JP7803487B2 - Rod-shaped member construction support method, rod-shaped member construction support system, and program - Google Patents

Description

Applicable under Article 30, Paragraph 2 of the Patent Act. Published on March 1, 2022 on the website: http://www.tc-iaip.org/dia/2022/

The present invention relates to a method for supporting the construction of rod-shaped members, a system for supporting the construction of rod-shaped members, and a program. Specifically, it relates to technology that enables the efficient and accurate identification of the position and angle of insertion holes for rod-shaped members during tunnel construction.

The NATM (New Austrian Tunneling Method) construction method, which utilizes the natural support function of the ground, is expanding its scope of application beyond traditional mountain tunnels by adopting new auxiliary construction methods as appropriate.
In this NATM method, concrete for the primary lining is sprayed onto the wall surface exposed by excavating and blasting the natural ground, followed by the installation of rock bolts and the pouring of concrete for the secondary lining. By carrying out this construction work, the natural ground and the sprayed concrete are integrated, and the holding power of the natural ground increases the tunnel strength.

Patent document 1 (Patent Document 1) is a conventional technology that aims to improve the efficiency of tunnel construction, including the installation of rock bolts. It is proposed to ensure safety by spraying concrete onto the tunnel face, while allowing construction workers to visually check the ground conditions on the surface of the sprayed concrete as they carry out their work.

This technology involves a method including a face photographing process in which an area including the entire first face, which is the exposed surface of the natural ground at the tunnel face, is photographed with a digital camera; a face information image data generation process in which data of an image portion of the first face obtained by photographing with the digital camera is extracted as face image data from the photographed image data of the area including the entire first face, and face information image data is generated based on the face image data, including at least a face image that is an actual image of the entire first face; a face spraying process in which concrete is sprayed onto the first face; and a face information image projection process in which at least the face image, which is an image indicated by the face information image data, is projected onto the second face, which is the surface of the concrete after spraying, at the actual size of the first face, in accordance with the shape and size of the second face.

Japanese Patent Application Laid-Open No. 2022-1732

The above-mentioned rock bolt installation involves inserting the rock bolt into a hole (which has already been filled with mortar) drilled into the wall surface through shotcrete. Inserting such a rock bolt is divided into three steps: finding the hole in the uneven wall surface, aligning the tip of the rock bolt with the hole position, and inserting the entire rock bolt into the hole.
Currently, these tasks are carried out manually by workers who go to the area around the hole, but this requires inserting long, heavy rock bolts into long, narrow holes filled with highly viscous mortar, which makes the work less easy and efficient.

On the other hand, various research is being conducted into methods for mechanizing the entire process of rock bolting - drilling holes in the wall, injecting mortar into the holes, and driving rock bolts - using a single arm. With these methods, for example, multiple rock drills can be used to drill holes and drive rock bolts simultaneously, thereby improving construction efficiency.

However, it has not yet been possible to efficiently and automatically identify the correct hole position and insertion angle on the wall surface, and automating the alignment of the rock bolt tip with the hole remains difficult. As a result, there have been cases where driving rock bolts while the hole position or insertion angle is not aligned with the correct hole position, leading to deformation or damage of the rock bolt inside the hole.

The present invention therefore aims to provide technology that makes it possible to efficiently and accurately determine the position and angle of insertion holes for rod-shaped members during tunnel construction.

The method for supporting installation of a rod-shaped member of the present invention, which solves the above-mentioned problem, is characterized by including the steps of: acquiring a photographic image of a wall surface into which a rod-shaped member is to be inserted; detecting an insertion hole already formed in the wall surface as an elliptical area based on brightness information indicated by the photographic image; and estimating the position and angle of the insertion hole on the wall surface based on the diameter and brightness information of the elliptical area indicated by the photographic image corresponding to the elliptical area.
Unlike conventional methods using visual servoing technology, which tend to have high implementation costs, this method uses a monocular camera that is easy to implement at tunnel construction sites in terms of both technology and cost, making it possible to efficiently and accurately identify the position and angle of holes for rock bolt installation, charging, etc. It is also expected to reduce the possibility of damage or deformation of rock bolts, etc., when they are inserted into holes.

