JP2019203865A - Diagnosis device, diagnosis method, and program - Google Patents

Abstract

To accurately diagnose the presence or absence of cracks on the surface of an object.SOLUTION: A diagnosis device includes: an area setting unit that sets a mesh for dividing at least part of an image area of a photographic image obtained by photographing an object into a plurality of areas and moves the position of the set mesh; and a diagnosis unit that diagnoses the presence or absence of cracks that have occurred on the object from the photographic image for each of the areas divided by the mesh set by the area setting unit and diagnoses the presence or absence of cracks from the photographic image for each of the areas divided by the mesh after movement.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a diagnostic device, a diagnostic method, and a program.

  A technique for detecting cracks generated on the surface of a concrete structure from an image of the concrete structure is disclosed (for example, Patent Document 1).

JP-A-2015-102382

  For example, when a photographed image is divided by a mesh and the presence or absence of cracks in each divided area is determined using AI (artificial intelligence), there is a crack near the corner in the area divided by the mesh. There was a possibility that the presence or absence of the crack could not be properly determined.

  SUMMARY An advantage of some aspects of the invention is that it provides a diagnostic device, a diagnostic method, and a program capable of accurately diagnosing the presence or absence of cracks generated on the surface of an object.

  The present invention has been made to solve the above problems, and one aspect of the present invention is to set a mesh that divides at least a part of an image area of a captured image obtained by capturing an object into a plurality of areas. And a region setting unit for moving the set mesh position, and diagnosing the presence or absence of cracks generated in the object from the captured image for each of a plurality of regions divided by the mesh set by the region setting unit And a diagnostic unit that diagnoses the presence or absence of the crack from the captured image for each of a plurality of regions divided by the mesh after movement.

  According to another aspect of the present invention, there is provided a diagnostic method in a diagnostic apparatus, wherein the region setting unit sets a mesh that divides at least a part of an image region of a captured image obtained by capturing an object into a plurality of regions. A step of diagnosing the presence or absence of cracks generated in the object from the captured image for each of a plurality of regions divided by the mesh set by the region setting unit, and the region setting unit A method of moving the set position of the mesh, and a step of diagnosing the presence or absence of cracks from the captured image for each of a plurality of regions divided by the mesh after the movement by the diagnosis unit It is.

  According to another aspect of the present invention, a step of setting a mesh that divides at least a part of an image area of a captured image obtained by capturing an object into a plurality of areas in the computer, and dividing by the set mesh is performed. A step of diagnosing the presence or absence of cracks generated in the object from the photographed image for each of a plurality of areas, a step of moving the set mesh position, and a plurality of divided by the mesh after the movement And a step of diagnosing the presence or absence of the crack from the photographed image for each region.

  According to the present invention, it is possible to accurately diagnose cracks generated on the surface of an object.

The block diagram which shows an example of the diagnostic apparatus 10 which concerns on embodiment. The figure which shows the example of a setting of the mesh which concerns on embodiment. The figure explaining the change of the position of a crack when moving a mesh to a horizontal direction. The figure which shows the 1st example of the example which the diagnostic accuracy improved after moving a mesh to a horizontal direction. The figure which shows the 2nd example of the example which the diagnostic accuracy improved after moving a mesh to a horizontal direction. The figure which shows the 3rd example of the example which the diagnostic accuracy improved after moving a mesh to a horizontal direction. The figure which shows the 4th example of the example which the diagnostic accuracy improved after moving a mesh to a horizontal direction. The figure which shows the 5th example of the example which the diagnostic accuracy improved after moving a mesh to a horizontal direction. The figure explaining the change of the position of a crack when moving a mesh to the vertical direction. The figure which shows the 1st example of the example which the diagnostic accuracy improved after moving a mesh to the vertical direction. The figure which shows the 2nd example of the example which the diagnostic accuracy improved after moving a mesh to the vertical direction. The figure which shows the 3rd example of the example which the diagnostic accuracy improved after moving a mesh to the vertical direction. The figure which shows the 4th example of the example which the diagnostic accuracy improved after moving a mesh to the vertical direction. The figure which shows the 5th example of the example which the diagnostic accuracy improved after moving a mesh to the vertical direction. The flowchart which shows an example of the diagnostic process which concerns on embodiment. The figure which shows the 1st example of the example of provision of the diagnostic result which concerns on embodiment. The figure which shows the 2nd example of the example of provision of the diagnostic result which concerns on embodiment. The figure which shows the 3rd example of the example of provision of the diagnostic result which concerns on embodiment. The figure explaining the production | generation method of the diagnostic result which combined three types of meshes. The figure which shows the provision example of the diagnostic result when combining 3 types of meshes.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

