JP2018169511A - Map data generation device, method for generating map data, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate three-dimensional map data for each lane while suppressing the amount of calculation. A map data generation device first generates vertical line network data including a vertical line link and a vertical line node based on road network data including a road link and a road node. Next, based on the upper and lower lane network data, two-dimensional lane network data including lane links and lane nodes is generated. Then, the three-dimensional lane network data is generated based on the two-dimensional lane network data. [Selection diagram] FIG. 4

Description

  The present invention relates to a map data generation method.

  In recent years, three-dimensional map data for each lane has been demanded as map data used for vehicle driving assistance and automatic driving. Patent Document 1 describes a map generation device for each lane (lane).

Japanese Patent Laying-Open No. 2015-4814

  However, in the method of Patent Document 1, a huge amount of calculation processing is required because the lane is detected by analyzing a local ortho image in which an image around the host vehicle is recorded.

  Examples of the problem to be solved by the present invention include the above. An object of the present invention is to provide a map data generation device capable of generating three-dimensional map data for each lane while suppressing a calculation amount.

  The invention according to claim 1 is a map data generation device, wherein a first process for generating vertical line network data including vertical link and vertical line nodes based on road network data including road links and road nodes A second processing unit for generating two-dimensional lane network data including lane links and lane nodes based on the vertical line network data, and generating three-dimensional lane network data based on the two-dimensional lane network data A third processing unit.

  The invention according to claim 12 is a map data generation method executed by the map data generation device, and based on road network data including road links and road nodes, vertical lines including vertical line links and vertical line nodes A first processing step for generating network data, a second processing step for generating two-dimensional lane network data including lane links and lane nodes based on the up-and-down line network data, and on the basis of the two-dimensional lane network data A third processing step of generating three-dimensional lane network data.

  The invention according to claim 13 is a program executed by a map data generation device including a computer, and based on road network data including road links and road nodes, vertical lines including vertical line links and vertical line nodes. A first processing unit that generates network data, a second processing unit that generates two-dimensional lane network data including lane links and lane nodes based on the up and down line network data, The computer is caused to function as a third processing unit that generates dimension lane network data.

The structure of the map data generation apparatus which concerns on an Example is shown. The outline of the map data generation method is shown. An example of each network data is shown. It is a flowchart of a data generation process. It is a flowchart of a vertical line network generation process. It is explanatory drawing of the production | generation method of an up-down line link / node. It is a flowchart of a two-dimensional lane network generation process.

  In a preferred embodiment of the present invention, the map data generation device includes a first processing unit that generates vertical line network data including vertical link and vertical line nodes based on road network data including road links and road nodes. A second processing unit for generating two-dimensional lane network data including a lane link and a lane node based on the up / down line network data; and a second processing unit for generating three-dimensional lane network data based on the two-dimensional lane network data. 3 processing units.

  The map data generation apparatus first generates vertical line network data including vertical link and vertical line nodes based on road network data including road links and road nodes. Next, two-dimensional lane network data including lane links and lane nodes is generated based on the up / down network data. Then, three-dimensional lane network data is generated based on the two-dimensional lane network data. Thereby, the three-dimensional lane network data can be efficiently generated using the existing road network data.

  In one aspect of the map data generation device, the first processing unit generates the vertical line link and the vertical line node based on the road link and road node and the number of lanes of the road link. In this aspect, the vertical link and the vertical line node are generated using the number of lanes of the road link. Preferably, the first processing unit shifts a point on the road link by a width corresponding to the width of the road link to generate the vertical line link.

  In another aspect of the map data generation device, the second processing unit may determine the lane link and the lane node based on the vertical line link and vertical line node and the number of lanes of the vertical line link. Generate. In this aspect, the lane link and the lane node are generated using the number of lanes of the up / down link.

  In another aspect of the map data generation device, the first processing unit generates an intra-intersection link in the unit of vertical line indicating a passable direction between the upper and lower line nodes included in the intersection. In this aspect, an intra-intersection link that indicates the direction in which traffic is allowed within the intersection is obtained. Preferably, the second processing unit generates an intra-intersection link in lane units from the intra-intersection link in vertical lines. Thereby, the link in an intersection in a lane unit can be obtained further.

