JP7812245B2 - Concrete damage determination system and concrete damage determination program - Google Patents

Description

The present invention relates to a concrete damage detection system and a concrete damage detection program for detecting damaged areas in concrete structures.

Conventionally, for example, by using a development drawing in which each surface of a structure is reflected on a single plane, it has been possible to easily determine the volume of damaged areas such as chipped parts of a structure. For this reason, technology that reflects multiple surfaces on a single plane, such as that shown in Patent Document 1, has attracted attention.

Patent Document 1 discloses a synthesis processing device that acquires supplementary synthesis information for synthesizing a first wide-area image corresponding to the wide-area inspection range of a structure based on multiple segmented images captured by an imaging device after dividing the structure to be inspected, reduces the image size of the segmented images, synthesizes the reduced segmented images based on the supplementary synthesis information, generates a second wide-area image, and outputs the multiple segmented images with the supplementary synthesis information added. In this way, Patent Document 1 discloses a synthesis processing device that is capable of efficiently synthesizing segmented images.

Republished Patent 2020/054584

However, with the technology disclosed in Patent Document 1, when damaged areas of a structure extend across multiple surfaces, if an image of a surface containing the damaged areas is analyzed using semantic segmentation or the like, there is a possibility that the damaged areas near the boundaries between surfaces may not be analyzed correctly when the damaged areas are reflected in the unfolded view. This may result in the damaged areas not being reflected in areas of the unfolded view where they would normally be reflected. For this reason, with the technology disclosed in Patent Document 1, there is a problem in that when damaged areas extend across two or more surfaces, it is difficult to accurately reflect the damaged areas in the unfolded view.

The present invention was devised in consideration of the above-mentioned problems, and its purpose is to provide a concrete damage detection system and a concrete damage detection program that can accurately reflect damage in an unfolded view even when the damage extends over two or more surfaces.

The concrete damage determination system of the first invention is characterized by comprising an acquisition means for acquiring an unfolded diagram including two or more surfaces of a structure and damaged areas on each of the two or more surfaces, and a determination means for determining that the damage line is a damaged area based on a damage line that connects two or more damaged areas on different surfaces of the unfolded diagram acquired by the acquisition means and intersects with the boundaries of the two or more surfaces of the unfolded diagram .

The concrete damage determination system of the second invention is characterized in that, in the first invention, the determination means determines that the damage line is a damaged area based on the length of the damage line that intersects with the boundaries of two or more faces of the unfolded view.

The concrete damage determination system of the third invention is characterized in that, in the first or second invention, it further comprises a division means for dividing the damaged area based on the positions of the pixels that constitute the damaged area determined by the determination means, and an area calculation means for calculating the area of each damaged area divided by the division means.

The concrete damage determination program of the fourth invention is characterized in that it has a computer execute an acquisition step of acquiring an unfolded diagram including two or more surfaces of a structure and damaged areas on each of the two or more surfaces, and a determination step of determining that a damage line is a damaged area based on a damage line that connects two or more damaged areas on different surfaces of the unfolded diagram acquired by the acquisition step and intersects with the boundaries of the two or more surfaces of the unfolded diagram.

According to the first to fourth inventions, a damaged area is newly determined based on the distance between two or more damaged areas on different surfaces of the developed view. This makes it possible to accurately reflect the damaged area in the developed view even when the damaged area continues across two or more surfaces. Furthermore, a damaged area is newly determined based on a damage line connecting two or more damaged areas on different surfaces of the developed view. This makes it possible to calculate the distance between two or more damaged areas from the length of the damage line, making it possible to accurately reflect the damaged area in the developed view.

In particular, according to the second aspect of the present invention, a damage line is determined as a damaged area based on the length of the damage line that intersects with the boundaries of two or more faces in the developed view. As a result, for example, a damage line that intersects with the boundaries of two or more faces in the developed view is a line that connects damaged areas on different faces, and therefore, the length of this damage line can be used to calculate the distance between damaged areas on different faces.

In particular, according to the third aspect of the present invention, the area of each divided damaged portion is calculated, which makes it possible to automatically calculate the area of the damaged portion.

FIG. 1 is a block diagram showing the overall configuration of a concrete damaged portion determination system according to this embodiment. FIG. 2( a ) is a schematic diagram showing an example of the configuration of a concrete damage determination system in this embodiment, and FIG. 2( b ) is a schematic diagram showing an example of the function of the concrete damage determination system in this embodiment. FIG. 3 is a flowchart showing the processing operations of the concrete damage determination system according to this embodiment. FIG. 4A is a schematic diagram showing an example of a structure in this embodiment, and FIG. 4B is an example of a development view in this embodiment. FIG. 5 is an example of a development that reflects the damage line. FIG. 6 is an example of a development that reflects the division of the damaged portion. FIG. 7 is an example of a development divided into grids.

