JP7217451B2 - Bolt detection method - Google Patents

Description

The present invention relates to a bolt detection method for detecting bolts used in existing structures.

Many power transmission towers (hereinafter simply referred to as steel towers) currently used in Japan are several decades old. No matter how strong a steel tower is, it is always exposed to weather phenomena after construction and is also subject to changes in the ground that accompany it, so it will change over time. Therefore, there is an increasing need to check the soundness of existing steel towers.

When repairing a steel tower, it is necessary to confirm the position of the bolts on the steel tower. Conventionally, the positions of bolts on steel towers have been confirmed by a method called member sketching. According to this, it is possible to suitably grasp the presence or absence of bolts added during repair after construction, that is, bolts not shown in the drawing at the time of construction. However, member sketching requires a skilled worker to manually measure the positions of the bolts, which places a heavy burden on the worker and takes a huge amount of time.

Therefore, in recent years, the detection of bolt positions using a laser scanner (LiDAR: Light Detection And Ranging) has been studied. A laser scanner is a device that uses the time difference and phase difference between emitted light and reflected light to measure the three-dimensional coordinates of an object in units of mm. In particular, in recent years, the performance of phase-difference terrestrial laser scanners has been dramatically improved. A terrestrial laser scanner irradiates a laser beam from a main body supported by a tripod or the like, and at the same time, the main body rotates, and acquires the three-dimensional coordinates of surrounding terrain and structures as a point group.

The inventors have already developed a steel tower tilt detection method using the above ground laser scanner (Patent Document 1). In the steel tower inclination detection method of Patent Document 1, an existing steel tower is scanned with a terrestrial laser scanner to acquire point cloud data, and from the point cloud data, each of the main columns provided in the steel tower represents the main column. Extract feature lines that are line segments. Then, the center line of the steel tower is calculated from the plurality of characteristic lines, and the inclination of the steel tower is calculated from the center line.

Japanese Patent Application No. 2017-141547

According to the steel tower tilt detection method of Patent Document 1, it is possible to directly and quantitatively measure the steel tower tilt without relying on the skill and experience of workers. For this reason, the inventors thought that the bolts used in the steel tower could be detected by utilizing the steel tower inclination detection method of Patent Document 1.

However, the terrestrial laser scanner acquires three-dimensional positional information of objects within the line of sight of the scanner as a huge amount of point cloud data. Therefore, no data is available for the shadowed parts of the object where the laser is occluded. Therefore, in the point cloud data obtained from the ground, the bolt portion has an incomplete shape, making it difficult to identify the bolt and estimate its position. In particular, when using a single terrestrial laser scanner, the point cloud data near the boundary of the shadowed area, which often has a shallow angle between the laser and the reflective surface, is used to identify the bolt and estimate its location. Then the error becomes large.

In order to reduce shadowed areas, it is conceivable to acquire point cloud data at multiple points using multiple terrestrial laser scanners. However, when a plurality of point cloud data acquired by a plurality of terrestrial laser scanners are synthesized, it becomes difficult to accurately identify the bolt position due to errors during synthesis.

Another possible method for recognizing the shape of bolts on a pylon is to use a handy 3D laser scanner. However, in this case the scanner needs to be close to the bolt. Therefore, it is not suitable for detecting bolts used in huge structures such as steel towers.

In view of such problems, the present invention provides a bolt detection method capable of accurately detecting bolts used in huge existing structures such as steel towers from point cloud data of a ground type laser scanner. intended to provide.

In order to solve the above problems, a representative configuration of the bolt detection method according to the present invention is a bolt detection method for detecting bolts used in an existing structure, wherein the structure is detected by a terrestrial laser scanner. Acquire point cloud data by scanning, detect the plane of the structure from the point cloud data, and from the plane of the structure, a first plane having an area larger than the first threshold and a second plane having an area smaller than the second threshold is detected, and among the second planes, the second plane parallel to the adjacent first plane is determined to be the upper surface of the bolt, and the point group constituting the second plane determined to be the bolt is the area of the bolt It is characterized by identifying that

In the above configuration, the plane of a structure such as a steel tower is first detected from the point cloud data acquired by the terrestrial laser scanner, and the first plane and the second plane are detected from that plane. As a result, the first plane can be identified as a member such as a large main leg member used in a steel tower, and the second plane can be identified as a bolt used to fix a member such as a main leg member. It can be identified as an object having a property.

