Breakline setting support system
The breakline setting support system directly extracts cross-sectional change points from three-dimensional measurement points to establish reliable breaklines, addressing the unreliability of conventional methods and reducing costs by eliminating the need for additional image processing.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- 北川 悦司
- Filing Date
- 2022-09-20
- Publication Date
- 2026-07-03
AI Technical Summary
Conventional breakline setting methods in three-dimensional models from point clouds rely on virtual coordinates, lacking reliability as they are not based on actual measurement points, and often require additional image processing, increasing costs.
A breakline setting support system that extracts cross-sectional change points from three-dimensional measurement points, utilizing a reference line, cross-sectional line segments, and extended regions to determine breaklines directly from actual measurement coordinates, optionally with image display and noise reduction.
Enables reliable breakline setting based on actual measurements, reducing costs by eliminating the need for additional image processing and improving the accuracy of three-dimensional models.
Smart Images

Figure 0007884410000001 
Figure 0007884410000002 
Figure 0007884410000003
Abstract
Description
Technical Field
[0005] ,
[0001] The present invention relates to a technology for creating a three-dimensional model of a ground object or the like. More specifically, it relates to a break line setting support system that can support the setting of break lines when setting break lines based on three-dimensional point cloud data.
Background Art
[0002] When performing measurements over a wide area to create a topographic map, conventionally, aerial photogrammetry using aerial photographs taken from an aircraft has been common. In recent years, however, airborne laser measurement has also been widely used. <0OO0011> Aerial photogrammetry is a method in which a pair of stereo-pair photographs, consisting of two different aerial photographs taken of the same location, is prepared, the same object included in both photographs is identified, and the coordinates of the object on the ground are obtained by using the difference (parallax) in the position of the object in the photographs. <OO00014> On the other hand, airborne laser measurement involves flying an aircraft over the terrain to be measured and measuring the reflected signal of the laser pulse irradiated onto the terrain. Since an aircraft is usually equipped with a positioning device such as a Global Navigation Satellite System (GNSS) and an inertial measurement device such as an Inertial Measurement Unit (IMU), the irradiation position (x, y, z) and irradiation attitude (ω, φ, κ) of the laser pulse can be grasped. As a result, the three-dimensional coordinates of the measurement point (the point where the laser pulse is reflected) can be obtained from the time difference between the irradiation time and the reception time. In recent years, in addition to airborne laser measurement in which a laser measuring device is mounted on an aircraft, ground-based laser measurement in which a laser measuring device is installed on the ground for measurement is also spreading.
[0005] Aerial photogrammetry, aerial laser measurement, and ground-based laser measurement yield a large number of measurement points with three-dimensional coordinates (hereinafter, these numerous measurement points are referred to as a "point cloud"). However, because it is difficult to grasp the relative positional relationships of the point cloud in its three-dimensional coordinates alone, a three-dimensional model created from this point cloud is usually used. A three-dimensional model is, so to speak, a three-dimensional reproduction of the shape of the object being measured based on the three-dimensional coordinates of the point cloud, and examples include "Digital Surface Models (DSMs)" and "Digital Elevation Models (DEMs)." These DSMs and DEMs are often composed of a large number of meshes, and the model is formed by assigning height to representative points in each mesh. Methods for assigning height to representative points in a mesh include the Triangulated Irregular Network (TIN) method, which uses an irregular triangular network formed from random data to determine height; the Nearest Neighbor method, which uses the closest measurement point; the Inverse Distance Weighting (IDW) method; the Kriging method; and the averaging method, among others.
[0006] When creating a 3D model, accurately defining break lines is extremely important. Here, a break line is a line connecting points of change in cross-section (points of transition in steepness and gradation) in the longitudinal direction, essentially the boundary line between two surfaces that intersect at a considerable angle. For example, in the case of an embankment slope, a break line is formed at the boundary between the shaped embankment slope and the road surface, and also at the boundary between the embankment slope and the berm. When creating a 3D model of such an embankment slope, it is important to accurately represent (reproduce) the break lines formed by the embankment slope and the road surface, and by the embankment slope and the berm.
[0007] Thus, breaklines are essential when forming surfaces using methods such as TIN, and are also indispensable when extracting points to retain during downsampling, when understanding shape changes such as settlement and uplift, or when ensuring consistency with CAD (Computer-Aided Design) drawings.
