas using the divided area editing unit 106, using the divided area image 1411 that displays the entire deformation 1410, without dividing the deformation 1410 into multiple divided areas.

[0069] As described above, according to this embodiment, it is possible to check and edit long deformations, which are important for checking the deterioration of a structure, without dividing them into multiple divided areas. This prevents misidentification of deformations and makes it easier for users to check and edit deformations.

[0070] <Embodiment 5> Next, a fifth embodiment of the present disclosure will be described. In the second embodiment, when it is difficult to divide the area without splitting the deformation, the area is divided so as to reduce the number of times the deformation is divided. In the fifth embodiment, attention is paid to the importance of each deformation. The deformation that requires attention when checking varies depending on the width, length, shape of the deformation, detection accuracy during image analysis, etc. By preventing the deformation that requires attention from being split when dividing the area, the efficiency of the user's inspection work can be improved. In this embodiment, the functional configuration example and the hardware configuration example of the image processing device are the same as those in the first embodiment, so explanations are omitted.

[0071] FIG. 15 is a diagram illustrating an example of a deformation importance evaluation table in this embodiment. 15(a) shows an example of an importance evaluation table for calculating the importance of a deformation based on length. The importance evaluation table 1501 is made up of conditions 1502 and scores 1503. The longer the length, the more important the deformation is that requires attention, and therefore the higher the score.

[0072] 15(b) shows an example of an importance evaluation table for calculating the importance of an abnormality based on width. The importance evaluation table 1511 is made up of conditions 1512 and scores 1513. A wide abnormality is easy to see and is unlikely to be overlooked, so the narrower the abnormality, the more likely it is to be overlooked, and the higher the score.

[0073] Figure 15(c) shows an example of an importance evaluation table for calculating the importance of an abnormality based on the reliability map. The importance evaluation table 1521 is made up of conditions 1522 and scores 1523. The reliability map is an index that shows the likelihood of an abnormality when an abnormality is detected by image analysis. The further the reliability map is from the threshold, the higher the reliability of the detected abnormality, and the closer it is to the threshold, the higher the possibility of a false positive. An abnormality that may be a false positive is one that requires attention during inspection, and so receives a high score.

[0074] Figure 15(d) shows an example of an importance evaluation table for calculating the importance of a deformation based on its shape. The importance evaluation table 1531 is made up of conditions 1532 and scores 1523. For example, if there is a branching or lattice-like deformation on the concrete surface, there is a high possibility that the surface will peel off, and so whether the deformation is worthy of attention depends on the shape of the deformation. Therefore, the score is changed depending on the shape of the deformation.

[0075] In the image processing method according to this embodiment, steps S501 to S506 are executed in sequence, similar to Fig. 5 shown in embodiment 1. In this embodiment, in step S503, which is the region division process, steps S601 to S604 are executed in sequence, similar to Fig. 6 shown in embodiment 1.

[0076] In the region division process S503, in S602, the region division unit 104 maps the calculated circumscribing rectangle to the deformation data and calculates the importance of the deformation. The importance L of each deformation can be determined as follows.

[0077]

number

[0078] 16 is a schematic diagram showing an example of deformation data 1601 in which the circumscribing rectangles are mapped and the deformation importance is calculated. The area dividing unit 104 calculates the importance of each deformation and assigns a weight to each circumscribing rectangle.

[0079] Consider a case where it is difficult to divide the area without dividing the deformation. In this embodiment, attention is paid to the importance of each deformation, and the division of deformations with as low importance as possible is permitted. In S604, the division line 1611 that divides the area of ​​the deformation data is set at a position that divides deformations with low importance. As a result, deformations with relatively high importance are not divided between the divided areas, and only deformations with low importance are divided, thereby minimizing the underestimation of deterioration due to deformation and the inefficiency of inspection work.

[0080] By calculating the importance of each anomaly and obtaining the divided areas as described above, it is possible to calculate the total importance of anomalies for each divided area. Therefore, when multiple users share the work of checking and editing the anomalies in the divided areas, it is possible to assign a work proficiency level to each user and assign the divided areas to be checked and edited according to their work proficiency. The allocation of divided areas can be performed automatically by the divided area editing unit 106 according to their work proficiency, for example. The work proficiency level can be determined appropriately based on the user's past performance (such as the number of times the work has been performed in the past). This more reliably prevents misidentification of anomalies and makes the inspection of anomalies more efficient.

[0081] As described above, according to this embodiment, when it is difficult to divide the area without dividing the anomaly, the area is divided so as to prevent the division of the anomaly with high importance while allowing the division of the anomaly with low importance as much as possible. This prevents the user from misidentifying the anomaly and makes it easier for the user to check and edit the anomaly.

[0082] <Other embodiments> Although various embodiments of the present disclosure have been described above in detail, the present disclosure can be embodied as, for example, a system, a device, a method, a program, a recording medium (storage medium), etc. Specifically, the present disclosure may be applied to a system configured from multiple devices (for example, a host computer, an interface device, an imaging device, a web application, etc.), or may be applied to an apparatus consisting of a single device.