In addition, the step of detecting the elliptical regions that are the insertion holes in the rod-shaped member installation support method of the present invention may also include the steps of detecting the contours of each region in the captured image, approximating the detected contours to an elliptical shape, and converting each region into an elliptical region, and identifying, based on the brightness information of each elliptical region, elliptical regions that are darker than other elliptical regions or a predetermined standard as the insertion holes.

This makes it possible to accurately identify insertion holes for rock bolts, etc., even if numerous elliptical areas are included in the captured image of the tunnel's wall surface. This in turn makes it possible to more efficiently and accurately identify the position and angle of insertion holes for rod-shaped members during tunnel construction. Note that while this image is essentially taken under conditions where only one insertion hole is the subject of the image capture, there is no limit to the number of insertion holes that can be identified, as any elliptical area darker than a predetermined standard can be identified as an insertion hole. Specifically, the above conditions correspond to an environment in which the capture conditions, such as the placement of the image capture device relative to the tunnel's wall surface, lens direction, focal length, and angle of view, have been adjusted in advance.

Furthermore, in the step of estimating the position and angle of the insertion hole in the rod-shaped member installation support method of the present invention, the position of the insertion hole may be identified based on the diameter of the elliptical region and the known diameter of the rod-shaped member, and the insertion hole may be identified as extending at an angle at which the brightness becomes darker based on the change in brightness in the minor axis direction of the elliptical region.

This allows for accurate estimation of the hole position by appropriately taking into account the similarity relationship between the shape in the image (the elliptical region corresponding to the hole) as viewed from the coordinate system of the camera device and the shape on the actual tunnel wall surface (of the actual hole), using the known hole diameter as a key. Furthermore, the angle of the hole (from the tunnel wall surface toward the natural ground) can be efficiently determined based on the relationship between the brightness gradation in the elliptical region and the extension direction of the hole. In other words, the position and angle of the insertion hole for the rod-shaped member during tunnel construction can be determined more efficiently and accurately.

Furthermore, the rod-shaped member construction support method of the present invention may further include a step of setting information about the position and angle of the insertion hole in the rod-shaped member construction device.

In this way, by inputting information about the position and angle of the insertion hole into a construction device such as a computer jumbo, it becomes possible to insert rock bolts and other items into the holes without relying on human power. This eliminates the need for manual work in high places or dangerous locations such as above the tunnel, reducing the burden and risks on workers. It is also expected to reduce the possibility of damage or deformation of rock bolts and other items when they are inserted into holes. Ultimately, it becomes possible to more efficiently and accurately determine the position and angle of the insertion hole for rod-shaped members during tunnel construction.

The rod-shaped member installation support system of the present invention is characterized by including an imaging device that captures an image of a wall surface into which a rod-shaped member is to be inserted, and an information processing device that acquires the image captured by the imaging device and executes a process of detecting an insertion hole already formed in the wall surface as an elliptical area based on brightness information indicated by the captured image, and a process of estimating the position and angle of the insertion hole on the wall surface based on the diameter and brightness information of the elliptical area indicated in the captured image corresponding to the elliptical area.

This makes it possible to provide a system that can efficiently and accurately identify the position and angle of the insertion hole for the rod-shaped member during tunnel construction.

In the rod-shaped member installation support system of the present invention, the information processing device may perform the process of detecting the elliptical regions that are the insertion holes by detecting the contours of each region in the captured image, approximating the detected contours to an elliptical shape, and converting each region into an elliptical region, and by identifying, based on the brightness information of each elliptical region, elliptical regions that are darker than other elliptical regions or a predetermined standard as the insertion holes.

This makes it possible to provide a system that can accurately identify insertion holes for rock bolts, etc., even if numerous elliptical areas are included in the captured image of the tunnel wall surface. Ultimately, this system makes it possible to more efficiently and accurately identify the position and angle of insertion holes for rod-shaped members during tunnel construction. Note that this image is basically assumed to be obtained under conditions where only one insertion hole is the subject of the image capture; however, since any elliptical area darker than a predetermined standard can be identified as an insertion hole, there is no limit to the number of insertion holes that can be identified. Specifically, the above conditions correspond to an environment in which the capture conditions, such as the placement of the image capture device relative to the tunnel wall surface, lens direction, focal length, and angle of view, have been adjusted in advance.