  In the other one aspect | mode of said map data generation apparatus, the said 1st process part produces | generates the said up-and-down line link which added the lane equivalent to the exclusive lane of a right turn or a left turn. In this aspect, when there is a dedicated lane for a right turn or a left turn, the information is added to the up / down line link. Preferably, the second processing unit generates the lane link corresponding to the right lane or the left turn lane from the up / down line link added with the lane corresponding to the right lane or the left turn lane. Thereby, the exclusive lane of the right turn or the left turn can further be added per lane.

  In the other one aspect | mode of said map data generation apparatus, the said 1st process part provides the stop line flag which shows the stop line which exists on the said up-and-down line link to the said up-and-down line link. In this aspect, stop line information can be added to the vertical link.

  In the other one aspect | mode of said map data generation apparatus, a said 2nd process part links | couples traffic signal information with the said lane node. In this aspect, traffic signal information can be added to the lane node.

  In the other one aspect | mode of said map data generation apparatus, a said 3rd process part adds the height information to the said lane link and the said lane node, and produces | generates the said three-dimensional lane network data. Thereby, lane network data having three-dimensional coordinates can be obtained.

  In another preferred embodiment of the present invention, the map data generation method executed by the map data generation apparatus is based on road network data including road links and road nodes, and vertical lines including vertical link and vertical line nodes. A first processing step for generating network data, a second processing step for generating two-dimensional lane network data including lane links and lane nodes based on the up-and-down line network data, and on the basis of the two-dimensional lane network data A third processing step of generating three-dimensional lane network data. Also by this method, the three-dimensional lane network data can be generated efficiently using the existing road network data.

  In another preferred embodiment of the present invention, the program executed by the map data generation device including a computer is based on road network data including road links and road nodes, and vertical lines including vertical link and vertical line nodes. A first processing unit that generates network data, a second processing unit that generates two-dimensional lane network data including lane links and lane nodes based on the up and down line network data, The computer is caused to function as a third processing unit that generates dimension lane network data. By executing this program on a computer, the above map data generating device can be realized. This program can be stored and handled in a storage medium.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
[Constitution]
FIG. 1 shows a configuration of a map data generation device 1 according to the embodiment. The map data generation device 1 includes a data generation unit 10, a road network (hereinafter “network” is described as “NW”) database 21 (hereinafter, “database” is described as “DB”), and an ortho image DB 22. And a point cloud DB 23, a vertical line NWDB 24, a 2D lane NWDB 25, and a 3D lane NWDB 26.

  The road NWDB 21 stores existing road NW data used in car navigation devices and the like. The ortho image DB 22 stores ortho images created based on aerial photographs and the like. Note that an “ortho image” is an image obtained by correcting an aerial photograph or the like taken from a certain area from above, by orthographic projection, and is a two-dimensional image when a certain area is viewed from above.

  The point cloud DB 23 stores 3D point cloud data measured by an external sensor such as a 3D lidar (LiDAR: Light Detection and Ranging) mounted on the vehicle.

  The data generation unit 10 includes a CPU, a ROM, a RAM, and the like, and performs data generation processing by executing a program prepared in advance. Specifically, the data generation unit 10 generates 3D lane NW data based on road NW data stored in the road NWDB 21.

  FIG. 2 shows an outline of processing by the data generation unit 10. As illustrated, the data generation unit 10 first generates vertical line NW data based on the road NW data stored in the road NWDB 21. At this time, the data generation unit 10 uses the ortho image in the ortho image DB 22 or the 3D point cloud data in the point cloud DB 23 as necessary. The generated vertical line NW data is stored in the vertical line NWDB 24.

  Next, the data generation unit 10 generates 2D lane NW data based on the vertical line NW data stored in the vertical line NWDB 24. At this time, the data generation unit 10 uses the ortho image in the ortho image DB 22 or the 3D point cloud data in the point cloud DB 23 as necessary. The generated 2D lane NW data is stored in the 2D lane NWDB 25.

  Next, the data generation unit 10 generates 3D lane NW data based on the 2D lane NW data stored in the 2D lane NWDB 25. At this time, the data generation unit 10 uses the ortho image in the ortho image DB 22 or the 3D point cloud data in the point cloud DB 23 as necessary. The generated 3D lane NW data is stored in the 3D lane NWDB 26. Thus, the data generation unit 10 generates 3D lane NW data based on the road NW data.

