JP7703170B2 - Point cloud acquisition simulation method for existing structures, and point cloud acquisition simulation system for existing structures - Google Patents

Description

The present invention relates to a simulation method and a simulation system used when carrying out repair and replacement work on existing structures such as bridges.

In recent years, existing structures such as bridges built decades ago have been showing signs of deterioration due to aging, and replacement work is being carried out on the RC decks on which vehicle loads are directly applied. In order to ensure the quality of design and construction, it is important to understand the shape of the existing structure before construction begins. Therefore, as shown in Figure 16, there are an increasing number of cases where, in order to understand the condition of existing bridges, point cloud data of existing bridges is obtained using measuring equipment, and the existing bridge is converted from this cloud data into a 3D model using a dedicated 3D CAD to create highly accurate design drawings of existing structures.

A 3D laser scanner is used as a measuring device to acquire point cloud data. This 3D laser scanner is composed of a mechanism that rotates a mirror that emits a laser vertically, and a mechanism that rotates the mirror horizontally, and is capable of measuring 360 degrees within the range directly visible from the device. However, this 3D laser scanner cannot acquire point cloud data in blind spots, so point cloud data for the required range is acquired by acquiring point cloud data at multiple locations while moving the 3D laser scanner and integrating the data.

When using a 3D laser scanner to measure civil engineering structures that span hundreds of meters, such as bridges, it is necessary to consider in advance where to install the 3D laser scanner based on the design drawings in order to obtain the necessary point cloud data. After consideration, scans are then carried out at the multiple determined locations, the obtained point clouds are integrated, and overall point cloud data is created, which is used in the design of repair and replacement work.

However, with this 3D laser scanner, it is difficult to determine in advance where to install measuring equipment to obtain the necessary point cloud data for structures such as steel bridges, which are made up of many thin-walled components, using only design drawings. This is because, in the case of steel bridges, the lower flanges of the main girders and the cross braces are in the blind spots of the 3D laser scanner, as shown in Figure 9.

Furthermore, if the necessary point cloud is not available for 3D modeling work after point cloud data acquisition, it is necessary to remeasure on-site, which can result in scanning more locations than necessary, increasing work time and costs.

The present invention aims to provide a point cloud acquisition simulation method and a point cloud acquisition simulation system for existing structures that can optimize the installation position of measuring equipment on-site and the number of scans, and eliminate the need to reacquire point cloud data due to missed points.

In order to solve the above problems, the present invention provides a point cloud acquisition simulation method for an existing structure, which is executed on a computer , using a pseudo 3D model created from a past design drawing or a BIM/CIM model of the existing structure, the method including: a step of dividing the existing structure on the 3D model into a plurality of polygonal meshes; a virtual laser setting step of, when a point on the 3D model away from the existing structure is designated as a virtual laser emission point, substituting an initial value for an irradiation angle of a virtual laser extending in a straight line from the virtual laser emission point; - an intersection coordinate calculation process for calculating intersection coordinates between a virtual laser at an irradiation angle set in the setting process or an irradiation angle changed by a predetermined angle from that angle and each mesh on the surface of the existing structure on the 3D model; an intersection coordinate recording process for recording the intersection coordinate with the shortest distance from the virtual laser emission point among the multiple intersection coordinates calculated in the intersection coordinate calculation process; and a virtual laser rotation process for changing the irradiation angle of the virtual laser by a predetermined angle and executing the intersection coordinate calculation process and the intersection coordinate recording process until the irradiation angle becomes 2π or more.

The invention according to claim 4 is a point cloud acquisition simulation system for an existing structure that is executed on a computer using a pseudo 3D model created from a past design drawing or a BIM/CIM model of an existing structure, and that when a location away from the existing structure on the 3D model is specified as a virtual laser emission point, an initial value is assigned to the irradiation angle of the virtual laser extending in a straight line from the virtual laser emission point. A virtual laser setting unit, an intersection coordinate calculation ...

In the present invention's point cloud acquisition simulation method for an existing structure and point cloud acquisition simulation system, a simulation is performed using a pseudo 3D model created from past design drawings or BIM/CIM models of the existing structure, so the acquisition range of point cloud data can be verified in advance on a computer, optimizing the installation position of on-site measuring equipment and the number of scans, and eliminating the need to reacquire data due to missed acquisitions.

