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

Abstract

A lane network generation device, a lane network generation method, and a computer program for lane network generation are provided. The lane network generation device has: a probability distribution calculation section (41) that calculates, based on a plurality of travel trajectories of vehicles that have traveled in a predetermined area connecting an entry lane and an exit lane, a probability distribution that indicates a probability that each point within the area is on a standard travel trajectory; a hypothetical boundary line generation section (42) that generates hypothetical boundary lines that link positions of both ends of the entry lane and the exit lane to each other; a control point setting section (43) that sets, as control points, at least two points among the points within the predetermined area that are between the two hypothetical boundary lines and have a variance of the probability distribution that is below a predetermined variance threshold; and a network generation section (44) that generates a Bezier curve based on the at least two control points and uses the generated Bezier curve as a line in the lane network that links the two lanes with respect to the predetermined area.

Description

Technical Field

[0001] This invention relates to a lane network generation apparatus, a lane network generation method, and a computer program for generating lane networks, wherein the lane network represents the connection relationship between the lanes of multiple interconnected roads. Background Technology

[0002] As one of the pieces of information contained in the high-precision map referenced by the vehicle's autonomous driving system for autonomous driving control, a lane network is provided, which represents the connection relationships between lanes and is referenced for setting the predetermined driving trajectory of the vehicle under autonomous driving control. Therefore, a technique for generating lane networks has been proposed (see Japanese Patent Application Publication No. 2010-26875 and Japanese Patent Application Publication No. 2018-105636).

[0003] The road network data generation apparatus disclosed in Japanese Patent Application Publication No. 2010-26875 calculates the intersection point of the extension line from the outer lane of the intersection connected to the entry point into the intersection and the extension line from the outer lane of the intersection connected to the exit point into the intersection, based on the combination of entry and exit points capable of travel at intersections. Furthermore, the apparatus sets a correction point between the entry point and the intersection point. Additionally, the apparatus sets an arc based on a quadratic Bézier curve for the three points—the intersection point, the correction point, and the exit point—as the inner lane of the intersection.

[0004] Furthermore, the path generation device disclosed in Japanese Patent Application Publication No. 2018-105636 obtains point data based on map information and the vehicle's own position. This point data includes multiple points corresponding to the centerline of the road on which the vehicle is traveling. The device then smooths the point data to obtain the road centerline, generating it as the vehicle's target travel path. In this process, the device extracts multiple point clusters, including adjacent points, from the point data and calculates the centroid positions of each extracted point cluster. Based on these centroid positions, the device performs smoothing using parametric curves such as Bézier curves, B-splines, NURBS, or spline interpolation curves. Summary of the Invention

[0005] In the aforementioned techniques, lane networks are generated by referencing map information. However, vehicle behavior is not considered in lane networks generated from map information; therefore, using this lane network as a target driving trajectory for vehicles can sometimes result in unnatural trajectories.

[0006] Therefore, the object of the present invention is to provide a lane network generation apparatus capable of generating a lane network suitable for vehicle driving.

[0007] According to one embodiment, a lane network generation apparatus is provided. This lane network generation apparatus includes: a probability distribution calculation unit that, with respect to a predetermined area connecting multiple roads, calculates a probability distribution for each point in the area, representing the probability that the point is located on a standard driving trajectory, based on multiple driving trajectories of vehicles traveling from an entry lane entering the area from one of the multiple roads to an exit lane exiting the area from another of the multiple roads; an imaginary boundary line generation unit that generates imaginary boundary lines connecting the two ends of the entry lane and the exit lane; a control point setting unit that sets at least two points in the area that are located between two imaginary boundary lines and whose variance of the probability distribution is below a predetermined variance threshold as control points; and a network generation unit that generates a Bézier curve based on the at least two control points, and uses the generated Bézier curve as a line connecting the entry lane and the exit lane with respect to the predetermined area.

[0008] In this lane network generation device, it is preferable that the imaginary boundary line generation unit generates a Bézier curve that connects the two ends of the entry lane and the exit lane to each other as the imaginary boundary line.