When the bolt is attached to a member such as a main leg member, the top surface of the bolt is parallel to the plane of the member. Therefore, of the second planes, the second plane parallel to the adjacent first plane can be determined to be the upper surface of the bolt. This makes it possible to accurately detect bolts used in huge existing structures such as steel towers from the point cloud data of the terrestrial laser scanner.

The size and location of the identified bolt may be determined by fitting the shape of a standard size bolt head to the area of the bolt. According to such a configuration, it is possible to grasp the size of the bolts to be used when repairing the structure, and it is possible to perform the repair work more efficiently.

The above fitting determines the size, position, and angle of rotation of the point cloud of the bolt area when the points forming the point cloud are closest to the standard size bolt head profile. and location are the size and location of the identified bolt. This makes it possible to more accurately grasp the size of the bolt.

A group of points existing between the first plane and the second plane may also be included in the area of the bolt. The point cloud existing between the first plane and the second plane is presumed to be the sides of the bolt. Therefore, by including the point cloud in the bolt region, the number of points that are close to the standard size bolt head profile during the fitting described above can be dramatically increased. This makes it possible to improve the accuracy of fitting.

The above fitting obtains the maximum width of the point cloud of the bolt region, sets a circle with the maximum width as the diameter, sets the standard size of the bolt head in this circle as the size of the identified bolt, and sets the point cloud The position of the circle that contains the most points that make up the position of the identified bolt may be determined. This makes it possible to better grasp the size and position of the bolt.

According to the present invention, it is possible to provide a bolt detection method capable of accurately detecting bolts used in a huge existing structure such as a steel tower from point cloud data of a ground type laser scanner. .

It is a figure explaining the detection apparatus used for the bolt detection method concerning this embodiment. 5 is a flowchart for explaining processing of a bolt detection method according to the embodiment; It is the figure which showed the outline|summary of the measurement by a ground type laser scanner. This is a subroutine showing the details of data editing processing. It is the figure which illustrated the outline|summary of a data editing process. It is a figure which illustrates the point group around the bolt in a steel tower. This is a subroutine showing the details of bolt detection processing. FIG. 11 is a diagram illustrating an image during bolt detection processing; It is a figure explaining a fitting process. FIG. 10 is a detailed diagram of the fitting process of FIG. 9(a); FIG. 10 is a detailed diagram of the fitting process of FIG. 9(a); It is a figure explaining the error of the fitting process of 1st Example and 2nd Example, and the conventional sketching process. 7 is a flowchart for explaining another example of the bolt detection method according to the embodiment; It is a figure explaining the intersection line of an angle material.

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The dimensions, materials, and other specific numerical values shown in these embodiments are merely examples for facilitating understanding of the invention, and do not limit the invention unless otherwise specified. In the present specification and drawings, elements having substantially the same functions and configurations are denoted by the same reference numerals to omit redundant description, and elements that are not directly related to the present invention are omitted from the drawings. do.

FIG. 1 is a diagram illustrating a detection device 160 used in the bolt detection method according to this embodiment. In the present embodiment, a case where the bolt detection method is executed using the detection device 160 shown in FIG. 1 will be described as an example. A bolt detection method may be implemented.

As shown in FIG. 1, the detection device 160 includes a memory 162 for storing programs and various data, a CPU 164 for performing central processing, and an input/output unit 166 for exchanging data with the laser scanner 142 . The CPU 164 also functions as a control unit 168 that controls program execution and the like, and a calculation unit 170 that performs various calculations.

FIG. 2 is a flowchart for explaining the processing of the bolt detection method according to this embodiment. FIG. 3 is a diagram showing an outline of measurement by the terrestrial laser scanner 142. As shown in FIG. 3A is a diagram showing an example of use of the laser scanner 142, and FIG. 3B is a diagram showing an example of point cloud data acquired by the laser scanner 142. FIG.

In addition, although the steel tower 140 is illustrated as an existing structure in this embodiment, it is not limited to this. The bolt detection method of this embodiment can also be applied to existing structures other than the steel tower 140 .

In the bolt detection method of this embodiment, the bolts 180 used in the steel tower 140 shown in FIG. 3(a) are detected. In the bolt detection method, as shown in FIG. 2, first, an existing steel tower 140 (see FIG. 3A) is scanned with a terrestrial laser scanner (laser scanner 142), and three-dimensional coordinates are obtained from the existing steel tower 140. Certain point cloud data is acquired (step S100).