This figure shows the design procedure for repair and replacement work on existing structures. FIG. 2 is a perspective view showing a pseudo 3D model. FIG. 2 is a diagram showing a flow of a point cloud acquisition simulation according to an embodiment of the present invention, out of the design procedure shown in FIG. 1 . FIG. 13 is a perspective view showing how a solid model of an existing structure drawing on a 3D model is divided into triangular meshes in the point cloud acquisition simulation. FIG. 13 is a diagram showing the change in the irradiation angle of a virtual laser in the point cloud acquisition simulation. FIG. 11 is an oblique view illustrating the intersection coordinates of a virtual laser and an existing structure in the point cloud acquisition simulation. FIG. 11 is a cross-sectional view showing the range of intersections that can be recorded by a virtual laser in the point cloud acquisition simulation. This is an image diagram of the integrated point cloud acquired in the point cloud acquisition simulation. FIG. 9 is an image diagram showing the integrated point cloud of FIG. 8 superimposed on a 3D model. FIG. 13 is a diagram showing a pre-processing of calculation of intersection coordinates in the point cloud acquisition simulation. FIG. 13 is a diagram showing a method of calculating intersection coordinates in the point cloud acquisition simulation. FIG. 13 is a diagram showing a method of calculating intersection coordinates in the point cloud acquisition simulation. FIG. 13 is a diagram showing how to speed up calculation of intersection coordinates in the point cloud acquisition simulation. FIG. 13 is a diagram showing how to speed up calculation of intersection coordinates in the point cloud acquisition simulation. FIG. 2 is a block diagram showing a configuration of a point cloud acquisition simulation system according to an embodiment of the present invention, in the design procedure of FIG. 1 . FIG. 1 is a diagram illustrating a conventional design procedure.

(Design procedures for repair and replacement work on existing structures)
Figure 1 shows the design procedure for repair and replacement work on an existing structure. As shown in this procedure, first, a pseudo 3D model 2 is created from past design drawings or BIM/CIM models of the existing structure to be repaired or replaced, and is loaded into simulation software executed on a computer (SS1). This 3D model 2 is created as a solid model, as shown in Figure 2.

Next, point cloud acquisition simulation 1 according to the present invention is executed (SS2), and based on point cloud acquisition simulation 1, the number of times to scan the measuring equipment and the placement locations of the measuring equipment are determined (SS3). Then, on-site surveys are repeated according to the placement locations of the multiple 3D laser scanners determined in step SS3 (SS4), and the point clouds obtained from these on-site surveys are integrated to create overall point cloud data (SS5). Finally, 3D modeling is performed using 3D CAD based on the overall point cloud data, and design drawings for repair and replacement work on existing structures are created (SS6).

(Point cloud acquisition simulation)
Next, the point cloud acquisition simulation 1 according to the present invention in step SS2 will be described with reference to Fig. 3 to Fig. 14. Fig. 3 shows the flow of the point cloud acquisition simulation 1. First, the 3D model 2 of the solid model created in step SS1 in Fig. 1 is divided into a plurality of triangular meshes as shown in Fig. 4 (ST1). In this embodiment, the model is divided into triangular meshes, but this is not limited to triangular meshes, and any polygonal mesh may be used.

Next, in step ST2, when a location that has been planned in advance and is away from the existing structure 7 on the 3D model 2 is specified as the virtual laser emission point, the initial values (planar direction θ1 = 0, longitudinal direction θ2 = 0) are substituted for the irradiation angle of the virtual laser. The virtual laser imitates the laser of the 3D laser scanner 8, and the virtual laser emission point imitates the laser emission point of the 3D laser scanner 8.

As shown in FIG. 5, the 3D laser scanner 8 irradiates a laser while the mirror rotates 360 degrees in the vertical direction and 360 degrees in the horizontal direction. Specifically, as shown in FIG. 5(a), the virtual laser 9 rotates around the Z axis in increments of Δθ (planar rotation) when the vertical direction of the virtual laser emission point 10 is the Z axis, thereby changing the direction of the virtual laser 9 in increments of Δθ. Also, as shown in FIG. 5(b), the planar direction of the virtual laser 9 is the U axis, the normal direction to the U axis, and the normal direction to the U axis, respectively, and the virtual laser 9 rotates around the V axis in increments of Δθ (longitudinal rotation) to change the direction of the virtual laser 9 in increments of Δθ. Δθ may be the same angle for planar rotation and longitudinal rotation, or may be different angles. In this embodiment, Δθ is set to 0.001 radians for both planar rotation and longitudinal rotation.

Next, in step ST5 of Fig. 3, coordinates are initialized. The coordinates of the virtual laser emission point are set as the origin (x = 0, y = 0, z = 0), and the coordinates of all triangular meshes are also transformed in accordance with the virtual laser emission point, from the coordinate system of the existing structure to the coordinate system of the virtual laser emission point. Then, in step ST6, a mesh variable i is set to the number of the first triangular mesh among the multiple triangular meshes divided in step ST1 (i = 0).