[0009] According to other embodiments, a lane network generation method is provided. The lane network generation method includes: for a predetermined area connecting multiple roads, based on multiple driving trajectories of vehicles traveling from an entry lane of one of the multiple roads into an exit lane of another of the multiple roads exiting the area, calculating a probability distribution for each point in the area representing the probability that the point is located on a standard driving trajectory; generating imaginary boundary lines connecting the two ends of the entry lane and the exit lane; setting at least two points in the predetermined area that are located between the two imaginary boundary lines and whose variance of the probability distribution is below a predetermined variance threshold as control points; generating a Bézier curve based on the at least two control points; and using the generated Bézier curve as a line connecting the entry lane and the exit lane with respect to the predetermined area.

[0010] According to another embodiment, a computer program for generating a lane network is provided. This computer program for generating a lane network includes commands for causing a computer to perform the following processing: with respect to a predetermined area connecting multiple roads, based on multiple travel trajectories of vehicles traveling from an entry lane of one of the multiple roads into an exit lane of another of the multiple roads, calculating for each point in the area a probability distribution representing the probability that the point is located on a standard travel trajectory; generating imaginary boundary lines connecting the two ends of the entry and exit lanes; setting at least two points in the predetermined area that are located between the two imaginary boundary lines and whose variance of the probability distribution is below a predetermined variance threshold as control points; generating a Bézier curve based on the at least two control points; and using the generated Bézier curve as a line connecting the entry and exit lanes with respect to the predetermined area.

[0011] The lane network generation apparatus disclosed herein achieves the effect of generating a lane network suitable for vehicle travel. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of a map generation system equipped with a lane network generation device.

[0013] Figure 2 This is a hardware structure diagram of a server as an example of a lane network generation device.

[0014] Figure 3 This is a functional block diagram of the processor associated with lane network generation processing.

[0015] Figure 4 This is a diagram showing an example of a Bézier curve that connects the endpoints of the entry lane and the exit lane.

[0016] Figure 5 This is a diagram illustrating an example of control point setting and lane network generation.

[0017] Figure 6 This is a flowchart of the lane network generation and processing steps. Detailed Implementation

[0018] Hereinafter, with reference to the accompanying drawings, the lane network generation apparatus, the lane network generation method executed in the lane network generation apparatus, and the computer program for lane network generation will be described. This lane network generation apparatus, for areas such as intersections where multiple roads connect, generates a lane network representing the connection relationships between lanes of the multiple roads connected to that area, based on actual driving trajectories collected from one or more vehicles.

[0019] Specifically, the lane network generation device calculates a probability distribution for each point within the intersection, representing the probability of being on a standard driving trajectory, based on multiple driving trajectories of vehicles traveling between two lanes at a predetermined intersection. Furthermore, the device generates two imaginary boundary lines connecting the two ends of each lane. Further, the device determines a column of points from the calculated probability distribution that are located between the two imaginary boundary lines and whose variance is below a predetermined variance threshold, and sets at least two points from this column as control points. Based on these control points, the device generates a Bézier curve, which serves as the line connecting the two lanes at the intersection in the lane network.

[0020] Figure 1 This is a schematic diagram of a map generation system equipped with a lane network generation device. In this embodiment, the map generation system 1 includes at least one vehicle 2 and a server 3, which is an example of a lane network generation device. Each vehicle 2 connects to the server 3, for example, by accessing a wireless base station 5 connected to a communication network 4 connected to the server 3 via a gateway (not shown) or the like, thereby connecting via the wireless base station 5 and the communication network 4. Furthermore, in Figure 1 For simplicity, only one vehicle 2 is shown in the diagram, but the map generation system 1 can also have multiple vehicles 2. Similarly, in Figure 1 The diagram only shows one wireless base station 5, but multiple wireless base stations 5 can also be connected to the communication network 4.

[0021] Vehicle 2 includes a GPS receiver, a driving trajectory recording device, and a wireless communication terminal. The GPS receiver receives GPS signals from GPS satellites at predetermined intervals and determines the vehicle 2's own position based on the received GPS signals. Furthermore, the GPS receiver outputs positioning information, representing the vehicle 2's own position based on GPS signals, to the driving trajectory recording device via an in-vehicle network at predetermined intervals. Alternatively, vehicle 2 may have a receiver that follows a satellite positioning system other than a GPS receiver. In this case, the receiver alone can determine the vehicle 2's own position.

