JP7621771B2 - Land feature height setting device - Google Patents

Description

The present invention relates to a technology for assigning feature heights, such as elevation, to features, and more specifically, to a feature height assignment device that assigns an appropriate "feature height" to each feature by performing statistical processing according to the feature's classification.

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. Meanwhile, in recent years, advances in measurement technology have made it easy to obtain large amounts of three-dimensional measurement points (hereafter referred to as "3D point clouds"), and advances in information technology have 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 used. 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.

The three-dimensional point cloud obtained by an MMS or the like is generally used as a three-dimensional model (hereinafter referred to as a "3D model") of the terrain to be measured. This 3D model represents the target terrain in three-dimensional coordinates, and is a terrain model such as a DSM (Digital Surface Model) or a DEM (Digital Elevation 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 assign height to the representative point of the small region. 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 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 variety of purposes compared to 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. Alternatively, recent advances in machine learning technology have made it possible to mechanically (automatically) extract features from aerial photographs or plan views and extract their shapes and attribute information. In this way, there are cases where a "two-dimensional terrain model" is prepared separately that does not include height information such as elevation, but does include attribute information of features. Using this two-dimensional terrain model 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

As mentioned above, a 3D model is composed of multiple small regions, and each small region is assigned height information (e.g., elevation value). Because the 3D point cloud is random data, multiple 3D measurement points may be placed within a small region, in which case the elevation value of the small region is determined by statistical processing using those 3D measurement points. For example, if there are five 3D measurement points within a small region, the average elevation value of those five points is calculated and this is determined as the elevation value of the small region.

Once the elevation value of the small area has been determined, it is possible to assign elevation values to the features as well. Specifically, a small area containing the feature is extracted, and the elevation value of that small area is used to assign an elevation value to the feature. However, depending on the type of feature, it may not always be appropriate to use the average elevation value within the small area. For example, if the feature is the edge of a road, it is considered more appropriate to use the lowest elevation value of the three-dimensional measurement points within the small area, and if the feature is the outer edge of a building (such as an office building), it is considered more appropriate to use the highest elevation value of the three-dimensional measurement points within the small area.

The objective of the present invention is to solve the problems of the conventional technology, that is, to provide a feature height assigning device that can assign a feature height (feature height) appropriate for its type to the feature.

The present invention focuses on the fact that multiple types of height information are prepared in advance by performing multiple types of statistical processing on 3D measurement points within a small area, and the height information appropriate to the type of feature is selected and assigned to that feature, and is an invention based on a previously unconventional idea.

The feature height assignment device of the present invention comprises a two-dimensional terrain model storage means, a point cloud storage means, a small area group storage means, an optimal statistics storage means, and a representative height determination means. Of these, the two-dimensional terrain model storage means is a means for storing a two-dimensional terrain model in which feature classifications are assigned to features in a target area, and the point cloud storage means is a means for storing a plurality of three-dimensional measurement points obtained by measuring the target area. The small area group storage means is a means for storing a small area group consisting of a plurality of small areas set in the target area, the optimal statistics storage means is a means for storing optimal statistics set according to the feature classification of the two-dimensional terrain model, and the representative height determination means is a means for determining a representative height of the small area. Note that a plurality of types of statistical heights obtained by a plurality of types of statistical processing using the three-dimensional measurement points in the small area are set for each small area. Then, the representative height determination means selects one statistical height from the multiple types of statistical heights set for the small area based on the optimal statistics of the feature classification of the two-dimensional terrain model corresponding to the small area, determines the selected statistical height as the representative height of the small area, and assigns a feature height to the feature based on the representative height.

The feature height assignment device of the present invention can also use two or more small area groups with different resolutions. In this case, two or more small area groups with different resolutions are stored in the small area group storage means. The representative height determination means selects a statistical height for each small area group, and determines, from the statistical heights for each selected small area group, the statistical height that is most suitable for the optimal statistics for the small area as the representative height of the small area.

The feature height assignment device of the present invention can also use the "minimum value" as the optimal statistic for the "road edge" feature classification of the two-dimensional terrain model. In this case, the optimal statistic storage means stores the minimum value as the optimal statistic for the road edge. The representative height determination means then selects the minimum statistical height from multiple types of statistical heights and determines this as the representative height, and assigns the feature height to the road edge based on the representative height.