A specific example of the laser scanner 142 is a phase difference laser scanner FARO Focus 3D X330 (hereinafter abbreviated as X330). The X330 is composed of a body portion 144 that emits laser light and a tripod 146 that supports the body portion 144 . The main unit 144 rotates the laser light in the vertical direction and in the horizontal direction to obtain the three-dimensional coordinates of the surrounding landforms and structures as point cloud data. In particular, the X330 has a measurement range of 0.6 to 330 m and a maximum measurement speed of 97,600 points/second, which is higher than that of conventional terrestrial laser scanners, making it suitable for this detection method.

In the bolt detection method of the present embodiment, point cloud data is acquired by scanning from the center of a plurality of legs 148 of the steel tower 140 . This makes it possible to efficiently acquire three-dimensional coordinates. It is conceivable to scan from the outside of the steel tower 140, but in order to acquire point cloud data from the outside of the steel tower 140, scanning must be performed from two or more locations. Then, an error occurs when combining a plurality of point cloud data. Therefore, it is important to scan from one place this time, and it is preferable to scan the inside of the steel tower 140, especially the center.

Here, the so-called raw data acquired by the laser scanner 142 includes, as shown in FIG. These point groups other than the pylon 140 increase the processing load and may lead to a decrease in bolt detection accuracy. Therefore, in the present embodiment, as shown in FIG. 2, data editing processing of acquired point cloud data is performed (step S102).

FIG. 4 is a subroutine showing details of the data editing process (step 102 in FIG. 2). FIG. 5 is a diagram illustrating an outline of data editing processing. The data editing process is performed to extract only the point group representing the steel tower 140 from the raw data acquired by the laser scanner 142 .

In the data editing process, the calculation unit 170 divides the point cloud data acquired by the laser scanner 142 into a predetermined horizontal grid pattern (step S110). Next, the calculation unit 170 extracts the installation area of the pylon 140 from each of the grid-like divided areas (step S112).

In this embodiment, since the laser scanner 142 is placed at the center of the steel tower 140, distant buildings such as condominiums are removed by acquiring only point groups within a certain distance from the center of the laser scanner 142. FIG. Therefore, the first grid is set in a range including a circle at a constant distance from the center of the laser scanner 142, as shown in FIG. 5(a).

Further, since the steel tower 140 tapers toward the top, the point cloud of only the portion below the power transmission line 150 is acquired to extract the region of the object. For extraction of the installation area, it is determined whether or not each area includes measurement points where the difference in height of the coordinate data (the difference in height between the highest point and the lowest point) is equal to or greater than a predetermined value. This is done by extracting an area in which measurement points with heights equal to or higher than the above exist.

As described above, a measurement point having a height difference equal to or greater than a predetermined value is a measurement point obtained from a structure having a certain height. For example, the minimum coordinate in the vertical direction in a certain area is regarded as the ground surface. Then, by extracting an area including measurement points with a height difference of 5 m or more from the ground surface, an area in which a structure having a height of 5 m or more exists can be obtained. Then, as shown in FIG. 5(b), the grid at the position where the steel frame exists can be extracted. Here, since only the point cloud of the portion below the transmission line 150 is targeted, there is a gap in the central portion.

When extracting the installation area (step S112), first, as shown in FIG. Next, processing is performed to fill in the concave portion of the outer contour. For this processing, for example, morphology calculations used in existing image processing can be applied. A morphology operation is a process of smoothing unevenness of an image and removing isolated points by performing dilation and contraction multiple times. As a result, as shown in FIG. 5(d), it is possible to obtain a substantially rectangular area without any concave portions.

After extracting the installation area of the steel tower, the point cloud data above the installation area is reacquired (including the area above the transmission line 150), and the measurement points near the ground surface are removed (step S114). FIG. 5(e) is a diagram illustrating only the point cloud data showing the steel tower 140. FIG. This completes the data editing process.

FIG. 6 is a diagram illustrating a point cloud around the bolt 180 in the pylon 140. As shown in FIG. Please refer to FIG. 2 again. After the data editing process described above is completed, the calculation unit 170 detects the plane of the steel tower 140 from the point cloud data (step S104). Thereby, as shown in FIG. 6A, the plane of the steel tower 140 including the leg 148 and the bolt 180 attached thereto is detected.

FIG. 7 is a subroutine showing the details of bolt detection processing. FIG. 8 is a diagram illustrating an image during bolt detection processing. Although FIG. 8A is the same image as FIG. 6A, it is shown again for easy understanding.