In the rod-shaped member installation support system of the present invention, the information processing device, in the process of estimating the position and angle of the insertion hole, may identify the position of the insertion hole based on the diameter of the elliptical region and the known diameter of the rod-shaped member, and, based on the change in brightness in the minor axis direction of the elliptical region, identify the insertion hole as extending at an angle where the brightness becomes dark.

This provides a system that can accurately estimate hole positions by appropriately taking into account the similarity relationship between the shape in the image (the elliptical region corresponding to the hole) as seen from the coordinate system of the camera and the shape on the actual tunnel wall surface (of the actual hole), using the known hole diameter as a key. Furthermore, this system can efficiently identify the angle of the hole (from the tunnel wall surface toward the interior of the natural ground) based on the relationship between the brightness gradation in the elliptical region and the extension direction of the hole.

In addition, in the rod-shaped member construction support system of the present invention, the information processing device may further perform a process of setting information on the position and angle of the insertion hole in the rod-shaped member construction device.

In this way, by having the system input information about the position and angle of the insertion hole into construction equipment such as a computer jumbo, it becomes possible to insert rock bolts and other items into the holes without relying on human labor. This eliminates the need for manual work in high places or dangerous locations such as above the tunnel, reducing the burden and risks on workers. It is also expected to reduce the possibility of damage or deformation of rock bolts and other items when they are inserted into holes. Ultimately, it becomes possible to more efficiently and accurately determine the position and angle of the insertion hole for rod-shaped members during tunnel construction.

The program of the present invention also causes an information processing device to perform the following processes: acquire, from an imaging device, a photographed image of the wall surface into which a rod-shaped member is to be inserted; detect an insertion hole already formed in the wall surface as an elliptical area based on brightness information indicated by the photographed image; and estimate the position and angle of the insertion hole on the wall surface based on the diameter and brightness information of the elliptical area indicated by the photographed image corresponding to the elliptical area.

This makes it possible to implement a function in the information processing system that efficiently and accurately identifies the position and angle of holes that are the target for rock bolt installation or charging. Furthermore, by linking an information processing system equipped with this function with construction equipment such as a computer jumbo, it becomes possible to efficiently automate the insertion of rock bolts into holes with appropriate precision. This can also have the effect of reducing the possibility of damage or deformation of rock bolts, etc. that occurs when they are inserted into holes.

This invention makes it possible to efficiently and accurately determine the position and angle of the insertion hole for the rod-shaped member during tunnel construction.

FIG. 1 is a diagram illustrating an example of an application environment of a construction support method according to an embodiment of the present invention. 1 is a diagram illustrating an example of the overall configuration of a construction support system according to an embodiment of the present invention. FIG. 1 is a diagram illustrating an example of a hardware configuration of an information processing device according to an embodiment of the present invention. 3A and 3B are diagrams illustrating an example of a coordinate system and definitions of vectors in the present embodiment. FIG. 2 is a diagram showing an example of a flow of a construction support method according to the present embodiment. 10A and 10B are diagrams showing gradation within an insertion hole in a captured image of the present embodiment. FIG. 2 is a top view showing a shooting position in this embodiment. 10A and 10B are diagrams showing examples of insertion holes in the present embodiment. FIG. 10 is a diagram showing checkerboard photography in this embodiment. 10A and 10B are diagrams showing insertion hole detection results in the present embodiment. FIG. 10 is a diagram showing the luminance situation (condition 1) on the minor axis of the ellipse in this embodiment. FIG. 10 is a diagram showing the luminance situation (condition 2) on the minor axis of the ellipse in this embodiment.

<Overview>
First, an environment in which the rod-shaped member construction support method of this embodiment is applied and an overview of the technology will be described. As illustrated in Figure 1, a rock bolt 1, which is a rod-shaped member, is driven into an insertion hole 5 provided in a tunnel wall surface 3 in the natural ground 2 that is the target of tunnel construction. This insertion hole 5 is drilled in the natural ground 2 through shotcrete 6 applied to the tunnel wall surface 3. Furthermore, prior to driving the rock bolt, the interior of the insertion hole 5 is filled with mortar.