[0022] The driving trajectory recording device, for example, includes a processor and a memory. The processor of the driving trajectory recording device arranges the positions of vehicle 2, represented by positioning information obtained from a GPS receiver, in chronological order to generate driving trajectory information representing the trajectory traveled by vehicle 2, and stores the generated driving trajectory information in the memory of the driving trajectory recording device. The driving trajectory recording device may also include vehicle 2 identification information in the driving trajectory information.

[0023] Furthermore, vehicle 2 may also have a camera for capturing images of the area surrounding vehicle 2. The camera may also generate images representing the area surrounding vehicle 2 at predetermined intervals and output these images to a driving trajectory recording device via an in-vehicle network. In this case, whenever the processor of the driving trajectory recording device receives an image from the camera, it detects ground features on the road or around the road from the image. Further, the processor of the driving trajectory recording device obtains distance measurement information of vehicle 2 from an electronic control device (not shown) that controls the movement of vehicle 2, and calculates the amount of movement and orientation displacement of vehicle 2 at each generation interval of each image. Furthermore, the processor of the driving trajectory recording device may also estimate the position of vehicle 2 at the time each image is generated based on the ground features detected from each image and the amount of movement and orientation displacement of vehicle 2 at each generation interval of each image, using a so-called Structure from Motion method. Furthermore, the processor of the driving trajectory recording device may also generate driving trajectory information by arranging the positions of vehicle 2 at the time each image is generated in chronological order.

[0024] When a predetermined time is reached, the driving trajectory recording device outputs the generated driving trajectory information to the wireless communication terminal. Furthermore, the predetermined time can be set, for example, when the ignition switch of vehicle 2 is turned off, or whenever a certain period of time (e.g., 30 minutes to 1 hour) elapses after the ignition switch of vehicle 2 is turned on. Alternatively, the vehicle 2 can be notified in advance of collection area information from server 3 via communication network 4 and wireless base station 5, indicating the area to be collected for the driving trajectory. In this case, the driving trajectory recording device can also refer to the collection area information and positioning information to determine the predetermined time when vehicle 2 moves out of the area to be collected.

[0025] The wireless communication terminal is a device that performs wireless communication processing according to a predetermined wireless communication standard. For example, it accesses the wireless base station 5 and connects to the server 3 via the wireless base station 5 and the communication network 4. The wireless communication terminal generates an uplink wireless signal containing driving trajectory information received from the driving trajectory recording device. The wireless communication terminal sends this uplink wireless signal to the wireless base station 5, thereby sending the driving trajectory information to the server 3. Additionally, the wireless communication terminal receives downlink wireless signals from the wireless base station 5 and transmits the collection area information from the server 3 contained in these signals to the driving trajectory recording device.

[0026] Next, server 3, which is an example of a lane network generation device, will be described.

[0027] Figure 2This is a hardware structure diagram of server 3, an example of a lane network generation device. Server 3 has a communication interface 31, a storage device 32, a memory 33, and a processor 34. The communication interface 31, storage device 32, and memory 33 are connected to processor 34 via signal lines. Server 3 may also have input devices such as a keyboard and mouse, and a display device such as an LCD.

[0028] Communication interface 31 is an example of a communication unit, having an interface circuit for connecting server 3 to communication network 4. Furthermore, communication interface 31 is configured to communicate with vehicle 2 via communication network 4 and wireless base station 5. Specifically, communication interface 31 transmits driving trajectory information received from vehicle 2 via wireless base station 5 and communication network 4 to processor 34. Additionally, communication interface 31 transmits collection area information received from processor 34 to vehicle 2 via communication network 4 and wireless base station 5.

[0029] Storage device 32 is an example of a storage unit, such as having a hard disk drive or an optical recording medium and its access device. Storage device 32 stores various data and information used in the lane network generation process. For example, storage device 32 stores a map for adding information to represent the lane network. Furthermore, storage device 32 stores driving trajectory information received from each vehicle 2. Furthermore, storage device 32 may also store a computer program executed on processor 34 for performing the lane network generation process.

