JP7650178B2 - Height information system - Google Patents

Description

This invention relates to a technology that assigns "height information" such as elevation or height difference from a reference surface, and more specifically, to a height information assignment system that can assign appropriate height information after extracting appropriate measurement points from multiple measurement points (three-dimensional coordinate points) based on the gradient between the measurement points.

Recently, there has been an increasing demand for topographical information (spatial information), and an increasing number of managers are requesting spatial information about facilities, such as their shapes and locations, in order to more thoroughly manage facilities installed on or along roads. At the same time, advanced maintenance and management of social infrastructure (hereinafter simply referred to as "social infrastructure") is considered an important issue in realizing Society 5.0, which is currently being promoted jointly by the public and private sectors. Furthermore, as autonomous driving technology becomes more practical, there is a strong demand from many quarters for various spatial information about roads, including road edges (road boundaries).

Traditionally, two-dimensional (2D) planar drawings (plan views) such as topographical maps have been the mainstream for showing spatial information. Although plan views may show "height information" such as contour lines and endpoint elevations, their main focus is on showing planar positions, making it difficult to grasp the target area as a three-dimensional (3D) space. On the other hand, in recent years, advances in measurement technology have made it easy to obtain large collections of three-dimensional measurement points (hereafter referred to as "three-dimensional point clouds"), and advances in information technology have also made it easy to handle these three-dimensional point clouds.

For example, to obtain a 3D point cloud of terrain including roads, measurement methods such as aerial photogrammetry, airborne laser measurement, terrestrial laser measurement, and MMS (Mobile Mapping System) are popular. Among these, MMS is a moving vehicle equipped with sensors such as a laser scanner, a camera, a satellite positioning system (GNSS: Global Navigation Satellite System) for acquiring the vehicle's own position, an IMU (Inertial Measurement Unit), and odometry, which allows the laser scanner to acquire a 3D point cloud while moving on the roadway.

Three-dimensional point clouds obtained by aerial photogrammetry or airborne laser measurement are generally used as a three-dimensional model (hereafter referred to as a "3D model") of the terrain being measured. This 3D model represents the target terrain in three-dimensional coordinates, and is a terrain model represented by DSM (Digital Surface Model), DEM (Digital Elevation Model), or DTM (Digital Terrain Model).

A 3D model is usually made up of small regions that divide the target planar range. These small regions are also called meshes, and are formed, for example, by dividing the area into orthogonal grids, and each small region has a representative point. Since the 3D point cloud obtained by measurement is often random data (data that is irregularly arranged on a plane), geometric calculations are often used to provide height information to the representative points of the small regions. This calculation method includes a TIN (Triangulated Irregular Network) method that finds height using an irregular triangular network formed from random data, a Nearest Neighbor method that uses the nearest laser measurement point, as well as the Inverse Distance Weighting (IDW), Kriging method, and averaging method.

3D models allow the target terrain to be grasped in both two and three dimensions, and can therefore be used for a wider variety of purposes than traditional floor plans. However, 3D models based on measurement results only provide spatial information based on three-dimensional coordinates, and cannot show the attributes of features. In other words, just by looking at a 3D model, it is not possible to understand where the outer edges of office buildings (the so-called edges) are, or where the edges of roads are.

Adding feature attribute information to a 3D model requires a survey of the features. This means that a manual survey is required, such as having workers go directly to the site and record the information they see on a drawing, or visually extracting the attributes of the features while examining aerial photographs. However, since roads, for example, generally have a considerable length, the amount of work required for the survey is enormous, and considering the labor and time involved, it is very costly.

As mentioned above, traditionally, plan views were mainly used. In some cases, these plan views were used as raster or vector data (i.e., digitized), and in other cases, features were converted into shapes (polylines or polygons) and given attribute information. With recent advances in machine learning technology, it has become possible to mechanically (automatically) extract features from aerial photographs or plan views and extract attribute information along with the shapes. In this way, there are many cases where plan views (hereafter referred to as "2D map information") that do not include height information such as elevation but do include attribute information of features are prepared separately. Using this 2D map information makes it possible to omit (or significantly reduce) manual surveys of features.

Therefore, Patent Document 1 proposes a technology for generating three-dimensional feature shape lines by adding elevations to the feature shape lines shown in two-dimensional map data using point clouds obtained by measurement.

JP 2020-013351 A

By aligning the 3D model and the 2D map data on a plane using their respective planar coordinates, it is possible to add attribute information to the 3D model, and height information can be added to the attribute information. In other words, height information can be added to road edges and the edges of office buildings, for example.

However, the inventors of the present application have found that the 3D point cloud (i.e., 3D model) obtained by measuring a road contains a lot of noise. In other words, they have confirmed that even if height information is added to the road edge using the above-mentioned method, the actual road edge tends not to be accurately reproduced. This is thought to be due to various facilities installed on the road. More specifically, 3D measurement points (hereinafter referred to as "3D measurement points") obtained by aerial photogrammetry, aerial laser measurement, etc. may not measure the road surface but electric wires and signs located above it, which correspond to noise for the road. When height information was actually added to the road edge without any special processing, in many cases there was a variation of several meters to 10 meters above and below.

The objective of the present invention is to solve the problems of the conventional technology, that is, to provide a height information assignment system that can assign height information to features while eliminating as many 3D measurement points as possible that cause noise.

The present invention focuses on the point that, from among 3D measurement points that are arranged on a line or within a narrow strip (hereinafter simply referred to as a "strip") and have vertical variation, 3D measurement points with relatively small height information are selected and height information is assigned to features, and is an invention based on a previously unseen idea.