[0083] Needless to say, the object of the present disclosure can be achieved by the following: That is, a recording medium (or storage medium) on which is recorded a software program code (computer program) that realizes the functions of the above-described embodiments is supplied to a system or device. Needless to say, such a storage medium is a computer-readable storage medium. Then, the computer (or CPU or MPU) of the system or device reads and executes the program code stored in the recording medium. In this case, the program code read from the recording medium is The program code itself will realize the functions of the above-described embodiments, and the recording medium on which the program code is recorded constitutes the present disclosure.

[0084] The disclosure of the various embodiments includes the following configurations and methods. (Configuration 1) an image analysis means for performing image analysis on an image of an object to be inspected and acquiring a state of change as a result of the image analysis; a division means for dividing the area of ​​the image analysis to obtain divided areas; It is equipped with The dividing means divides the area of ​​the image analysis based on the state of the deformation. Image processing device. (Configuration 2) The dividing means Divide the image analysis area so that the deformation is not divided as much as possible; 2. The image processing device according to claim 1. (Configuration 3) The dividing means acquiring a candidate dividing line region that is a candidate for a dividing point in the region of the image analysis; setting a division line in all or part of the division line candidate region, and dividing the region for image analysis along the division line; 3. The image processing device according to configuration 1 or 2. (Configuration 4) The dividing means Obtain a deformation surrounding area surrounding the deformation; Acquire the dividing line candidate area that does not overlap with the deformation surrounding area. 4. The image processing device according to configuration 3. (Configuration 5) The dividing means Obtain a deformation surrounding area surrounding the deformation; Acquire a dividing line candidate area so that the number of overlaps with the deformation surrounding area is reduced in accordance with the state of the deformation. 5. The image processing device according to configuration 3 or 4. (Configuration 6) The dividing means When setting the division line in the division line candidate region, a priority is assigned to the selection of the division line candidate region in which the division line is to be set. 6. The image processing device according to configuration 5. (Configuration 7) The dividing means The priority of the dividing line candidate area including the overlapping portion of the deformation surrounding areas is set low. 7. The image processing device according to configuration 6. (Configuration 8) The dividing means Dividing the image into divided regions so that adjacent divided regions overlap each other. 5. The image processing device according to any one of the first to fourth configurations. (Configuration 9) The dividing means The deformation included in the overlapping portion is assigned to one of the divided regions. 9. The image processing device according to configuration 8. (Configuration 10) Regarding the first and second deformations of the deformations, The overlapping portion in one of the divided regions includes the entire first deformation and a part of the second deformation, The overlapping portion in the other divided region includes the entire second deformation and a part of the first deformation, 10. The image processing device according to configuration 8 or 9. (Configuration 11) The dividing means Dividing the image analysis area by ignoring predetermined deformations among the deformations; The predetermined deformation is assigned to a plurality of the divided regions. 5. The image processing device according to any one of the first to fourth configurations. (Configuration 12) The dividing means Excluding the predetermined deformation, a deformation surrounding area surrounding the deformation of the image analysis result is obtained, Acquire a division line candidate area that does not overlap with the deformation surrounding area. 12. The image processing device according to claim 11. (Configuration 13) The image processing device further includes a display unit that displays the image analysis result obtained by the image analysis unit, The display means a plurality of divided regions including the entirety of the predetermined deformation are simultaneously displayed; 13. The image processing device according to configuration 11 or 12. (Configuration 14) Further provided is an editing means for reflecting a result of editing the deformation in the divided area in the image analysis result, The editing means assigning an index to a boundary portion of the predetermined deformation in the divided region; 14. The image processing device according to any one of configurations 11 to 13. (Configuration 15) The dividing means Setting importance to the deformation, Dividing the image analysis area based on the importance. 5. The image processing device according to any one of the first to fourth configurations. (Configuration 16) The dividing means Calculating the importance based on the width, length, shape and reliability map of the deformation. 16. The image processing device according to claim 15. (Configuration 17) The dividing means setting the division lines in the division line candidate regions based on the number of divisions designated by a user; 8. The image processing device according to any one of configurations 3 to 7. (Configuration 18) The dividing means setting the division line in the division line candidate area based on a division size designated by a user; 8. The image processing device according to any one of configurations 3 to 7. (Configuration 19) The dividing means The division lines are set so that the vertical and horizontal widths of the divided regions are equal to or greater than a certain value. 8. The image processing device according to any one of configurations 3 to 7. (Configuration 20) The image analysis device further includes a management unit for managing the image analysis results obtained by the image analysis unit. 19. The image processing device according to any one of the first to nineteenth aspects. (Configuration 21) The image processing device further includes a display unit that displays the image analysis result obtained by the image analysis unit. 21. The image processing device according to any one of configurations 1 to 20. (Configuration 22) The display means enlarge and display the predetermined divided area; 22. The image processing device according to claim 21. (Configuration 23) The display means highlighting a predetermined divided region; 23. The image processing device according to configuration 21 or 22. (Configuration 24) Further provided is an editing means for reflecting the result of editing the deformation in the divided area on the image analysis result. 24. The image processing device according to any one of configurations 1 to 23. (Configuration 25) The editing means The addition and correction of the deformation performed on the divided region are reflected in the image analysis result. 25. The image processing device according to claim 24. (Method 1) A first step of performing image analysis on the image of the inspection target and obtaining the deformation state as a result of the image analysis; A second step of dividing the image analysis area to obtain divided areas; It is equipped with The second step divides the area of ​​the image analysis based on the state of the deformation. Image processing methods. (Configuration 26) A program for causing a computer to execute each step of method 1. [Explanation of symbols]