The driving of such a rock bolt 1 involves inserting the rock bolt 1 into the insertion hole 5 (filled with mortar) drilled as described above. This work would traditionally have been performed manually, but in an environment where the construction support method of this embodiment is applied, it can be automated with sufficient precision using construction equipment such as a computer jumbo 60.

In other words, by appropriately coordinating with the construction equipment, it is possible to automate the steps that make up the process of inserting the rock bolt 1: finding the insertion hole 5 from the uneven tunnel wall surface 3, aligning the tip of the rock bolt 1 with the opening position of the insertion hole 5, and inserting the entire rock bolt into the insertion hole 5.
<System Configuration>
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail below with reference to the accompanying drawings. Fig. 2 is a diagram showing an example of the configuration of a construction support system 100 (hereinafter the same) for a rod-shaped member 1 according to this embodiment.

The construction support system 100 for rod-shaped members 1 in this embodiment is a system that can efficiently and accurately identify the position and angle of the insertion hole 5 for a rod-shaped member, such as a rock bolt 1, during tunnel construction.

As shown in Figure 2, this construction support system 100 is composed of at least a digital camera 20 and an information processing device 50, out of the digital camera 20, an information processing device 50, and a computer jumbo 60.

The digital camera 20 is a commercially available monocular camera, and there is no need to use a special camera unit such as a depth camera. However, due to the influence of spring water and mortar near the tunnel's surrounding wall surface 3, appropriate care must be taken when installing a general digital camera 20 (which does not have strict dustproof or waterproof features).

As a result, it is necessary to take photographs from a distance of 5 to 10 m from the tunnel wall surface 3. Therefore, it is preferable to use a variable-focus lens with a variable focal length (for example, 70 to 200 mm) as the lens 21 attached to the digital camera 20, in order to be able to accommodate a variety of shooting conditions.

This digital camera 20 has a storage medium 22 that stores and holds image data, and can distribute the image data 23 held on this storage medium 22 to an external device via a near-field wireless communication means 24 such as Bluetooth (registered trademark). In this embodiment, the external device to which this image data is distributed is an information processing device 50.

Of course, the method for transmitting image data captured by the digital camera 20 to the information processing device 50 is not limited to these wireless/wired communication methods.

If it is difficult to use either wireless or wired communication methods due to factors such as the tunnel environment or the various resources available, a method may be used in which a worker removes the storage medium 22 from the digital camera 20 and inserts it into the reader 25, allowing the information processing device 50 to read the captured data. The reader 25 is connected to the information processing device 50 via an appropriate interface and is capable of reading and writing data to and from the storage medium 22.

The information processing device 50 is a general computer device, and as shown in FIG. 4, has a storage device 51, memory 53, a calculation device 54, an input device 55, an output device 56, and a communication device 57.

Of these, the storage device 51 is composed of a non-volatile storage element such as an HDD (Hard Disk Drive) or SSD (Solid State Drive), and stores the program 52 and data required for processing, as well as processing results.

The program 52 stored in the storage device 51 is a program for implementing the functions required by the information processing device 50 that constitutes the construction assistance system 100 of this embodiment. Therefore, the program 52 implements algorithms that perform various analyses using the formulas (such as formulas 1 to 10) described below, and settings processing for the computer jumbo 60.

Memory 53 is also composed of a volatile memory element such as RAM (Random Access Memory).

The calculation device 54 is a CPU (Central Processing Unit) that executes the program 52 stored in the storage device 51 by reading it into memory 53, thereby performing overall control of the device itself and carrying out various judgments, calculations, and control processes.

The input device 55 is a device such as a keyboard, microphone, or mouse that accepts key input or voice input from the user.

The output device 56 is a device such as a display or speaker that displays data processed by the computing device 54.

The communication device 57 is also assumed to be the short-range wireless communication means 24 of the digital camera 20, or an interface card that connects to the computer jumbo 60 via an appropriate network and handles communication processing.