The height information assignment system of the present invention is a system that assigns height information to a baseline (or band area) based on multiple target measurement points arranged on the baseline (or inside the band area), and is equipped with a point-to-point gradient calculation means and a suitable point extraction means. The baseline (or band area) has a starting point side and an end point side set, and the multiple target measurement points are arranged in a forward direction (direction from the starting point side to the end point side). The point-to-point gradient calculation means is a means for calculating the point-to-point forward gradient (forward gradient between the target measurement points) based on the three-dimensional coordinates related to the two target measurement points, and the suitable point extraction means is a means for extracting a target measurement point on the end side related to the point-to-point forward gradient as a suitable point when the point-to-point forward gradient satisfies a predetermined gradient requirement. Then, height information is assigned to the baseline (or band area) based on the three-dimensional coordinates related to the suitable point.

The height information providing system of the present invention can also be configured to extract a target measurement point on the end point side related to the point-to-point forward gradient as a suitable point when the point-to-point forward gradient falls below a predetermined forward gradient threshold. In this case, the point-to-point gradient calculation means calculates the point-to-point forward gradient with the suitable point as the start point side, and the point-to-point gradient calculation means repeatedly calculates the point-to-point forward gradient while changing the target measurement point in the forward direction until a suitable point is extracted by the suitable point extraction means.

The height information assignment system of the present invention can also extract appropriate points based on the point-to-point forward gradient and point-to-point reverse gradient calculated by the point-to-point gradient calculation means. Here, the point-to-point reverse gradient is a gradient in the reverse direction (from the end point side to the start point side) between the target measurement points, with the appropriate point on the end point side. In this case, when the point-to-point reverse gradient falls below a predetermined reverse gradient threshold, the appropriate point extraction means extracts the target measurement point on the start point side related to the point-to-point reverse gradient as an appropriate point, and the point-to-point gradient calculation means repeatedly calculates the point-to-point forward gradient while changing the target measurement point in reverse order until the appropriate point is extracted by the appropriate point extraction means. Then, height information is assigned to the baseline (or band area) using the three-dimensional coordinates related to the appropriate point extracted based on the point-to-point forward gradient and extracted based on the point-to-point reverse gradient.

The height information assignment system of the present invention can also assign height information to an endless circular baseline (or band area) whose starting point and end point coincide. In this case, one direction of the baseline (or band area) (e.g., clockwise) is the forward direction, and the opposite direction (e.g., counterclockwise) is the reverse direction.

The height information providing system of the present invention may further include a branch point setting means and a target range dividing means. The branch point setting means is a means for setting a target measurement point (i.e., the lowest point) having the smallest height information among a plurality of target measurement points as a branch point, while the target range dividing means is a means for dividing the range into two baselines (or band areas) with the branch point as the boundary. In this case, the point-to-point gradient calculation means calculates the point-to-point gradient for each baseline (or band area), and the appropriate point extraction means extracts appropriate points for each baseline (or band area).

The height information providing system of the present invention may further include a base point setting means and a target measurement point selection means. The base point setting means is a means for setting a starting point and a destination point from among a plurality of target measurement points and setting the starting point and the destination point as appropriate points, while the target measurement point selection means is a means for selecting one of the target measurement points located between the two appropriate points as the target measurement point. In this case, the point-to-point gradient calculation means calculates a point-to-point forward gradient based on the appropriate point on the end point side and the three-dimensional coordinates related to the target point, and calculates a point-to-point reverse gradient based on the appropriate point on the start point side and the three-dimensional coordinates related to the target point. The appropriate point extraction means extracts the target point as an appropriate point when the point-to-point forward gradient related to the target point is below the forward gradient threshold and the point-to-point reverse gradient related to the target point is below the reverse gradient threshold. The target measurement point selection means repeatedly selects the target measurement point as the target measurement point while changing the target measurement point in ascending order of height information until the appropriate point is extracted by the appropriate point extraction means.

The height information assignment system of the present invention can also select the target measurement point having the smallest height information (i.e., the lowest point) among the target measurement points as the new measurement point of interest after excluding the appropriate points already extracted by the appropriate point extraction means.

The height information providing system of the present invention has the following effects.
(1) It is possible to add height information to target features such as roads after removing 3D measurement points that are noise for the target features. As a result, it is possible to reproduce, for example, actual road edges more accurately.
(2) Since no judgment by the user is required, the effort required by the user is reduced and human errors due to subjectivity, etc. are avoided, making it possible to obtain efficient and highly accurate results.

1A is a plan view showing two-dimensional map information having two-dimensional information (plane coordinates and attribute information) of road edges, and FIG. 1B is a longitudinal section formed of three-dimensional measurement points arranged at road edges. FIG. 2A is a plan view showing three-dimensional measurement points arranged on a baseline, and FIG. 2B is a plan view showing three-dimensional measurement points arranged in a band region. FIG. 1 is a block diagram showing the main configuration of a height information providing system according to the present invention. FIG. 1A is a model diagram illustrating the process of arranging target measurement points related to an endless circular road edge, and FIG. 1B is a model diagram illustrating the process of dividing the road edge into two at a branch point as the boundary. FIG. 4 is a schematic longitudinal section showing nine target measurement points arranged by a measurement point arrangement means. FIG. 2 is a flow chart showing the flow of main processes when the height information providing system of the present invention is used. (a) is a longitudinal section arranged by the measurement point arrangement means and without any special noise processing, (b) is a longitudinal section after median filter processing, and (c) is a longitudinal section showing a solid line connecting the appropriate points superimposed with the true value of the road edge elevation. FIG. 11 is a flow diagram showing the main process flow for extracting appropriate points using the forward pass method. (a) is a longitudinal section diagram showing a schematic diagram of the situation in which an appropriate point is extracted using the forward pass method with the starting point as the reference point, and (b) is a longitudinal section diagram showing a schematic diagram of the situation in which an appropriate point is extracted using the forward pass method with the already extracted appropriate point as the reference point. FIG. 11 is a flow diagram showing the main process flow for extracting the appropriate points using a round-trip method. FIG. 13 is a longitudinal sectional view showing a schematic diagram of a situation in which a reverse direction appropriate point is extracted by a round trip method with a destination point as a reference point. FIG. 1 is a block diagram showing the main configuration of a height information providing system according to the present invention. FIG. 2 is a flow chart showing the flow of main processes when the attention-type height information providing system of the present invention is used. (a) is a longitudinal section diagram showing a schematic diagram of a situation in which the focus-based height information assigning system extracts an appropriate point based on a starting point and a destination point, and (b) is a longitudinal section diagram showing a schematic diagram of a situation in which the focus-based height information assigning system extracts an appropriate point based on an already extracted appropriate point and a destination point. A longitudinal section in which the four extracted appropriate points are connected by dashed lines.

