Method for Detecting Positions and Radii of Multiple Holes

Abstract

To provide a technique capable of accurately detecting a position and a radius of a plurality of holes drilled on a wall surface of a concrete structure.SOLUTION: Two-dimensional images are acquired from different directions with respect to a drilled wall surface. The acquired two-dimensional image is shape-restored to three-dimensional point group data. A depth point deeper than the wall surface by a predetermined dimension is extracted based on the three-dimensional point group data. A plurality of group point groups are acquired formed by a set of depth points in a horizontal and a vertical directions of the wall surface. With respect to each group point group, the group point group is assumed to form a circle and its center coordinates and a radius are calculated. Thus, it is possible to accurately detect the position and radius of a plurality of the drilled holes.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to a method for detecting the positions and radii of a plurality of holes, and particularly to a method for detecting the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure using a drilling device.

Background Art

[0002] As one of the reinforcement (seismic reinforcement) methods for existing concrete structures such as concrete structures in contact with the ground above ground, underground, semi-underground, etc., and concrete structures constructed above ground close to railways, roads, etc., there is a post-construction shear reinforcement method.

[0003] The post-construction shear reinforcement method is a method of improving the shear strength of the structure by drilling holes in the wall surface of an existing concrete structure, filling and injecting a fixing material (mortar) into the holes, and then inserting and curing post-construction shear reinforcement bars (hereinafter referred to as "shear reinforcement bars") to integrate the shear reinforcement bars with the structure.

[0004] In such a post-construction shear reinforcement method, since the number of holes for inserting shear reinforcement bars is extremely large (reaching several thousand or more), it is required to improve the efficiency of measuring the drilled shape (drilling position).

[0005] Here, Patent Document 1 (Japanese Patent Application Laid-Open No. 2002-288678) discloses a technique of photographing a drilled wall surface and obtaining the hole position from the brightness difference between the hole and the wall surface. Further, Patent Document 2 (Japanese Patent Application Laid-Open No. 2014-163898) discloses a technique of obtaining the hole position from the surface shape obtained by scanning a drilled wall surface with laser light.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Patent Document 2

SUMMARY OF THE INVENTION

PROBLEMS TO BE SOLVED BY THE INVENTION

[0007] However, in the technique described in Patent Document 1, when the concrete is wet and discolored black or the holes are underground, the brightness difference between the holes and the wall surface becomes unclear during photography, making it difficult to accurately identify the drilling positions.

[0008] Also, in the technique described in Patent Document 2, at the start of drilling, due to vibrations of the drilling device or the like, the tip shakes finely or the concrete at the hole periphery is damaged, resulting in a detected hole diameter larger than the actual hole diameter (the internal hole diameter).

[0009] The present invention has been made in view of the above technical background, and an object thereof is to provide a technique capable of accurately detecting the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure.

MEANS FOR SOLVING THE PROBLEMS

[0010] To solve the above problems, a method for detecting the positions and radii of a plurality of holes according to the invention described in claim 1 is a method for detecting the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure using a drilling device, comprising obtaining two-dimensional images from different directions with respect to the wall surface that has been drilled, By sequentially executing a first step of extracting feature points in the two-dimensional image, a second step of associating the extracted feature points between two-dimensional images from different directions, and a third step of restoring a three-dimensional shape using the coordinate values of the associated feature points restoring the shape of the two-dimensional images to three-dimensional point cloud data, and based on the three-dimensional point cloud data , a point that is deeper than the depth of the defect at the hole periphery when the wall surface is drilled and is located in the vicinity of the wall surface extracting depth points, obtaining a plurality of group point clouds formed by the set of depth points in the horizontal and vertical directions of the wall surface, and calculating the center coordinates and radius of a circle assuming that each group point cloud forms a circle.

[0013] Claim 2 The method for detecting the positions and radii of a plurality of holes according to the invention described in Claim 1In the described invention, the group of depth points is formed by a set of a reference depth point that is any one of the depth points and depth points within a predetermined range from the reference depth point, which is characterized in that.

[0014] Claim 3 The method for detecting the positions and radii of a plurality of holes according to the invention described in is the above Claim 1 or 2 In the described invention, the two-dimensional image is obtained by flying a drone equipped with two-dimensional image acquisition means along the wall surface, running a traveling device equipped with two-dimensional image acquisition means along the wall surface, or manually by the two-dimensional image acquisition means, which is characterized in that.