[0085] 101: Image processing device 102: Structure image management department 103: Image analysis unit 104: Area division part 105: Divided area display area 106: Division Area Editorial Department 107: Structure image storage section 108:Analysis result storage section

Claims

1. an image analysis means for performing image analysis on an image of an object to be inspected and acquiring a state of change as a result of the image analysis; a division means for dividing the area of ​​the image analysis to obtain divided areas; It is equipped with The dividing means divides the area of ​​the image analysis based on the state of the deformation. Image processing device.

2. The dividing means Divide the image analysis area so that the deformation is not divided as much as possible; The image processing device according to claim 1 .

3. The dividing means acquiring a candidate dividing line region that is a candidate for a dividing point in the region of the image analysis; setting a division line in all or part of the division line candidate region, and dividing the region for image analysis along the division line; The image processing device according to claim 1 .

4. The dividing means Obtain a deformation surrounding area surrounding the deformation; Acquire the dividing line candidate area that does not overlap with the deformation surrounding area. The image processing device according to claim 3 .

5. The dividing means Obtain a deformation surrounding area surrounding the deformation; Acquire a dividing line candidate area so that the number of overlaps with the deformation surrounding area is reduced in accordance with the state of the deformation. The image processing device according to claim 3 .

6. The dividing means When setting the division line in the division line candidate region, a priority is assigned to the selection of the division line candidate region in which the division line is to be set. The image processing device according to claim 5 .

7. The dividing means The priority of the dividing line candidate area including the overlapping portion of the deformation surrounding areas is set low. The image processing device according to claim 6 .

8. The dividing means Dividing the image into divided regions so that adjacent divided regions overlap each other. The image processing device according to claim 1 .

9. The dividing means The deformation included in the overlapping portion is assigned to one of the divided regions. The image processing device according to claim 8 .

10. Regarding the first and second deformations of the deformations, The overlapping portion in one of the divided regions includes the entire first deformation and a part of the second deformation, The overlapping portion in the other divided region includes the entire second deformation and a part of the first deformation, The image processing device according to claim 8 .

11. The dividing means Dividing the image analysis area by ignoring predetermined deformations among the deformations; The predetermined deformation is assigned to a plurality of the divided regions. The image processing device according to claim 1 .

12. The dividing means Excluding the predetermined deformation, a deformation surrounding area surrounding the deformation of the image analysis result is obtained, Acquire a division line candidate area that does not overlap with the deformation surrounding area. The image processing device according to claim 11 .

13. The image processing device further includes a display unit that displays the image analysis result obtained by the image analysis unit, The display means a plurality of divided regions including the entirety of the predetermined deformation are simultaneously displayed; The image processing device according to claim 11 .

14. Further provided is an editing means for reflecting a result of editing the deformation in the divided area in the image analysis result, The editing means assigning an index to a boundary portion of the predetermined deformation in the divided region; The image processing device according to claim 11 .

15. The dividing means Setting importance to the deformation, Dividing the image analysis area based on the importance. The image processing device according to claim 1 .

16. The dividing means Calculating the importance based on the width, length, shape and reliability map of the deformation. The image processing device according to claim 15.

17. The dividing means setting the division lines in the division line candidate regions based on the number of divisions designated by a user; The image processing device according to claim 3 .

18. The dividing means setting the division line in the division line candidate area based on a division size designated by a user; The image processing device according to claim 3 .

19. The dividing means The division lines are set so that the vertical and horizontal widths of the divided regions are equal to or greater than a certain value. The image processing device according to claim 3 .

20. The image analysis device further includes a management unit for managing the image analysis results obtained by the image analysis unit. The image processing device according to claim 1 .

21. The image processing device further includes a display unit that displays the image analysis result obtained by the image analysis unit. The image processing device according to claim 1 .

22. The display means enlarge and display the predetermined divided area; The image processing device according to claim 21 .

23. The display means highlighting a predetermined divided region; The image processing device according to claim 21 .

24. Further provided is an editing means for reflecting the result of editing the deformation in the divided area on the image analysis result. The image processing device according to claim 1 .

25. The editing means The addition and correction of the deformation performed on the divided region are reflected in the image analysis result. The image processing device according to claim 24.

26. A first step of performing image analysis on an image of an object to be inspected and acquiring a state of change as a result of the image analysis; a second step of dividing the image analysis area to obtain divided areas; It is equipped with The second step divides the area of ​​the image analysis based on the state of the deformation. Image processing methods.

27. A program for causing a computer to execute the steps recited in claim 26.

Images