An example of an embodiment of the height information assignment system of the present invention will be described with reference to the drawings. The height information assignment system of the present invention is particularly effective in cases where height information is assigned to features with shapes that are linear or band-like, with prominent axial dimensions, and can be applied to various features with such shapes, but for convenience, an example will be described here in which a "road edge (road boundary line)" is the target feature. The "height information" assigned by the information assignment system of the present invention also includes various information related to height, such as the relative height difference from an arbitrary reference surface, but for convenience, an example will be described here in which "altitude" is used as the height information.

1. Overall Overview Figure 1(a) is a plan view showing "two-dimensional map information" that includes two-dimensional information (plane coordinates and attribute information) of road edges, and Figure 1(b) is a longitudinal section consisting of "three-dimensional measurement points" placed on road edges. As mentioned above, the two-dimensional map information is a plan view that does not include height information such as altitude, but has plane coordinates and attribute information of features, and the three-dimensional measurement points are three-dimensional measurement points obtained by aerial photogrammetry or airborne laser measurement.

The longitudinal section of FIG. 1(b) shows three-dimensional measurement points directly above the road edge or three-dimensional measurement points around the road edge (for example, representative points of a mesh that hangs over the road edge) lined up from the starting point side to the end point side, with the horizontal axis showing the distance from the starting point side to each three-dimensional measurement point (distance between points, X-axis coordinate difference, Y-axis coordinate difference, etc.), and the vertical axis showing the elevation of the three-dimensional measurement point. In other words, the longitudinal section of FIG. 1(b) shows a plurality of three-dimensional measurement points lined up in order from the three-dimensional measurement point closest to the starting point (hereinafter referred to as the "starting point") to the three-dimensional measurement point closest to the end point (hereinafter referred to as the "arrival point"), with each point plotted up and down according to its elevation. Note that the road edge shown in FIG. 1(a) is an endless ring (so-called loop-shaped), so in FIG. 1(b), the road edge is cut at a specified three-dimensional measurement point and then developed. Therefore, in a case like FIG. 1, the starting point and the arrival point coincide.

As described above, the height information assignment system of the present invention is particularly effective in cases where height information is assigned to linear features (such as road edges) as shown in FIG. 2(a) or strip-shaped features as shown in FIG. 2(b). In the example shown in FIG. 2(a), the linear feature is treated as a "baseline," and the longitudinal section of FIG. 1(b) is created using three-dimensional measurement points placed on or around this baseline. On the other hand, in the example shown in FIG. 2(b), the strip-shaped feature is treated as a "strip area," and the longitudinal section of FIG. 1(b) is created using three-dimensional measurement points placed within this strip area.

As already explained, 3D point clouds (i.e., 3D models) obtained by measuring roads tend to contain a lot of noise. Therefore, when the multiple 3D measurement points that make up the 3D point cloud are represented on the longitudinal section shown in Figure 1 (b), it becomes clear that there is considerable variation up and down. Therefore, the present invention does not use 3D measurement points located on the upper part of the longitudinal section (i.e., high altitude), but instead uses only 3D measurement points located on the lower part of the longitudinal section (i.e., low altitude) to provide height information for features. In other words, 3D measurement points thought to represent power lines, etc. are excluded, and only 3D measurement points thought to represent road edges are selected and used.

2. Appropriate Image Selection System Next, the height information assignment system of the present invention will be described in detail with reference to the drawings. Fig. 3 is a block diagram showing the main components of the height information assignment system 100 of the present invention. As shown in this figure, the height information assignment system 100 of the present invention is configured to include a point-to-point gradient calculation means 101 and an appropriate point extraction means 102, and can also be configured to include a branch point setting means 103, a target range division means 104, a feature extraction means 105, a target measurement point extraction means 106, a measurement point arrangement means 107, a height information assignment means 108, a 3D topographical model storage means 109, a 2D map information storage means 110, a height information storage means 111, and the like.

The point-to-point gradient calculation means 101, the appropriate point extraction means 102, the branch point setting means 103, the target range division means 104, the feature extraction means 105, the target measurement point extraction means 106, the measurement point arrangement means 107, and the height information assignment means 108 that constitute the height information assignment system 100 can be manufactured as dedicated devices, or a general-purpose computer device can be used. This computer device is equipped with a processor such as a CPU, storage and memory such as ROM and RAM, and some also include input means such as a mouse and keyboard and a display. Examples of such devices include tablet computers (such as iPad (registered trademark)) and mobile terminal devices such as smartphones, or personal computers (PCs) and servers.

The 3D terrain model storage means 109, 2D map information storage means 110, and elevation information storage means 111 can be stored in a computer device or can be built in a database server. When built in a database server, they can be placed on a local network (LAN: Local Area Network) or can be a cloud server that stores data via the Internet (e.g., wireless communication).

Below, we will explain in detail each of the main elements that make up the height information provision system 100.

(2D map information storage means and feature extraction means)
The 2D map information storage means 110 is a means for storing 2D map information, and the feature extraction means 105 is a means for extracting a target feature (in this case, a road edge) from the 2D map information. This feature extraction means 105 can be designed to automatically extract a type of feature (road edge) specified by an operator, or can be designed to extract the feature by operating a pointing device or a keyboard while visually viewing the 2D map information.

(3D terrain model storage means and target measurement point extraction means)
The 3D topographical model storage means 109 is a means for storing a large number of 3D measurement points (i.e., a 3D point cloud). The target measurement point extraction means 106 is a means for extracting 3D measurement points (hereinafter referred to as "target measurement points") located on a baseline (around the baseline) or within a band area. More specifically, the 3D point cloud and 2D map information are aligned in plan based on the plane coordinates, and 3D measurement points related to the baseline or band area are extracted as target measurement points.