Next, the coordinates of the intersection between the triangular mesh that matches the number of the mesh variable i and the virtual laser extending from the origin in a specified direction are calculated (ST7). Then, as a result of the calculation, it is determined whether there is an intersection (ST8), and if there is no intersection, the process proceeds to step ST11. If there is an intersection, it is determined whether the coordinates of the intersection are at the shortest distance from the virtual laser emission point (origin) (ST9). As shown in FIG. 6, the virtual laser 9 intersects with multiple triangular meshes (j, k), so for example, in the case of an intersection (j) with a triangular mesh (j) that is not at the shortest distance, the process proceeds to step ST11, and in the case of an intersection (k) with a triangular mesh (k) that is at the shortest distance, the intersection coordinates are recorded (ST10) and the process proceeds to step ST11.

In step ST11, it is determined whether the value of mesh variable i is the last number of the triangular mesh. If it is not the number of the last triangular mesh, mesh variable i is incremented by one to calculate the intersection coordinates for all triangular meshes, and the process returns to step ST7. If the value of mesh variable i is the number of the last triangular mesh, the address of the coordinate value recorded in step ST10 is incremented, and mesh variable i is set to the number of the first triangular mesh, and the process proceeds to step ST12.

In step ST12, it is determined whether the irradiation angle θ2 of the virtual laser in the longitudinal direction is θ2 >= 2π, and if it is 2π or more, the process proceeds to step ST13. If it is less than 2π, the virtual laser is rotated longitudinally, and the irradiation angle θ2 of the virtual laser in the longitudinal direction is incremented by Δθ (θ2 = θ2 + Δθ), as shown in FIG. 5(b). Then, the process returns to step ST5.

In step ST13, it is determined whether the irradiation angle θ1 of the virtual laser in the planar direction is θ1>=2π, and if it is 2π or more, the process proceeds to step ST14. If it is less than 2π, the virtual laser is rotated in the plane, and the irradiation angle θ1 of the virtual laser in the planar direction is incremented by Δθ (θ1=θ1+Δθ) as shown in FIG. 5(a). Then, the process returns to step ST4.

Finally, in step ST14, the multiple coordinate values recorded in step ST10 are converted into the coordinate system of the virtual laser emission point in step ST5, so a coordinate conversion is performed to return them to the coordinate values in the coordinate system of the existing structure, and the process is terminated.

As described above, multiple coordinate values are recorded in the point cloud acquisition simulation, but as shown in Figure 7, coordinate values are not recorded in areas of the existing structure on the 3D model that the virtual laser does not reach. Therefore, it is necessary to consider whether the areas that the virtual laser does not reach are necessary for creating blueprints for repair and replacement work on the existing structure, and if so, the location of the virtual laser emission point is added and the point cloud acquisition simulation (ST1-14) is performed again to obtain point cloud data for the areas necessary for creating the blueprints.

Figure 8 is an image of the point clouds acquired by performing point cloud acquisition simulations multiple times and integrating the acquired point clouds, and Figure 9 is an image of the integrated point clouds of Figure 8 superimposed on 3D model 2. As shown in Figure 9, even if the point cloud acquisition simulation is performed multiple times, there will be some locations where the point cloud cannot be acquired due to being blocked by the lower flange or cross beam. In this case, the position of the virtual laser emission point will be considered and the point cloud acquisition simulation will be performed again so that there are no locations where the point cloud cannot be acquired.

After the point cloud acquisition simulation (ST1-14) is repeated and point cloud data acquisition for the locations required for creating the design drawings is completed, the number of measurements and placement locations of the 3D laser scanner in step SS3 of Figure 1 are determined from the point cloud acquisition simulation, and the on-site survey in step SS4 is performed according to the number of measurements and placement locations of the 3D laser scanner in this point cloud acquisition simulation.

The point cloud acquisition simulation according to the embodiment of the present invention makes it possible to know the measurement locations required for 3D modeling of an existing structure in its current state without going to the site where the existing structure is located to measure it. Therefore, as in the past, scanning locations were determined from the design drawings, and the existing structure in its current state was scanned, which resulted in missing point cloud data and the need for remeasurement.

In addition, by placing virtual laser emission points in various locations in the point cloud acquisition simulation, it is possible to determine which locations will minimize the number of virtual laser emission points, allowing the installation locations of measuring equipment on-site and the number of scans to be optimized.

(Calculation of intersection coordinates with virtual laser)
Next, the calculation of the intersection coordinate with the virtual laser in step ST7 of FIG. 3 will be described in detail with reference to FIGS.

