Lane network generation device, lane network generation method, and computer program for generating lane network

Abstract

To provide a lane network generation device capable of generating a lane network suitable for traveling of a vehicle.SOLUTION: A lane network generation device includes: a probability distribution calculation unit 41 that calculates, on the basis of a plurality of travel tracks of vehicles that have traveled in a predetermined region connecting an entrance lane and an exit lane, a probability distribution expressing a probability that an individual point in the region is positioned on a traveling track serving as a reference; a virtual boundary line generation unit 42 that generates a virtual boundary line connecting positions of opposite ends of the entrance lane and the exit lane; a control point setting unit 43 that sets, from among the points in the predetermined region, at least two points which are positioned between two virtual boundary lines and at which a variance of the probability distribution is equal to a predetermined scattering threshold value or less as control points; and a network generation unit 44 that generates a Bezier curve on the basis of the at least two control points and sets the generated Bezier curve as a line connecting two vehicle lanes for the predetermined region in a lane network.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to a lane network generating device, a lane network generating method, and a computer program for generating a lane network that generates a lane network that represents the connection relationships between the respective lanes of a plurality of roads that are connected to each other. [Background technology]

[0002] One of the pieces of information contained in a high-precision map that an autonomous driving system refers to in order to control the autonomous driving of a vehicle is a lane network that represents the connection relationships between lanes and is referred to in order to set a planned driving trajectory of the vehicle under autonomous driving control. Therefore, a technology for generating a lane network has been proposed (see Patent Documents 1 and 2).

[0003] The road network data generation device disclosed in Patent Document 1 determines the intersection of an extension line of the outside intersection lane connected to the entrance point into the intersection and an extension line of the outside intersection lane connected to the exit point into the intersection for a combination of an entrance point and an exit point at an intersection. Furthermore, this road network data generation device sets a correction point between the entrance point and the intersection. Then, this road network data generation device sets a circular arc using a quadratic Bezier curve for the three points consisting of the intersection point, the correction point, and the exit point, and sets it as an inside intersection lane.

[0004] The route generation device disclosed in Patent Document 2 acquires point sequence data including a plurality of points corresponding to the lane centerline of the lane on which the vehicle is traveling, based on map information and the position of the vehicle. The route generation device then performs smoothing based on the point sequence data to acquire the lane centerline, and generates the acquired lane centerline as the target travel route of the vehicle. In this case, the route generation device extracts a plurality of point sequence groups including a plurality of points adjacent to each other from the point sequence data, and calculates the center of gravity position of each of the extracted plurality of point sequence groups. The route generation device then performs smoothing processing using a parametric curve such as Bezier, B-spline, NURBS, or a spline interpolation curve, based on each center of gravity position. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2010-26875 A [Patent Document 2] JP 2018-105636 A Summary of the Invention [Problem to be solved by the invention]

[0006] In the above technology, a lane network is generated by referring to map information. However, since the lane network generated from map information does not take into account the behavior of the vehicle, the lane network may result in an unnatural trajectory when used as a target driving trajectory of the vehicle.

[0007] Therefore, an object of the present invention is to provide a lane network generation device capable of generating a lane network suitable for vehicle travel. [Means for solving the problem]

[0008] According to one embodiment, there is provided a lane network generation device, which includes: a probability distribution calculation unit that calculates, for a given region connecting a plurality of roads, a probability distribution representing a probability that each point in the region is located on a standard travel trajectory based on a plurality of travel trajectories of a vehicle that traveled from an entrance lane on one of the plurality of roads to an exit lane on another of the plurality of roads, a virtual boundary generation unit that generates a virtual boundary line connecting the positions of both ends of the entrance lane and the exit lane, a control point setting unit that sets, as control points, at least two points among the given points that are located between the two virtual boundary lines and have a variance of the probability distribution equal to or less than a predetermined variance threshold, and a network generation unit that generates a Bezier curve based on the at least two control points and sets the generated Bezier curve as a line connecting the entrance lane and the exit lane for the given region.

[0009] In this lane network generating device, it is preferable that the virtual boundary generating unit generates, as the virtual boundary, a Bezier curve connecting the positions of both ends of the entrance lane and the exit lane.