After performing the plane detection process shown in FIG. 2 (step S104), the calculation unit 170 subsequently performs bolt detection process (step S106). In the bolt detection process, as shown in FIG. 7, the calculation unit 170 detects the first plane (step S202), and then detects the second plane (step S204).

When detecting the first plane, the calculation unit 170 determines whether the area of the plane of the steel tower 140 (structure) is larger than the first threshold. For example, as shown in FIGS. 6A and 8A, if the area is larger than the first threshold value T1, the area is determined to be the first area. Thereby, the leg portion 148, which is a large member, can be identified in the steel tower 140. FIG.

When detecting the second plane, the calculation unit 170 determines whether the area of the plane of the steel tower 140 structure is smaller than the second threshold. For example, as shown in FIGS. 6A and 8A, if the area is smaller than the second threshold value T2, the area is determined to be the second area. This allows the identification of objects that may be bolts 180 .

By performing the process of detecting the first plane (step S202) and the process of detecting the second plane (step S204) for the image shown in FIG. A first plane P1 and a second plane P2 are detected at .

Here, when the bolt 180 is attached to a member such as the leg 148, its top surface is parallel to the plane of that member. Therefore, if the second plane P2 is parallel to the first plane P1 identified as the leg portion 148, it can be determined that the second plane P2 is the upper surface of the bolt 180. FIG. Therefore, after detecting the first plane P1 and the second plane, the calculation unit 170 determines that the second plane P2 parallel to the adjacent first plane P1 among the detected second planes P2 is the upper surface of the bolt 180. (step S206).

After determining the upper surface of the bolt 180, the calculation unit 170 identifies the group of points forming the second plane determined to be the bolt 180 as the region of the bolt 180 (step S208). As a result, the upper surface 180b of the bolt 180 is detected from the image shown in FIG. 8(b) as shown in FIG. 8(c). As described above, according to the bolt detection method of the present embodiment, it is possible to accurately detect bolts used in a huge existing structure such as the steel tower 140 from the point cloud data of the laser scanner 142. Become.

Here, when acquiring point cloud data by scanning from the center of the plurality of legs 148 of the steel tower 140 as in the present embodiment, the point cloud of the bolt 180 is captured by a laser scanner as shown in FIG. The side facing 142 is dense, but the opposite side is sparse. For this reason, the contour of the bolt 180 is partially missing on the opposite side, and it is difficult to grasp the size of the bolt 180 . Therefore, in the bolt detection method of the present embodiment, once the region of the bolt 180 is identified, the computing unit performs fitting processing (step S210).

FIG. 9 is a diagram for explaining fitting processing. 10 and 11 are detailed diagrams of the fitting process of FIG. 9(a). As a first example of the fitting process, as shown in FIG. Fit the standard size bolt head shape S to the contour of the

In the fitting of the first embodiment, first, neighboring points in the point group recognized as the area of the bolt 180 shown in FIG. 10(a) are connected, and converted into the mesh model shown in FIG. 10(b). Then, from the image shown in FIG. 10(b), the boundary B on the side facing the laser scanner 142 is detected as shown in FIG. 10(c). Then, as shown in FIG. 10(d), the diameter r, the center position (Cx, Cy), and the rotation angle θ of the hexagon that best fits the boundary B are calculated.

Specifically, while slightly changing the center position (Cx, Cy) of the hexagon S and the rotation angle θ, the number of boundary points whose distance from the side of the hexagon S is equal to or less than the threshold ε is maximized. Calculate (Cx, Cy) and θ.

The size of the bolt 180 is determined by standards, for example, 27 mm shown in FIG. 11(a), 30 mm shown in FIG. 11(b), and 32 mm shown in FIG. 11(c). Thus, a hexagon of multiple bolt sizes can be applied and the size that fits best can be the size of the bolt 180 in question. In the example shown in FIG. 11, the size of the bolt 180 is 30 mm in FIG. 11(b).

When determining the size of the bolt 180, let {pi} be the point detected as the boundary, and let ds(pi) be the shortest distance between the point pi and the hexagon whose diameter is s. Then, it is preferable to determine s as the diameter of the bolt 180, which minimizes the following formula 1.

By performing the fitting process, it can be determined that the size and position of the hexagon S determined in this process are the size and position of the identified bolt 180 shown in FIG. 8(d) (step S212). ). This completes the bolt detection process. By including the fitting process as in the present embodiment, the size of the bolts 180 used in repairing the structure can be grasped, and the repair work can be performed more efficiently.