First, an outline of the present embodiment will be described.
In recent years, verification of image diagnosis using an ultra-high resolution camera has been performed as an alternative means of close visual inspection for inspecting cracks and the like generated on a bridge in an actual site. For example, by using an image of 100 million pixels photographed by an ultra-high resolution camera, an inspection environment almost equivalent to a close visual inspection performed by a person is constructed. In particular, in places where it is difficult for people to approach and where close visual inspection is difficult, it is possible to easily obtain a photographed image to be inspected by using a UAV (Unmanned Aerial Vehicle) equipped with a camera. The diagnosis that was present is effective. In addition, since there is a shortage of personnel who have undergone inspection (inspectors), there is a method of performing image diagnosis using AI (artificial intelligence) on a captured image. In this case, for example, a method of dividing a captured image with a mesh and diagnosing the presence or absence of cracks for each divided area using AI (artificial intelligence) is conceivable. However, if there is a crack near the corner in the area divided by the mesh, there is a possibility that the presence or absence of the crack cannot be properly determined. Therefore, the diagnostic apparatus according to the present embodiment is configured to improve the diagnostic accuracy by moving the mesh when diagnosing the presence or absence of cracks generated in the object.

(Configuration of diagnostic device)
Hereinafter, the configuration of the diagnostic apparatus according to the present embodiment will be described in detail. Here, “crack” is a general term for things like cracks, scratches, dents and the like generated on the surface of an object. The object is an object to be inspected, for example, a concrete structure such as a bridge or a building. The object is not limited to a concrete structure, and may be any object as long as cracking occurs.

  FIG. 1 is a block diagram illustrating an example of a diagnostic apparatus 10 according to the present embodiment. The diagnostic apparatus 10 includes a communication unit 110, an input unit 120, a display unit 130, a storage unit 150, and a control unit 170.

  The communication unit 110 may include hardware (for example, an antenna and a transmission / reception device) for connecting to the Internet using a cellular network, a Wi-Fi network, or the like. The communication unit 110 may be a connector connected to an external storage medium, or may be a connector connected to an external device such as an imaging device via a cable.

  The input unit 120 includes some or all of various keys, buttons, dial switches, a mouse, and the like. The input unit 120 may be a touch panel formed integrally with the display unit 130, for example.

  The display unit 130 is configured to include a display such as an LCD (Liquid Crystal Display) or an organic EL (Electroluminescence).

  The storage unit 150 is, for example, a flash memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), an SSD (Solid State Drive), or an HDD (Hard Disk Drive). For example, the storage unit 150 stores captured image data 151, teaching data 152, a learned model 153, and the like.

  The photographed image data 151 is photographed image data obtained by photographing an object to be inspected (for example, a concrete structure such as a bridge).

  The teaching data 152 is teaching data for generating a learned model of AI used when diagnosing the presence or absence of cracks from a captured image. For example, image data including cracks is stored as teaching data 152. The image data including cracks is, for example, image data obtained by photographing an arbitrary object (for example, a concrete structure such as a bridge pier) that is actually cracked. Further, image data that does not include cracks may be further stored as the teaching data 152. These image data may be image data in which the presence or absence of cracks has been identified in the past by a person (for example, a person who has undergone inspection), or image data that has been diagnosed by the diagnostic device 10 for the presence or absence of cracks. Alternatively, it may be image data in which the presence or absence of cracks is diagnosed by another diagnostic device. In addition, image data including an object (such as grass) that may be included in the captured image may be further stored as the teaching data 152. In addition, by including image data including an object (such as grass) that may be included in the captured image in the teaching data, it is possible to suppress erroneous diagnosis due to a crack. For example, a plurality of pieces of image data including cracks, image data not including cracks, and image data including grass are stored as teaching data 152.

  The learned model 153 stores a learned model that the learning unit 172 described below generates based on the teaching data 152 described above.

  The control unit 170 includes, for example, an acquisition unit 171, a learning unit 172, a region setting unit 173, as a functional configuration realized by a hardware processor such as a CPU (Central Processing Unit) executing a program (software). , A diagnosis unit 174 and a result providing unit 175 are provided.

  The acquisition unit 171 acquires a captured image via the communication unit 110 and stores it in the storage unit 150 as captured image data 151. The photographed image is a photographed image obtained by photographing an object to be inspected (for example, a concrete structure such as a bridge). For example, a captured image of an object captured using UAV or a captured image of an object captured by a person.