The feature height assignment device of the present invention may further include a template storage means, a template selection means, and a template arrangement means. The template storage means is a means for storing a plurality of types of cross-sectional templates, the template selection means is a means for selecting a desired cross-sectional template from the cross-sectional templates stored in the template storage means, and the template arrangement means is a means for arranging the template selected by the template selection means so that it is in the cross-sectional direction of the road. The cross-sectional template is formed by two or more line segments set on a vertical plane. When the difference between the height of the cross-sectional template at the road edge and the height of the road edge assigned by the representative height determination means exceeds a predetermined threshold value, the height of the cross-sectional template is adopted as the height of the road edge.

The feature height determination device of the present invention has the following effects.
(1) By performing statistical processing according to the type of feature and assigning a feature height appropriate to that feature, it is possible to generate a 3D model that is more in line with the current situation than before.
(2) Since a 3D model that reflects the current situation can be obtained, it is possible to manage road facilities, etc. in a more advanced manner and provide map information that is more useful for automated driving.
(3) By using a two-dimensional terrain model with feature classification, manual surveys of features can be omitted or significantly reduced.

FIG. 1 is a block diagram showing the main configuration of a feature height determination device according to the present invention. FIG. 1A is a model diagram showing a low-resolution small region group generated by a small region of relatively large size, and FIG. 1B is a model diagram showing a high-resolution small region group generated by a small region of relatively small size. A model diagram showing a 3D point cloud and small region groups overlaid with their planar positions aligned. A model diagram showing a schematic diagram of a situation in which multiple types of statistical amounts are set for each small area. FIG. 11 is a model diagram showing 19 relevant small areas extracted for the left road edge and 23 relevant small areas extracted for the right road edge; (a) is a model diagram showing a schematic crossing template representing a road crossing with a bowing slope, (b) is a model diagram showing a schematic crossing template representing a road crossing with a bowing slope and a central reservation, and (c) is a model diagram showing a schematic crossing template representing a road crossing with a bowing slope and different sidewalk heights. 1A is a vertical cross-sectional view showing a road where the difference in height between the feature height at the road edge and the template height is below a threshold value, and FIG. 1B is a vertical cross-sectional view showing a road where the difference in height between the feature height at the road edge and the template height is above a threshold value. FIG. 2 is a flow chart showing the flow of main processes performed by the feature height assignment device of the present invention. 4 is a flowchart showing a main process flow for correcting feature heights of road edges by a feature height correction means.

An example of the feature height assignment device of the present invention is explained with reference to the drawings.

Figure 1 is a block diagram showing the main components of the feature height assignment device 100 of the present invention. As shown in this figure, the feature height assignment device 100 is configured to include a representative height determination means 101, a two-dimensional terrain model storage means 109, a point cloud storage means 110, a small area group storage means 111, and an optimal statistics storage means 112, and can also be configured to include a template selection means 102, a template arrangement means 103, a statistical height calculation means 104, a feature extraction means 105, a corresponding small area extraction means 106, a feature height setting means 107, a feature height correction means 108, a template storage means 113, and a statistical height storage means 114.

The representative height determination means 101, statistical height calculation means 104, feature extraction means 105, relevant small area extraction means 106, feature height setting means 107, and feature height correction means 108 that make up the feature height assignment device 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, memories such as ROM and RAM, and some also include input means such as a mouse and keyboard, and a display, and can be configured, for example, as a personal computer (PC) or a server.

The two-dimensional terrain model storage means 109, point cloud storage means 110, small area group storage means 111, optimal statistics storage means 112, template storage means 113, and statistical height storage means 114 can use the storage device of a general-purpose computer (for example, a personal computer), 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 (i.e., wireless communication).

Below, we will explain in detail each of the main elements that make up the feature height assignment device 100 of the present invention.

(2D terrain model storage means)
The two-dimensional terrain model storage means 109 stores a "two-dimensional terrain model" of a target planar range (hereinafter referred to as a "target area"). The two-dimensional terrain model is two-dimensional terrain information such as a topographical map that does not include height information, and is a terrain model that includes information on features (hereinafter referred to as "feature information"). The feature information includes information on the planar position and shape of features such as polylines and polygons (hereinafter referred to as "feature figures"), and attribute information on the features. The attribute information on the features further includes the type of feature, such as road edges, building edges, parks, and schools (hereinafter referred to as "feature classification"). In other words, feature figures are arranged on a plane within the target area, and attribute information on the features (including feature classification) is associated with these feature figures.

(Point cloud storage means)
The point cloud storage means 110 stores a "three-dimensional point cloud." Here, a three-dimensional point cloud is a collection of many three-dimensional measurement points acquired as a result of measuring a target area by aerial photogrammetry, airborne laser measurement, terrestrial laser measurement, MMS, or other methods, and these three-dimensional measurement points can naturally be represented by three-dimensional coordinates consisting of planar position and height information.