First, when the virtual laser emission point is specified in step ST2, a preliminary preparation as shown in FIG. 10 is performed in the point cloud acquisition simulation. In this preliminary preparation, a small coordinate transformation matrix and a large coordinate transformation matrix are created for each triangular mesh. Then, in the small coordinate transformation matrix, a vector with the virtual laser emission point as its start point and one vertex of the triangular mesh as its end point is transformed, and it is determined whether each triangular mesh intersects with the virtual laser. Triangular meshes that are determined not to intersect with the virtual laser are excluded from the calculation of the intersection coordinate with the virtual laser in step ST7.

In a point cloud acquisition simulation according to an embodiment of the present invention, the intersection coordinates with the virtual laser may be calculated for all triangular meshes, or triangular meshes that are determined not to intersect may be deleted, the remaining triangular meshes may be renumbered, and the intersection coordinates with the virtual laser may be calculated for all intersecting triangular meshes.

Next, as shown in Figures 11 and 12, in calculating the intersection coordinate with the virtual laser in step ST7, the direction point S2 of the virtual laser determined by the irradiation angles θ1 and θ2 of the virtual laser is used, and a vector with one vertex of the triangular mesh as its start point and the virtual laser direction point S2 as its end point is transformed with a small coordinate transformation matrix. Then, the Z component of vector R1 with the virtual laser emission point transformed in advance as its start point and one vertex of the triangular mesh as its end point is compared with the Z component of vector R2 with one vertex of the triangular mesh as its start point and the virtual laser direction point S2 as its end point. If they are the same, the triangular mesh and the virtual laser are parallel, so there is no intersection.

If the Z component of vector R1, which starts at the virtual laser emission point and ends at one vertex of the triangular mesh, is different from the Z component of vector R2, which starts at one vertex of the triangular mesh and ends at virtual laser direction point S2, the intersection coordinates of the triangular mesh and the virtual laser are calculated. Then, it is determined whether the intersection coordinates are within the triangular mesh. If they are within the triangular mesh, the intersection coordinates have been transformed using the small coordinate transformation matrix, so by transforming the intersection coordinates using the large coordinate transformation matrix, the intersection coordinates are returned to the original coordinate system.

(Faster calculation of intersection coordinates)
Although the calculation of the intersection coordinate with the virtual laser in step ST7 is performed as described above, it is also possible to speed up the calculation of the intersection coordinate. Speeding up the calculation of the intersection coordinate will be described below with reference to FIGS.

First, as shown in Figure 13, as a preliminary step, when initializing coordinates in step ST5 of Figure 3, after converting to the coordinate system of the laser emission point, the center coordinate values and circumscribing circles are created for all objects of existing structures on the 3D model. In addition, center coordinate values and circumscribing circles are created for all triangular meshes of existing structures on the 3D model.

Once these preliminary preparations are complete, a process is performed to skip intersection coordinate calculation for objects that are unlikely to intersect with the virtual laser. Specifically, the distance from the virtual laser's plane vector to the object's center and the length of the circumscribing circle's radius are compared, and if the distance from the virtual laser's plane vector to the object's center is long, there is no possibility of intersection, so intersection coordinate calculation is skipped for that object. Next, a judgment is made about the model in the opposite direction to the virtual laser, and if the distance from the virtual laser's normal vector to the object's center is shorter than the length of the circumscribing circle's radius, there is no possibility of intersection, so intersection coordinate calculation is skipped for that object as well.

Once the process of skipping intersection coordinate calculations for objects is complete, for the remaining objects that were not skipped, a process is performed to skip intersection coordinate calculations for triangular meshes that are unlikely to intersect with the virtual laser, on an individual triangular mesh basis, as shown in Figure 14. As with objects, this process involves a determination based on the virtual laser's plane vector and the distance to the center of the triangular mesh, and a determination of a model in the opposite direction to the virtual laser. Next, a determination is made between the virtual laser's plane vector and the slope of the triangular mesh, and between the z component of the virtual laser during longitudinal rotation and the slope of the triangular mesh, and for triangular meshes that are determined not to intersect, the difference coordinate calculation is skipped.

The virtual laser setting process of the present invention corresponds to the processing of steps ST3 and ST4 in FIG. 3, the intersection coordinate calculation process corresponds to step ST7 in FIG. 3, the intersection coordinate recording process 5 corresponds to steps ST8 to ST10 in FIG. 3, and the virtual laser rotation process 6 corresponds to the processing of steps ST12 and ST13 in FIG. 3.