(Means for arranging measurement points)
The measurement point arrangement means 107 is a means for selecting a starting point and a destination point from among the target measurement points, and for arranging the target measurement points in order from the starting point side to the destination point side (i.e., from the starting point to the destination point). When arranging the target measurement points, they are arranged in order on a line segment connecting the starting point and the destination point, and when selecting the starting point (or the destination point), it is advisable to select the one with the lowest elevation (hereinafter referred to as the "minimum elevation point") among the target measurement points. For example, in the case of an endless circular (so-called loop-shaped) road edge shown in FIG. 4(a), it is advisable to select the minimum elevation point among the target measurement points as the starting point and the destination point, and then arrange the target measurement points on the line segment connecting the starting point and the destination point. At this time, the target measurement points are arranged from the starting point side (the left side in the figure) in order of proximity to the starting point.

When there is a minimum elevation point along the way, like the road edge shown in Figure 4(b), this cannot be selected as the starting point or the destination point. In this case, it is advisable to set the minimum elevation point as a branch point and divide the road edge into two with the branch point as the boundary. That is, as shown in Figure 4(b), the branch point is set as the starting point and the other end as the destination point, and the subsequent processing is performed on each of the two divided road edges. Note that the branch point is set by the branch point setting means 103 (Figure 3), and the road edge is divided into two by the target range dividing means 104 (Figure 3).

Figure 5 is a schematic longitudinal section showing nine target measurement points (PA1 to PA9) arranged by the measurement point arrangement means 107. In this figure, the target measurement point PA1 is selected as the starting point, and the target measurement point PA9 is selected as the destination point, and the target measurement points PA2 to PA8 are arranged in order from the starting point side. For convenience, the direction from the starting point side to the end point side (i.e., from the starting point to the destination point) (to the right in the figure) is referred to as the "forward direction", the direction from the end point side to the starting point side (i.e., from the destination point to the starting point) (to the left in the figure) is referred to as the "reverse direction", and the gradient in the forward direction is referred to as the "forward gradient", and the gradient in the reverse direction is referred to as the "reverse gradient". For example, if we focus on the target measurement point PA4 in Figure 5, the gradient from the target measurement point PA4 to the target measurement point PA5 is the "forward gradient", and the gradient from the target measurement point PA4 to the target measurement point PA3 is the "reverse gradient". Of course, the forward and reverse gradients are calculated based on the three-dimensional coordinates of the target measurement point (three-dimensional measurement point).

(Point-to-point gradient calculation means)
The point-to-point gradient calculation means 101 is a means for calculating a forward gradient between a focused target measurement point and another target measurement point located on the end side of the focused target measurement point in a longitudinal section connecting the target measurement points shown in Fig. 5 (hereinafter referred to as "point-to-point forward gradient"). In addition to calculating the point-to-point forward gradient, the point-to-point gradient calculation means 101 can also be a means for calculating an inverse gradient between the focused target measurement point and another target measurement point located on the start side of the focused target measurement point (hereinafter referred to as "point-to-point reverse gradient").

(Appropriateness point extraction means)
The appropriate point extraction means 102 is a means for extracting an "appropriate point" based on the point-to-point forward gradient and point-to-point reverse gradient of a target measurement point. For example, when the point-to-point forward gradient of a target measurement point satisfies a predetermined "gradient requirement", the target measurement point can be extracted as an appropriate point. Here, the gradient requirement is a requirement defined by an upper gradient limit and a lower gradient limit, such as a range of -3% to 3% or a range of 1% to 3%. Therefore, when the point-to-point forward gradient is within the range defined by the gradient requirement, the appropriate point extraction means 102 extracts the target measurement point related to the point-to-point forward gradient as an appropriate point.

The appropriate point extraction means 102 can also be configured to extract a target measurement point as an appropriate point when the point-to-point forward gradient of the target measurement point falls below a predetermined threshold (hereinafter, "forward gradient threshold"). For example, in a case where the forward gradient threshold is 3%, the appropriate point extraction means 102 extracts the target measurement point related to the point-to-point forward gradient as an appropriate point when the point-to-point forward gradient falls below 3%. Furthermore, the appropriate point extraction means 102 can also be configured to extract the target measurement point as an appropriate point when the point-to-point reverse gradient of the target measurement point falls below a predetermined threshold (hereinafter, "reverse gradient threshold"). For example, in a case where the reverse gradient threshold is 2%, the appropriate point extraction means 102 extracts the target measurement point related to the point-to-point reverse gradient as an appropriate point when the point-to-point reverse gradient falls below 2%. The forward gradient threshold and reverse gradient threshold can be set to different values, or both can be set to the same value, such as 3%. The starting point and destination point can also be set as appropriate points regardless of the conditions described here.

(Height information providing means and height information storage means)
The altitude information assigning means 108 is a means for assigning an elevation to the road edge based on the appropriate point extracted by the appropriate point extracting means 102. At this time, the altitude information assigning means 108 assigns the elevation of the appropriate point (three-dimensional coordinate point) to a position corresponding to the plane coordinates of the appropriate point. The elevation assigned to the road edge (or the three-dimensional coordinates of the appropriate point) is stored in the altitude information storage means 111.

(Processing flow)
Next, the main process flow when using the height information providing system 100 of the present invention will be described with reference to Fig. 6. In Fig. 6 and the flow diagram of Fig. 8 described later, the action to be performed is shown in the center column, the things necessary for that action are shown in the left column, and the things resulting from that action are shown in the right column.

First, the feature extraction means 105 extracts the desired road edge from the 2D map information stored in the 2D map information storage means 110 (Step 210 in FIG. 6). As described above, the feature extraction means 105 can be designed to automatically extract road edges specified by an operator, or to have the operator extract the desired road edge by operating a pointing device or keyboard while visually viewing the 2D map information. Note that if the desired road edge having two-dimensional coordinates is prepared, this process (Step 210 in FIG. 6) can be omitted (skip).