The Computer Jumbo 60 is a construction device that uses the same arm to mechanize the drilling of insertion holes 5 in the tunnel's peripheral wall surface 3, the injection of mortar into the insertion holes 5, and the driving of rock bolts 1 into the insertion holes 5.

This computer jumbo 60 is equipped with multiple arms and mechanisms such as rock drills and rock bolt gripping units at the ends of the arms, as well as a control unit 61 that controls the mechanisms and a communications unit 62 that handles communications with external devices.

The computer jumbo 60 receives information settings from the information processing device 50 regarding the position and angle of the insertion hole 5, and the control unit 61 automatically performs operations such as moving and fixing the arm, starting/stopping the rock drill, grasping and inserting the rock bolt according to the contents of this information setting.
<Hole detection in captured images>
Based on the above, in this embodiment, assuming that the rock bolt installation work is mechanized using a computer jumbo 60, which is a machine dedicated to installing rock bolts 1, a technology is described in which the position and angle of an already drilled insertion hole 5 in the tunnel wall surface 3 is estimated by photographing with a digital camera 20 from a distance of at least a predetermined distance from the tunnel wall surface 3.

In this embodiment, driving a rock bolt can be achieved by aligning the tip of the rock bolt with the image coordinates (uc, vc) of the detected insertion hole center (hereinafter referred to as the hole center) in the camera coordinate system of the digital camera 20, and by aligning the central axis of the rock bolt 1 with the central axis of the insertion hole 5. Therefore, it is necessary to measure the image coordinates (uc, vc) of the hole center in the above-mentioned camera coordinate system and the central axis of the insertion hole 5.

Based on these assumptions, the following describes a method for detecting insertion holes 5 from images of the tunnel's peripheral wall surface 3 captured by a digital camera 20, and estimating the position and angle of the insertion holes 5 using the detected hole diameter and a known hole diameter (the specified hole diameter for the insertion holes 5).

The diameter of the drilled insertion hole 5 is assumed to be known, as it depends on the diameter of the rod used for drilling. The drilled insertion hole 5 is assumed to be perpendicular to the tunnel's peripheral wall surface 3 and to appear as a perfect circle on the tunnel's peripheral wall surface 3. In this embodiment, it is assumed that the imaging range of the digital camera 20 is set in advance so that only one insertion hole 5 is included in the captured image of the tunnel's peripheral wall surface 3. However, there is no limit to the number of insertion holes that can be included in the captured image. Any elliptical area that is darker than a predetermined standard can be identified as an insertion hole, so there is no need to set any conditions for the number of insertion holes to be identified.

The procedure for identifying the insertion hole 5 in the captured image is shown below based on the flow in Figure 5. Note that the information processing device 50 acquires captured image data of the tunnel's peripheral wall surface 3 from the digital camera 20 and stores this data in the storage device 51 or memory 53 as the processing target.

Therefore, the information processing device 50 applies an existing threshold selection method (e.g., N. Otsu, "A threshold selection method from gray-level histograms," IEEE Transactions on Systems, Man, and Cybernetics, vol. 9, no. 1, pp. 62-66, 1979) to the above-mentioned captured image to create a binary image (s1).
Next, the information processing device 50 performs a contour detection process (e.g., S. Suzuki and K. Abe, "Topological structural analysis of digitized binary images by border following," Computer Vision, Graphics, and Image Processing, vol. 30, no. 1, pp. 32-46, 1985) on the binary image obtained in s1 (s2).

The information processing device 50 also performs ellipse fitting on all of the contours detected in s2 and selects one of the resulting ellipses as the insertion hole 5 (s3).

The above-mentioned ellipse fitting is performed by the following procedure: When the coefficients of the ellipse equation are a1, a2, a3, a4, and a5, a-[a1, a2, a3, a4, a5] _T is minimized using the least squares method as shown in equation (1).
[Formula 1]
Here, _Nj is the total number of points that make up the j-th contour.