(Small area group storage means)
The small area group storage means 111 stores "small area groups". Here, a small area is a divided area (so-called mesh) formed by dividing the target area by, for example, orthogonal grids, and a small area group is a collection of these small areas. Note that the small area and the small area group are composed of information indicating a two-dimensional position (planar position), and do not have height information such as altitude. In other words, a small area group represents the planar range of the entire target area by a plurality of small areas. The small area can be set with any shape and size, and the smaller the size (area) of the small area, the higher the resolution of the small area group generated, while the larger the size (area) of the small area, the lower the resolution of the small area group generated. For example, in FIG. 2(a), a low-resolution small area group MG1 is generated by a small area with a relatively large size, and in FIG. 2(b), a high-resolution small area group MG2 is generated by a small area with a size half that of FIG. 2(a). In this way, multiple small area groups with different resolutions can be generated for the same target area, and the small area group storage means 111 can store these multiple types of small area groups.

Three-dimensional terrain models (3D models) such as DSM and DEM can be generated from the above-mentioned three-dimensional point cloud and small area group. 3D models are usually composed of multiple small areas, and are generated by assigning height information (e.g., altitude) to the representative points of each small area. Since three-dimensional point clouds obtained by measurement (e.g., laser measurement) are often random data, geometric calculations are often used to assign altitude to the representative points of the small areas. As for the calculation method, as already mentioned, the TIN method and nearest neighbor method, inverse distance weighting method, Kriging method, averaging method, etc. can be mentioned.

(Statistics calculation method)
The statistical height calculation means 104 is a means for calculating a "statistical height" for each small region. As shown in FIG. 3, when the three-dimensional point cloud and the small region group MG are superimposed after aligning the planar positions, In some cases, multiple three-dimensional measurement points are placed in a small area. In this case, by performing statistical processing using these three-dimensional measurement points, height information such as an elevation value can be obtained at a representative point of the small area. will be granted.

Values (hereinafter referred to as "statistical values") obtained by statistical processing using multiple values include the average value, minimum value, maximum value, mode, median, etc. Therefore, multiple types of statistical values can be obtained from multiple three-dimensional measurement points arranged in a small area, namely, height information as the average value (hereinafter referred to as "average height"), height information as the minimum value (hereinafter referred to as "minimum height"), height information as the maximum value (hereinafter referred to as "maximum height"), height information as the mode value (hereinafter referred to as "mode"), height information as the median value (hereinafter referred to as "median height"), etc. can be obtained. Of course, as in conventional statistical processing, it is also possible to obtain the average height, minimum height, median height, etc. by removing some of the upper or lower values as noise. For convenience, the statistical values obtained by these various statistical processing (i.e., average height, minimum height, maximum height, mode, median height, etc.) are referred to as "statistical heights".

The statistical height calculation means 104 calculates multiple types of statistical heights for each small area. For example, in FIG. 4, the statistical height calculation means 104 calculates five types of statistical heights for each of small areas MS11 to MS33, and an average height, minimum height, maximum height, most frequent height, and median height are set for each of the nine small areas. Of course, the statistical heights set for each small area can be appropriately selected, such as "minimum height and maximum height." The statistical heights calculated by the statistical height calculation means 104 are stored for each small area by the statistical height storage means 114 (FIG. 1).

As mentioned above, multiple small area groups with different resolutions can be generated for the same target area, and the small area group storage means 111 can store these multiple types of small area groups. As can be seen from FIG. 2, even if the three-dimensional point cloud is the same, the number of three-dimensional measurement points placed in each small area will differ if the small area groups are different. In this case, it is preferable that the statistical height calculation means 104 calculates the statistical height for each of the multiple small area groups, and the statistical height storage means 114 stores the statistical height for each of the multiple small area groups. Specifically, the statistical height shown in FIG. 4 is calculated and stored for the small area of the small area MG1 shown in FIG. 2(a), and the statistical height shown in FIG. 4 is also calculated and stored for the small area of the small area group MG2 shown in FIG. 2(b).

(Optimal Statistics Storage Means)
The optimal statistics storage means 112 stores feature classifications and optimal statistics in association with each other. Here, optimal statistics refers to the most suitable statistical processing for that feature classification (i.e., statistical height). For example, if the feature classification is "road edge", the optimal statistics is "statistical processing to obtain the minimum height", if the feature classification is "outer edge of a building", the optimal statistics is "statistical processing to obtain the maximum height", and if the feature classification is "park", the optimal statistics is "statistical processing to obtain the average height". Specifically, a table that associates feature classifications with optimal statistics is created in advance, and the optimal statistics storage means 112 stores the "feature classification-optimal statistics table".