Once the road edges have been extracted, the target measurement point extraction means 106 extracts the 3D measurement points located on the road edges as "target measurement points" (Step 220 in FIG. 6). The measurement point arrangement means 107 then selects a starting point and a destination point from the target measurement points, and arranges the multiple target measurement points in order from the starting point side to the end point side (i.e., from the starting point to the destination point), generating a longitudinal section that connects the target measurement points shown in FIG. 1(b) and FIG. 5 (Step 230 in FIG. 6).

When the longitudinal section is generated, the point-to-point gradient calculation means 101 calculates the "point-to-point forward gradient" and "point-to-point reverse gradient" for the target measurement point (Step 240 in FIG. 6). Then, when the point-to-point forward gradient and point-to-point reverse gradient are obtained, the appropriate point extraction means 102 extracts appropriate points based on the point-to-point forward gradient and point-to-point reverse gradient for the target measurement point (Step 250 in FIG. 6), and the height information assignment means 108 assigns an elevation to the road edge based on the three-dimensional coordinates of the appropriate points (Step 260 in FIG. 6).

As described above, the longitudinal section generated by the measurement point arrangement means 107 can be used as is for subsequent processing, or the longitudinal section generated by the measurement point arrangement means 107 can be subjected to noise processing such as median filter processing before subsequent processing. FIG. 7(a) shows a longitudinal section generated by the measurement point arrangement means 107 without any particular noise processing. In contrast, FIG. 7(b) shows a longitudinal section (dashed line in the figure) that has been subjected to median filter processing, and also shows a solid line connecting the appropriate points extracted by the appropriate point extraction means 102. In addition to the solid line connecting the appropriate points, FIG. 7(c) shows the true (actually measured) road edge elevation with a dashed line. As can be seen from FIG. 7(c), the road edge elevation given by the height information giving system 100 of the present invention is roughly the same as the true road edge elevation. In other words, the information provision system 100 can provide elevation information for road edges with a high degree of accuracy that can be used for a variety of purposes.

Now, the appropriate points are extracted according to a prescribed procedure. Below, the process of extracting appropriate points from the target measurement points will be explained in detail. The process of extracting appropriate points can be broadly divided into the "outbound method", in which extraction is performed sequentially from the starting point to the destination point, and the "round-trip method", in which extraction is performed provisionally from the starting point to the destination point and also provisionally from the destination point to the starting point. Therefore, the outbound method and the round-trip method will be explained in turn.

In the forward path method, the next appropriate point is extracted based on the appropriate point extracted by the appropriate point extraction means 102. More specifically, the appropriate point already extracted is set as the starting point side point of the inter-point sequential gradient (hereinafter referred to as the "gradient starting point"), and the target measurement point located on the end point side of the appropriate point (i.e., the gradient starting point) is set as the end point side point of the inter-point sequential gradient (hereinafter referred to as the "gradient end point"). The appropriate point extraction means 102 extracts a new appropriate point based on the inter-point sequential gradient calculated from the gradient starting point and the gradient end point. Therefore, the inter-point gradient calculation means 101 repeatedly calculates the inter-point sequential gradient while fixing the gradient starting point (the already extracted appropriate point) and changing the target measurement point toward the end point side (in the forward order) until an appropriate point is extracted, and the appropriate point extraction means 102 extracts the gradient end point of the inter-point sequential gradient as a new appropriate point when the requirements are satisfied.

Figure 8 is a flow diagram showing the main process flow for extracting appropriate points using the outbound method. As shown in this figure, when a longitudinal section connecting the target measurement points is generated by the measurement point arrangement means 107 (Step 230), an appropriate point that has already been extracted is set as the gradient starting point (Step 241 in Figure 8). However, when the first appropriate point is extracted, naturally no appropriate point has been extracted, so it is advisable to first set the starting point as an appropriate point and then set this as the gradient starting point.

When an already extracted suitable point (or starting point) is set as the gradient starting point, a target measurement point located on the end point side of the gradient starting point is searched for and set as the gradient end point (Step 242 in FIG. 8). Then, the point-to-point gradient calculation means 101 calculates the point-to-point forward gradient based on the gradient starting point and the gradient end point (Step 243 in FIG. 8), and the suitable point extraction means 102 compares the point-to-point forward gradient with the forward gradient threshold (Step 244 in FIG. 8). If the point-to-point forward gradient exceeds the forward gradient threshold (No in Step 244 in FIG. 8), a target measurement point different from this gradient end point is set as a new gradient end point (Step 242 in FIG. 8), and the calculation of the point-to-point forward gradient (Step 243 in FIG. 8) and the comparison of the point-to-point forward gradient with the forward gradient threshold (Step 244 in FIG. 8) are repeated. On the other hand, when the point-to-point forward gradient falls below the forward gradient threshold (Yes in Step 244 in FIG. 8), this gradient end point is extracted as a suitable point (Step 250), and the suitable point extracted here (i.e., the gradient end point up to this point) is set as a new gradient start point (Step 241 in FIG. 8), and the process from setting the gradient end point (Step 242 in FIG. 8) to extracting the suitable point (Step 250) is repeated.

Figure 9 is a vertical cross-sectional view showing a schematic diagram of the situation in which the appropriate point is extracted by the forward route method. In Figure 9(a), the starting point PST is first set as the gradient starting point. Then, the target measurement point PA11 is searched for as the gradient end point, and the point-to-point gradient is calculated. However, since the point-to-point gradient consisting of the starting point PST (gradient starting point) and the target measurement point PA11 (gradient end point) exceeds the gradient threshold, the target measurement point PA12 on the end point side is searched for and the point-to-point gradient is calculated. Since the point-to-point gradient consisting of the starting point PST (gradient starting point) and the target measurement point PA12 (gradient end point) also exceeds the gradient threshold, the target measurement point PA13 on the end point side is searched for and the point-to-point gradient is calculated. Then, since the point-to-point gradient consisting of the starting point PST (gradient starting point) and the target measurement point PA13 (gradient end point) falls below the gradient threshold, the target measurement point PA13 is extracted as a new appropriate point.