[0015] Claim 4 The method for detecting the positions and radii of a plurality of holes according to the invention described in is the above Claim 3 In the described invention, the two-dimensional image acquisition means is a 2D camera, which is characterized in that.

[0016] The method for detecting the positions and radii of a plurality of holes according to the invention described in claim 7 is in the invention described in any one of claims 1 to 6 above, wherein the center coordinates and radius of the circle are calculated by the least squares method, the Hough transform or RANSAC, which is characterized in that.

Advantages of the Invention

[0017] According to the present invention, two-dimensional images are obtained from different directions for a wall surface with drilled holes, the obtained two-dimensional images are shape-restored into three-dimensional point cloud data, depth points that are a predetermined dimension deeper than the wall surface are extracted based on the three-dimensional point cloud data, a plurality of group point clouds formed by sets of depth points in the horizontal and vertical directions of the wall surface are obtained, and for each group point cloud, the center coordinates and radius are calculated assuming that the group point cloud forms a circle. Therefore, it becomes possible to accurately detect the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure.

Brief Description of the Drawings

[0018]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Embodiments for Carrying Out the Invention

[0019] Hereinafter, embodiments as an example of the present invention will be described in detail with reference to the drawings. In the drawings for explaining the embodiments, the same reference numerals are generally given to the same components, and the repeated description thereof will be omitted.

[0020] The method for detecting the positions and radii of a plurality of holes in this embodiment is for detecting the positions and radii of a plurality of holes drilled in the post-construction shear reinforcement method, which is one of the methods for reinforcing (seismic reinforcement) existing concrete structures, and for measuring the drilled shape (drilling position).

[0021] Here, the post-construction shear reinforcement method is, for example, a method of drilling holes in the wall surface of an existing concrete structure S in contact with the ground G as shown in FIG. 1, or an existing concrete structure S constructed on the ground close to a building such as a railway or a road, using a drilling device (not shown), inserting a post-construction shear reinforcement bar (hereinafter referred to as "shear reinforcement bar") R into the drilled hole H, filling and injecting a fixing material (mortar) M and curing it to integrate the shear reinforcement bar R and the structure S, thereby improving the shear resistance of the structure S.

[0022] As the shear reinforcement bar R, for example, a bar in which one side of a generally used reinforcing bar R1 is threaded, obliquely cut, and a hexagonal nut R2 is attached as a fixing body at the tip is applicable.

[0023] An example of a plurality of holes drilled in the wall surface of the structure S is shown in FIG. 2. In this embodiment, a case where holes with a diameter of 40 mm are drilled at intervals of 510 mm (hole center intervals) in the horizontal and vertical directions is shown. The drilling depth is 500 mm. However, in the present invention, the hole diameter, hole interval, and drilling depth are not limited to these values, but are set to necessary values according to the state of the structure S and the like.

[0024] Now, if these multiple holes are drilled in the wall surface of the structure S, prior to the injection of the fixing material, measurement of the drilled shape, that is, detection of the positions and radii of the drilled multiple holes is performed to confirm whether holes of appropriate size are drilled at appropriate positions.

[0025] Here, as described above, in the technique of photographing the drilled wall surface and obtaining the hole position from the brightness difference between the hole and the wall surface, if the brightness difference between the hole and the wall surface becomes unclear during photographing, it becomes difficult to surely specify the drilling position.

[0026] Further, in the technique of obtaining the hole position from the surface shape obtained by scanning the drilled wall surface with laser light, since the tip of the drilling device shakes or the concrete at the peripheral part of the hole H is damaged at the start of drilling (see FIG. 3), it is detected as a hole diameter larger than the actual hole diameter (the internal hole diameter).

[0027] Therefore, in the present embodiment, in view of these problems, the positions and radii of a plurality of holes H drilled in the wall surface of the concrete structure S are detected by the method shown in FIG. 4. Here, FIG. 4 is a flowchart showing a method for detecting the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure according to an embodiment of the present invention.

[0028] In FIG. 4, first, a two-dimensional image of the wall surface of the drilled structure S is acquired (step S1). Specifically, a 2D camera (a camera that images an object in a 2D (Two-Dimensions: two-dimensional) manner), which is two-dimensional image acquisition means, is used to acquire two-dimensional images of the wall surface of the structure S from different directions. The imaging directions may be two different directions from each other, or may be three or more directions.