[0008] Conventionally, methods for setting breaklines from point clouds have included using the intersection lines of planes, using the normal vectors of each point cloud, and using the contour lines of planes. Furthermore, Patent Document 1 discloses a new technique in which a provisional breakline is set from the 3D coordinates of the point cloud, a partial image including that breakline is displayed, and then the final breakline is set. [Prior art documents] [Patent Documents]
[0009] [Patent Document 1] Japanese Patent Publication No. 2020-17109 [Overview of the Initiative] [Problems that the invention aims to solve]
[0010] According to the invention disclosed in Patent Document 1, breaklines can be accurately set even in areas with a low density of point clouds. On the other hand, in the prior art, including the invention of Patent Document 1, the coordinates of the breakline (or the points that make up the breakline) were determined by calculations based on measurement points (actual measured coordinates). In other words, the breakline was represented by virtual coordinates without directly using the coordinates of the actually measured points, and therefore it could not necessarily be said to be a reliable breakline based on actual measurements.
[0011] The object of the present invention is to solve the problems of the conventional method, namely, to provide a breakline setting support system that can set breaklines based on the coordinates of measured points. [Means for solving the problem]
[0012] The present invention focuses on setting a cross-sectional line segment perpendicular to a reference line, extracting cross-sectional constituent points from measurement points around that cross-sectional line segment, and further extracting cross-sectional change points (points of transition and points of gradual change) that constitute the break line from the cross-sectional constituent points. This invention is based on an idea that has not been seen before.
[0013] The breakline setting support system of the present invention is a system that supports the setting of breaklines from three-dimensional measurement points, and comprises a reference line setting means, a cross section line segment setting means, a first extended region setting means, a provisional cross section line setting means, a second extended region setting means, a cross section component point extraction means, and a cross section change point extraction means. Of these, the reference line setting means is a means for setting a reference line on a horizontal projection plane, the cross section line segment setting means is a means for setting a cross section line segment perpendicular to the reference line, the first extended region setting means is a means for setting a first extended region (a region extended by a first extended width along the reference line with respect to the cross section line segment), and the provisional cross section line setting means is a means for extracting a first measurement point (a measurement point included in the first extended region) and setting a provisional cross section line composed of line segments on a vertical plane based on the first measurement point. Furthermore, the second expansion region setting means is a means for setting the second expansion region (a region expanded by the second expansion width along the reference line with respect to the cross section line segment), the cross section component point extraction means is a means for extracting the second measurement point (a measurement point included in the second expansion region) and for extracting cross section component points that are located near the provisional cross section line when the second measurement point is projected onto a vertical plane, and the cross section change point extraction means is a means for extracting cross section change points (points of transition and points of gradation on the vertical plane) from among the cross section component points. The second expansion width is set to a dimension shorter than the first expansion width, and the cross section component point extraction means extracts cross section component points on the condition that the distance from the provisional cross section line is below a predetermined separation threshold. Then, a break line can be set based on multiple cross section change points.
[0014] The breakline setting support system of the present invention may further include a display means for displaying an image such as an aerial photograph of the target area (however, an image associated with three-dimensional coordinates). In this case, the operator can set a reference line using the reference line setting means while checking the image. Furthermore, the three-dimensional coordinates of the set reference line (or points constituting the reference line) are determined by performing calculation processing based on the three-dimensional coordinates corresponding to the image.
[0015] The breakline setting support system of the present invention may further include cluster generation means and noise reduction means. The cluster generation means is a means for generating clusters of cross-sectional points based on the separation between cross-sectional points, and the noise reduction means is a means for removing cross-sectional points as noise from clusters in which the total number of cross-sectional points constituting the cluster falls below a predetermined cluster threshold. In this case, the cross-sectional change point extraction means extracts cross-sectional change points from the cross-sectional points from which the noise has been removed.
[0016] The breakline setting support system of the present invention may further include a coordinate transformation means. This coordinate transformation means rotates a reference line so that it is parallel to an axis set in a horizontal plane, and transforms the planar coordinates of the measurement points in accordance with the rotation of the reference line. In this case, the provisional section line setting means extracts a first measurement point based on the measurement point whose planar coordinates have been transformed, and the section component point extraction means extracts a second measurement point based on the measurement point whose planar coordinates have been transformed. [Effects of the Invention]
[0017] The breakline setting support system of the present invention has the following effects: (1) Breaklines can be set based on actually measured coordinates, meaning that more reliable breaklines can be obtained compared to conventional technology. (2) If a point cloud with three-dimensional coordinates is available, it is not necessarily required to prepare images such as orthophotos, and breaklines can be set at a lower cost compared to conventional techniques.