[0030] Memory 33 is another example of a storage unit, such as a non-volatile semiconductor memory and a volatile semiconductor memory. Furthermore, memory 33 temporarily stores various data generated during the execution of lane network generation processing.

[0031] The processor 34 has one or more CPUs (Central Processing Units) and their peripheral circuits. The processor 34 may also have other arithmetic circuits such as logic units or numerical arithmetic units. Furthermore, whenever the processor 34 receives driving trajectory information from any of the vehicles 2, it saves the received driving trajectory information to the storage device 32. Further, the processor 34 performs lane network generation processing.

[0032] Figure 3This is a functional block diagram of the processor 34 associated with lane network generation processing. The processor 34 includes a probability distribution calculation unit 41, a hypothetical boundary line generation unit 42, a control point setting unit 43, a network generation unit 44, and a map generation unit 45. These units of the processor 34 are, for example, functional modules implemented by a computer program operating on the processor 34. Alternatively, these units of the processor 34 may also be provided in a dedicated arithmetic circuit of the processor 34.

[0033] The probability distribution calculation unit 41 determines the lanes in which vehicle 2 can enter the intersection from one of the roads connected to the intersection that is the target of the lane network generation, and the lanes in which vehicle 2 can exit the intersection from the intersection from the other roads it is the target of the generation. Hereinafter, the lanes in which vehicle 2 can enter the intersection are sometimes referred to as entry lanes. In addition, the lanes in which vehicle 2 exits the intersection are sometimes referred to as exit lanes. Furthermore, the probability distribution calculation unit 41 calculates a predetermined probability distribution for each point within the intersection based on multiple driving trajectory information of vehicle 2 when it travels from the target entry lane toward the exit lane. The predetermined probability distribution represents the likelihood that a vehicle traveling from the target entry lane toward the exit lane is located on a standard driving trajectory. In addition, the standard driving trajectory is conceived as the trajectory of a vehicle traveling from the target entry lane toward the exit lane within the intersection at a standard speed without any evasive maneuvers.

[0034] The probability distribution calculation unit 41 determines the driving trajectory information stored in the storage device 32, which represents a driving trajectory that passes through a range shown on a map corresponding to the entry lane and exit lane of the target intersection. Furthermore, the probability distribution calculation unit 41 determines the driving trajectory represented by each driving trajectory information as the driving trajectory of vehicle 2 when it travels from the target entry lane towards the target exit lane. The probability distribution calculation unit 41 reads the determined driving trajectory information from the storage device 32.

[0035] The probability distribution calculation unit 41 calculates a predetermined probability distribution for each point within the intersection based on the driving trajectory represented by the driving trajectory information read from the storage device 32. In this embodiment, the probability distribution for each point is set to follow a Gaussian process. Therefore, the predetermined probability distribution can be set to a two-dimensional Gaussian distribution parallel to the road surface.

[0036] The probability distribution calculation unit 41 calculates the probability distribution for each point within the intersection, referring to each driving trajectory, using a Gaussian process regression model calculation method, such as Bayesian inference. Specifically, for each point within the intersection, the probability distribution calculation unit 41 sets a two-dimensional Gaussian distribution with a predetermined variance-covariance matrix as the initial probability distribution. Furthermore, using the set Gaussian distribution as the prior probability distribution, the probability distribution calculation unit 41 calculates the subsequent probability distribution for each point within the intersection based on each point on each driving trajectory. The probability distribution calculation unit 41 updates the variance-covariance matrix for each point within the intersection according to the calculated subsequent probability distribution, thereby calculating the probability distribution for each point within the intersection. The average position of the calculated probability distribution for each point, i.e., the probability value at that point, represents the likelihood that the point lies on the standard driving trajectory. Furthermore, the variance-covariance matrix of this probability distribution represents the degree of observation error when the point lies on a standard driving trajectory. The larger the value of each element in the variance-covariance matrix, the greater the observation error. Therefore, the more driving trajectories a point has traversed, the higher the probability value at the average position, and the smaller the values ​​of each element in the variance-covariance matrix.