When the target measurement point PA13 is extracted as a suitable point, as shown in FIG. 9(b), the target measurement point PA13 is then set as a new gradient start point. Then, the target measurement point PA14 is searched for as the gradient end point, and the point-to-point forward gradient is calculated. However, since the point-to-point forward gradient consisting of the target measurement point PA13 (gradient start point) and the target measurement point PA14 (gradient end point) exceeds the forward gradient threshold, the target measurement point PA15 on the end point side is searched for and the point-to-point forward gradient is calculated. Since the point-to-point forward gradient consisting of the target measurement point PA13 (gradient start point) and the target measurement point PA15 (gradient end point) falls below the forward gradient threshold, the target measurement point PA15 is extracted as a new suitable point. In this way, the process proceeds in the forward direction (end point side) while extracting suitable points, and when a suitable point is no longer extracted, the extraction process of the suitable point is terminated, and the height information assignment means 108 assigns an elevation to the road edge based on the three-dimensional coordinates of the suitable point (Step 260).

In the round trip method, the appropriate point is first extracted by the outward method described in Figures 8 and 9. However, the appropriate point extracted here is only a provisional appropriate point (hereinafter referred to as the "forward appropriate point"). After the outward method processing, the same processing as the outward method is performed in the opposite direction to the outward path (i.e., the return path) to extract the appropriate point. More specifically, the already extracted appropriate point is set as the end point side point of the point-to-point reverse gradient (also referred to as the "gradient end point" here), and the target measurement point located on the starting point side of the appropriate point (i.e., the gradient end point) is set as the starting point side point of the point-to-point reverse gradient (also referred to as the "gradient starting point" here), and the appropriate point extraction means 102 extracts a new appropriate point based on the point-to-point reverse gradient calculated from these gradient end points and gradient starting points. Therefore, the point-to-point gradient calculation means 101 repeatedly calculates the point-to-point reverse gradient while changing the target measurement point toward the starting point (in the reverse direction) while fixing the gradient end point (the already extracted appropriate point), until an appropriate point is extracted, and the appropriate point extraction means 102 extracts the gradient starting point related to the point-to-point reverse gradient as a new appropriate point when the requirements are satisfied. However, the appropriate points extracted on the return path are also only provisional appropriate points (hereinafter referred to as "reverse direction appropriate points"). Then, the target measurement point where the forward direction appropriate point and the reverse direction appropriate point match (i.e., the target measurement point that is both a forward direction appropriate point and a reverse direction appropriate point) is determined to be the final appropriate point.

Figure 10 is a flow diagram showing the main process flow for extracting appropriate points using the round trip method. As shown in this diagram, first, a forward appropriate point is extracted using the outward path method described above (Steps 241 to 251). Once a forward appropriate point has been extracted (Step 251), the appropriate point that has already been extracted is set as the gradient end point (Step 245 in Figure 10). However, since no appropriate point has been extracted when extracting the first appropriate point on the outward path, it is best to first set the arrival point as the appropriate point and then set this as the gradient end point.

When an already extracted suitable point (or arrival point) is set as the gradient end point, a target measurement point located closer to the start point than the gradient end point is searched for and set as the gradient start point (Step 246 in FIG. 10). Then, the point-to-point gradient calculation means 101 calculates the point-to-point inverse gradient based on the gradient end point and the gradient start point (Step 247 in FIG. 10), and the suitable point extraction means 102 compares the point-to-point inverse gradient with the inverse gradient threshold (Step 248 in FIG. 10). If the point-to-point inverse gradient exceeds the inverse gradient threshold (No in Step 248 in FIG. 10), a target measurement point different from this gradient start point is set as a new gradient start point (Step 246 in FIG. 10), and the calculation of the point-to-point inverse gradient (Step 246 in FIG. 10) and the comparison of the point-to-point inverse gradient with the inverse gradient threshold (Step 248 in FIG. 10) are repeated. On the other hand, when the point-to-point reverse gradient falls below the reverse gradient threshold (Yes in Step 248 in FIG. 10), this gradient start point is extracted as a reverse direction appropriate point (Step 252 in FIG. 10), and the extracted reverse direction appropriate point (i.e., the gradient start point up to this point) is set as a new gradient end point (Step 245 in FIG. 10), and the process from setting the gradient start point (Step 246 in FIG. 10) to extracting the reverse direction appropriate point (Step 252) is repeated.

Figure 11 is a vertical cross-sectional view showing a schematic diagram of the situation in which the reverse direction appropriate point is extracted by the round trip method. In this figure, the arrival point PEN is set as the gradient end point. Then, the target measurement point PA25 is searched for as the gradient start point, and the point-to-point reverse gradient is calculated. However, since the point-to-point reverse gradient formed by the arrival point PEN (gradient end point) and the target measurement point PA25 (gradient start point) exceeds the reverse gradient threshold, the target measurement point PA24 on the start point side is searched for and the point-to-point reverse gradient is calculated. Since the point-to-point reverse gradient formed by the arrival point PEN (gradient end point) and the target measurement point PA24 (gradient start point) also exceeds the reverse gradient threshold, the target measurement point PA23 on the start point side is searched for and the point-to-point reverse gradient is calculated. Then, since the point-to-point reverse gradient formed by the arrival point PEN (gradient end point) and the target measurement point PA23 (gradient start point) falls below the reverse gradient threshold, the target measurement point PA23 is extracted as a new appropriate point.

In this way, the process proceeds in the reverse direction (towards the starting point) while extracting appropriate points, and the extraction process of appropriate points ends when no more appropriate points are extracted. The forward appropriate points are then compared with the backward appropriate points (Step 253 in FIG. 10), and the target measurement point where they match is determined as the final appropriate point (Step 254 in FIG. 10). Once the final appropriate point has been determined, the height information assignment means 108 assigns an elevation to the road edge based on the three-dimensional coordinates of the appropriate point (Step 260).