Brief Description of the Drawings
[0018] [Figure 1] Block diagram showing the main configuration of the break line setting support system of the present invention. [Figure 2] Plan view showing an example of a reference line set near the top surface of the levee. [Figure 3] Plan view showing a reference line generated parallel to the Y-axis direction. [Figure 4] Plan view showing a plurality of cross-section line segments set along the reference line. [Figure 5] (a) Cross-sectional view showing a plurality of first measurement points projected onto a vertical plane including a cross-section line segment, (b) Cross-sectional view showing a provisional cross-section line composed of line segments on the vertical plane. [Figure 6] Plan view showing a second expansion area set by a second expansion width having a dimension shorter than the first expansion width. [Figure 7] (a) Cross-sectional view showing a vicinity range set based on the provisional cross-section line, (b) Cross-sectional view showing a second measurement point within the vicinity range. [Figure 8] Model diagram showing the situation of removing noise from a cluster generated by a cluster generation means. [Figure 9] Flow chart showing an example of the main processing flow until the break line setting support system of the present invention generates a break line. [Figure 10] Flow chart showing an example of the main processing flow until the break line setting support system of the present invention generates a break line after setting an expansion area in three stages.
Mode for Carrying Out the Invention
[0019] An example of an embodiment of the break line setting support system of the present invention will be described based on the drawings.
[0020] The breakline setting support system of the present invention utilizes measurement points having three-dimensional coordinates (hereinafter simply referred to as "3D measurement points"). Here, three-dimensional coordinates are coordinates that have information on planar position and height, and examples include coordinates (X, Y, Z) in a three-axis orthogonal coordinate system, or latitude, longitude, and elevation in a geodetic coordinate system. Furthermore, conventional surveying techniques such as aerial photogrammetry, aerial laser measurement, ground-based laser measurement, and surveying using a TS (Total Station) can be used to acquire 3D measurement points. For convenience, this explanation will describe an example in which the 3D measurement point has coordinates (X, Y, Z) in a three-axis orthogonal coordinate system.
[0021] Furthermore, one of the features of the breakline setting support system of the present invention is that it extracts "section change points" using a large number of 3D measurement points (hereinafter simply referred to as "3D point cloud data"). Here, section change points are so-called "points of transition" or "points of transition" that show a significant change in the shape obtained by projecting the 3D point cloud data onto a vertical plane (hereinafter referred to as "section shape"), and examples include boundary points where the slope gradient changes significantly, boundary points between an embankment slope and a road surface, boundary points between an embankment slope and a berm, and boundary points between the inclined surface and the top surface of a concrete retaining wall. When multiple section shapes are arranged in the longitudinal direction (horizontal direction roughly perpendicular to the direction of the section shape), multiple section change points will be arranged in the longitudinal direction, and a breakline can be set by connecting similar (generally the same elevation) section change points in the longitudinal direction. In other words, the breakline setting support system of the present invention provides section change points that constitute a breakline and can support the setting of a breakline.
[0022] Figure 1 is a block diagram showing the main configuration of the breakline setting support system 100 of the present invention. As shown in this figure, the breakline setting support system 100 is configured to include a reference line setting means 101, a cross section line segment setting means 102, a first extended area setting means 103, a provisional cross section line setting means 104, a second extended area setting means 105, a cross section constituent point extraction means 106, and a cross section change point extraction means 107. It can also be configured to include a display means 108, a cluster generation means 109, a noise reduction means 110, a reference line calculation means 111, a coordinate transformation means 112, a breakline generation means 113, an image data storage means 114, a point cloud data storage means 115, and the like.
[0023] The breakline setting support system 100 comprises a reference line setting means 101, a cross-section line segment setting means 102, a first extended area setting means 103, a provisional cross-section line setting means 104, a second extended area setting means 105, a cross-section component point extraction means 106, a cross-section change point extraction means 107, a cluster generation means 109, a noise reduction means 110, a reference line calculation means 111, a coordinate transformation means 112, and a breakline generation means 113. These components can be manufactured as dedicated components or utilize a general-purpose computer device. In other words, each component performs its own specific processing by having the computer device execute calculations according to a predetermined program. This computer device includes a processor such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit), memory such as ROM or RAM, and may also include input means such as a mouse or keyboard and a display. For example, it can be configured using a personal computer (PC) or a server. When using a computer device equipped with a display, it is preferable to use that display as the display means 108.
[0024] Furthermore, the image data storage means 114 and the point cloud data storage means 115 can utilize the storage devices of a general-purpose computer (e.g., a personal computer), or they can be built on a database server. When built on a database server, it can be located on a local network (LAN: Local Area Network), or it can be a cloud server that stores data via the internet.
[0025] The following describes in detail each of the main elements constituting the breakline setting support system 100 of the present invention.