[0029] Here, as an example of the method for acquiring the two-dimensional image, for example, a method of acquiring while flying along the wall surface using an unmanned aerial vehicle (UAV) called a drone equipped with a 2D camera, a method of acquiring while running a running device equipped with a 2D camera along the wall surface, a method of manually acquiring with a 2D camera, and the like. When using a running device, the running device equipped with a 2D camera can be configured to be able to run on a rail laid parallel to the wall surface, or to be able to run automatically parallel to the wall surface by installing wheels.

[0030] In addition, the two-dimensional image acquisition means is not limited to a 2D camera. That is, any device capable of acquiring a two-dimensional image other than a 2D camera may be used.

[0031] Here, if it is possible to restore the shape from the two-dimensional image obtained by imaging the entire wall surface to three-dimensional point cloud data, the two-dimensional image may be obtained by imaging the entire wall surface from different directions. However, if the area of the entire wall surface is large and the shape cannot be restored to three-dimensional point cloud data from the two-dimensional image obtained by imaging the entire wall surface (for example, the image becomes unclear in the two-dimensional image obtained by imaging the entire wall surface), the wall surface is divided into a plurality of regions, and two-dimensional images are obtained from different directions for each region. Note that the regions to be divided may partially overlap.

[0032] Now, if a two-dimensional image of the wall surface has been obtained, the obtained two-dimensional image is subjected to shape restoration to three-dimensional point cloud data (step S2). Here, the shape restoration to three-dimensional point cloud data will be described with reference to FIG. 5. FIG. 5 is a flowchart showing a method for restoring the shape of a two-dimensional image to three-dimensional point cloud data.

[0033] In FIG. 5, first, as a first step, feature points in the obtained two-dimensional image are extracted (step S2-1). Feature points are edges, corners, etc. in the two-dimensional image. For feature points that are the same in a plurality of two-dimensional images obtained from different directions, the three-dimensional position can be obtained, and the shape can be grasped. Therefore, a feature point is a point having a feature that can be associated with the same point in a plurality of two-dimensional images.

[0034] Note that the feature points can be given manually, but this not only takes a long time for the operation but also risks overlooking some points. Therefore, SIFT (Scale-Invariant Feature Transform) is known as a technology for automatically detecting feature points. SIFT is an algorithm that detects feature points based on the luminance changes in an image (keypoint detection) and calculates feature quantities based on the luminance gradients and luminance directions around the feature points (feature quantity description). However, the feature points may be given manually as described above, or may be automatically detected by a technology other than SIFT.

[0035] If the feature points in the two-dimensional image are extracted in step S2-1, then as the second step, for the extracted feature points, association is performed between two-dimensional images from different directions (step S2-2). Association means detecting which position in another two-dimensional image obtained from a different direction corresponds to the feature points given in a certain two-dimensional image. It is possible to perform the association by manually giving the feature points, but in addition to the work itself becoming very difficult, it is also conceivable that the subjectivity of the operator will enter into the association. In contrast, if the above-described SIFT is used, it becomes possible to automatically perform the association by using the similarity of the feature quantities, so the reliability of the association is ensured.

[0036] And if the association is performed in step S2-2, then as the third step, the three-dimensional shape is restored using the coordinate values of the associated feature points (step S2-3). Specifically, using the principle of triangulation, for two-dimensional images from different directions, if straight lines are drawn respectively for the feature points that correspond to each other from the captured 2D cameras, the intersection point of the drawn straight lines becomes the restored three-dimensional point. And by performing this operation for each detected feature point, the three-dimensional shape can be restored.

[0037] In this embodiment, the horizontal direction of the wall surface is set as the x-axis (horizontal axis coordinate), the vertical direction is set as the y-axis (vertical axis coordinate), and the depth direction of the hole is set as the z-axis (vertical axis coordinate with respect to the xy plane: the depth direction of the hole is negative) (see FIGS. 3, 8, and 9).

[0038] If the two-dimensional image of the drilled wall surface is restored to a three-dimensional point cloud data by sequentially executing the first to third steps described above, return to FIG. 4 and extract a point 10 mm deep from the wall surface (depth point: point on the z-axis) based on the three-dimensional point cloud data (step S3). As described above, when extracting the depth point, if the peripheral portion of the drilled hole is missing (see FIG. 3) and the data on the surface of the wall is extracted, the missing hole shape (that is, a hole that is not the original shape) will be extracted. However, if a non-missing point (depth point) is extracted, the original shape of the hole can be extracted.