Therefore, the center point (uc, vc) of the ellipse to be found is expressed by equations (2) and (3), and the angle θ of the major axis from the u axis is expressed by equation (4).
[Formula 2]
[Formula 3]
[Formula 4]
Furthermore, the insertion hole 5 is determined from among the multiple ellipses by utilizing the fact that the inside of the insertion hole 5, i.e., the elliptical region, appears dark in the image. Therefore, the information processing device 50 calculates the average brightness value within each of all detected ellipses. Next, the information processing device 50 adopts the elliptical region with the lowest calculated average brightness as the insertion hole 5.
<Location estimation using known hole diameter>
The coordinate system assumed in this embodiment and the definitions of each vector are shown in Fig. 4. The assumed coordinate systems are the image coordinate system (u, v) and the camera coordinate system (x, y, z), as shown in Fig. 4. Furthermore, the endpoints of the major axis of the ellipse determined in s3 above, which are on the Z=S plane, are designated x1 and x2, and the endpoints on the Z=1 plane are designated r1 and r2, and the center points of the ellipse are designated Xc (on the Z=S plane) and rc (on the Z=1 plane), respectively.

Of these, the center point Xc of the ellipse is defined as the hole position vector, and vector a, which is coaxial with the central axis of the insertion hole 5 and points in the depth direction of the insertion hole 5, is defined as the direction vector of the insertion hole 5.

Here, since the diameter of the insertion hole 5 is known, the information processing device 50 calculates (s4) the hole position vector Xc of the insertion hole 5 from the relationship between the diameter of the insertion hole 5 in the image and the actual diameter of the insertion hole 5. Here, since the corresponding ellipses on the two planes are similar to each other as shown in Figure 4, the relationships of formulas (5), (6), and (7) hold.
[Formula 5]
[Formula 6]
[Formula 7]
Furthermore, from the similarity relationship described above, Zr=S can be obtained by equation (8).
[Formula 8]
Therefore, the information processing device 50 can calculate the hole position vector Xc by substituting the formula (8) into the formula (7).
<Hole angle estimation>
As already mentioned, when the insertion hole 5 drilled in the tunnel peripheral wall surface 3 is perpendicular to the tunnel peripheral wall surface 3, it forms a perfect circle on the tunnel peripheral wall surface 3. In this case, the shape of the insertion hole 5 becomes an ellipse that varies depending on the shooting angle of the digital camera 20. Therefore, the information processing device 50 estimates the angle at which the insertion hole 5 extends in the natural ground 2 based on these conditions (s5). Note that the angle at which the insertion hole 5 extends in the natural ground 2 refers to the angle between the position of the opening of the insertion hole 5 and the axis of the hole in the camera coordinate system.

The angle θ (0deg≦θ≦180deg) from the u-axis, which is the major axis of the ellipse (identified in s3 as the insertion hole 5 in the captured image), and the angle at which the ellipse becomes a circle (= appears to be a circle) due to the shape change caused by the above-mentioned shooting angle when the ellipse is rotated around the major axis is φ (-90deg≦φ≦90deg). The information processing device 50 calculates the direction vector a of the insertion hole 5 based on equation (9).
[Formula 9]
Here, R is a rotation matrix calculated from the normal and angle using Rodrigues' formula. The angle φ is calculated as shown in Equation (10) using the length a of the major axis and the length b of the minor axis of the ellipse.
[Formula 10]
Here, the angle φ can be either positive or negative depending on the direction of rotation, so it must be uniquely determined for angle estimation. Figure 6 shows an oblique view of the insertion hole 5 opened in the peripheral wall surface 3 of the tunnel.

As shown in Figure 6, when the insertion hole 5 is viewed obliquely, it appears as an ellipse whose major axis is the diameter of the hole. In this case, one side of the minor axis appears bright because the wall inside the hole is continuous with the wall outside the hole, while the other side appears dark because no light reaches the inside of the hole, creating a gradation.

Therefore, by considering the direction vector a of the insertion hole 5 to be pointing toward the darker side of the gradation on the minor axis of the ellipse, the angle φ can be uniquely determined as follows:

First, the information processing device 50 obtains the brightness information (attached to the captured image data) of the pixels on the minor axis of the ellipse identified in s3, and obtains a linear approximation equation using the least squares method. If the slope of the line indicated by the approximated equation is positive, φ is positive; conversely, if the slope of the line is negative, φ is negative.