[0026] (Point cloud data storage means and image data storage means) The point cloud data storage means 115 is a means for storing 3D point cloud data (i.e., a large number of 3D measurement points) of the target area, while the image data storage means 114 is a means for storing aerial photographs taken of that target area. It is desirable that the aerial photographs stored by the image data storage means 114 be associated with 3D coordinates. For example, by assigning 3D coordinates to each pixel that makes up the aerial photograph, it is possible to create an "aerial photograph corresponding to 3D coordinates". In assigning 3D coordinates to an aerial photograph, conventional photogrammetry techniques or SfM (Structure from Motion), which has been widely used in recent years, can be used to assign 3D coordinates. If 3D coordinates are prepared independently of the aerial photograph, it is also possible to assign 3D coordinates to the aerial photograph using ICP (Iterative Closest Point).
[0027] (Means for setting reference lines) The reference line setting means 101 is a means for setting a "reference line". Here, a reference line is a line segment formed on a horizontal projection plane (a plane set at the same elevation), that is, a line segment on a plane formed by coordinates that have different planar coordinates (X,Y) but the same elevation (Z) value (for example, 0m). Furthermore, this reference line indicates the longitudinal direction for the cross-sectional shape described above, in other words, multiple cross-sectional shapes are set along the reference line. Therefore, the reference line LS is set in a location where a break line (i.e., a cross-sectional change point) is likely to exist, as shown in Figure 2, and is set to be roughly parallel to the direction of the predicted break line. Figure 2 is a plan view showing an example of a reference line LS set near the top surface of an embankment, and in this figure, the reference line LS is shown as a white line. For convenience, here we will refer to the direction of the reference line LS as the "longitudinal direction", and the direction perpendicular to the reference line LS (i.e., the direction when the cross-sectional shape is projected onto a horizontal plane) as the "cross-sectional direction".
[0028] When setting the reference line LS using the reference line setting means 101, it is possible to automatically set (automatically interpret) the reference line LS from an aerial photograph by utilizing conventional image recognition technology or machine learning (ML) technology such as deep learning. Alternatively, the aerial photograph read from the image data storage means 114 can be displayed on a display means 108 such as a display, and the operator can set the desired reference line LS by visually viewing the aerial photograph and operating the reference line setting means 101, which consists of a pointing device (mouse, touch panel, pen tablet, touchpad, trackpad, trackball, etc.) or keyboard.
[0029] When the reference line LS is set by the reference line setting means 101, the reference line calculation means 111 calculates the coordinates of the reference line LS. More specifically, the coordinates of the reference line LS are calculated by determining the two-dimensional coordinates (X,Y) of the points that make up the reference line LS (especially the points at both ends) and by calculating the linear equation (general equation in a two-dimensional coordinate system) that represents the reference line LS. At this time, if the reference line LS is formed by multiple line segments (i.e., a polyline) as shown in Figure 2, the coordinates are calculated for each line segment (i.e., the reference line LS).
[0030] (Coordinate transformation means) The coordinate transformation means 112 is a means for transforming the coordinates of the reference line LS. As shown in Figure 2, when the reference line LS is formed by multiple line segments indicating different directions, subsequent processing is carried out while maintaining each direction, that is, while taking those directions into consideration. Alternatively, the transformation process can be performed so that all line segments are in the same direction, in other words, so that all reference lines LS (polylines) are parallel to each other, before subsequent processing is carried out. In this case, it is preferable to transform the coordinates of the reference line LS so that it is parallel to one of the three-dimensional coordinate axes that constitute a plane set in the target range. For example, in Figure 3, by transforming the coordinates of the reference line LS, a reference line LS parallel to the Y-axis direction is generated. By having the coordinate transformation means 112 transform the reference line LS so that it is parallel to the Y-axis (or X-axis), the number of steps for the transformation process is increased, but the subsequent calculation process becomes easier, which is preferable.
[0031] (Means for setting cross-sectional line segments) The section line segment setting means 102 is a means for setting "section line segments". Here, a section line segment is a line segment that is constructed on a horizontal projection plane, similar to the reference line LS, that is, a line segment on a plane formed by coordinates that have different planar coordinates (X,Y) but the same elevation (Z). Furthermore, this section line segment is a line segment perpendicular to the reference line LS, and therefore the direction of the section line segment coincides with the direction when the cross-sectional shape is projected onto the horizontal plane (i.e., the cross-sectional direction). The section line segment setting means 102 sets multiple (three in the figure) section line segments at regular intervals (or irregular intervals) along the reference line LS, as shown in Figure 4.