  Based on the teaching data 152, the learning unit 172 generates an AI learned model 153 for diagnosing the presence or absence of cracks. For example, the learning unit 172 performs machine learning using any or all of a plurality of image data including cracks, image data not including cracks, image data including grass, etc. as teaching data, and determines whether or not there are cracks. A learned model 153 for determination is generated as a learning result.

  The area setting unit 173 sets a mesh that divides at least a part of an image area of a captured image obtained by capturing an object into a plurality of areas. FIG. 2 is a diagram illustrating a mesh setting example according to the present embodiment. In the example shown in the figure, the object to be inspected is the surface of a concrete pier, and a mesh M is set to divide the captured image obtained by photographing the object into a rectangular region of 6 × 19. The presence or absence of cracks is diagnosed for each area divided with respect to the range of the mesh M. Note that the number of divisions, the shape of the divided areas, the mesh setting range, and the like are examples, and are not limited to the illustrated examples.

  The region setting unit 173 moves the set mesh position. For example, the region setting unit 173 moves in the horizontal direction or the vertical direction at a distance narrower than the mesh interval. The horizontal direction is a direction (horizontal direction in FIG. 2) parallel to the horizontal line (longitudinal line) of the mesh M of 6 × 19 horizontally shown in FIG. The vertical direction is a direction orthogonal to the horizontal direction, and is a direction (vertical direction in FIG. 2) parallel to the vertical line (short-side line) of the mesh 6 of length 6 × width 19 shown in FIG. .

  For example, the region setting unit 173 may move the set mesh position in the lateral direction (left or right direction) by a distance narrower than the mesh interval. The region setting unit 173 may move the set mesh position in the vertical direction (up or down) by a distance narrower than the mesh interval. The region setting unit 173 may move the set mesh position in the horizontal direction (left or right direction) and the vertical direction (up or down direction) by a distance narrower than the mesh interval. Here, the distance moved is, for example, a distance corresponding to half of the mesh interval.

  The diagnosis unit 174 uses the AI learned model 153 to diagnose the presence or absence of cracks from the captured image in which the region setting unit 173 has set a mesh. First, the diagnosis unit 174 diagnoses the presence or absence of cracks from the captured image for each of a plurality of regions divided by the mesh set by the region setting unit 173. Furthermore, the diagnosis unit 174 diagnoses the presence or absence of cracks from the captured image for each of a plurality of regions divided by the mesh after movement. For example, the diagnosis unit 174 performs a diagnosis for each region divided by the initially set mesh, and then further halves the mesh interval in the horizontal direction (left or right direction) or the vertical direction (up or down direction). Diagnosis is performed for each area divided by the moved mesh that has moved a distance corresponding to. As a result, the crack present in the corner of the area divided by the initially set mesh moves toward the center in the area divided by the moved mesh, so the presence or absence of cracks can be accurately diagnosed. become able to.

  The result providing unit 175 uses the display unit 130 to provide the user with the diagnosis result by the diagnosis unit 174. For example, the result providing unit 175 causes the display unit 130 to display information based on the probability of including a crack for each of a plurality of regions divided by the mesh of the captured image. The information based on the probability including a crack may be a probability value, or may be a symbol, color, brightness, or the like corresponding to the probability.

(Example of improved diagnostic accuracy)
Next, with reference to FIGS. 3 to 14, an example in which diagnosis accuracy is actually improved by moving the mesh will be described. First, an example in which diagnosis accuracy is improved when the mesh is moved in the lateral (left / right) direction will be described with reference to FIGS. FIG. 3 is a diagram for explaining a change in the position of a crack when the mesh is moved in the horizontal direction. Here, an example is shown in which the initially set mesh is moved in the horizontal direction by a distance corresponding to half the mesh interval. The mesh M1a indicates an image in one area among the areas divided by the initially set mesh. The mesh M1b shows an image in the region when the mesh M1a is moved in the horizontal direction by a distance that is half the mesh interval. The crack CR1 present at the right end in the mesh M1a relatively moves to the vicinity of the center in the mesh M1b after the movement. This improves the accuracy of crack diagnosis. Note that the arrows shown in the figure indicate that the distance of half the mesh interval is moved to the right, but even in the case of movement to the left, the area immediately to the right of the mesh M1a is the same position as the area of the mesh M1b. Similarly, the diagnostic accuracy is improved.