(Modification)
The height information imparting system 100 of the present invention can also be designed to extract a suitable point based on the point-to-point forward gradient and point-to-point inverse gradient related to a target measurement point of interest (hereinafter, for convenience, referred to as the "attention-type height information imparting system 100a"). Fig. 12 is a block diagram showing the main configuration of the height information imparting system 100a of the present invention. As shown in this figure, the attention-type height information imparting system 100a is configured to further include a base point setting means 112 and a target measurement point selection means 113 in addition to the respective means constituting the height information imparting system 100 shown in Fig. 3. These base point setting means 112 and the target measurement point selection means 113 can also be manufactured as dedicated means, or a general-purpose computer device can be used.

The base point setting means 112 is a means for setting the starting point and the destination point described above from among a plurality of target measurement points, and for setting these starting point and destination points as appropriate points. On the other hand, the target measurement point selection means 113 is a means for selecting one target measurement point as a "target measurement point" from among the target measurement points located between the starting point and the destination point.

Figure 13 is a flow diagram showing the main process flow when using the attention-based height information assignment system 100a. In this diagram, the center column shows the action to be performed, the left column shows what is necessary for that action, and the right column shows what results from that action.

First, the base point setting means 112 sets a starting point and a destination point from among the target measurement points (Step 301 in FIG. 13). Once the starting point and destination point have been set, the gradient starting point and gradient end point are set from among the appropriate points that have already been extracted (Step 302 in FIG. 13). However, when setting the first gradient starting point and gradient end point, naturally no appropriate points have been extracted, so it is advisable to set the starting point as the gradient starting point and the destination point as the gradient end point. Then, the focus measurement point selection means 113 selects a focus point from among the target measurement points that are located between the gradient starting point and the gradient end point (Step 303 in FIG. 13). The focus measurement point selection means 113 selects the target measurement point that is the minimum elevation point from among the target measurement points (target measurement points between the gradient starting point and the gradient end point) that are the subject of the target measurement points (target measurement points between the gradient starting point and the gradient end point) as the focus point.

Once the point of interest is selected, the point-to-point gradient calculation means 101 calculates the point-to-point forward gradient and point-to-point inverse gradient for the point of interest. More specifically, the point-to-point inverse gradient for the point of interest is calculated based on the point of interest and the gradient start point (Step 304 in FIG. 13), and the point-to-point forward gradient for the point of interest is calculated based on the point of interest and the gradient end point (Step 305 in FIG. 13).

Once the point-to-point forward gradient and point-to-point reverse gradient for the point of interest have been obtained, the point-to-point forward gradient is compared to the forward gradient threshold, and the point-to-point reverse gradient is compared to the reverse gradient threshold (Step 306 in FIG. 13). If the point-to-point forward gradient exceeds the forward gradient threshold or the point-to-point reverse gradient exceeds the reverse gradient threshold (No in Step 306 in FIG. 13), a target measurement point other than the point of interest is set as a new point of interest (Step 303 in FIG. 13), and the calculation of the point-to-point reverse gradient (Step 304 in FIG. 13) and the comparison with each gradient threshold (Step 306 in FIG. 13) are repeated. On the other hand, when the point-to-point forward gradient falls below the forward gradient threshold and the point-to-point reverse gradient falls below the reverse gradient threshold (Yes in Step 306 in FIG. 13), this point of interest is extracted as a suitable point (Step 250), and the suitable point extracted here (i.e., the point of interest up to this point) is set as a new gradient start point or gradient end point (Step 302 in FIG. 13), and the selection of the point of interest (Step 303 in FIG. 13) to the extraction of the suitable point (Step 250) are repeated.

Figure 14 is a vertical cross-sectional view showing a schematic diagram of the situation in which the focus-type height information assignment system 100a extracts appropriate points. In Figure 14(a), the starting point PST is initially set as the gradient start point, and the end point PEN is set as the gradient end point. Then, the target measurement point PA31, which is the minimum elevation point among the target measurement points located between the starting point PST (gradient start point) and the end point PEN (gradient end point), is selected as the focus point. However, although the point-to-point forward gradient consisting of the target measurement point PA31 (focus point) and the end point PEN (gradient end point) is below the forward gradient threshold, the point-to-point reverse gradient consisting of the target measurement point PA31 (focus point) and the starting point PST (gradient start point) exceeds the reverse gradient threshold. Therefore, the target measurement point PA33 is selected as the new focus point. Excluding target measurement point PA31, target measurement point PA33 is the lowest elevation point (i.e., the target measurement point with the second lowest elevation) among the target measurement points located between starting point PST and end point PEN. Then, since the point-to-point forward gradient formed by target measurement point PA33 (point of interest) and end point PEN (gradient end point) is below the forward gradient threshold, and the point-to-point reverse gradient formed by target measurement point PA33 (point of interest) and starting point PST (gradient start point) is below the reverse gradient threshold, target measurement point PA33 is extracted as the new appropriate point.

When the target measurement point PA33 is extracted as a suitable point, as shown in FIG. 14(b), the target measurement point PA33 is then set as a new gradient start point or gradient end point. Then, first, the target measurement point PA35, which is the minimum elevation point among the target measurement points located between the target measurement point PA33 (gradient start point) and the end point PEN (gradient end point), is selected as the focus point. However, although the point-to-point reverse gradient consisting of the target measurement point PA35 (focus point) and the target measurement point PA33 (gradient start point) is below the reverse gradient threshold, the point-to-point forward gradient consisting of the target measurement point PA35 (focus point) and the end point PEN (gradient end point) exceeds the forward gradient threshold. Next, the target measurement point PA32, which is the minimum elevation point among the target measurement points located between the starting point PST (gradient end point) and the target measurement point PA33 (gradient start point), is selected as the focus point. The point-to-point reverse gradient formed by the target measurement point PA32 (point of interest) and the starting point PST (gradient start point) is below the reverse gradient threshold, and the point-to-point forward gradient formed by the target measurement point PA32 (point of interest) and the target measurement point PA33 (gradient end point) is also below the forward gradient threshold, so that the target measurement point PA32 is extracted as a new appropriate point. Figure 15 shows a longitudinal section in which four appropriate points (start point PST, target measurement point PA32, target measurement point PA33, and end point PEN) are connected by dashed lines. In this way, the process continues while changing the target measurement point in ascending order of elevation while extracting appropriate points, and when no appropriate points are extracted, the extraction process of appropriate points is terminated. Then, the height information assigning means 108 assigns elevation to the road edge based on the three-dimensional coordinates of the appropriate points (Step 260).