[0039] In the present embodiment, a point 10 mm deep from the wall surface is set as the depth point. However, the depth point does not have to be a point 10 mm deep from the wall surface and can be freely set as long as it is a point at a dimension deeper than the missing depth of the hole peripheral portion. However, it is desirable that the depth point be as close to the wall surface as possible as long as it satisfies the condition of being deeper than the missing depth of the hole peripheral portion. This is because if the depth point is too deep, the inner peripheral surface of the hole is steep, so that the number of feature points extracted from the two-dimensional image decreases for some holes, and sufficient three-dimensional point cloud data cannot be obtained.

[0040] FIG. 6 shows an image obtained by extracting the data of the acquired depth points. Since the three-dimensional point cloud data is restored from two-dimensional images from different directions, the angle when imaging with the 2D camera will be inclined with respect to the axial direction of the hole. Therefore, since the depth point is not the entire circumference but a part of the inner peripheral surface of the hole, as shown in the figure, the acquired image is in an arc shape.

[0041] Now, if the depth points are extracted, a plurality of group point clouds formed by the set of depth points in the horizontal and vertical directions of the wall surface are acquired (step S4). That is, for the extracted depth points, the point cloud of the depth points within the range set for the coordinates (x, y) of each point is searched, and these point clouds are acquired for each hole as a group point cloud formed by one set.

[0042] Here, the acquisition of the group point cloud will be described with reference to FIGS. 7 and 8. FIG. 7 is a flowchart showing the method for acquiring the group point cloud, and FIG. 8 is an explanatory diagram showing the setting of the range in the acquisition of the group point cloud.

[0043] In FIG. 7, first, a reference depth point (reference depth point P1 in FIG. 8) is determined (step S4-1). As shown in FIG. 8, the reference depth point P1 is the depth point of any one of the extracted depth points.

[0044] If the reference depth point P1 is determined, depth points within the range of ±50 mm from the reference depth point P1 are searched for (step S4-2). In FIG. 8, the reference depth point P1 is a point on the circumference. As described above, since the diameter of the hole in this embodiment is 40 mm, the range of ±50 mm is set including an error of 10 mm. That is, Lx and Ly in FIG. 8 are each 50 mm. As a result, depth points within the range of ±50 mm in the x-axis direction and the y-axis direction (within the range surrounded by the outer broken line in FIG. 8) centered on the reference depth point P1 are searched for. It is desirable that the search for depth points within the range be automatically performed by software.

[0045] And if depth points within the above-mentioned range are searched for, the points formed by the set of the reference depth point and the searched depth points (that is, the points shown in FIG. 8) are acquired as the group point cloud (step S4-3).

[0046] Now, if the group point cloud is acquired in this way, returning to FIG. 4, for each group point cloud, assuming that the group point cloud forms a circle, the center coordinates and radius of the circle are calculated by the least squares method (step S5). That is, for the acquired group point cloud, the circle is expressed as follows, and the center coordinates and radius of the circle are calculated by the least squares method (an algorithm for obtaining the coefficients of a function that minimizes the sum of the squares of the errors between the data points and the function).

[0047] (xi - a) 2 +(yi - b) 2 = r2

[0048] Here, xi is the x-axis coordinate value of an arbitrary point on the circumference, yi is the y-axis coordinate value of an arbitrary point on the circumference, a is the x-axis coordinate value of the center of the circle, b is the y-axis coordinate value of the center of the circle, and r is the radius of the circle.

[0049] The image data shown in FIG. 6 is read, and the group point clouds obtained for each hole, and the circles approximated by the least squares method and the values of their center coordinates are shown in FIG. 9. Also, the values of the center coordinates and radii of the circles in FIG. 9 are shown in FIG. 10.

[0050] In FIG. 9, the symbols 1 to 4 indicate the circle numbers. Also, separately from the circle numbers, the drilling numbers c, a, d, b corresponding to the respective circle numbers are attached. In these drilling numbers c, a, d, b, the thick lines are the group point clouds, and the thin lines are the lines that supplement the thick lines to draw the circles obtained by the least squares method. Note that the numbers to the right of the drilling numbers c, a, d, b are the center coordinates (x-axis coordinate value, y-axis coordinate value) of the circles.

[0051] For the circles with circle numbers 1 to 4 (drilling numbers c, a, d, b), the radii and center coordinates (x-axis coordinate value, y-axis coordinate value) are shown in FIG. 11.