  4 to 8 show an example in which the diagnostic accuracy is improved after the movement when the mesh shown in FIG. 3 is moved in the horizontal (left and right) direction than before the movement. When the mesh is moved in the horizontal (left / right) direction, it is effective in improving diagnostic accuracy mainly for cracks that are long in the vertical direction. Here, the diagnosis unit 174 indicates “probability including a crack” as a diagnosis result using the AI learned model 153. The probability of including a crack is an index of the probability that an image in each region divided by a mesh includes a crack, and the minimum value is 0% and the maximum value is 100%.

  In the example shown in FIG. 4, CR2 is present at the left end in the initially set mesh M2a, but CR2 is cracked near the center in the mesh M2b after the movement by moving the mesh in the horizontal direction. Is moving. In the diagnosis result for the initially set mesh M2a, the probability that the image in the mesh M2a includes a crack was 61%. On the other hand, in the diagnosis result for the mesh M2b after movement, the probability that the image in the mesh M2b includes a crack was diagnosed as 99.7%. That is, in the diagnosis result after the mesh movement, the 38.7 point diagnosis accuracy was improved.

  In the example shown in FIG. 5, CR3 is cracked toward the right end in the initially set mesh M3a, but in the mesh M3b after the movement by moving the mesh in the lateral direction, CR3 is cracked near the center. Is moving. In the diagnosis result for the initially set mesh M3a, the probability that the image in the mesh M3a includes a crack was diagnosed as 2.9%. On the other hand, in the diagnosis result for the mesh M3b after movement, the probability that the image in the mesh M3b includes cracks was diagnosed as 99.5%. That is, 96.6 points of diagnostic accuracy improved in the diagnostic result after the mesh movement.

  In the example shown in FIG. 6, CR4 is cracked toward the right end in the initially set mesh M4a, but the mesh CR4 is cracked near the center in the moved mesh M4b by moving the mesh in the horizontal direction. Is moving. In the diagnosis result for the initially set mesh M4a, the probability that the image in the mesh M4a includes a crack was diagnosed as 11.2%. On the other hand, in the diagnosis result for the mesh M4b after movement, the probability that the image in the mesh M4b includes a crack was diagnosed as 99.9%. That is, in the diagnosis result after moving the mesh, the 88.7 point diagnosis accuracy improved.

  In the example shown in FIG. 7, CR5 is present at the right end in the initially set mesh M5a, but CR5 is cracked near the center in the mesh M5b after the movement by moving the mesh in the horizontal direction. Is moving. In the diagnosis result for the initially set mesh M5a, the probability that the image in the mesh M5a includes cracks was diagnosed as 5.8%. On the other hand, in the diagnosis result for the mesh M5b after movement, the probability that the image in the mesh M5b includes a crack was diagnosed as 99.1%. That is, 93.3 point diagnosis accuracy improved in the diagnosis result after moving the mesh.

  In the example shown in FIG. 8, there is a crack CR6 toward the right end in the initially set mesh M6a, but in the mesh M6b after the movement by moving the mesh in the horizontal direction, the CR6 is cracked near the center. Is moving. In the diagnosis result for the initially set mesh M6a, the probability that the image in the mesh M6a includes a crack was diagnosed as 17.1%. On the other hand, in the diagnosis result for the mesh M6b after movement, the probability that the image in the mesh M6b includes a crack was diagnosed as 56.5%. That is, in the diagnosis result after the mesh movement, the 39.4 point diagnosis accuracy improved.

  Next, an example in which the diagnostic accuracy is improved when the mesh is moved in the vertical (vertical) direction will be described with reference to FIGS. FIG. 9 is a diagram for explaining a change in the position of a crack when the mesh is moved in the vertical direction. Here, an example is shown in which the initially set mesh is moved in the vertical direction by a distance corresponding to half the mesh interval. The mesh M7a shows an image in one area among the areas divided by the initially set mesh. A mesh M7b shows an image in an area when the mesh M7a is moved in the vertical direction by a distance that is half of the mesh interval. The crack CR7 existing at the lower end in the mesh M7a relatively moves to the vicinity of the center in the vertical direction in the mesh M7b after the movement. This improves the accuracy of crack diagnosis. Note that the arrow shown in the figure indicates that the half of the mesh interval is moved downward, but the area below the mesh M7a is the same position as the area of the mesh M7b even when moving upward. The same diagnosis result is obtained.