The information processing device 50 transmits the thus obtained information on the position and angle of the insertion hole 5 in the tunnel peripheral wall surface 3 to the computer jumbo 60 via the communication device 57, and sets the information in the computer jumbo 60 (s6). The computer jumbo 60 receives this information via the communication unit 62 and passes it on to the control unit 61. The control unit 61 sets this information as control information for mechanisms such as an arm and a rock drill, and guides the tip of the rock bolt 1 to the opening of the insertion hole 5 in the tunnel peripheral wall surface 3, and then inserts it into the hole at an appropriate angle.
<Experimental conditions>
To verify the effectiveness of the construction support method of this embodiment, an experiment was conducted in a tunnel under construction. Fig. 7 shows the positional relationship between the digital camera 20 and the insertion hole 5 under each condition, and Fig. 8 shows the measured insertion hole 5. As shown in Fig. 7, images were taken with the digital camera 20 from positions of 45 degrees and 90 degrees for the insertion hole 5, which had a diameter of 47 mm, in the tunnel peripheral wall surface 3.

Furthermore, based on the assumptions already stated, the experiment was conducted under the condition that only one insertion hole 5 for the lock bolt 1 was always within the angle of view of the digital camera 20.

In order to obtain true values of the experimental results, multiple images were taken under each condition, with the origin of the checkerboard 70 aligned with the center of the insertion hole 5, as shown in Figure 9. The position and direction of the checkerboard 70 were estimated from the image capture results, and the hole position vector and the direction vector of the insertion hole 5 calculated from each captured image were averaged to obtain the true value.
<Experimental Results>
10 shows the detection of the insertion hole 5 using the method of this embodiment. After contour detection and ellipse fitting of the acquired photographed image, the ellipse with the darkest average brightness was detected as the insertion hole 5. As a result, it was confirmed that the correct insertion hole 5 could be detected under all conditions.

Figure 11 shows the brightness on the minor axis of the ellipse detected under condition 1 and its linear approximation, and Figure 12 shows the brightness on the minor axis of the ellipse detected under condition 2 and its linear approximation.

From Figures 11 and 12, it can be seen that a luminance gradation occurs on the minor axis of the ellipse detected as the insertion hole 5. Furthermore, from the slope of the straight line linearly approximated using the least squares method for the luminance, it was possible to estimate that the angle φ for condition 1 shown in Figure 11 is positive, and the angle φ for condition 2 shown in Figure 12 is negative.

Table 1 shows the error between the distance estimated using the detection result of the insertion hole 5 and the distance calculated from the checkerboard image.
[Table 1]
From this, it can be seen that of the hole position vectors estimated using the detection results of the insertion holes 5, the error in the x-component, which is considered important for driving the rock bolt 1, is within ±20 mm. On the other hand, it can be seen that the error in the y-component is larger than the target. This is thought to be due to the fact that when the insertion holes 5 were detected, the processed shape of the hole opening was not ideal, causing the outline to be detected as being large.

In addition, the error in the z component, which is not used when driving rock bolt 1, is thought to have increased because it is heavily dependent on the contour detection results.

Table 2 also shows the error between the value of the direction vector of the insertion hole 5 estimated using the detection result of the insertion hole 5 and the direction vector of the insertion hole 5 calculated from the checkerboard image.
[Table 2]
Thus, an error occurred between the estimated value and the true value of the direction vector of the insertion hole 5. The reason for this is that the measured insertion hole 5 was not perpendicular to the tunnel wall surface 3, and the shape of the insertion hole 5 on the tunnel face 5 was not a perfect circle.

As shown in the results above, although errors occurred in each component of the hole position vector and the direction vector of the insertion hole 5, we were able to successfully identify the direction of the insertion hole 5.