The concrete damage detection system to which the present invention is applied will be described in detail below with reference to the drawings.

Figure 1 is a block diagram showing the overall configuration of a concrete damage assessment system 100 to which the present invention is applied. The concrete damage assessment system 100 assesses damaged areas 21 in a structure 20. The concrete damage assessment system 100 includes an imaging device 2 and an assessment device 1 connected to the imaging device 2.

The structure 20 is, for example, a bridge, viaduct, building, or part of any of these made of concrete. The structure 20 is a three-dimensional structure having two or more surfaces. The structure 20 is a three-dimensional structure having, for example, surfaces 20a, 20b, and 20c. The structure 20 may also include a damaged portion 21 indicating damage such as chipping that is connected to two or more surfaces. The structure 20 includes, for example, a damaged portion 21 that is connected from surface 20a to surface 20c.

The imaging device 2 is a device that captures images of the surface of the structure 20. The imaging device 2 outputs the captured image data to the determination device 1. The imaging device 2 may also be connected to the determination device 1 via a wireless communication network such as the Internet (not shown).

The determination device 1 is composed of an electronic device such as a PC (personal computer), smartphone, tablet device, or wearable device. This determination device 1 determines damaged areas 21 in a structure 20 based on image data output from an imaging device 2.

As shown in FIG. 2(a), the determination device 1 includes a housing 10, a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a storage unit 104, and I/Fs 105 to 107. The CPU 101, ROM 102, RAM 103, storage unit 104, and I/Fs 105 to 107 are connected by an internal bus 110.

The CPU 101 controls the entire determination device 1. The ROM 102 stores the operation code of the CPU 101. The RAM 103 is a working area used when the CPU 101 is operating. The storage unit 104 stores various information such as image data and development drawings. The storage unit 104 may be, for example, a data storage device such as an HDD (Hard Disk Drive), SSD (Solid State Drive), SD card, or miniSD card. Note that the determination device 1 may also have a GPU (Graphics Processing Unit), not shown, for example.

I/F 105 is an interface for sending and receiving various information to and from the imaging device 2, etc. I/F 106 is an interface for sending and receiving information to and from the input unit 108. The input unit 108 may be, for example, a keyboard, and a user using the determination device 1 inputs various information or control commands for the determination device 1 via the input unit 108. I/F 107 is an interface for sending and receiving various information to and from the display unit 109. The display unit 109 outputs various information such as the determination results stored in the storage unit 104, or the processing status of the determination device 1. A display is used as the display unit 109, and may be, for example, a touch panel type.

The storage unit 104 stores, for example, image data and development diagrams acquired from the imaging device 2, as well as algorithms used to assess damage.

The display unit 109 displays various information. For example, the display unit 109 displays the judgment results, etc.

Figure 2(b) is a schematic diagram showing an example of the functions of the determination device 1. The determination device 1 includes an acquisition unit 11, an expansion unit 12, a division unit 13, a memory unit 14, an output unit 15, a calculation unit 16, and a reflection unit 17. Note that the acquisition unit 11, expansion unit 12, classification unit 13, memory unit 14, output unit 15, calculation unit 16, and reflection unit 17 shown in Figure 2(b) are realized by the CPU 101 using RAM 103 as a working area to execute a program stored in the storage unit 104, etc., and may be controlled by artificial intelligence, for example.

The acquisition unit 11 acquires multiple image data including each surface of the structure 20 that is the target of damage assessment. The acquisition unit 11 acquires image data from an imaging device 2, etc., or may acquire image data from an imaging device 2 built into the assessment device 1, for example. The frequency and period at which the acquisition unit 11 acquires various types of information are arbitrary.

The unfolding unit 12 generates a development in which two or more surfaces of a structure are reflected on a single plane based on the image data acquired by the acquisition unit 11. The unfolding unit 12 generates a development from multiple image data including each surface of the structure acquired by the acquisition unit 11, for example.

The reflection unit 17 reflects the damaged area in the unfolded view generated by the unfolding unit 12. The reflection unit 17 reflects the damaged area in the unfolded view generated by the unfolding unit 12, for example, based on the image data acquired by the acquisition unit 11.

The division unit 13 clusters and divides the damaged areas in the development drawing in which the damaged areas have been reflected by the reflection unit 17.

The calculation unit 16 calculates the area of each damaged area on the development divided by the division unit 13.

The memory unit 14 retrieves various pieces of information stored in the storage unit 104 as needed. The memory unit 14 stores various pieces of information acquired or output by the acquisition unit 11, expansion unit 12, division unit 13, calculation unit 16, and reflection unit 17 in the storage unit 104.