  FIGS. 10 to 14 show an example in which the diagnostic accuracy is improved after movement when the mesh shown in FIG. 9 is moved in the vertical (vertical) direction. When the mesh is moved in the vertical (vertical) direction, it is effective in improving diagnostic accuracy mainly for cracks that are long in the horizontal direction. In the example shown in FIG. 10, there is a crack CR8 toward the upper end in the initially set mesh M8a, but in the mesh M8b after the movement by moving the mesh in the vertical direction, the crack CR8 is relative to the center. Is moving. In the diagnosis result for the initially set mesh M8a, the probability that the image in the mesh M8a includes a crack was 94.6%. On the other hand, in the diagnosis result for the mesh M8b after movement, the probability that the image in the mesh M8b includes cracks was diagnosed as 97.6%. That is, the three-point diagnostic accuracy is improved in the diagnostic result after the mesh is moved.

  In the example shown in FIG. 11, there is a crack CR9 toward the lower end in the initially set mesh M9a, but in the mesh M9b after the movement by moving the mesh in the vertical direction, the crack CR9 is relative to the center. Is moving. In the diagnosis result for the initially set mesh M9a, the probability that the image in the mesh M9a includes a crack was 6.1%. On the other hand, in the diagnosis result for the mesh M9b after movement, the probability that the image in the mesh M9b includes a crack was diagnosed as 57.2%. In other words, the diagnostic accuracy after moving the mesh improved the 51.1 point diagnostic accuracy.

  In the example shown in FIG. 12, the CR10 is cracked toward the lower end in the initially set mesh M10a. However, the CR10 is cracked near the center in the moved mesh M10b by moving the mesh in the vertical direction. Is moving. In the diagnosis result for the initially set mesh M10a, the probability that the image in the mesh M10a includes a crack was 51%. On the other hand, in the diagnosis result for the mesh M10b after movement, the probability that the image in the mesh M10b includes a crack was diagnosed as 99.7%. That is, in the diagnosis result after moving the mesh, the 48.7 point diagnosis accuracy improved.

  In the example shown in FIG. 13, the CR11 is cracked toward the upper end (and the right end) in the initially set mesh M11a, but in the mesh M11b after the movement by moving the mesh in the vertical direction, CR11 is relatively moved by cracking near the center (and the right end). In the diagnosis result for the initially set mesh M11a, the probability that the image in the mesh M11a includes a crack was diagnosed as 42.7%. On the other hand, in the diagnosis result for the mesh M11b after movement, the probability that the image in the mesh M11b includes cracks was diagnosed as 99.8%. That is, in the diagnosis result after moving the mesh, the 57.1 point diagnosis accuracy improved.

  In the example shown in FIG. 14, the CR12 is cracked toward the lower end in the initially set mesh M12a. However, the CR12 is cracked near the center in the moved mesh M6b by moving the mesh in the vertical direction. Is moving. In the diagnosis result for the initially set mesh M12a, the probability that the image in the mesh M12a includes a crack was 68.3%. On the other hand, according to the diagnosis result for the mesh M12b after movement, the probability that the image in the mesh M12b includes cracks was diagnosed as 94.3%. That is, the diagnostic accuracy after moving the mesh improved 26-point diagnostic accuracy.

(Diagnosis process operation)
Next, an operation of a diagnostic process in which the diagnostic apparatus 10 according to the present embodiment diagnoses cracks generated on the surface of the object will be described.
FIG. 15 is a flowchart illustrating an example of a diagnostic process according to the present embodiment.

  First, the acquisition unit 171 acquires a captured image of an object via the communication unit 110 and stores it in the storage unit 150 as captured image data 151 (step S101).

  Next, the region setting unit 173 sets a mesh in at least a part of the image region of the captured image of the captured image data 151 (step S103).

  Then, the diagnosis unit 174 diagnoses the presence or absence of cracks from the captured image for each of a plurality of regions divided by the mesh set by the region setting unit 173 (step S105).

  Next, the region setting unit 173 moves the set mesh position in the horizontal direction (left or right direction). Specifically, the region setting unit 173 moves the set mesh position by a distance corresponding to half the mesh interval in the horizontal direction (left or right direction) (step S107).

  Then, the diagnosis unit 174 diagnoses the presence or absence of cracks from the captured image for each of a plurality of regions divided by the moved mesh moved by the region setting unit 173 (step S109).

  Subsequently, the region setting unit 173 moves the set mesh position in the vertical direction (up or down). Specifically, the region setting unit 173 moves the set mesh position in the vertical direction (up or down) by a distance corresponding to half of the mesh interval (step S111).

  Then, the diagnosis unit 174 diagnoses the presence or absence of cracks from the captured image for each of a plurality of regions divided by the moved mesh moved by the region setting unit 173 (step S113).