Figure 15 is a block diagram showing the system of the point cloud acquisition simulation 1. This point cloud acquisition simulation system 11 is realized by a computer, and is composed of a CPU 12 that performs calculations, a HDD 13 in which a point cloud acquisition simulation program 131 is stored, a RAM 14 in which the point cloud acquisition simulation program 131 is deployed when the point cloud acquisition simulation is executed, and an input device 15. The point cloud acquisition simulation program 131 includes a virtual laser setting unit 3 that assigns an initial value to the irradiation angle of the virtual laser with the input device 15, an intersection coordinate calculation unit 4 that calculates the intersection coordinate of the virtual laser and the surface of the existing structure, an intersection coordinate recording unit 5 that records the intersection coordinate of the shortest distance from the virtual laser emission point, and a virtual laser rotation unit 6 that changes the irradiation angle by a predetermined angle.

The virtual laser setting unit 3 corresponds to the processing of steps ST3 and ST4 in FIG. 3, the intersection coordinate calculation unit 4 corresponds to step ST7 in FIG. 3, the intersection coordinate recording unit 5 corresponds to steps ST8 to ST10 in FIG. 3, and the virtual laser rotation unit 6 corresponds to the processing of steps ST12 and ST13 in FIG. 3.

REFERENCE SIGNS LIST 1 Point cloud acquisition simulation 2 3D model 3 Virtual laser setting unit 4 Intersection coordinate calculation unit 5 Intersection coordinate recording unit 6 Virtual laser rotation unit 7 Existing structure 8 3D laser scanner 9 Virtual laser 10 Virtual laser emission point 11 Point cloud acquisition simulation system 12 CPU
13 HDD
131 Point cloud acquisition simulation program 14 RAM
15 Input device i Mesh variables

Claims (4)

A point cloud acquisition simulation method for an existing structure executed on a computer using a pseudo 3D model created from past design drawings or BIM/CIM models of the existing structure,
Dividing the existing structure on the 3D model into a plurality of polygonal meshes;
a virtual laser setting step of, when a location on the 3D model away from an existing structure is designated as a virtual laser emission point, substituting an initial value for an irradiation angle of a virtual laser extending in a straight line from the virtual laser emission point;
An intersection coordinate calculation process for calculating intersection coordinates between a virtual laser at an irradiation angle set in the virtual laser setting process or an irradiation angle changed by a predetermined angle from the angle and each mesh of the surface of the existing structure on the 3D model;
an intersection coordinate recording step of recording the intersection coordinate having the shortest distance from the virtual laser emission point among the plurality of intersection coordinates calculated in the intersection coordinate calculating step;
A point cloud acquisition simulation method for an existing structure, comprising: a virtual laser rotation process in which the irradiation angle of the virtual laser is changed by a predetermined angle and the intersection coordinate calculation process and the intersection coordinate recording process are executed until the irradiation angle becomes 2π or more. In the intersection coordinate calculation step, an intersection coordinate between the virtual laser and each mesh of the existing structure is calculated,
The point cloud acquisition simulation method for an existing structure executed on a computer as described in claim 1 , characterized in that in the intersection coordinate recording process, the intersection coordinate with the shortest distance from the virtual laser emission point is recorded among the multiple calculated intersection coordinates. The point cloud acquisition simulation method for an existing structure executed on a computer according to claim 1, characterized in that the polygonal mesh is a triangular mesh , and in the virtual laser setting process, it is determined whether the virtual laser hits each triangular mesh, and if it is determined that the virtual laser does not hit, calculations are not performed in the intersection coordinate calculation process for the triangular mesh that is not hit by the virtual laser. A point cloud acquisition simulation system for an existing structure that is executed on a computer using a pseudo 3D model created from past design drawings or BIM/CIM models of the existing structure,
a virtual laser setting unit that, when a location on the 3D model away from an existing structure is specified as a virtual laser emission point, assigns an initial value to an irradiation angle of a virtual laser that extends in a straight line from the virtual laser emission point;
an intersection coordinate calculation unit that calculates intersection coordinates between a virtual laser at an irradiation angle set by the virtual laser setting unit or an irradiation angle changed by a predetermined angle from the irradiation angle and each mesh of a surface of an existing structure on the 3D model divided into a plurality of polygonal meshes;
an intersection coordinate recording unit that records the intersection coordinate of the intersection that is the shortest distance from the virtual laser emission point among the plurality of intersection coordinates calculated by the intersection coordinate calculation unit;
A point cloud acquisition simulation system for an existing structure comprising: a virtual laser rotation unit that changes the irradiation angle of the virtual laser by a predetermined angle until the irradiation angle becomes 2π or more, and causes the intersection coordinate calculation unit and the intersection coordinate recording unit to perform intersection coordinate calculation and intersection coordinate recording .

Images