[0037] The probability distribution calculation unit 41 notifies the control point setting unit 43 of the probability distribution calculated for each point.

[0038] The hypothetical boundary line generation unit 42 determines, with reference to a map, the positions of the two ends of the intersection of the target entry lane and the target exit lane in the width direction of the lane. Furthermore, the hypothetical boundary line generation unit 42 pairs the endpoints of one side of the target entry lane and the endpoint of one side of the target exit lane, and pairs the endpoints of the other side of the target entry lane and the other side of the target exit lane. At this time, the hypothetical boundary line generation unit 42 sets each pair in such a way that the lines connecting the endpoints of the entry lane and the endpoint of the exit lane in each pair do not intersect each other.

[0039] The imaginary boundary line generation unit 42 generates a Bézier curve connecting the endpoints of the entering lane and the exit lane. Simultaneously, the imaginary boundary line generation unit 42 extends the boundary of the entering lane (defined by lane markings or curbs, for example) passing through the endpoint of the entering lane into the intersection, and extends the boundary of the exit lane (which passes through the endpoint of the exit lane) into the intersection. The imaginary boundary line generation unit 42 calculates the intersection point of the line obtained by extending the boundary on the entering lane side and the line obtained by extending the boundary on the exit lane side, and sets control points at the midpoint of the intersection point and the midpoint of the endpoint on the entering lane side, respectively. Based on the set control points, the imaginary boundary line generation unit 42 generates a Bézier curve connecting the endpoints of the entering lane side and the exit lane side as the imaginary boundary line.

[0040] Figure 4 This is a diagram illustrating an example of a Bézier curve connecting the endpoints of an entry lane and an exit lane. Figure 4 In this system, roads connect to intersection 400 from the top, bottom, left, and right. Let the entry lane 402 of road 401, which connects to intersection 400 from below, and the exit lane 404 of road 403, which connects to intersection 400 from the right, be the focus of this system, respectively. In this case, a Bézier curve 411 connecting the right endpoint 402a of entry lane 402 to the lower endpoint 404a of exit lane 404, and a Bézier curve 412 connecting the left endpoint 402b of entry lane 402 to the upper endpoint 404b of exit lane 404 are generated as imaginary boundary lines. Imaginary boundary lines 411 and 412 connect the endpoints of entry lane 402 and exit lane 404 to each other, respectively. Therefore, the area enclosed by imaginary boundary lines 411 and 412 is presumed to be the area corresponding to the passage of vehicles from the entry lane 402 toward the exit lane 404. That is, imaginary boundary lines 411 and 412 are respectively presumed to represent the boundaries of the area envisioned as the passage of vehicles from the entry lane 402 toward the exit lane 404 (hereinafter sometimes referred to as the imaginary passage area).

[0041] The imaginary boundary line generation unit 42 notifies the control point setting unit 43 of the two imaginary boundary lines that have been generated.

[0042] The control point setting unit 43 determines a column of points within the intersection that are located between two imaginary boundary lines and whose probability distribution variance is below a predetermined variance threshold. Furthermore, the control point setting unit 43 sets at least two points from the determined column of points as control points.

[0043] The control point is preferably located at a position where the deviation from the position of each travel trajectory in a direction orthogonal to the direction of travel of vehicle 2 moving from the observed entry lane towards the exit lane is minimal. The control point setting unit 43 then calculates a baseline connecting the midpoints of the observed entry lane and exit lane. This baseline can be, for example, a straight line or a convex arc relative to the center of the intersection. The control point setting unit 43 calculates the position on the nearest baseline to each point within the intersection and determines the tangent direction of the baseline at that calculated position. The control point setting unit 43 then performs an affine transformation on the probability distribution of each point within the intersection to an orthogonal coordinate system, where the tangent direction at the position on the nearest baseline and the direction orthogonal to the tangent direction are respectively used as axes. Furthermore, for each point within the intersection, the control point setting unit 43 uses the element values ​​of the variance covariance matrix in the direction orthogonal to the aforementioned tangent direction in the probability distribution after affine transformation as the variance value to be compared with the variance threshold.