  The result providing unit 175 uses the display unit 130 to provide the user with the diagnosis result by the diagnosis unit 174. For example, the result providing unit 175 causes the display unit 130 to display information based on the probability of including a crack for each of a plurality of regions divided by the mesh of the captured image (step S115).

(Example of providing diagnosis results)
16 to 18 are diagrams illustrating examples of providing diagnosis results according to the present embodiment. FIG. 16 shows a diagnosis result by the initially set mesh. Further, FIG. 17 shows a diagnosis result by the moved mesh obtained by moving the initially set mesh in the horizontal direction by a distance corresponding to half of the mesh interval. Further, FIG. 18 shows a diagnosis result by the moved mesh obtained by moving the initially set mesh in the vertical direction by a distance corresponding to half of the mesh interval. In each of FIGS. 16 to 18, three types of diagnosis results for each region divided by the mesh are a region having a probability of including a crack of less than 30%, a region of 30 to 60%, and a region of 60% or more. It is displayed with symbols. An area where the probability of including a crack is less than 30% is a diagnosis that “the probability of occurrence of a crack is low”. The region where the probability of including a crack is 30 to 60% is a diagnosis that “the probability of occurrence of cracking is medium”. An area having a probability of including cracks of 60% or more is a diagnosis that “the probability of occurrence of cracks is high”.

  As described above, when cracks are present in the corners of the area divided by the mesh, the cracks are not always properly diagnosed. In particular, when many images of the teaching data are cracked images that pass through the vicinity of the center of the image, the cracks present in the corners are not easily diagnosed. For example, it is conceivable to improve the recognition accuracy by including an image with cracks in the corners of the image in the teaching data, but it is difficult to recognize images of cracks in the corners in the first place. It is not easy to learn the optimal teaching data. On the other hand, by moving the mesh in the horizontal or vertical direction, the cracks that existed in the corners of the area divided by the mesh are moved relatively to the center, which can easily improve the diagnostic accuracy. it can.

  For example, if the mesh is moved laterally as shown in FIG. 17, the crack on the left side of the pier (vertical crack) that was not correctly diagnosed with the initially set mesh shown in FIG. 16 is properly diagnosed. It was. In addition, by moving the mesh in the vertical direction as shown in FIG. 18, the crack at the center of the pier (lateral crack) that could not be accurately diagnosed with the initially set mesh shown in FIG. Diagnosed.

  Also, the diagnosis result with the initially set mesh shown in FIG. 16, the diagnosis result with the mesh moved in the horizontal direction shown in FIG. 17, and the diagnosis result with the mesh moved in the vertical direction shown in FIG. A diagnostic result may be generated by combining the three types of diagnostic results. For example, the crack occurrence probability of a region where three meshes overlap may be an average of the crack generation probabilities of the three overlapping meshes. Details will be described with reference to FIG.

  FIG. 19 is a diagram illustrating a method for generating a diagnosis result obtained by combining the diagnosis results of three types of meshes. Here, for ease of explanation, a 4 × 4 mesh is extracted and shown. The mesh MA shown in (A) is the mesh set first. The mesh MB shown in (B) is obtained by moving the mesh MA in the horizontal direction by a distance that is half the mesh interval. The mesh MC shown in (C) is obtained by moving the mesh MA in the longitudinal direction by a distance that is half the mesh interval. (D) is the figure which accumulated mesh MA, mesh MB, and mesh MC. In (D), the region MD1 is a region where the region MA1 of the mesh MA, the region MB1 of the mesh MB, and the region MC1 of the mesh MC overlap. The diagnosis result of the overlapping area MD1 is a crack occurrence probability obtained by averaging the crack occurrence probability of the area MA1 of the mesh MA, the crack occurrence probability of the area MB1 of the mesh MB, and the crack occurrence probability of the area MC1 of the mesh MC. .

  FIG. 20 is a diagram illustrating an example of providing diagnosis results when the diagnosis results of three types of meshes are combined. In this way, by taking the average of the diagnostic results of the three types of meshes, the vertical cracks that could not be accurately diagnosed with the mesh results set at the beginning were The cracks in the horizontal direction, which are complemented by the diagnosis result, can be complemented by the diagnosis result with the mesh moved in the vertical direction, and the diagnostic accuracy can be improved. Further, by combining the diagnosis results of the three types of meshes, the diagnosis can be performed for each of the more finely divided areas, so that the position of the crack can be specified in more detail. As shown in FIG. 20, the diagnosis result when the diagnosis results of the three types of meshes are combined is more accurate in identifying cracks than the diagnosis result of the initially set mesh shown in FIG. There is more detail.