(Feature Extraction Means)
The feature extraction means 105 is a means for extracting a predetermined feature figure (polyline, etc.) based on feature information possessed by the two-dimensional topographical model. When extracting feature figures, it is possible to sequentially extract all feature figures possessed by the two-dimensional topographical model, or to extract feature figures of a feature classification designated by an operator. Alternatively, it is possible to extract a desired feature figure by using a pointing device (mouse, touch panel, pen tablet, touch pad, track pad, track ball, etc.) or a keyboard while the operator visually checks the two-dimensional topographical model displayed on a display means (such as a liquid crystal display).

(Means for extracting small areas of interest)
The relevant small area extraction means 106 is a means for extracting small areas (hereinafter referred to as "relevant small areas") that include feature figures (polylines, etc.) extracted by the feature extraction means 105. For example, in Fig. 5, the feature figures of the left road edge and the right road edge are extracted by the feature extraction means 105, and the relevant small area extraction means 106 extracts 19 relevant small areas (shaded small areas) corresponding to the left road edge and 23 relevant small areas (shaded small areas) corresponding to the right road edge.

(Means for determining representative height)
The representative height determination means 101 is a means for determining a height (hereinafter referred to as a "representative height") that represents a relevant small area extracted by the relevant small area extraction means 106. Specifically, the means obtains optimal statistics by querying the optimal statistics storage means 112 with the feature classification related to the extracted relevant small area, and obtains a statistical height (e.g., a minimum height) set for the relevant small area from the statistical height storage means 114 based on the optimal statistics (FIG. 1).

As already mentioned, multiple small area groups with different resolutions can be generated for the same target area, and the statistical height storage means 114 can be configured to store a statistical height for each of these multiple small area groups. In this case, the relevant small area extraction means 106 can be configured to extract relevant small areas for each of the multiple small area groups, and the representative height determination means 101 can be configured to determine the statistical height based on the multiple small area groups.

For example, in Figure 2, small area group MG1 and small area group MG2 with different resolutions are generated, and road edges are extracted as feature figures. Then, the relevant small area extraction means 106 extracts five relevant small areas based on small area group MG1 and eight relevant small areas based on small area group MG2, and a representative height is determined for each relevant small area. Looking at Figure 2, a road edge (feature figure) at a certain position is included in both the relevant small area of small area group MG1 and the relevant small area of area group MG2. In other words, a road edge at a certain position corresponds to two types of relevant small areas and two types of representative heights.

In this way, when the same feature figure (e.g. road edge) corresponds to two or more representative heights, it is advisable to select from these representative heights the one that is most suitable for that feature. In other words, from the two or more representative heights that correspond to the same feature figure, one representative height is selected based on the optimal statistics corresponding to the feature classification of that feature. Specifically, in a case where the feature classification is "road edge" and the optimal statistics is "statistical processing to obtain the minimum height," the smallest representative height is selected from the two or more representative heights. Also, in a case where the optimal statistics is "statistical processing to obtain the average height (or most frequent height or median height)," the average value (or most frequent height or median height) obtained from the three-dimensional measurement points contained in two or more corresponding small areas can be used as the representative height, or the one closest to the average value can be selected as the representative height from the two or more representative heights.

(Feature height setting means)
The feature height setting means 107 is a means for assigning height information (hereinafter referred to as "feature height") to features. Specifically, the representative height of a relevant small area including a feature figure is set as the feature height and assigned to the feature corresponding to the position of the relevant small area. For example, in the case of Fig. 5, the representative heights of 19 relevant small areas are assigned as the feature height to the left road edge (feature), and the representative heights of 23 relevant small areas are assigned as the feature height to the right road edge (feature).

(Template storage means)
The template storage means 113 stores "crossing templates." Here, a crossing template indicates a typical cross-sectional shape of the road, and is graphic data formed by two or more line segments set on a vertical plane. For example, in FIG. 6, three types of crossing templates are set, where (a) is a crossing template representing a road crossing with a bowing gradient, (b) is a crossing template representing a road crossing with a median strip with a bowing gradient, and (c) is a crossing template representing a road crossing with a bowing gradient and different sidewalk heights. As shown in FIG. 6, it is preferable to form the crossing template so as to include the center of the road.