[0032] (First extension area setting means) The first expansion area setting means 103 is a means for setting the "first expansion area". Here, the first expansion area is an area configured on a horizontal projection plane, that is, a planar range formed on a plane of the same elevation (Z). Furthermore, this first expansion area is set based on the cross-sectional line segment LC, and therefore the first expansion area setting means 103 sets the first expansion area for each cross-sectional line segment LC. For example, in Figure 3, the first expansion area RA01 is set by extending it by a predetermined width (hereinafter referred to as the "first expansion width") along the reference line LS, with the cross-sectional line segment LC as the center. Although Figure 3 shows an example where the cross-sectional line segment LC is extended by the first expansion width on both sides, the first expansion area RA01 can also be set by extending it on only one side of the cross-sectional line segment LC, or the left and right sides of the cross-sectional line segment LC can be extended by different widths (lengths) to set the first expansion area RA01. The value of the first expansion width and the manner in which the cross-sectional line segment LC is extended based on the cross-sectional line segment LC should be determined in advance.
[0033] (Temporary section line setting means) The provisional section line setting means 104 is a means for extracting a "first measurement point" and setting a "provisional section line". Here, the first measurement point is a 3D measurement point whose planar coordinates (X,Y) are included in the first extended region RA01, that is, a 3D measurement point that is included within the first extended region RA01 when projected onto a horizontal projection plane. On the other hand, the provisional section line is a line segment formed by the first measurement point relating to the same first extended region RA01. The procedure for the provisional section line setting means 104 to extract a first measurement point and set a provisional section line based on that first measurement point will be described in detail below.
[0034] First, the provisional section line setting means 104 reads 3D measurement points from the point cloud data storage means 115 and extracts the first measurement points included in the first extended region RA01 by comparing the 3D measurement points with the first extended region RA01. However, if the coordinates of the reference line LS have been transformed by the coordinate transformation means 112, the 3D measurement points are also subjected to the same coordinate transformation process before the first extended region RA01 and the 3D measurement points are compared. Of course, in this case, only the planar coordinates of the 3D measurement points are transformed. The provisional section line setting means 104 also extracts the first measurement points for each of the first extended regions RA01. Therefore, if different coordinate transformations are performed on multiple reference lines LS, such as when the reference line LS is formed by multiple polylines, the coordinate transformation of the 3D measurement points is performed according to each reference line LS before they are compared with the first extended region RA01.
[0035] When a first measurement point is extracted for each first extended region RA01, the provisional section line setting means 104 projects the first measurement point onto a vertical plane in the cross-sectional direction (for example, a vertical plane containing the section line segment LC). Figure 5(a) is a cross-sectional view showing multiple first measurement points P01 projected onto a vertical plane containing the section line segment LC. When the first measurement points P01 included in the first extended region RA01 are projected onto the same vertical plane, the point density of the first measurement points P01 increases in areas that are actually flat, such as the top surface and inclined surface of the embankment. Then, a line segment is generated on the same vertical plane based on the first measurement points P01 arranged on the vertical plane. The line segment generated here is the "provisional section line LCP" shown in Figure 5(b). When generating the provisional section line LCP based on the first measurement points P01 on the vertical plane, conventional calculation methods such as the RANSAC (Random Sampling Consensus) method and the least squares method can be used.
[0036] (Second extension area setting means) The second expansion area setting means 105 is a means for setting the "second expansion area". Here, the second expansion area is an area configured on a horizontal projection plane, similar to the first expansion area RA01, that is, a planar range formed on a plane of the same elevation (Z). Furthermore, this second expansion area is set based on the cross-sectional line segment LC, and therefore the second expansion area setting means 105 sets the second expansion area for each cross-sectional line segment LC. For example, the second expansion area can be set by extending by a predetermined width (hereinafter referred to as the "second expansion width") along the reference line LS with the cross-sectional line segment LC as the center. However, as shown in Figure 6, the second expansion area RA02 is set by a second expansion width that is shorter in dimension (length) than the first expansion width used to set the first expansion area RA01. For example, if the first expansion width is set to 1m, the second expansion width can be set to 0.1m. It is advisable to pre-determine the value of the second expansion width and the manner in which the expansion is performed based on the cross-sectional line segment LC.
[0037] (Cross section constituent point extraction means) The cross-sectional point extraction means 106 is a means for extracting "second measurement points" and "cross-sectional point components". Here, a second measurement point is a 3D measurement point whose planar coordinates (X,Y) are included in the second extended region RA02, that is, a 3D measurement point that is included within the second extended region RA02 when projected onto a horizontal projection plane. On the other hand, a cross-sectional point component is a point that constitutes a typical cross-sectional shape of the second extended region RA02. The procedure by which the cross-sectional point extraction means 106 extracts second measurement points and extracts cross-sectional point component components based on those second measurement points will be described in detail below.