[0010] According to another embodiment, there is provided a lane network generation method, which includes: calculating a probability distribution representing a probability that each point in a predetermined region connecting a plurality of roads is located on a standard travel trajectory based on a plurality of travel trajectories of a vehicle traveling from an entrance lane on one of the plurality of roads to an exit lane on another of the plurality of roads to exit the region; generating a virtual boundary line connecting both ends of the entrance lane and the exit lane; setting at least two points among the points in the predetermined region that are located between the two virtual boundary lines and have a variance of the probability distribution equal to or less than a predetermined variance threshold as control points; generating a Bezier curve based on the at least two control points; and setting the generated Bezier curve as a line connecting the entrance lane and the exit lane for the predetermined region.

[0011] According to yet another embodiment, there is provided a lane network generating computer program, the lane network generating computer program including instructions for causing a computer to execute the following: for a given region connecting a plurality of roads, based on a plurality of travel trajectories of a vehicle traveling from an entrance lane on any one of the plurality of roads to an exit lane on another of the plurality of roads to exit the given region, calculate a probability distribution representing a probability that each point in the given region is located on a standard travel trajectory, generate virtual borders connecting the positions of both ends of the entrance lane and the exit lane, set at least two points among the given region that are located between the two virtual borders and have a variance of the probability distribution equal to or less than a predetermined variance threshold as control points, generate a Bezier curve based on the at least two control points, and set the generated Bezier curve as a line connecting the entrance lane and the exit lane for the given region. Effect of the Invention

[0012] The lane network generation device according to the present disclosure has an effect of being able to generate a lane network suitable for vehicle travel. [Brief description of the drawings]

[0013] [Figure 1] 1 is a schematic configuration diagram of a map generation system in which a lane network generation device is implemented. [Diagram 2] FIG. 2 is a hardware configuration diagram of a server which is an example of a lane network generating device. [Diagram 3] FIG. 13 is a functional block diagram of a processor related to lane network generation processing. [Figure 4] FIG. 2 is a diagram showing an example of a Bezier curve connecting an end point of an entrance lane and an end point of an exit lane. [Diagram 5] FIG. 13 is a diagram showing an example of setting control points and generating a lane network. [Figure 6] 13 is an operational flowchart of a lane network generation process. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] Hereinafter, a lane network generation device, a lane network generation method executed by the lane network generation device, and a computer program for generating a lane network will be described with reference to the drawings. The lane network generation device generates a lane network that represents the connection relationship between the lanes of multiple roads connected to an area, such as an intersection, based on actual driving trajectories collected from one or more vehicles.

[0015] Specifically, the lane network generation device calculates a probability distribution representing the probability that each point in the intersection is located on a standard travel trajectory based on a plurality of travel trajectories of vehicles that have traveled between two lanes of interest at a given intersection. The lane network generation device also generates two virtual borders connecting the positions of both ends of the two lanes. Furthermore, the lane network generation device identifies a sequence of points that are located between the two virtual borders and have a variance of the probability distribution equal to or less than a predetermined variance threshold among the individual points for which the probability distribution has been calculated, and sets at least two points included in the identified sequence of points as control points. The lane network generation device then generates a Bézier curve based on the at least two control points, and sets the generated Bézier curve as a line connecting the two lanes for the intersection in the lane network.

[0016] FIG. 1 is a schematic diagram of a map generation system in which a lane network generation device is implemented. In this embodiment, the map generation system 1 has at least one vehicle 2 and a server 3 which is an example of a lane network generation device. Each vehicle 2 is connected to the server 3 via the wireless base station 5 and the communication network 4 by accessing, for example, a wireless base station 5 connected to the communication network 4 to which the server 3 is connected via a gateway (not shown). Note that, for simplification, only one vehicle 2 is illustrated in FIG. 1, but the map generation system 1 may have multiple vehicles 2. Similarly, only one wireless base station 5 is illustrated in FIG. 1, but multiple wireless base stations 5 may be connected to the communication network 4.

[0017] The vehicle 2 has a GPS receiver, a travel locus recording device, and a wireless communication terminal. The GPS receiver receives GPS signals from GPS satellites at predetermined intervals and measures the vehicle 2's own position based on the received GPS signals. The GPS receiver outputs positioning information representing the positioning result of the vehicle 2's own position based on the GPS signals to the travel locus recording device at predetermined intervals via the in-vehicle network. The vehicle 2 may have a receiver that complies with a satellite positioning system other than the GPS receiver. In this case, the receiver may measure the vehicle 2's own position.