(Template Selection Means and Template Placement Means)
The template selection means 102 is a means by which an operator uses a pointing device, a keyboard, or the like to select a desired crossing template from among the crossing templates stored in the template storage means 113. The template selection means 102 can be designed to select a desired crossing template by an operator operation, or can be designed to automatically select from the "funnel classification" on the cross section. The template placement means 103 is a means for placing the selected crossing template so that it is in the cross direction of the road. The process by which the template placement means 103 places the crossing template will be described in detail below.

The feature height is assigned to the road in advance. Specifically, the feature extraction means 105 extracts the feature figure (polygon, etc.) of the road, and the corresponding small area extraction means 106 extracts the corresponding small area, and the representative height determined by the representative height determination means 101 for each corresponding small area is assigned to the road as the feature height. Then, based on the assigned feature height, a cross-sectional view of the road projected on a vertical plane (hereinafter referred to as a "measured cross-sectional view") is created, and the road center (feature) and road edge (feature) of the 2D terrain model are set on the measured cross-sectional view. The measured cross-sectional view can be created at a position specified by the operator, or can be automatically generated at regular intervals (for example, 20 m pitch) in the road extension direction. When the measured cross-sectional view is created, the road surface height at the road centerline and the road center is adjusted as shown in FIG. 7, and then the cross-sectional template is placed. At this time, the cross-sectional template is placed on the measured cross-sectional view, that is, in the cross-sectional direction of the road. Therefore, it is advisable for a two-dimensional terrain model to include road centers, road edges, and road cross directions as features.

(Means of correcting feature height)
The object height correction means 108 is a means for correcting the object height of the road edge set by the object height setting means 107 as necessary. Specifically, when the difference (hereinafter referred to as "differential height") between the object height at the road edge and the height of the cross-sectional template at the road edge (hereinafter referred to as "template height") exceeds a predetermined threshold value, the object height correction means 108 corrects the object height of the road edge. In other words, instead of the object height of the road edge set by the object height setting means 107, the template height is set as a new object height of the road edge. On the other hand, when the differential height is below the threshold value, the object height correction means 108 maintains the object height of the road edge as it is.

In the example of FIG. 7(a), the difference between the feature height at the road edge and the template height is below the threshold, so the feature height correction means 108 maintains the feature height of the road edge as is. In contrast, in the example of FIG. 7(b), the difference between the feature height at the right road edge and the template height is above the threshold, so the feature height correction means 108 sets the template height as the new feature height of the right road edge. In actual road edges, retaining walls may be installed along the way, and in that case, the retaining walls are obtained as the feature height, so it is possible that the result will not grasp the entire road edge. Therefore, if the shape is significantly different from the crossing template that represents the road crossing, the template height of the crossing template is used rather than the feature height of the road edge set based on the 3D point cloud, so that the entire road edge is grasped.

(Processing flow)
The main processing of the feature height assignment device 100 will be described in detail below with reference to Fig. 8. Fig. 8 is a flow chart showing the flow of the main processing of the feature height assignment device 100 of the present invention. In this flow chart, 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.

As shown in FIG. 8, first, a three-dimensional point cloud is read from the point cloud storage means 110, and a small area group is read from the small area group storage means 111, and the statistical height calculation means 104 calculates a statistical height for each small area (Step 201). At this time, the statistical height calculation means 104 calculates all statistical heights (average height, minimum height, maximum height, most frequent height, median height, etc.) of preselected types for each small area, and if the small area group storage means 111 stores multiple types of small area groups, it calculates a statistical height for each small area group.

Next, the feature extraction means 105 is used to extract a specific feature figure (such as a polyline) based on the feature information contained in the two-dimensional topographical model (Step 202), and the relevant small area extraction means 106 extracts a relevant small area for that feature figure (Step 203). Once the relevant small area has been extracted, the representative height determination means 101 obtains optimal statistics from the optimal statistics storage means 112 based on the feature classification related to the relevant small area (Step 204), and obtains a statistical height (e.g., minimum height) set for the relevant small area from the statistical height storage means 114 based on the optimal statistics, and determines it as the representative height of the relevant small area (Step 205).

The series of processes (Step 204 to Step 205) up to determining the representative height of the relevant small area are repeatedly performed for the relevant small area extracted by the relevant small area extraction means 106, and the feature height setting means 107 assigns a feature height to the feature based on these representative heights (Step 206).