As a second example of the fitting process, we refer to the maximum width of the point cloud in the area of the bolt 180 . Specifically, as shown in FIG. 9B, the computing unit 170 first acquires the maximum width W of the point group in the area of the bolt 180, and sets a circle C having a diameter equal to the maximum width. Next, the calculation unit 170 determines the size of the bolt 180 to be the standard size in which the head of the bolt fits in this circle. Furthermore, the position of the circle C that contains the largest number of points forming the point group is determined to be the position of the bolt. Such a configuration also allows the size and position of the bolt 180 to be properly grasped.

Note that when performing the above fitting process, it is preferable to include the point group existing between the first plane and the second plane in the region of the bolt. Since the point group existing between the first plane and the second plane is presumed to be the side surface of the bolt 180, by including the point group in the region of the bolt 180, the standard size can dramatically increase the number of points close to the bolt head profile. Therefore, it becomes possible to improve the precision at the time of fitting.

FIG. 12 is a diagram for explaining the error between the fitting process of the first and second embodiments and the conventional sketching process, showing the distance (mm) from the valley line of the leg material. As shown in FIG. 12, when the size of the bolt is actually measured with a vernier caliper and when the size of the bolt is determined by the fitting process, the error in the fitting process of the first embodiment is smaller. From this, it can be understood that it is preferable to employ the fitting process of the first embodiment.

FIG. 13 is a flow chart illustrating another example of the bolt detection method of this embodiment. It should be noted that hereinafter, the same reference numerals are given to processes that overlap with the bolt detection method described above, and the description thereof will be omitted. FIG. 14 is a diagram for explaining intersection lines of angle members.

In the bolt detection method shown in FIG. 13, between steps S208 and S210 described above, processing is performed to identify the area of the bolt 180 using the line of intersection of the first plane (trough line of the leg member) ( step S214). Specifically, when the steel tower is an angle steel tower, the legs 148 of the steel tower 140 use L-shaped angle members. Therefore, in step 214, first, the trough line of the internal angle of the angle member used for the leg portion 148 is extracted.

The calculation unit 170 extracts adjacent planes from among the first planes detected in step S202. Through this processing, a plurality of intersecting planes including the inner two surfaces of each angle member can be acquired. Subsequently, the calculation unit 170 determines the boundary of the intersecting first planes as the intersection line L of the first planes shown in FIG. 14 . Then, after determining the line of intersection L of the first plane (trough line of the leg), the calculation unit 170 calculates the area of the bolt 180 from the point group of a predetermined distance from the line of intersection L of the first plane (trough line of the leg). and discriminate.

The bolt 180 is a member mainly used for connecting angle members. Therefore, since the bolt 180 is arranged within a predetermined distance from the line of intersection L of the first plane detected as the angle material used for the leg portion 148, the bolt 180 is within the range of the predetermined distance from the line of intersection L of the first plane. can be determined to be the bolt 180 . This makes it possible to simplify the processing in the calculation unit 170 and to determine the bolt 180 more accurately.

Although the preferred embodiments of the present invention have been described above with reference to the accompanying drawings, it goes without saying that the present invention is not limited to such examples. It is obvious that a person skilled in the art can conceive of various modifications or modifications within the scope described in the claims, and these also belong to the technical scope of the present invention. Understood.

INDUSTRIAL APPLICABILITY The present invention can be used as a bolt detection method for detecting bolts used in existing structures.

DESCRIPTION OF SYMBOLS 140... Steel tower 142... Laser scanner 144... Main body part 146... Tripod 148... Leg part 150... Power transmission line 160... Detecting device 162... Memory 166... Input/output part 168... Control part 170... Operation unit, 180... Volt

Claims (2)

A bolt detection method for detecting a bolt used in an existing structure,
Scanning the structure with a terrestrial laser scanner to acquire point cloud data,
detecting a plane of the structure from the point cloud data;
detecting a first plane having an area larger than a first threshold and a second plane having an area smaller than a second threshold from the planes of the structure;
determining that a second plane parallel to the adjacent first plane is the upper surface of the bolt;
Identifying a group of points forming the second plane determined to be the bolt as the area of the bolt ,
fitting a standard size bolt head shape to the bolt area;
The fitting determines the size, position, and angle of rotation of the point cloud of the bolt area when the points constituting the point cloud are closest to the standard size bolt head profile. is the size and position of the identified bolt. 2. The bolt detection method according to claim 1, wherein a group of points existing between the first plane and the second plane are also included in the area of the bolt.

Images