[0018] The travel locus recording device includes, for example, a processor and a memory. The processor of the travel locus recording device chronologically arranges the positions of the vehicle 2 represented by the positioning information acquired from the GPS receiver to generate travel locus information representing the locus traveled by the vehicle 2, and stores the generated travel locus information in the memory of the travel locus recording device. The travel locus recording device may include identification information of the vehicle 2 in the travel locus information.

[0019] The vehicle 2 may have a camera for photographing the surrounding area of ​​the vehicle 2. The camera may generate an image showing the surrounding area of ​​the vehicle 2 at a predetermined cycle and output the generated image to the travel trajectory recording device via the in-vehicle network. In this case, the processor of the travel trajectory recording device detects features on or around the road on which the vehicle 2 is traveling from the image every time the processor receives an image from the camera. Furthermore, the processor of the travel trajectory recording device acquires odometry information of the vehicle 2 from an electronic control device (not shown) that controls the traveling of the vehicle 2, and obtains the amount of movement and the amount of displacement of the direction of the vehicle 2 at each generation interval of each image. The processor of the travel trajectory recording device may estimate the position of the vehicle 2 at the time of generating each image according to a so-called Structure from Motion technique based on the features detected from each image and the amount of movement and the amount of displacement of the direction of the vehicle 2 at each generation interval of each image. The processor of the travel trajectory recording device may generate travel trajectory information by arranging the positions of the vehicle 2 at the time of generating each image in chronological order.

[0020] The travel locus recording device outputs the generated travel locus information to the wireless communication terminal at a predetermined timing. The predetermined timing may be, for example, the timing when the ignition switch of the vehicle 2 is turned off, or the timing every time a certain time (for example, 30 minutes to 1 hour) has elapsed since the ignition switch of the vehicle 2 is turned on. Alternatively, collection area information indicating an area to be collected for travel loci may be notified in advance to the vehicle 2 from the server 3 via the communication network 4 and the wireless base station 5. In this case, the travel locus recording device may refer to the collection area information and the positioning information and set the timing when the vehicle 2 moves outside the area to be collected as the predetermined timing.

[0021] The wireless communication terminal is a device that executes wireless communication processing conforming to a predetermined wireless communication standard, and for example, by accessing the wireless base station 5, it is connected to the server 3 via the wireless base station 5 and the communication network 4. The wireless communication terminal generates an uplink wireless signal including the travel trajectory information received from the travel trajectory recording device. The wireless communication terminal transmits the uplink wireless signal to the wireless base station 5 to transmit the travel trajectory information to the server 3. The wireless communication terminal also receives a downlink wireless signal from the wireless base station 5, and passes the collection area information from the server 3, which is included in the wireless signal, to the travel trajectory recording device.

[0022] Next, the server 3, which is an example of a lane network generating device, will be described. 2 is a hardware configuration diagram of a server 3 which is an example of a lane network generating device. The server 3 has a communication interface 31, a storage device 32, a memory 33, and a processor 34. The communication interface 31, the storage device 32, and the memory 33 are connected to the processor 34 via signal lines. The server 3 may further have input devices such as a keyboard and a mouse, and a display device such as a liquid crystal display.

[0023] The communication interface 31 is an example of a communication unit, and has an interface circuit for connecting the server 3 to the communication network 4. The communication interface 31 is configured to be able to communicate with the vehicle 2 via the communication network 4 and the wireless base station 5. That is, the communication interface 31 passes the traveling trajectory information received from the vehicle 2 via the wireless base station 5 and the communication network 4 to the processor 34. The communication interface 31 also transmits the collection area information received from the processor 34 to the vehicle 2 via the communication network 4 and the wireless base station 5.

[0024] The storage device 32 is an example of a storage unit, and includes, for example, a hard disk device or an optical recording medium and an access device therefor. The storage device 32 stores various data and information used in the lane network generation process. For example, the storage device 32 stores a map for adding information representing the lane network. Furthermore, the storage device 32 stores driving trajectory information received from each vehicle 2. Furthermore, the storage device 32 may store a computer program to be executed on the processor 34 for executing the lane network generation process.