The output unit 15 outputs various information. The output unit 15 transmits various information to the display unit 109 via the I/F 107.

The operation of the concrete damage detection system 100 configured as described above will now be explained.

In the concrete damage determination system 100, damaged areas on each of two or more surfaces of a structure are reflected in a development in which two or more surfaces are reflected on a single plane, damage lines connecting the reflected damaged areas are reflected in the development, and new damaged areas are reflected in the development based on the reflected damage lines. The processing operation flow of this concrete damage determination system 100 is shown in Figure 3. Detailed processing at each step in Figure 3 will be described below.

Figure 4(a) is a schematic diagram showing an example of a structure 20 in this embodiment, and Figure 4(b) is an example of an unfolded view 30 in this embodiment. First, in step S11, the concrete damage determination system 100 acquires the unfolded view 30. In such a case, the acquisition unit 11 may acquire image data captured by the imaging device 2, and the unfolded view 30 may be acquired by the expansion unit 12 generating the unfolded view 30 based on the image data acquired by the acquisition unit 11. The image data of the structure 20 captured by the imaging device 2 preferably uses, for example, multiple image data captured from the perpendicular direction of each surface (20a, 20b, 20c) of the structure 20, but this is not limited to this. The expansion unit 12 generates the unfolded view 30 from the images including each surface acquired by the acquisition unit 11.

The unfolded view 30 is a plan view including two or more surfaces (20a, 20b, 20c) of the structure 20 and the damaged areas 21 on each of the two or more surfaces (20a, 20b, 20c). The unfolded view 30 is a view in which two or more surfaces (20a, 20b, 20c) of the structure 20 are connected and unfolded onto a single plane. In this case, surface 30a of the unfolded view 30 represents surface 20a of the structure 20, surface 30b of the unfolded view 30 represents surface 20b of the structure 20, and surface 30c of the unfolded view 30 represents surface 20c of the structure 20. It is preferable that the unfolded view 30 reflects image data captured from each perpendicular direction of the surfaces (20a, 20b, 20c) of the building 20, but this is not a limitation.

In addition, in step S11, the concrete damage determination system 100 may acquire, for example, image data captured in advance and stored on a server, and a pre-stored unfolded view 30, via the Internet, etc.

Next, in step S12, the damaged areas 21 are reflected in the unfolded view 30 acquired in step S11. In this case, the reflection unit 17 may perform image analysis of the image data acquired by the acquisition unit 11, for example, using semantic segmentation, extract the damaged areas 21 on each surface (20a, 20c) of the structure 20 from the image data, and reflect the damaged areas 31 in the unfolded view 30.

Semantic segmentation is a technique for distinguishing between different objects contained in an image and identifying the category and class to which each object belongs. Semantic segmentation involves processing an image in which each pixel in the image is assigned a category and class, which may refer to an object, person, etc. Semantic segmentation may also be performed using, for example, a convolutional neural network (CNN). In such cases, images of the damaged area 21 may be acquired in advance and used as training data.

In addition, at this time, the damage to corner 22 caused by surfaces 20a and 20c in Figure 4(a) is not reflected in damaged area 31 in the unfolded view 30, and a missing area 32 is generated.

Next, in step S13, as shown in FIG. 5, the reflection unit 17 reflects the damage line 33 on, for example, the unfolded view 30. The damage line 33 is a line reflected on the unfolded view 30 to determine the missing portion 32 generated in step S12. The damage line 33 is, for example, a straight line extending from the damaged portion 31 in any direction on the unfolded view 30. The damage line 33 is preferably, for example, a straight line reflected starting from the damaged portion 31 and aligned parallel to each other pixel by pixel in the x and y directions that intersect perpendicularly on the unfolded view 30, but this is not limited thereto. The damage line 33 is, for example, a line connecting two or more damaged portions 31 on different faces. The damage line 33 may be, for example, a line 33a connecting the damaged portion 31a on face 30a and the damaged portion 31c on face 30c.

Next, in step S13, the reflection unit 17 newly determines the damaged areas 31 based on the distance between two or more damaged areas 31 on different faces of the unfolded drawing 30. The reflection unit 17 further reflects the damaged areas 31 on the unfolded drawing 30, for example, based on the damage lines 33 reflected in step S13. In such a case, the reflection unit 17 further reflects the damaged areas 31 on the unfolded drawing 30 based on the length of the damage lines 33 for the two or more damaged areas 31 on the unfolded drawing 30 and the number of faces reflected in the unfolded drawing 30 that intersect with the damage lines 33. For example, the reflection unit 17 further reflects the damaged areas 31 on the unfolded drawing 30 based on damage lines 33 that intersect with two or more faces reflected in the unfolded drawing 30 and have a length less than a reference value. Specifically, if the length of damage line 33a connecting damaged area 31a on surface 30a of developed drawing 30 and damaged area 31c on surface 30c of developed drawing 30 is shorter than a reference value, the portion of developed drawing 30 where damage line 33a is reflected is considered to be a hole 32, and damaged area 31 is further reflected in the portion of developed drawing 30 where damage line 33a is reflected. The reference value may be, for example, twice the depth of the chipping of damaged area 21. In this way, by reflecting damaged area 31 on damage line 33, for example, it is possible to accurately reflect damaged area 31 in developed drawing 30 even if the chipped portion continues to the side.