  As described above, the diagnostic apparatus 10 according to the present embodiment sets a mesh that divides at least a part of an image area of a captured image obtained by capturing an object into a plurality of areas, and is divided by the set mesh. The presence or absence of cracks generated on the object is diagnosed from the captured image for each of a plurality of areas. Further, the diagnostic device 10 moves the set mesh position, and diagnoses the presence or absence of the crack from the captured image for each of a plurality of areas divided by the moved mesh.

  As a result, the diagnosis device 10 moves the mesh again and makes a diagnosis again even if it cannot be diagnosed accurately due to the presence of cracks at the corners of the region divided by the initially set mesh. Cracks generated on the surface of an object can be diagnosed with high accuracy, and diagnostic accuracy can be improved.

  For example, the diagnostic apparatus 10 moves the set mesh position in a horizontal direction or a vertical direction by a distance narrower than the mesh interval.

  Thereby, when there is a crack toward the corner of the area divided by the initially set mesh, the diagnostic device 10 can relatively move the position of the crack within the area divided by the mesh. . Therefore, since the position of the crack that exists near the corner before the movement of the mesh moves relatively toward the center after the movement of the mesh, the crack generated on the surface of the object can be accurately diagnosed.

  More specifically, the distance for moving the mesh may be a distance corresponding to half of the mesh interval.

  As a result, when there is a crack near the corner of the region divided by the initially set mesh, the diagnostic apparatus 10 moves the position of the crack relative to the position near the center in the region divided by the mesh. Since it can be moved, the crack diagnosis accuracy can be improved.

  The distance for moving the mesh is not limited to a distance corresponding to half of the mesh interval, and may be an arbitrary distance within a range of distances narrower than the mesh interval. Moreover, when the distance which moves a mesh is shorter than the half distance of the space | interval of a mesh, you may increase the frequency | count of moving. For example, when the distance to move the mesh is half the distance between the meshes, the mesh is moved once. When the distance to move the mesh is a quarter distance between the meshes, the mesh is 2 It may be moved once.

  In addition, the diagnostic device 10 moves the set mesh position in each of a first direction (for example, the horizontal direction) and a second direction (for example, the vertical direction) perpendicular to the first direction. Also good. The diagnostic apparatus 10 then captures the captured images for each of the plurality of regions divided by the initially set mesh and the plurality of regions divided by the moved mesh that has moved in the first direction (for example, the horizontal direction). The presence or absence of cracks generated on the object is diagnosed using each of the captured image and the captured image for each of the plurality of regions divided by the moved mesh moved in the second direction (for example, the vertical direction). May be.

  Thereby, the diagnostic device 10 moves the mesh in two directions perpendicular to each other (for example, the horizontal direction and the vertical direction), thereby making a left and right corner within the region divided by the initially set mesh. Since it is possible to make a diagnosis by moving both the cracks present in the center and the cracks present in the upper and lower corners relatively to the vicinity of the center, the crack diagnosis accuracy can be improved. Further, by combining the diagnosis results of the three types of meshes, the diagnosis can be performed for each of the more finely divided areas, so that the position of the crack can be specified in more detail.

  In addition, the diagnostic device 10 learns a plurality of images including cracks as at least teaching data, and is generated from the captured image to the object for each region divided by the mesh based on the learned model of AI based on the learning result. Diagnose for cracks.

  Thereby, the diagnostic apparatus 10 can accurately diagnose the presence / absence of a crack using AI even if there is no experienced person or skilled person who can visually identify cracks. Therefore, as more and more aging piers are increasing in the future, it is expected that there will be a shortage of experienced and skilled personnel as the population declines. Inspection is possible and can contribute to the maintenance of the pier.

  The embodiments of the present invention have been described in detail above with reference to the drawings. However, the specific configuration is not limited to the above-described one, and various design changes and the like can be made without departing from the scope of the present invention. Is possible.

  In the above-described embodiment, the example in which the mesh is moved in the horizontal direction and the vertical direction has been described. For example, the horizontal direction corresponds to the horizontal direction with respect to the object, and the vertical direction corresponds to the vertical direction with respect to the object. However, the relationship between the directions is not limited to the horizontal direction and the vertical direction because it changes depending on the tilt of the camera and the mesh setting method when the captured image is captured.