[0025] The memory 33 is another example of a storage unit, and includes, for example, a non-volatile semiconductor memory and a volatile semiconductor memory. The memory 33 temporarily stores various data generated during the execution of the lane network generation process.

[0026] The processor 34 has one or more central processing units (CPUs) and their peripheral circuits. The processor 34 may further have other arithmetic circuits such as a logical arithmetic unit or a numerical arithmetic unit. Each time the processor 34 receives travel trajectory information from any of the vehicles 2, the processor 34 stores the received travel trajectory information in the storage device 32. Furthermore, the processor 34 executes a lane network generation process.

[0027] 3 is a functional block diagram of the processor 34 related to the lane network generation process. The processor 34 has a probability distribution calculation unit 41, a virtual boundary generation unit 42, a control point setting unit 43, a network generation unit 44, and a map generation unit 45. Each of these units in the processor 34 is a functional module realized by, for example, a computer program running on the processor 34. Alternatively, each of these units in the processor 34 may be a dedicated arithmetic circuit provided in the processor 34.

[0028] The probability distribution calculation unit 41 specifies a lane on one road of interest that the vehicle 2 can enter the intersection and a lane on another road of interest that the vehicle 2 can exit from the intersection. Hereinafter, the lane on which the vehicle 2 can enter the intersection may be simply referred to as an entrance lane. Also, the lane on which the vehicle 2 exits the intersection may be simply referred to as an exit lane. The probability distribution calculation unit 41 calculates a predetermined probability distribution for each point in the intersection based on a plurality of travel trajectory information when the vehicle 2 travels from the entrance lane of interest to the exit lane. The predetermined probability distribution indicates the likelihood that the point is located on a standard travel trajectory of the vehicle traveling from the entrance lane of interest to the exit lane. Also, the standard travel trajectory is assumed as a trajectory when the vehicle travels at a standard speed and without any particular evasive action when traveling through the intersection from the entrance lane of interest to the exit lane of interest.

[0029] The probability distribution calculation unit 41 identifies, from among the pieces of travel trajectory information stored in the storage device 32, travel trajectory information that represents a travel trajectory that passes through an area that corresponds to the entrance lane and the exit lane of the intersection of interest shown on the map. Then, the probability distribution calculation unit 41 identifies the travel trajectory represented by each piece of identified travel trajectory information as the travel trajectory when the vehicle 2 traveled from the entrance lane of interest to the exit lane of interest. The probability distribution calculation unit 41 reads out the identified travel trajectory information from the storage device 32.

[0030] The probability distribution calculation unit 41 calculates a predetermined probability distribution for each point in the intersection based on the travel trajectory represented in each travel trajectory information read from the storage device 32. In this embodiment, the probability distribution for each point follows a Gaussian process. Therefore, the predetermined probability distribution can be a two-dimensional Gaussian distribution parallel to the road surface.

[0031] The probability distribution calculation unit 41 calculates a probability distribution for each point in the intersection by referring to each travel trajectory, according to a calculation method of a Gaussian process regression model, for example, a method using Bayesian estimation. That is, the probability distribution calculation unit 41 sets a two-dimensional Gaussian distribution having a predetermined variance-covariance matrix as an initial probability distribution for each point in the intersection. Then, the probability distribution calculation unit 41 calculates a posterior probability distribution for each point in the intersection based on each point on each travel trajectory, using the set Gaussian distribution as a priori probability distribution. Then, the probability distribution calculation unit 41 may calculate a probability distribution for each point in the intersection by updating the variance-covariance matrix according to the calculated posterior probability distribution for each point in the intersection. The average position of the probability distribution calculated for each point, that is, the probability value at that point, represents the likelihood that that point is located on a standard travel trajectory. In addition, the variance-covariance matrix of the probability distribution represents the degree of observation error when that point is located on a standard travel trajectory, and the larger the value of each element of the variance-covariance matrix, the larger the observation error. Therefore, the more a point passes through on the travel trajectory, the higher the probability value at the average position becomes, and the smaller the value of each element of the variance-covariance matrix becomes.

[0032] The probability distribution calculation unit 41 notifies the control point setting unit 43 of the probability distribution calculated for each point.