[NW data]
Next, each NW data will be described. 3A to 3C show examples of each NW data. Note that the example of FIG. 3 shows NW data near a certain intersection.

(Road NW data)
FIG. 3A shows an example of road NW data. The road NW data includes a road link La corresponding to a road and a road node Na corresponding to an intersection or an end of the road link La. The subscript “a” is used for the sign of the road NW data. The road link La includes, as attribute data, traffic regulation data indicating one-way traffic, entry prohibition, etc., regulation speed data indicating a speed limit, and the like.

  Furthermore, for high-standard roads such as automobile roads, national roads, and major prefectural roads, the road link La includes the number of lanes and width information as attribute data. In the example of FIG. 3A, the notation “(3, 3)” indicates that the road is a three-lane road on one side. The notation “(?)” Indicates that there is no information on the number of lanes. The width information is classified into a plurality of classes such as “less than 3 m”, “3 m or more and less than 5.5 m”, “5.5 m or more and less than 13 m”, “13 m or more”, and the like.

(Up and down line NW data)
FIG. 3B shows an example of the vertical line NW data. The up / down line NW data includes an up / down line link Lb corresponding to an up line or down line of a road and an up / down line node Nb corresponding to an intersection or an end of the up / down line link Lb. Note that the subscript “b” is used for the sign of the vertical line NW data. The up / down line link Lb includes information on the number of lanes and the width as attribute data. In the example of FIG. 3B, the notation “(3)” indicates a three-lane road.

  The vertical line NW data includes an intra-intersection link CLb which is a link within the intersection. The intra-intersection link CLb indicates a direction in which the vehicle can pass at the intersection. The up / down line link Lb includes a stop line flag indicating a stop line provided near the intersection. In the example of FIG. 3B, the stop line is indicated by the symbol “STb”.

(Lane NW data)
FIG. 3C shows an example of lane NW data. The lane NW data includes a lane link Lc corresponding to a road lane and a lane node Nc corresponding to an intersection or an end of the lane link Lc. The subscript “c” is used for the sign of the lane NW data.

  The lane NW data includes an intra-intersection link CLc, which is a link within the intersection. The intra-intersection link CLc indicates a direction in which the vehicle can pass at the intersection. The lane link Lc includes a stop line flag indicating a stop line provided near the intersection. In the example of FIG. 3C, the stop line is indicated by the symbol “STc”.

[Data generation processing]
Next, data generation processing by the data generation unit 10 will be described in detail.
(Main process)
FIG. 4 is a flowchart of the data generation process. This processing is performed by the data generation unit 10. Specifically, the data generation unit 10 implements data generation processing by executing a program prepared in advance.

  First, the data generation unit 10 performs vertical line NW generation processing (step S11). The vertical line NW generation process is a process for generating vertical line NW data based on road NW data. Next, the data generation unit 10 performs 2D lane NW generation processing (step S12). The 2D lane NW generation process is a process of generating 2D lane NW data based on the vertical line NW data generated in step S11. Next, the data generation unit 10 performs 3D lane NW generation processing (step S13). The 3D lane NW generation process is a process of generating 3D lane NW data based on the 2D lane NW data generated in step S12. Then, the data generation process ends.

Next, each process will be described in detail.
(Up / down line NW generation processing)
First, the vertical line NW generation process in step S11 will be described in detail. FIG. 5 is a flowchart of the vertical line NW generation process. First, the data generation unit 10 generates vertical line links and vertical line nodes based on the road NW data (step S21).

  FIG. 6A shows a method for generating up / down line links and up / down line nodes. As described above, the road link La of the main road includes the number of lanes and the width information as attribute data. In FIG. 6A, when one vertical line link is generated from the road link La, the data generation unit 10 configures the road link La, that is, nodes Na1 and Na2 at both ends of the road link La. Then, normal vectors are obtained at two points Pa1 and Pa2 between them. Specifically, the data generation unit 10 obtains a normal vector NV1 for the line segment 41 constituting the road link La at the node Na1. Further, in the node Na2, a normal vector NV4 for the line segment 43 constituting the road link La is obtained. In addition, the data generation unit 10 sets a normal vector obtained by adding the normal vectors of the line segments on both sides of the points Pa1 and Pa2 in the middle of the road link La. That is, for the point Pa1, the sum of the normal vector for the line segment 41 and the normal vector for the line segment 42 is defined as a normal vector NV2. Similarly, for the point Pa2, the sum of the normal vector for the line segment 42 and the normal vector for the line segment 43 is defined as a normal vector NV3.