[0038] First, the cross-sectional point extraction means 106 reads 3D measurement points from the point cloud data storage means 115 and extracts the second measurement points included in the second extended region RA02 by comparing the 3D measurement points with the second extended region RA02. However, if the coordinates of the reference line LS have been transformed by the coordinate transformation means 112, the 3D measurement points are also subjected to the same coordinate transformation process before being compared with the second extended region RA02. Of course, in this case, only the planar coordinates of the 3D measurement points are transformed. The cross-sectional point extraction means 106 also extracts the second measurement points for each second extended region RA02. Therefore, if different coordinate transformations are performed on multiple reference lines LS, such as when the reference line LS is formed by multiple polylines, the coordinate transformation of the 3D measurement points is performed according to each reference line LS before being compared with the second extended region RA02.
[0039] When a second measurement point is extracted for each second extended region RA02, the cross-sectional point extraction means 106 projects the second measurement point onto a vertical plane in the cross-sectional direction (for example, a vertical plane containing the cross-sectional line segment LC). Then, from among the second measurement points projected onto the vertical plane, the second measurement point that is in the vicinity of the provisional cross-sectional line LCP on that vertical plane is extracted as a "cross-sectional point". Specifically, as shown in Figure 7(a), a range based on the provisional cross-sectional line LCP (hereinafter referred to as the "neighborhood range RN") is set, and as shown in Figure 7(b), the second measurement point P02 within the neighborhood range RN is extracted as a cross-sectional point. The neighborhood range RN can be set by extending the provisional cross-sectional line LCP by a predetermined threshold (hereinafter referred to as the "separation threshold WD"), as shown in Figure 7(b). In this case, the cross-sectional point extraction means 106 extracts cross-sectional points on the condition that the distance from the provisional cross-sectional line LCP is less than the separation threshold WD.
[0040] (Cluster generation means) The cluster generation means 109 is a means for generating "clusters" based on the cross-sectional constituent points. This cluster is a conventionally known concept, and is, so to speak, a collection (block) of cross-sectional constituent points. In generating clusters, various conventional techniques can be used, such as generating clusters based on the condition that the separation between cross-sectional constituent points is small (i.e., the distance between points is below a threshold). Of course, the cluster generation means 109 generates clusters for each second extended region RA02 (i.e., for each cross-sectional line segment LC).
[0041] (Noise reduction means) The noise reduction means 110 is a means of removing cross-sectional constituent points (noise) that are not suitable for forming the cross-sectional shape among the cross-sectional constituent points. Figure 8 is a model diagram showing the situation in which noise NS is removed from cluster CL generated by the cluster generation means 109. In this figure, the noise reduction means 110 extracts the cross-sectional constituent points related to the cluster CL enclosed by the circle in the upper right as noise NS (upper part of the figure) and removes that noise NS (lower part of the figure). When extracting noise NS, it is possible to select cluster CLs in which the total number of cross-sectional constituent points constituting the cluster CL is less than a predetermined number (hereinafter referred to as the "cluster threshold"), and then extract the cross-sectional constituent points included in that cluster CL as noise NS. In this case, it is possible to specify that cluster CLs are selected in which the total number of cross-sectional constituent points is less than or equal to the cluster threshold, or to specify that cluster CLs are selected in which the total number of cross-sectional constituent points is less than or equal to the cluster threshold. Of course, the noise reduction means 110 removes noise NS for each second extended region RA02 (i.e., each cross-sectional line segment LC).
[0042] (Means for extracting cross-sectional change points) The section change point extraction means 107 is a process that extracts "section change points" from among the section constituent points. As previously described, these section change points are points that show significant changes, such as "points of transition" and "points of transition," in the cross-sectional shape obtained by projecting the section constituent points onto a vertical plane. The procedure by which the section change point extraction means 107 extracts section change points will be explained in detail below.
[0043] First, the cross-section change point extraction means 107 generates an arbitrary number of line segments (hereinafter referred to as "contour lines") based on the cross-section constituent points extracted by the cross-section constituent point extraction means 106. If the noise NS is removed by the noise reduction means 110, the cross-section change point extraction means 107 will naturally generate contour lines based on the cross-section constituent points from which the noise NS has been removed. In generating these contour lines, various conventional techniques can be utilized, such as connecting points with distances below a threshold, or employing a "contour line extraction" method using the ConcaveHull function.