[0033] The virtual boundary generating unit 42 refers to the map and identifies the positions of both ends of the lane width direction at the connection position of each intersection of the entrance lane of interest and the exit lane of interest. The virtual boundary generating unit 42 then sets one end point of the entrance lane of interest and one end point of the exit lane of interest as one pair, and sets the other end point of the entrance lane of interest and the other end point of the exit lane of interest as another pair. At this time, the virtual boundary generating unit 42 sets each pair so that the lines connecting the end points of the entrance lane and the end points of the exit lane in each pair do not intersect with each other.

[0034] The virtual boundary generating unit 42 generates a Bezier curve connecting the end point of the entrance lane and the end point of the exit lane for each pair. At this time, the virtual boundary generating unit 42 extends the boundary of the entrance lane (the boundary is defined by, for example, a lane dividing line or a curb) that passes through the end point of the entrance lane into the intersection for each pair, and extends the boundary of the exit lane that passes through the end point of the exit lane into the intersection. The virtual boundary generating unit 42 obtains an intersection point between a line obtained by extending the boundary on the entrance lane side and a line obtained by extending the boundary on the exit lane side for each pair, and sets control points at the midpoint between the intersection point and the end point of the entrance lane side and the midpoint between the intersection point and the end point of the exit lane side. Then, the virtual boundary generating unit 42 generates a Bezier curve connecting the end point of the entrance lane side and the end point of the exit lane side as a virtual boundary for each bear based on the set control points.

[0035] FIG. 4 is a diagram showing an example of a Bezier curve connecting an end point of an entrance lane and an end point of an exit lane. In FIG. 4, roads are connected to an intersection 400 from above, below, left and right. Of these roads, an entrance lane 402 of a road 401 connected to the intersection 400 from below and an exit lane 404 of a road 403 connected to the intersection 400 from the right side are respectively assumed to be an entrance lane of interest and an exit lane of interest. In this case, a Bezier curve 411 connecting a right end point 402a of the entrance lane 402 and a lower end point 404a of the exit lane 404, and a Bezier curve 412 connecting a left end point 402b of the entrance lane 402 and an upper end point 404b of the exit lane 404 are generated as virtual boundaries. The virtual boundary 411 and the virtual boundary 412 connect the end points of the entrance lane 402 and the exit lane 404, respectively. Therefore, the area between the imaginary boundary line 411 and the imaginary boundary line 412 is estimated to correspond to an area through which vehicles travel from the entrance lane 402 to the exit lane 404. In other words, the imaginary boundary line 411 and the imaginary boundary line 412 are estimated to represent the boundaries of areas through which vehicles travel from the entrance lane 402 to the exit lane 404 are expected to travel (hereinafter, sometimes referred to as imaginary traffic areas).

[0036] The virtual boundary generating unit 42 notifies the control point setting unit 43 of the two virtual boundary lines that have been generated.

[0037] The control point setting unit 43 identifies a sequence of points that are located between two virtual borders and have a probability distribution variance equal to or smaller than a predetermined variance threshold among the individual points in the intersection. Then, the control point setting unit 43 sets at least two points included in the identified sequence of points as control points.

[0038] It is preferable that the control point is a position where the variation in position for each travel trajectory is small in a direction perpendicular to the traveling direction of the vehicle 2 traveling from the entrance lane of interest to the exit lane of interest. Therefore, the control point setting unit 43 obtains a reference line connecting the midpoints of the entrance lane and the exit lane of interest. The reference line can be, for example, a straight line or a circular arc that is convex with respect to the center of the intersection. The control point setting unit 43 calculates the position on the reference line closest to each point in the intersection and obtains the tangent direction of the reference line at the calculated position. For each point in the intersection, the control point setting unit 43 performs an affine transformation on the probability distribution of the point into an orthogonal coordinate system whose axes are the tangent direction and the direction perpendicular to the tangent direction at the position on the reference line closest to the point. Then, the control point setting unit 43 may set the element value of the variance-covariance matrix in the direction perpendicular to the tangent direction in the probability distribution after the affine transformation for each point in the intersection as the variance value to be compared with the variance threshold value.

[0039] The control point setting unit 43 may divide the area of ​​the intersection into a plurality of sections (for example, 5 to 10) along a straight line connecting the midpoint of the entrance lane and the midpoint of the exit lane of interest. The control point setting unit 43 may then set, as control points, two or more points included in different sections among the points in the series of points. This prevents the intervals between the control points from becoming too narrow.