  When the normal vectors at the nodes Na1 and Na2 and the points Pa1 and Pa2 are obtained, the data generation unit 10 shifts each point in the direction of the normal vector by a distance RW corresponding to the width of the road link La. , Generate corresponding nodes or points. That is, the position obtained by shifting the road node Na1 by the distance RW in the direction of the normal vector NV1 is defined as the vertical line node Nb1. Similarly, a position obtained by shifting the road node Na2 by the distance RW in the direction of the normal vector NV4 is defined as an up / down line node Nb2. In addition, the data generation unit 10 sets a position obtained by shifting the point Pa1 by the distance RW in the direction of the normal vector NV2 as a point Pb1 on the vertical line link Lb. Similarly, a position obtained by shifting the point Pa2 by the distance RW in the direction of the normal vector NV3 is defined as a point Pb2 on the vertical line link Lb. Then, the vertical line link Lb is generated based on the vertical line nodes Nb1 and Nb2 and the points Pb1 and Pb2.

  As described above, the data generation unit 10 generates one (one of the vertical lines) vertical link Lb and the vertical line node Nb from the road link La and the road node Na. In addition, the data generation part 10 produces | generates the other of an up-and-down line link by producing | generating the same process to the other side (lower side in FIG. 6 (A)). Thus, the vertical line link Lb and the vertical line node Nb shown in FIG. 3B are generated. In addition, the data generation unit 10 may add traveling direction information (that is, information indicating whether it is up or down) as attribute data to the generated vertical line link Lb, or the direction of the vertical line link Lb. The traveling direction information may be given by (for example, a start point and an end point that configure the up-down line link Lb).

  As shown in the example of FIG. 3A, when the road NW data does not include information on the number of lanes or the width, the data generation unit 10 stores the image of the ortho image stored in the ortho image DB 22. Information on the number of lanes and the width can be acquired by analysis or analysis of 3D point cloud data stored in the point cloud DB 23. Moreover, the data generation part 10 is good also as acquiring information on the number of lanes and width from the outside, for example, referring other DB or accepting a user's manual input.

  Next, the data generation unit 10 generates an intra-intersection link (step S22). Specifically, the data generation unit 10 extracts the vertical line node Nb existing in the intersection from the vertical line node Nb generated in step S21. Next, the data generation unit 10 acquires all the vertical line links Lb connected to the extracted vertical line node Nb, and based on the traveling direction information, either the inflow link or the outflow link to the intersection Categorize.

  Next, the data generation unit 10 generates the intra-intersection link CLb by connecting the vertical line node Nb connected to the inflow link and the vertical line node Nb connected to the outflow link. At this time, the data generation unit 10 determines the direction of the intra-intersection link CLb based on the direction of the inflow link to the intersection. For example, the data generation unit 10 classifies the intra-intersection link CLb into straight, right turn, and left turn according to the rules shown in FIG. In this example, for a certain vertical line node Nb in an intersection, if the connected vertical line node Nb is within a range of ± 30 ° with respect to the direction of the inflow link connected thereto, the straight intra-intersection link CLb If it is within the range of 30 ° to 180 °, the right turn intersection link CLb is generated, and if it is within the range of −30 ° to −180 °, the left turn intersection link CLb is generated. In the above example, the intra-intersection link CLb is classified as straight, right turn, or left turn based on the angle between the vertical line node Nb connected to the inflow link and the vertical line node Nb connected to the outflow link. However, the present invention is not limited to this. Based on the angle formed by the direction indicated by the inflow link and the direction indicated by the link at the connection destination, straight travel, right turn, and left turn may be classified.