[0044] Furthermore, the control point setting unit 43 can divide the area within the intersection into multiple sections (e.g., 5 to 10) along a straight line connecting the midpoint of the approach lane and the midpoint of the exit lane. Also, the control point setting unit 43 can set two or more points belonging to the aforementioned point series that are contained within distinct sections as control points. This prevents the spacing between control points from becoming too narrow.

[0045] By using points whose variance is below a predetermined variance threshold as control points, the control point setting unit 43 sets control points at locations where the vehicle is highly likely to pass through whenever it travels from the observed entry lane toward the exit lane. Furthermore, by setting control points within the hypothetical passage area, the control point setting unit 43 can prevent the vehicle from using locations on abnormal travel trajectories, such as when the vehicle 2, for some reason, makes a wide turn or takes a shortcut from the observed entry lane to the exit lane, as control points.

[0046] The control point setting unit 43 notifies the network generation unit 44 of the location of each set control point.

[0047] The network generation unit 44 generates Bézier curves for intersections that are the targets of lane network generation, based on at least two control points set by the control point setting unit 43. Furthermore, the network generation unit 44 uses the generated Bézier curves as lines connecting the target entry and exit lanes in the lane network for that intersection. By using the control points set as described above, the network generation unit 44 can smoothly connect the target entry and exit lanes, and generate a lane network that closely approximates the trajectory of a vehicle during actual travel.

[0048] Figure 5 This is a diagram illustrating an example of control point setting and lane network generation. Figure 5 In, with Figure 4 Similarly, in the intersection shown, roads connect to intersection 500 from the top, bottom, left, and right. Let the approach lane 502 of road 501, which connects to intersection 500 from below, and the exit lane 504 of road 503, which connects to intersection 500 from the right, be the considered approach lane and exit lane, respectively. The variance of the probability distribution of points 531 and 532 within intersection 500, located between the two imaginary boundary lines 511 and 512 connecting the endpoints of approach lane 502 and exit lane 504, is below a variance threshold. Therefore, points 531 and 532 are designated as control points. Furthermore, the Bézier curve 541 generated based on control points 531 and 532 becomes the line connecting approach lane 502 and exit lane 504, constituting the lane network.

[0049] The network generation unit 44 outputs information representing the generated lane network to the map generation unit 45.

[0050] The map generation unit 45 generates a map containing information representing the lane network for each intersection area. Specifically, the map generation unit 45 adds information representing the lane network generated by the network generation unit 44 to each intersection shown on the map read from the storage device 32. Thus, a map containing the lane network is generated.

[0051] The map generation unit 45 saves the generated map to the storage device 32. Alternatively, the map generation unit 45 may output the generated map to other devices via the communication interface 31.

[0052] Figure 6 This is the flowchart of the lane network generation process in server 3. The processor 34 of server 3 executes the lane network generation process according to the following flowchart, based on the groups of entering and exiting lanes in the intersections that are the objects of lane network generation.

[0053] The probability distribution calculation unit 41 of the processor 34 calculates the probability distribution representing the likelihood of being located on the driving trajectory for each point within the intersection based on each driving trajectory from the observed entry lane to the observed exit lane (step S101).

[0054] In addition, the imaginary boundary line generation unit 42 of the processor 34 generates a Bézier curve that connects the two ends of the lane in the width direction at the intersection of the entrance lane and the exit lane to each other as an imaginary boundary line (step S102).

[0055] The control point setting unit 43 of the processor 34 determines a column of points within the intersection that are located between two imaginary boundary lines and whose variance of probability distribution is below a predetermined variance threshold (step S103). Furthermore, the control point setting unit 43 sets at least two points from the determined column of points as control points (step S104).

[0056] The network generation unit 44 of the processor 34 generates a Bézier curve based on at least two set control points, and uses the generated Bézier curve as the line connecting the observed entry lane and exit lane in the lane network (step S105). Furthermore, the map generation unit 45 of the processor 34 adds information to the map representing the line connecting the observed entry lane and exit lane (step S106). Then, the processor 34 ends the lane network generation process.