Next, the process by which the feature height correction means 108 corrects the feature height of the road edge will be described in detail with reference to Figure 9. Figure 9 is a flow diagram showing the main process flow for correcting the feature height of the road edge by the feature height correction means 108. Note that in this flow diagram, the action to be performed is shown in the center column, what is necessary for that action is shown in the left column, and what results from that action is shown in the right column.

As shown in Figure 9, when the feature height setting means 107 assigns feature heights to the road edges (Step 206), the operator uses a pointing device, keyboard, etc. to select the desired cross-sectional template from the template storage means 113 (Step 207). Then, the template placement means 103 aligns the road centerline with the road surface height at the road center as shown in Figure 7, and places the cross-sectional template over the measured cross-sectional view (Step 208). At this time, the cross-sectional template is placed so that it is in the cross-sectional direction of the road.

When the cross-sectional template is placed in the specified position, the feature height correction means 108 calculates the difference in height between the feature height at the road edge and the cross-sectional template (Step 209) and compares this difference in height with a threshold value (Step 210). If the difference in height exceeds the threshold value (Yes in Step 210), the feature height correction means 108 sets the template height as the new feature height of the road edge instead of the feature height of the road edge set by the feature height setting means 107 (Step 211). On the other hand, if the difference in height is below the threshold value (No in Step 210), the feature height correction means 108 maintains the feature height of the road edge as it is.

The feature height assignment device 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 feature height assignment device of the present invention is not only applicable to industry, but is also expected to make a significant contribution to society.

100 Feature height assigning device of the present invention 101 Representative height determination means (of feature height assigning device) 102 Template selection means (of feature height assigning device) 103 Template arrangement means (of feature height assigning device) 104 Statistical height calculation means (of feature height assigning device) 105 Feature extraction means (of feature height assigning device) 106 Corresponding small area extraction means (of feature height assigning device) 107 Feature height setting means (of feature height assigning device) 108 Feature height correction means (of feature height assigning device) 109 2D topographical model storage means (of feature height assigning device) 110 Point cloud storage means (of feature height assigning device) 111 Small area group storage means (of feature height assigning device) 112 Optimal statistics storage means (of feature height assigning device) 113 Template storage means (of feature height assigning device) 114 Statistical height storage means (of feature height assigning device) MG Small Area Group

Claims (4)

a two-dimensional topographical model storage means for storing a two-dimensional topographical model in which a topographical object classification is assigned to the topographical objects of a target area;
a point cloud storage means for storing a plurality of three-dimensional measurement points obtained by measuring a target area;
a small area group storage means for storing a small area group consisting of a plurality of small areas set for a target area;
A statistical amount calculation means for calculating a plurality of types of statistical amounts for each of the small regions;
an optimal statistics storage means for storing optimal statistics that are set in advance in association with the feature classification of the two-dimensional topographical model;
a representative height determining means for determining a representative height of the small region,
the statistical height calculation means calculates a plurality of types of the statistical height for each small region by performing a plurality of types of statistical processing using a plurality of the three-dimensional measurement points within the small region;
the representative height determination means selects one statistical height from a plurality of types of statistical heights related to the small region based on the optimal statistics of the feature classification of the two-dimensional topographical model corresponding to the small region, and determines the selected statistical height as the representative height of the small region;
assigning a feature height, which is the height of the feature, to the feature based on the representative height;
A feature height imparting device. The small region group storage means stores two or more small region groups having different resolutions,
the representative height determination means selects the statistical height of each of the small regions for each of the small region groups, and determines, from the statistical heights related to the small regions for each of the selected small region groups, the statistical height that is most suitable for the optimal statistics of the small region as the representative height of the small region;
2. The feature height determination device according to claim 1. the optimal statistics storage means stores a "minimum value" for the "road edge" of the feature classification of the two-dimensional terrain model as the optimal statistics;
the representative height determination means selects a minimum statistical height from among a plurality of types of the statistical heights and determines the minimum statistical height as the representative height, and assigns the feature height to the road edge based on the representative height.
3. The feature height determination device according to claim 1 or 2. A template storage means for storing a plurality of types of crossing templates;
a template selection means for selecting a desired transverse template from the transverse templates stored in the template storage means;
a template arrangement means for arranging the crossing template selected by the template selection means so as to be in a crossing direction of a road,
The transverse template is formed by two or more line segments set on a vertical plane;
When a difference between a height of the cross-sectional template at the road edge and a height of the road edge assigned by the representative height determining means exceeds a predetermined threshold, the height of the cross-sectional template is adopted as the height of the road edge.
4. The feature height determination device according to claim 3.

Images