  Next, the data generation unit 10 generates a lane in a section where the lane increases or decreases (step S23). Specifically, the data generation unit 10 detects a portion where a right turn dedicated lane (lane) or a left turn dedicated lane is added at an intersection, and creates a corresponding vertical line link Lb and vertical line node Nb. In addition, the data generation unit 10 detects a portion where the lane is decreased due to a narrow road width or the like, and corrects the corresponding vertical line link Lb and vertical line node Nb.

  For example, in the example of FIG. 3B, a right turn lane is provided in front of the intersection on the road on the left side in the figure, so that the number of lanes increases by one for the up / down link Lbx immediately before the intersection. 4 lanes. In this case, the data generation unit 10 newly establishes a new up / down line node Nbx at the point where the right turn dedicated lane starts, and the three lane up / down line link Lb and the four lane up / down line link including the right turn dedicated lane. Lbx.

  When increasing or decreasing the lane, the data generation unit 10 increases or decreases the number of lanes by analyzing the image of the ortho image stored in the ortho image DB 22 or analyzing the 3D point cloud data stored in the point cloud DB 23. Can be detected. Moreover, the data generation part 10 is good also as acquiring the information of increase / decrease in the number of lanes from the outside, for example, referring another DB or accepting a user's input.

  Next, the data generation unit 10 assigns a stop line flag (step S24). Specifically, the data generation unit 10 detects the presence or absence of a stop line in the inflow link with respect to the vertical line node Nb in the intersection. In this case, the data generation unit 10 can determine that there is a stop line at an intersection where there is a traffic light by referring to the data on the presence / absence of the traffic signal included in the road NW data. The data generation unit 10 may acquire stop line information by image analysis of an ortho image, analysis of 3D point cloud data, user input, and the like. When the stop line is detected, the data generation unit 10 assigns a stop line flag as attribute data of the vertical line link Lb where the stop line exists. Note that the stop line flag may include the position coordinates of the stop line.

  Then, the data generation unit 10 stores the generated vertical line NW data in the vertical line NWDB 24 (step S25). Thereby, the up-and-down line NW generation process ends.

(2D lane NW generation process)
Next, the 2D lane NW generation process in step S12 will be described. FIG. 7 is a flowchart of 2D lane NW generation processing. First, the data generation unit 10 generates a lane link Lc and a lane node Nc based on the vertical line NW data (step S31). Specifically, the data generation unit 10 generates lane links Lc corresponding to the number of lanes from one vertical line link Lb in the vertical line NW data. Here, the method of generating the lane link Lc from the up / down line link Lb is basically the same as the method of generating the up / down line link Lb from the road link La in step S21 of the above-described up / down line NW generation process. That is, as described with reference to FIG. 6A, the normal vectors at the node Nb and the point P constituting the vertical link Lb are obtained, and each point corresponds to the width of the lane in the normal vector direction. The lane link Lc and the lane node Nc are generated with a shift by the distance. The lane width information can be obtained by using the width information included in the road NW data, or by analyzing the ortho image or analyzing the 3D point cloud data.

  Next, the data generation unit 10 generates an intra-intersection link (step S32). Specifically, as shown in FIGS. 3B and 3C, the data generation unit 10 changes the intra-intersection link CLc in units of lanes from the intra-intersection link CLb in units of vertical lines included in the vertical line NW data. Generate.

  Next, the data generation unit 10 performs a stop line flag process (step S33). Specifically, as shown in FIGS. 3B and 3C, the data generation unit 10 generates a lane-by-lane stop line flag from the up-and-down line unit stop line flag included in the up-and-down line NW data. . Thereby, the stop line flag STc for each lane shown in FIG. 3C is generated from the stop line flag STb for each vertical line shown in FIG.

  Next, the data generation unit 10 associates traffic lights (step S34). Specifically, the data generation unit 10 acquires node-traffic association information from an external signal table, and associates the traffic signal with the corresponding lane node Nc. Thereby, the corresponding traffic signal information is associated with the lane node Nc. The traffic light table is prepared in advance based on road NW data and the like. Alternatively, the data generation unit 10 may detect a traffic signal by image analysis of an ortho image, analysis of 3D point cloud data, or the like to create a traffic signal table.

  And the data generation part 10 memorize | stores the produced | generated 2D lane NW data in 2D lane NWDB25 (step S35). Thereby, the 2D lane NW generation process ends.