  In the above-described embodiment, an example of the process of performing the diagnosis by moving the mesh in the horizontal direction and the vertical direction has been described. However, the process may be performed by moving the mesh in either one of the directions. Further, the diagnosis may be performed by moving the mesh in three or more directions.

  For example, in the above-described embodiment, in the case of a mesh divided into a rectangular area of 6 × 19 rectangles, an example in which the mesh is moved in the horizontal direction or the vertical direction according to the shape of the mesh has been described, but the present invention is not limited thereto. The mesh may be moved in an oblique direction instead of or in addition to the horizontal direction or the vertical direction. Also, the shape of the mesh (the shape of the divided area and the shape of the entire mesh) is not limited to a rectangle, and may be a square, a triangle, or a polygonal shape of a pentagon or more, or a curve. The shape may be included. Further, the shape of the area divided by the mesh may be different from the shape of the entire mesh.

  In addition, you may make it implement | achieve part or all of the control part 170 in embodiment mentioned above with a computer. In that case, a program for realizing a part or all of the functions of the control unit 170 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed. It may be realized by. Here, the “computer system” is a computer system built in the diagnostic device 10 and includes an OS and hardware such as peripheral devices. The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” is a medium that dynamically holds a program for a short time, such as a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line, In such a case, a volatile memory inside a computer system serving as a server or a client may be included and a program that holds a program for a certain period of time. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  Moreover, you may implement | achieve part or all of the control part 170 in embodiment mentioned above as integrated circuits, such as LSI (Large Scale Integration). Further, for example, a part or all of the control unit 170 may be integrated into a processor. Further, the method of circuit integration is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. In addition, when an integrated circuit technology that replaces LSI appears due to the advancement of semiconductor technology, an integrated circuit based on the technology may be used.

  DESCRIPTION OF SYMBOLS 10 Diagnostic apparatus, 110 Communication part, 120 Input part, 130 Display part, 150 Storage part, 151 Captured image data, 152 Teaching data, 153 Learned model, 170 Control part, 171 Acquisition part, 172 Learning part, 173 Area setting part 174 Diagnosis Department, 175 Results Provision Department

Claims (8)

An area setting unit that sets a mesh that divides at least a part of an image area of a captured image obtained by capturing an object into a plurality of areas, and moves the position of the set mesh;
While diagnosing the presence or absence of cracks generated in the object from the captured image for each of a plurality of regions divided by the mesh set by the region setting unit, for each of the plurality of regions divided by the mesh after movement A diagnostic unit for diagnosing the presence or absence of cracks from a captured image;
A diagnostic device comprising: The region setting unit
Move the set mesh position in the first direction by a distance narrower than the mesh interval,
The diagnostic device according to claim 1. The first direction is a horizontal direction or a vertical direction.
The diagnostic device according to claim 2. The region setting unit
Moving the set mesh position in each of a first direction and a second direction orthogonal to the first direction;
The diagnostic unit
The captured image for each of the plurality of regions divided by the mesh set by the region setting unit; the captured image for each of the plurality of regions divided by the mesh after movement moved in the first direction; Diagnosing the presence or absence of cracks using each of the captured images for each of a plurality of regions divided by the moved mesh moved in the second direction;
The diagnostic apparatus as described in any one of Claims 1-3. The distance that the area setting unit moves the mesh is a distance corresponding to half of the mesh interval.
The diagnostic device according to any one of claims 1 to 4. The diagnostic unit
A learning unit for learning a plurality of images including the cracks as at least teaching data;
Based on a learned model based on a learning result by the learning unit, diagnose the presence or absence of the crack from the captured image for each region divided by the mesh,
The diagnostic device according to any one of claims 1 to 5. A diagnostic method in a diagnostic device,
An area setting unit that sets a mesh that divides at least a part of an image area of a captured image obtained by capturing an object into a plurality of areas;
A step of diagnosing the presence or absence of cracks generated in the object from the captured image for each of a plurality of regions divided by the mesh set by the region setting unit;
The region setting unit moving the set mesh position;
The diagnostic unit diagnoses the presence or absence of cracks from the captured image for each of a plurality of areas divided by the mesh after the movement;
A diagnostic method comprising: On the computer,
Setting a mesh that divides at least a part of an image area of a photographed image obtained by photographing an object into a plurality of areas;
Diagnosing the presence or absence of cracks generated in the object from the captured image for each of a plurality of regions divided by the set mesh;
Moving the set mesh position;
Diagnosing the presence or absence of cracks from the captured image for each of a plurality of areas divided by the mesh after the movement;
A program for running