[0057] As explained above, this lane network generation device calculates the probability distribution of points located on the driving trajectories from the entry lane to the exit lane of the intersection being the target of lane network generation, based on multiple driving trajectories. Furthermore, the device sets at least two points within the intersection that are located within a hypothetical passage area and whose variance of the probability distribution is below a variance threshold as control points. The device then uses a Bézier curve generated based on the set control points as the line connecting the target entry lane and exit lane in the lane network. By generating the lane network in this way, the device can create a lane network suitable for vehicle traffic.

[0058] According to a variation, the imaginary boundary line generation unit 42 may also generate an imaginary boundary line connecting the endpoints of the considered entry lane and exit lane as a line other than a Bézier curve. For example, the imaginary boundary line generation unit 42 may also use an arc or spline curve as the imaginary boundary line, which connects the endpoints of the considered entry lane and exit lane and is convex relative to the center of the intersection. Alternatively, the imaginary boundary line generation unit 42 may also generate an imaginary boundary line connecting the endpoints of the considered entry lane and exit lane as a spiral curve that is convex relative to the center of the intersection and has a curvature that increases with distance from the endpoints of the entry lane and exit lane.

[0059] Furthermore, the area of ​​the road that becomes the object of lane network generation in the above-described embodiments or variations is not limited to intersections, but can also be an area where one road connects to other roads, such as a branching point or merging point. In this case, the lane network generation device can also generate a lane network by performing the same processing as the above-described embodiments or variations according to the groups of entering and exiting lanes.

[0060] The computer program that enables the computer to perform the functions of the processor of the lane network generation apparatus according to the above-described embodiments or variations may also be provided in the form of a recording medium that can be read by a computer. Furthermore, the recording medium that can be read by a computer may be, for example, a magnetic recording medium, an optical recording medium, or a semiconductor memory.

[0061] As described above, those skilled in the art can make various modifications within the scope of this invention according to the implemented form.

Claims

1. A lane network generation device, comprising: The probability distribution calculation unit calculates a probability distribution for each point in a predetermined area connecting multiple roads, based on multiple driving trajectories of vehicles that have traveled from an entry lane in one of the multiple roads to an exit lane in another of the multiple roads. This probability distribution represents the probability that a point is located on a standard driving trajectory. An imaginary boundary line generation unit generates imaginary boundary lines that connect the positions of the two ends of the entry lane and the exit lane to each other. The control point setting unit sets at least two points among the points in the region that are located between two hypothetical boundary lines and whose variance of the probability distribution is below a predetermined variance threshold as control points. as well as The network generation unit generates a Bézier curve based on the at least two control points, and uses the generated Bézier curve as a line connecting the entry lane and the exit lane with respect to the region.

2. The lane network generation device according to claim 1, The imaginary boundary line generation unit generates a Bézier curve that connects the positions of the two ends of the entry lane and the exit lane to each other as the imaginary boundary line.

3. A method for generating a lane network, comprising: Regarding a predetermined area connecting multiple roads, based on multiple driving trajectories of vehicles traveling from an entry lane in one of the multiple roads to an exit lane in another of the multiple roads, a probability distribution representing the probability that a point is located on a standard driving trajectory is calculated for each point in the area. Generate an imaginary boundary line that connects the positions of the two ends of the entry lane and the exit lane to each other; At least two points within the region that are located between the two hypothetical boundary lines and whose variance of the probability distribution is below a predetermined variance threshold are designated as control points. Generate a Bézier curve based on the at least two control points; as well as The generated Bézier curve is used as a line connecting the entry lane and the exit lane with respect to the region.

4. A computer program for generating lane networks, used to cause a computer to execute: Regarding a predetermined area connecting multiple roads, based on multiple driving trajectories of vehicles traveling from an entry lane into the area on any of the multiple roads to an exit lane exiting the area on other roads of the multiple roads, a probability distribution is calculated for each point in the area, representing the probability that the point is located on a standard driving trajectory. Generate an imaginary boundary line that connects the positions of the two ends of the entry lane and the exit lane to each other; At least two points within the region that are located between the two hypothetical boundary lines and whose variance of the probability distribution is below a predetermined variance threshold are designated as control points. Generate a Bézier curve based on the at least two control points; as well as The generated Bézier curve is used as a line connecting the entry lane and the exit lane with respect to the region.

Images