The height information providing system of the present invention can be particularly useful for managing various facilities, including road facilities, and as map information used in automated driving. Furthermore, the present invention can provide highly accurate step information that is useful for elderly people and wheelchair users, and can also be effectively used in disaster prevention planning. The height information providing system of the present invention is not only applicable to industry, but is also expected to make a significant contribution to society.

100 Elevation information assignment system of the present invention 100a Focus-type elevation information assignment system of the present invention 101 Point-to-point gradient calculation means (of the elevation information assignment system of the present invention) 102 Appropriate point extraction means (of the elevation information assignment system of the present invention) 103 Branch point setting means (of the elevation information assignment system of the present invention) 104 Target range division means (of the elevation information assignment system of the present invention) 105 Land feature extraction means (of the elevation information assignment system of the present invention) 106 Target measurement point extraction means (of the elevation information assignment system of the present invention) 107 Measurement point arrangement means (of the elevation information assignment system of the present invention) 108 Elevation information assignment means (of the elevation information assignment system of the present invention) 109 3D topographical model storage means (of the elevation information assignment system of the present invention) 110 2D map information storage means (of the elevation information assignment system of the present invention) 111 Elevation information storage means (of the elevation information assignment system of the present invention) 112 (of the focus-type height information providing system of the present invention) Reference point setting means 113 (of the focus-type height information providing system of the present invention) Focus measurement point selection means

Claims (5)

A system for assigning height information to a baseline or a band area based on a plurality of target measurement points arranged in a forward direction from the starting point side to the end point side, the system comprising:
a point-to-point gradient calculation means for calculating a point-to-point forward gradient, which is the gradient in the forward direction between the two target measurement points , based on three-dimensional coordinates of the two target measurement points, and for calculating a point-to-point reverse gradient, which is the gradient in the reverse direction between the target measurement points from the end point side to the start point side, based on the three-dimensional coordinates of the two target measurement points;
an appropriate point extraction means for extracting the target measurement point on the end point side related to the point-to-point forward gradient as an appropriate point when the point-to-point forward gradient satisfies a predetermined gradient requirement, and for extracting the target measurement point on the start point side related to the point-to-point reverse gradient as the appropriate point when the point-to-point reverse gradient falls below a predetermined reverse gradient threshold value ;
the point-to-point gradient calculation means calculates the point-to-point forward gradient with the appropriate point as a starting point side, and calculates the point-to-point inverse gradient with the appropriate point as an end point side;
The point-to-point gradient calculation means calculates the point-to-point forward gradient while changing the target measurement point in the forward direction until the appropriate point is extracted by the appropriate point extraction means, and calculates the point-to-point reverse gradient while changing the target measurement point in the reverse direction until the appropriate point is extracted by the appropriate point extraction means;
assigning height information to the baseline or the band region based on three-dimensional coordinates of the appropriate points extracted by the point-to-point forward gradient and extracted by the point-to-point inverse gradient;
A height information providing system. The base line or the band region is an endless ring whose starting point side and ending point side coincide with each other,
One direction of the baseline or the band region is defined as the forward direction, and the opposite direction of the forward direction is defined as the reverse direction.
2. The height information providing system according to claim 1 . a branch point setting means for setting a target measurement point having minimum height information among the plurality of target measurement points as a branch point;
and a target range dividing means for dividing the base line or the band area into two base lines or band areas with the branch point as a boundary,
The point-to-point gradient calculation means calculates the point-to-point forward gradient for each of the base lines or the band regions;
The appropriate point extraction means extracts the appropriate points for each of the baselines or band regions.
2. The height information providing system according to claim 1 . A system for assigning height information to a baseline or a band area based on a plurality of target measurement points arranged in a forward direction from the starting point side to the end point side, the system comprising:
a point-to-point gradient calculation means for calculating a point-to-point forward gradient, which is a gradient in the forward direction between the two target measurement points, based on three-dimensional coordinates of the two target measurement points;
an appropriate point extraction means for extracting the target measurement point on the end point side related to the point-to-point sequential gradient as an appropriate point when the point-to-point sequential gradient satisfies a predetermined gradient requirement;
a base point setting means for setting a starting point and a destination point from among the plurality of target measurement points and setting the starting point and the destination point as the appropriate point;
and a target measurement point selection means for selecting one of the target measurement points disposed between the two appropriate points as a target measurement point,
the point-to-point gradient calculation means calculates the point- to-point forward gradient based on the appropriate point on the end point side and the three-dimensional coordinates of the measurement point of interest, and calculates a point-to-point reverse gradient, which is a gradient in the reverse direction from the end point side to the start point side, based on the appropriate point on the start point side and the three-dimensional coordinates of the measurement point of interest ;
the appropriate point extraction means extracts the target measurement point as the appropriate point when the point-to-point forward gradient associated with the target measurement point falls below a predetermined forward gradient threshold and the point-to-point reverse gradient associated with the target measurement point falls below a predetermined reverse gradient threshold;
the target measurement point selection means selects the target measurement point as the target measurement point while changing the target measurement point in ascending order of height information until the appropriate point is extracted by the appropriate point extraction means;
A height information providing system. the target measurement point selection means selects, as a new target measurement point, a target measurement point having the smallest height information among the target measurement points excluding the appropriate points already extracted by the appropriate point extraction means;
5. The height information providing system according to claim 4 .

Images