[0044] Once a contour line is generated based on the cross-sectional constituent points, the cross-sectional change point extraction means 107 simplifies the vertices of the contour line and extracts the vertices that constitute the contour line. Various conventional techniques can be used to simplify the vertices of the contour line, such as employing the "Douglous-Pecker" method. The cross-sectional change point extraction means 107 then extracts the vertices of the contour line as "cross-sectional change points." Of course, the cross-sectional change point extraction means 107 extracts cross-sectional change points for each second extended region RA02 (i.e., for each cross-sectional line segment LC).
[0045] (Breakline generation means) The breakline generation means 113 is a means for generating breaklines using cross-sectional change points extracted by the cross-sectional change point extraction means 107. Specifically, it generates breaklines by connecting the cross-sectional change points extracted for each cross-sectional line segment LC in the longitudinal direction (i.e., the direction of the reference line LS). Multiple types of cross-sectional change points may be extracted for one cross-sectional line segment LC, but the breakline generation means 113 selects similar cross-sectional change points (those with roughly the same elevation) and connects them in the longitudinal direction.
[0046] (Process flow) The main processes of the breakline setting support system 100 of the present invention will be described in detail below with reference to Figure 9. Figure 9 is a flowchart showing an example of the main processing flow of the breakline setting support system 100 of the present invention until a breakline is generated. The center column shows the processes to be executed, the left column shows what is necessary for those processes, and the right column shows what results from those processes.
[0047] To extract cross-sectional change points and generate breaklines, first, a reference line LS is set by the reference line setting means 101 as shown in Figure 9 (Step 201 in Figure 9), and the coordinates of the reference line LS are calculated by the reference line calculation means 111. Furthermore, if the breakline setting support system 100 is equipped with a coordinate transformation means 112, the coordinates of the reference line LS are transformed (Step 202 in Figure 9).
[0048] When the reference line LS is set (or when the coordinates of the reference line LS are transformed), the section line segment LC is set by the section line segment setting means 102 (Step 203 in Figure 9). Next, the first extended region RA01 is set by the first extended region setting means 103 (Step 204 in Figure 9), the first measurement point P01 is extracted by the provisional section line setting means 104 (Step 205 in Figure 9), and the provisional section line LCP is set (Step 206 in Figure 9).
[0049] Once the provisional cross-section line LCP is set, the second extended region RA02 is set by the second extended region setting means 105 (Step 207 in Figure 9), the second measurement point P02 is extracted by the cross-section component point extraction means 106 (Step 208 in Figure 9), and the cross-section component points are extracted (Step 209 in Figure 9).
[0050] Once the cross-sectional points are extracted, the cluster generation means 109 generates clusters CL (Step 210 in Figure 9), and the noise removal means 110 removes noise NS from the cross-sectional points (Step 211 in Figure 9). Then, the cross-sectional change point extraction means 107 extracts cross-sectional change points from the cross-sectional points (Step 212 in Figure 9), and the breakline generation means 113 generates breaklines (Step 213 in Figure 9).
[0051] Up to this point, we have described a specification in which the first extended region RA01 is set, and then the second extended region RA02 is set, that is, the extended regions are set in two stages before the cross-sectional change points are extracted. However, the breakline setting support system 100 of the present invention can also be specified to set the extended regions in three or more stages before the cross-sectional change points are extracted. Figure 10 is a flowchart showing an example of the main processing flow of the breakline setting support system 100 of the present invention from setting the extended regions in three stages until a breakline is generated. The center column shows the processes to be executed, the left column shows what is necessary for those processes, and the right column shows what results from those processes. Note that even in the specification in which the extended regions are set in three or more stages, Step 201 (setting the reference line) to Step 209 (extracting cross-sectional constituent points) and Step 210 (clustering) to Step 213 (setting the breakline) shown in Figure 9 are the same processes, so these processes are omitted in Figure 10, and similarly in the following explanation, these processes will also be omitted.
[0052] When setting the expansion region in three stages, once the first cross-sectional point is extracted (Step 209 in Figure 9), the provisional cross-sectional line setting means 104 sets the second provisional cross-sectional line LCP based on the first cross-sectional point (Step 214 in Figure 10). The second provisional cross-sectional line LCP is set using the same procedure as the first provisional cross-sectional line LCP, and of course, it is set for each second expansion region RA02 (i.e., each cross-sectional line segment LC). Furthermore, the third expansion region is set for each second expansion region RA02 (i.e., each cross-sectional line segment LC) using the same procedure as the first expansion region RA01 and the second expansion region RA02 (Step 215 in Figure 10). However, the third expansion region is set with a third expansion width that is shorter in dimension (length) than the second expansion width used to set the second expansion region RA02.