[0052] Also, for the circles corresponding to the drilled holes, the distances between the center coordinates in adjacent circles are shown in FIG. 12. It can be seen from FIG. 12 that each of the drilled holes is formed at intervals of approximately 510 mm (hole center intervals) in the substantially horizontal and vertical directions.

[0053] As described above, according to the present embodiment, two-dimensional images are acquired from different directions for the wall surface with drilled holes, the acquired two-dimensional images are shape-restored into three-dimensional point cloud data, depth points that are a predetermined dimension (10 mm in the present embodiment) deeper than the wall surface are extracted based on the three-dimensional point cloud data, and a plurality of group point clouds formed by sets of depth points are acquired in the horizontal and vertical directions of the wall surface. For each group point cloud, the center coordinates and radius are calculated on the assumption that the group point cloud forms a circle. Therefore, it is possible to accurately detect the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure.

[0054] Therefore, when determining the hole position from the brightness difference between the photographed wall surface and the hole, it is difficult to identify the hole position when the brightness difference is unclear, or when determining the hole position from the surface shape obtained by scanning the wall surface with a laser beam, when the concrete at the hole periphery is missing, the detected hole diameter will not be larger than the actual one.

[0055] The invention made by the present inventor has been specifically described based on the embodiments. However, the embodiments disclosed in this specification are illustrative in all respects and are not limited to the disclosed technology. That is, the technical scope of the present invention should not be construed restrictively based on the description in the above embodiments, but should be construed according to the description in the claims. All changes are included as long as they are equivalent to the technology described in the claims and do not deviate from the gist of the claims.

[0056] For example, in the present embodiment, the least squares method is used to calculate the center coordinates and radius of the circle, but it is not limited to this. Various known methods such as the Hough transform (an algorithm that extracts data points many times, obtains a function for them, aggregates the coefficients, and adopts the combination of coefficients that is most suitable as a candidate) and RANSAC (RANdom SAmple Consensus: an algorithm that samples points randomly many times and searches for the function with the least error) can be applied.

Industrial Applicability

[0057] In the above description, the present invention has been shown to be used for detecting the position and radius of holes drilled for inserting shear reinforcement bars into an existing concrete structure in the post-construction shear reinforcement method. However, it is not limited to the holes for inserting shear reinforcement bars. For example, during the RC additional casting method for seismic reinforcement, it can be widely applied to detecting the position and radius of holes (anchor holes) for post-construction anchors (anti-slip anchors) constructed to enhance the integrity of the new and old concrete surfaces, such as for various purposes like detecting the position and radius of holes drilled in concrete structures.

Explanation of Symbols

[0058] H Hole R Shear Reinforcement Bar R1 Bar R2 Hexagonal Nut S Structure

Claims

1. A method for detecting the positions and radii of a plurality of holes drilled in the wall surface of a concrete structure using a hole drilling device, comprising: acquiring two-dimensional images from different directions with respect to the wall surface after drilling; a first step of extracting feature points in the two-dimensional image, a second step of performing association between the two-dimensional images from different directions for the extracted feature points, and a third step of restoring a three-dimensional shape using the coordinate values of the associated feature points, thereby restoring the shape of the two-dimensional image to three-dimensional point cloud data; extracting depth points that are points deeper than the defect depth of the hole periphery when drilling the wall surface and are located near the wall surface based on the three-dimensional point cloud data; acquiring a plurality of group point clouds formed by the set of depth points in the horizontal and vertical directions of the wall surface; calculating the center coordinates and radius of a circle assuming that each group point cloud forms a circle; A method for detecting the positions and radii of a plurality of holes, characterized by the above.

2. The group point cloud is formed by a set of a reference depth point which is any one of the depth points and the depth points within a predetermined range from the reference depth point. A method for detecting the positions and radii of a plurality of holes according to Claim 1, characterized by the above.

3. The two-dimensional image is acquired by flying a drone equipped with two-dimensional image acquisition means along the wall surface, running a traveling device equipped with two-dimensional image acquisition means along the wall surface, or manually by the two-dimensional image acquisition means. A method for detecting the positions and radii of a plurality of holes according to Claim 1 or 2, characterized by the above.

4. The two-dimensional image acquisition means is a 2D camera. A method for detecting the positions and radii of a plurality of holes according to Claim 3, characterized by the above.

5. The center coordinates and radius of the circle are calculated by the least squares method, Hough transform or RANSAC. A method for detecting the positions and radii of a plurality of holes according to any one of Claims 1 to 4, characterized by the above.

Images