[0040] By setting a point where the variance is equal to or less than a predetermined variance threshold as a control point, the control point setting unit 43 sets the control point at a position where the vehicle is likely to pass every time the vehicle travels from the entrance lane of interest to the exit lane of interest. Also, by setting a point within the virtual traffic area as a control point, the control point setting unit 43 can prevent setting a position on an abnormal travel trajectory, such as when the vehicle 2 travels from the entrance lane of interest to the exit lane of interest in a wide detour or shortcut, as a control point for some reason.

[0041] The control point setting unit 43 notifies the network generation unit 44 of the positions of each of the set control points.

[0042] The network generation unit 44 generates a Bezier curve for an intersection for which a lane network is to be generated, based on at least two control points set by the control point setting unit 43. The network generation unit 44 then sets the generated Bezier curve as a line connecting an entrance lane and an exit lane of interest for that intersection in the lane network. By using the control points set as described above, the network generation unit 44 can generate a lane network that smoothly connects an entrance lane and an exit lane of interest, and that forms a curve that is close to the trajectory of an actual vehicle travel.

[0043] FIG. 5 is a diagram showing an example of setting of control points and generation of a lane network. In FIG. 5, roads are connected to an intersection 500 from above, below, left and right, as in the intersection shown in FIG. 4. Of these roads, an entrance lane 502 of a road 501 connected to the intersection 500 from below, and an exit lane 504 of a road 503 connected to the intersection 500 from the right side are considered to be an entrance lane of interest and an exit lane of interest, respectively. Among the individual points in the intersection 500, the variance of the probability distribution with the points 531 and 532 as the average positions among the points located between two imaginary boundaries 511 and 512 connecting the end point of the entrance lane 502 and the end point of the exit lane 504 is equal to or less than the variance threshold. Therefore, the points 531 and 532 are set as control points. A Bezier curve 541 generated based on the control points 531 and 532 becomes a line connecting the entrance lane 502 and the exit lane 504, which constitutes the lane network.

[0044] The network generation unit 44 outputs information representing the generated lane network to the map generation unit 45.

[0045] The map generating unit 45 generates a map including information representing a lane network for each intersection area. That is, the map generating unit 45 adds information representing the lane network generated by the network generating unit 44 to each intersection shown on the map read from the storage device 32. In this way, a map including the lane network is generated.

[0046] The map generating unit 45 stores the generated map in the storage device 32. Alternatively, the map generating unit 45 may output the generated map to another device via the communication interface 31.

[0047] 6 is an operational flowchart of the lane network generation process in the server 3. The processor 34 of the server 3 may execute the lane network generation process according to the operational flowchart shown below for each pair of an entrance lane and an exit lane of interest at an intersection for which a lane network is to be generated.

[0048] The probability distribution calculation unit 41 of the processor 34 calculates a probability distribution representing the likelihood that each point in the intersection is located on a driving trajectory from a target entrance lane to a target exit lane, based on the driving trajectory (step S101).

[0049] In addition, the virtual boundary generating unit 42 of the processor 34 generates a Bezier curve connecting the two ends of the lane width direction at the connecting position of the intersection of the entrance lane of interest and the exit lane of interest as a virtual boundary (step S102).

[0050] The control point setting unit 43 of the processor 34 identifies a sequence of points that are located between two imaginary boundaries and have a probability distribution variance equal to or smaller than a predetermined variance threshold (step S103), and sets at least two points included in the identified sequence of points as control points (step S104).

[0051] The network generation unit 44 of the processor 34 generates a Bezier curve based on at least two set control points, and sets the generated Bezier curve as a line connecting the entrance lane and the exit lane of interest in the lane network (step S105). Then, the map generation unit 45 of the processor 34 adds information representing the line connecting the entrance lane and the exit lane of interest to the map (step S106). Then, the processor 34 ends the lane network generation process.

[0052] As described above, the lane network generation device calculates a probability distribution of each point in the intersection on the driving trajectory based on a plurality of driving trajectories from an entrance lane of interest to an exit lane of the intersection for which the lane network is to be generated. Furthermore, the lane network generation device sets at least two points among the individual points in the intersection that are located within a virtual traffic area and have a probability distribution variance equal to or less than a variance threshold as control points. The lane network generation device then sets a Bézier curve generated based on the set control points as a line connecting the entrance lane and exit lane of interest in the lane network. By generating a lane network in this manner, the lane network generation device can generate a lane network suitable for vehicle driving.