[0053] Once the third extension region is set, a third measurement point is extracted for each second extension region RA02 (i.e., for each section line LC) using the same procedure as for the first measurement point P01 and the second measurement point P02 (Step 216 in Figure 10). Specifically, 3D measurement points that are included within the third extension region when projected onto the horizontal projection plane are extracted as third measurement points. Once the third measurement points are extracted, a second set of section constituent points is extracted for each second extension region RA02 (i.e., for each section line segment LC) using the same procedure as for the first set of section constituent points (Step 209 in Figure 9) (Step 217 in Figure 10). Specifically, among the third measurement points projected onto the vertical plane in the section direction, the third measurement points that are near the second provisional section line LCP are extracted as section constituent points.
[0054] Then, clusters are generated based on the second set of cross-sectional points (Step 210 in Figure 9), and as explained below using Figure 9, noise NS is removed from the cross-sectional points (Step 211 in Figure 9), cross-sectional change points are extracted from the cross-sectional points after noise NS removal (Step 212 in Figure 9), and break lines are generated based on the cross-sectional change points (Step 213 in Figure 9). [Industrial applicability]
[0055] The breakline setting support system of the present invention can be used to create 3D models of civil engineering structures such as bridges, retaining walls, cut slopes, and embankment slopes, as well as artificial structures such as apartment buildings and office buildings, and can also be used to create 3D models of natural terrain. According to the present invention, accurate 3D models of terrain can be created, and as a result, terrain changes can be appropriately grasped by comparing 3D models from two different periods, and thus effective disaster countermeasures can be implemented. Therefore, the present invention is not only industrially applicable but also has the potential to make a significant contribution to society. [Explanation of Symbols]
[0056] 100 Breakline setting support system of the present invention 101 Reference line setting means 102 Cross-sectional line segment setting means 103 First extended area setting means 104 Temporary section line setting means 105 Second extended area setting means 106 Section constituent point extraction means 107 Cross-sectional change point extraction means 108 Display means 109 Cluster generation means 110 Noise reduction means 111 Reference line calculation means 112 Coordinate transformation means 113 Breakline generation means 114 Image data storage means 115 Point cloud data storage means CL cluster LC section line segment LCP provisional cross section line LS reference line NS Noise P01 First measurement point P02 Second measurement point RA01 First Expansion Area RA02 Second Expansion Area RN Neighborhood Range WD separation threshold
Claims
1. A system that assists in setting break lines from three-dimensional measurement points. A means for setting a reference line on a horizontal projection plane, A means for setting a cross-sectional line segment perpendicular to the aforementioned reference line, A first expansion region setting means for setting a first expansion region that is expanded by a first expansion width along the reference line with respect to the aforementioned cross-sectional line segment, A provisional section line setting means for extracting a first measurement point which is a measurement point included in the first extended region, and setting a provisional section line composed of line segments on a vertical plane based on the first measurement point, A second expansion region setting means for setting a second expansion region that is expanded by a second expansion width along the reference line with respect to the aforementioned cross-sectional line segment, A means for extracting a second measurement point which is a measurement point included in the second extended region, and for extracting cross-sectional constituent points which are located near the provisional cross-sectional line when the second measurement point is projected onto a vertical plane, The system includes a means for extracting cross-sectional change points that are transition points or transition points in the vertical plane from among the aforementioned cross-sectional constituent points, The second expansion width is shorter than the first expansion width. The cross-sectional point extraction means extracts the cross-sectional points on the condition that the distance from the provisional cross-sectional line is below a predetermined separation threshold, Based on a plurality of cross-sectional change points, the break line can be set. A breakline setting support system characterized by the following features.
2. The system further includes a display means for displaying an image associated with three-dimensional coordinates, The operator can set the reference line using the reference line setting means while checking the image. The breakline setting support system according to feature 1.
3. A cluster generation means that generates clusters of cross-sectional constituent points based on the separation between the aforementioned cross-sectional constituent points, The system further comprises noise reduction means for removing as noise cross-sectional points related to a cluster in which the total number of cross-sectional points constituting the cluster falls below a predetermined cluster threshold, The cross-sectional change point extraction means extracts the cross-sectional change points from the cross-sectional constituent points excluding the noise. The breakline setting support system according to feature 1.
4. The system further includes a coordinate transformation means that rotates the reference line so that it is parallel to an axis set on a horizontal plane, and transforms the planar coordinates of the measurement points in accordance with the rotation of the reference line, The provisional section line setting means extracts the first measurement point based on the measurement point whose planar coordinates have been transformed, The cross-sectional point extraction means extracts the second measurement point based on the measurement point whose planar coordinates have been transformed. The breakline setting support system according to feature 1.