(3D lane NW generation process)
Next, the 3D lane NW generation process in step S13 will be described. The 2D lane NW data generated in step S12 is two-dimensional data, and the lane link Lc and the lane node Nc do not include height information. Therefore, the data generation unit 10 adds height information (z coordinate) to the lane link Lc and the lane node Nc. Height information can be added by the following method.

  The first method uses height information included in road NW data. That is, the height information of the road link La and road node Na corresponding to each lane link Lc and lane node Nc is acquired from the road NW data, and is given to each lane link Lc and lane node Nc.

  The second method uses an ortho image stored in the ortho image DB 22. That is, when height information is included in the ortho image, the height information is given to each lane link Lc and lane node Nc. The third method uses 3D point cloud data stored in the point cloud DB 23. That is, by analyzing the 3D point cloud data, the height of the point on each lane link Lc or the point of each lane node Nc is estimated, and the value is given to each lane link Lc and lane node Nc. In addition, you may utilize combining said 1st-3rd method arbitrarily.

  The 3D lane NW data generated in this way has three-dimensional coordinates at the lane link Lc and the lane node Nc. Then, the data generation unit 10 stores the generated 3D lane NW data in the 3D lane NWDB 26. Thus, the data generation process ends.

  As described above, according to the present embodiment, the lane NW data is generated using the existing road NW data, so that the lane NW data can be generated efficiently.

1 Map data generator 10 Data generator 21 Road NWDB
22 Ortho Image DB
23 point cloud DB
24 Vertical Line NWDB
25 2D Lane NWDB
26 3D Lane NWDB

Claims (14)

A first processing unit that generates vertical line network data including vertical link and vertical line nodes based on road network data including road links and road nodes;
A second processing unit for generating two-dimensional lane network data including a lane link and a lane node based on the vertical line network data;
A third processing unit for generating three-dimensional lane network data based on the two-dimensional lane network data;
A map data generation device comprising:   The map data generation device according to claim 1, wherein the first processing unit generates the vertical line link and the vertical line node based on the road link and road node and the number of lanes of the road link.   The map data generation device according to claim 2, wherein the first processing unit generates the vertical line link by shifting a point on the road link by a width corresponding to a width of the road link.   4. The map data generation according to claim 2, wherein the second processing unit generates the lane link and the lane node based on the vertical line link and the vertical line node and the number of lanes of the vertical line link. apparatus.   5. The map data generation device according to claim 1, wherein the first processing unit generates an intra-intersection link in a unit of vertical line indicating a passable direction between the vertical line nodes included in the intersection.   The map data generation device according to claim 5, wherein the second processing unit generates an in-intersection link in lane units from the in-intersection link in vertical lines.   The map data generation device according to any one of claims 1 to 6, wherein the first processing unit generates the up-and-down line link to which a lane corresponding to a dedicated lane for a right turn or a left turn is added.   8. The map data according to claim 7, wherein the second processing unit generates the lane link corresponding to a right turn or left turn dedicated lane from the up / down line link to which a lane corresponding to the right turn or left turn dedicated lane is added. Generator.   The map data generation device according to any one of claims 1 to 8, wherein the first processing unit assigns a stop line flag indicating a stop line existing on the vertical line link to the vertical line link.   The map data generation device according to any one of claims 1 to 9, wherein the second processing unit associates traffic signal information with the lane node.   The map data generation device according to any one of claims 1 to 10, wherein the third processing unit generates the three-dimensional lane network data by adding height information to the lane link and the lane node. A map data generation method executed by a map data generation device,
A first processing step of generating vertical line network data including vertical link and vertical line nodes based on road network data including road link and road node;
A second processing step of generating two-dimensional lane network data including a lane link and a lane node based on the vertical line network data;
A third processing step of generating three-dimensional lane network data based on the two-dimensional lane network data;
A map data generation method comprising: A program executed by a map data generation device including a computer,
A first processing unit for generating vertical line network data including vertical link and vertical line nodes based on road network data including road links and road nodes;
A second processing unit for generating two-dimensional lane network data including a lane link and a lane node based on the vertical line network data;
A third processing unit for generating three-dimensional lane network data based on the two-dimensional lane network data;
A program for causing the computer to function as   A storage medium storing the program according to claim 13.

Images