[0053] According to a modified example, the virtual boundary generating unit 42 may generate a virtual boundary connecting the respective end points of the entrance lane and the exit lane of interest as a line other than a Bezier curve. For example, the virtual boundary generating unit 42 may set the virtual boundary as a circular arc or a spline curve that connects the respective end points of the entrance lane and the exit lane of interest and is convex with respect to the center of the intersection. Alternatively, the virtual boundary generating unit 42 may generate a virtual boundary connecting the respective end points of the entrance lane and the exit lane of interest as a clothoid curve that is convex with respect to the center of the intersection and has a larger curvature as the distance from the end points of the entrance lane and the exit lane increases.

[0054] In addition, the area of ​​the road for which the lane network is generated according to the above embodiment or modification is not limited to an intersection, but may be an area where one road is connected to another road, such as a branch point or a junction. In this case, the lane network generating device can generate a lane network by executing the same process as in the above embodiment or modification for each pair of an entrance lane and an exit lane of interest.

[0055] A computer program that causes a computer to realize the functions of each unit of the processor of the lane network generating device according to each of the above embodiments or modifications may be provided in a form stored in a computer-readable recording medium. Note that the computer-readable recording medium may be, for example, a magnetic recording medium, an optical recording medium, or a semiconductor memory.

[0056] As described above, those skilled in the art can make various modifications to the embodiments within the scope of the present invention. [Explanation of symbols]

[0057] 1. Map Generation System 2 Vehicles 3 Server 31 Communication Interface 32 Storage Device 33 Memory 34 processors 41 Probability distribution calculation unit 42 Virtual boundary generation unit 43 Control point setting section 44 Network Generation Section 45 Map Generation Section 4. Communication Network 5. Radio base station

Claims

1. a probability distribution calculation unit that calculates, for a given area connecting a plurality of roads, a probability distribution representing the probability that each point in the area is located on a standard travel trajectory based on a plurality of travel trajectories of a vehicle that traveled from an entrance lane of one of the plurality of roads to the area to an exit lane of another of the plurality of roads to exit the area; a virtual boundary generating unit that generates a virtual boundary connecting the positions of both ends of the entrance lane and the exit lane; a control point setting unit that sets, as control points, at least two points that are located between the two virtual boundary lines and have a variance of the probability distribution equal to or smaller than a predetermined variance threshold value among the individual points in the region; a network generating unit that generates a Bezier curve based on the at least two control points and sets the generated Bezier curve as a line connecting the entrance lane and the exit lane for the area; A lane network generating device having:

2. The lane network generating device according to claim 1 , wherein the virtual boundary generating unit generates, as the virtual boundary, a Bezier curve connecting positions of both ends of the entrance lane and the exit lane.

3. calculating a probability distribution representing the probability that each point in a given area connecting a plurality of roads is located on a standard travel trajectory based on a plurality of travel trajectories of a vehicle that traveled from an entrance lane of one of the plurality of roads to enter the area to an exit lane of another of the plurality of roads to exit the area; generating a virtual boundary line connecting the positions of both ends of the entrance lane and the exit lane; setting at least two points, among the individual points in the region, that are located between the two virtual boundary lines and have a variance of the probability distribution that is equal to or smaller than a predetermined variance threshold, as control points; generating a Bezier curve based on the at least two control points; The generated Bezier curve is a line connecting the entrance lane and the exit lane for the area. A lane network generating method comprising:

4. calculating a probability distribution representing the probability that each point in a given area connecting a plurality of roads is located on a standard travel trajectory based on a plurality of travel trajectories of a vehicle that traveled from an entrance lane of one of the plurality of roads to enter the area to an exit lane of another of the plurality of roads to exit the area; generating a virtual boundary line connecting the positions of both ends of the entrance lane and the exit lane; setting at least two points, among the individual points in the region, that are located between the two virtual boundary lines and have a variance of the probability distribution that is equal to or smaller than a predetermined variance threshold, as control points; generating a Bezier curve based on the at least two control points; The generated Bezier curve is a line connecting the entrance lane and the exit lane for the area. A computer program for generating a lane network, which causes a computer to execute the above-mentioned steps.

Images