As described above, this embodiment makes it possible to efficiently and accurately identify the position and angle of holes into which rod-shaped elements such as rock bolts and explosive charges will be inserted during tunnel construction. Furthermore, by inputting information about the hole position and angle into a construction device such as a computer jumbo, it becomes possible to insert rock bolts and other elements into holes without relying on human labor. This eliminates the need for manual work in high places or dangerous locations, such as above tunnels, and reduces the burden and risks on workers. It is also expected to have the effect of reducing the possibility of damage or deformation of rock bolts and other elements when they are inserted into holes.

The present invention has been specifically described based on the embodiments, but it is not limited to these and various modifications are possible without departing from the spirit of the invention.

1. Rock bolt (rod-shaped member)
2. Ground 3. Tunnel wall surface (wall)
5 (Lock bolt) insertion hole 20 Digital camera (photography device)
21 Lens 22 Storage medium 23 Photography data 24 Short-distance wireless communication means 25 Reader 50 Information processing device 51 Storage device 52 Program 53 Memory 54 Arithmetic device 55 Input device 56 Output device 57 Communication device 60 Computer Jumbo (rod-shaped member construction device)
61 Control unit 62 Communication unit 100 Construction support system

Claims (9)

acquiring a photographed image of a wall surface into which a rod-shaped member is to be inserted;
detecting an insertion hole already formed in the wall surface as an elliptical region based on brightness information indicated by the captured image;
estimating a position and an angle of the insertion hole on the wall surface based on a diameter and brightness information of the elliptical region shown in the photographed image corresponding to the elliptical region;
A method for supporting construction of a rod-shaped member, comprising: In the step of detecting the elliptical region that is the insertion hole,
a step of detecting an outline of each region in the photographed image, approximating the detected outline to an elliptical shape, and converting each region into an elliptical region; and a step of identifying, among the elliptical regions, an elliptical region that is darker than another elliptical region or a predetermined reference, as the insertion hole, based on brightness information of each of the elliptical regions.
The method for supporting installation of a rod-shaped member according to claim 1 . In the step of estimating the position and angle of the insertion hole,
a position of the insertion hole is identified based on a diameter of the elliptical region and a known diameter of the rod-shaped member, and based on a transition in brightness in the minor axis direction of the elliptical region, the insertion hole is identified as extending at an angle at which the brightness becomes darker;
The method for supporting installation of a rod-shaped member according to claim 1 . Further includes a step of setting information about the position and angle of the insertion hole in the installation device for the rod-shaped member.
The method for supporting installation of a rod-shaped member according to claim 1 . an imaging device that images a wall surface into which the rod-shaped member is to be inserted;
An information processing device that acquires a photographed image obtained by photographing with the photographing device, and detects an insertion hole formed in the wall surface as an elliptical region based on brightness information indicated by the photographed image, and executes a process of estimating a position and an angle of the insertion hole on the wall surface based on a diameter and brightness information of the elliptical region indicated by the photographed image corresponding to the elliptical region;
A construction support system for rod-shaped members, comprising: The information processing device includes:
In the process of detecting the elliptical region that is the insertion hole,
The method detects the contour of each area in the captured image, approximates the detected contour to an elliptical shape, and converts each area into an elliptical area; and, based on brightness information of each of the elliptical areas, identifies an elliptical area that is darker than another elliptical area or a predetermined standard as the insertion hole.
6. The rod-shaped member construction support system according to claim 5. The information processing device includes:
In the process of estimating the position and angle of the insertion hole,
The position of the insertion hole is identified based on the diameter of the elliptical region and the known diameter of the rod-shaped member, and the insertion hole is identified as extending at an angle at which the brightness becomes dark based on the transition of brightness in the minor axis direction of the elliptical region.
6. The rod-shaped member construction support system according to claim 5. The information processing device includes:
The processing further includes setting information about the position and angle of the insertion hole in the installation device for the rod-shaped member.
6. The rod-shaped member construction support system according to claim 5. In the information processing device,
A process of acquiring, from an imaging device, an image of a wall surface into which a rod-shaped member is to be inserted;
A process of detecting an insertion hole already formed in the wall surface as an elliptical region based on brightness information indicated by the captured image;
A process of estimating a position and an angle of the insertion hole on the wall surface based on a diameter and brightness information of the elliptical area shown in the photographed image corresponding to the elliptical area;
A program that executes the following.