Furthermore, in step S13, the reflection unit 17 may newly reflect the damaged area 31 based on the length of the damage line 33 that intersects with the boundary 34 of two or more faces in the unfolded view 30. In such a case, the reflection unit 17 may newly determine the damaged area 31 based on, for example, the damage line 33a that intersects with the boundary 34 between faces 30a and 30c in the unfolded view 30. For example, if a damage line 33 whose length is equal to or less than a reference value intersects with two of the classified faces (30a, 30b, 30c), the reflection unit 17 may newly determine this damage line 33 as the damaged area 31.

Next, in step S14, as shown in FIG. 6, the division unit 13 divides the damaged area 31 into different damaged areas 31 according to the positions of the pixels that make up the damaged area 31 in the development view 30 to which the new damaged area 31 has been reflected in step S13. For example, as shown in FIG. 6, the damaged area 31 is divided into different damaged areas 31a and 31b. In this case, the division unit 13 divides the damaged area 31 using, for example, the k-means algorithm, so that the sum of squares of the distances between the center of gravity of the damaged area 31 and the positions of the pixels that make up the damaged area 31 is minimized for each divided damaged area 31.

The k-means algorithm first randomly selects K (where K is an integer greater than 1) objects (first feature values) as initial cluster centers from the first feature values of multiple pixels within the target region. The number of selected cluster centers may be the same as the number of clusters set in advance. Next, the distance between each object and the initial cluster centers is calculated, and each object is assigned to the cluster center closest to it. The cluster centers and their assigned objects represent a single cluster. After all objects have been assigned, the cluster centers of each cluster are recalculated based on the objects currently in the cluster. This process is repeated until a certain termination condition is met. The termination condition may be that no (or a minimum number of) objects are reassigned to a different cluster, and no (or a minimum number of) cluster centers are re-assigned. Using the above method, clustering of multiple pixels is completed, resulting in the acquisition of multiple clusters equal to the set number of clusters. After the clustering process using the k-means algorithm, multiple clusters may be acquired and their cluster centers may be determined.

Next, in step S15, the calculation unit 16 calculates the area of each damaged area 31 divided in step S14. In this case, the calculation unit 16 may reflect a grid such as that shown in FIG. 7 in the developed view 30 and calculate the area according to the number of grids representing the damaged area 31. The grids may be straight lines extending in both the x and y directions, for example. The area per grid may also be set to any value.

By performing each of the steps described above, the operation of the concrete damage assessment system 100 is completed. This makes it possible to automatically calculate the area of the damaged area 31, which can be used to calculate the amount of cross-section repair material to be used, etc.

1: Determination device 2: Imaging device 10: Housing 11: Acquisition unit 12: Expansion unit 13: Division unit 14: Storage unit 15: Output unit 16: Calculation unit 17: Reflection unit 20: Structure 21: Damaged part 22: Corner 30: Expansion diagram 31: Damaged part 32: Holes 33: Damage line 34: Boundary 100: Concrete damaged part determination system 101: CPU
102: ROM
103: RAM
104: Storage section 105: I/F
106: I/F
107: I/F
108: Input unit 109: Display unit 110: Internal bus

Claims (4)

an acquisition means for acquiring a development diagram including two or more surfaces of a structure and damaged portions of each of the two or more surfaces;
and a determination means for determining , based on a damage line that connects two or more damaged areas on different faces of the development drawing acquired by the acquisition means and intersects with a boundary between two or more faces of the development drawing, that the damage line is a damaged area . The concrete damaged portion determination system according to claim 1, wherein the determination means determines the damage line as a damaged portion based on a length of the damage line intersecting with boundaries of two or more faces of the developed view. a dividing means for dividing the damaged area based on the positions of pixels constituting the damaged area determined by the determining means;
3. The system for determining damaged portions of concrete according to claim 1, further comprising: area calculation means for calculating an area of each damaged portion divided by the dividing means. an acquisition step of acquiring a development diagram including two or more surfaces of the structure and damaged portions of each of the two or more surfaces;
and a determining step of determining , based on the damage line that connects two or more damaged areas on different faces of the developed drawing acquired by the acquiring step and intersects with the boundary of the two or more faces of the developed drawing, that the damage line is a damaged area .