JP7779194B2 - Travel path estimation method and travel path estimation device - Google Patents

Description

The present invention relates to a roadway estimation method and a roadway estimation device.

Patent Document 1 discloses a lane map generation device that generates a highly accurate lane-level map by detecting a group of center points for each driving lane from an image captured by a camera mounted on a vehicle, storing the detected group of center points in association with each link on a road network, and generating center lines for each driving lane corresponding to each link. At intersections, this lane map generation device associates entrance points to the intersection with exit points from the intersection, and estimates the driving lanes within the intersection using spline curves connecting the entrance points, midpoints, and exit points.

JP 2015-4814 A

In the conventional lane map generation device described above, the direction of the entrance point is determined to establish a correspondence between the entrance point and exit point of an intersection. To determine the direction of the entrance point, a statistical analysis of the driving trajectory that passes through both the entrance point and the exit point may be performed. In this case, in order to estimate the driving lane within the intersection, multiple driving trajectories that actually pass through both the entrance point and the exit point of the driving lane must be acquired.

The present invention was made in consideration of the above problems, and aims to provide a roadway estimation method and roadway estimation device that can estimate the shape of a roadway even in areas at intersections where no driving trajectory has been acquired.

To solve the above-mentioned problems, a lane estimation method and device according to one aspect of the present invention acquires data indicating the positions and directions of multiple road arrow markings just before an intersection, acquires data indicating the travel trajectory of a vehicle traveling through the intersection, calculates the distance in the lane width direction from the travel trajectory to each of the multiple road arrow markings, determines the destination of the travel trajectory based on the direction of a reference road arrow marking, which is a road arrow marking whose distance is less than a first threshold, generates data indicating a duplicated travel trajectory by translating the travel trajectory to the position of a road arrow marking that indicates the same direction as the destination of the travel trajectory and is different from the reference road arrow marking, and estimates the shape of the lane within the intersection based on the data indicating the duplicated travel trajectory.

According to the present invention, it is possible to estimate the shape of a road at an intersection even in areas where no driving trajectory has been acquired.

FIG. 1 is a block diagram showing an example of the configuration of a roadway estimation device according to an embodiment of the present invention. FIG. 2 is a diagram illustrating an example of road arrow markings and the positions of the road arrow markings. FIG. 3 is a flowchart showing an example of the flow of operations performed by the path estimation unit of the path estimation device according to one embodiment of the present invention. FIG. 4 is a flowchart showing the processing steps of a travel locus destination determination process performed by the travel path estimation unit of the travel path estimation device according to one embodiment of the present invention. FIG. 5 is a diagram for explaining a method for determining the destination of a travel path. FIG. 6 is a diagram for explaining a method for determining the destination of a travel locus. FIG. 7 is a flowchart showing the processing steps of a travel trajectory duplication process performed by the travel path estimation unit of the travel path estimation device according to one embodiment of the present invention. FIG. 8 is a diagram for explaining a method for duplicating a travel locus. FIG. 9 is a diagram for explaining a method for duplicating a travel locus. FIG. 10A is a diagram for explaining a method for adjusting the turning radius of a traveling locus. FIG. 10B is a diagram for explaining a method for adjusting the turning radius of the traveling locus.

Embodiments of the present invention will be described below with reference to the drawings. In the drawings, identical parts will be designated by the same reference numerals and their descriptions will be omitted.

[Configuration example of a roadway estimation device]
An example of the configuration of a roadway estimation device according to this embodiment will be described with reference to Fig. 1. As shown in Fig. 1, the roadway estimation device 1 includes a recognition unit 10, a storage unit 11, a car navigation system 12, a roadway estimation unit 13, a camera 20, a Light Detection and Ranging (LiDAR) 30, and a Global Navigation Satellite System (GNSS) receiver 40, and is configured using, for example, an IC, an LSI, or the like.

The path estimation device 1 is composed of general-purpose electronic circuits including a microcomputer, microprocessor, and CPU (Central Processing Unit), as well as peripheral devices such as memory. Each function of this path estimation device 1 can be implemented by one or more processing circuits. The processing circuit includes a programmed processing device, such as a processing device including electrical circuits, and also includes devices such as application specific integrated circuits (ASICs) and conventional circuit components arranged to perform the functions described in the embodiments. The path estimation device 1 according to this embodiment is mounted on a vehicle.

The camera 20 has an imaging element such as a CCD (Charge-Coupled Device) or CMOS (Complementary Metal Oxide Semiconductor). The camera 20 is mounted on the vehicle and captures images of the area ahead of the vehicle at predetermined intervals. The camera 20 outputs the captured images of the area ahead of the vehicle to the recognition unit 10.

LiDAR 30 is a sensor that detects environmental information around the vehicle using LiDAR, an optical sensor technology. Specifically, LiDAR 30 scans a laser beam within a certain angular range, receives the reflected light, and measures the time difference between the time the laser is emitted and the time the reflected light is received to detect the relative distance and direction of other vehicles to the vehicle (i.e., their relative position to the vehicle). LiDAR 30 outputs the detected data to the recognition unit 10. Note that LiDAR 30 may be any sensor that detects the relative position of other vehicles around the vehicle, and may use sensors other than LiDAR, such as laser radar or sonic radar. LiDAR and multiple different types of sensors other than LiDAR may also be used.

The GNSS receiver 40 calculates position information indicating the vehicle's latitude, longitude, and altitude based on signals received from multiple navigation satellites, and detects the vehicle's position on the ground. The GNSS is, for example, a GPS receiver in the Global Positioning System (GPS). The GNSS receiver 40 outputs the vehicle's position information to the recognition unit 10 and the car navigation system 12.

The recognition unit 10 is connected to the camera 20, LiDAR 30, and GNSS receiver 40, and includes an arrow marking detection unit 101 and a driving trajectory detection unit 102.

The arrow marking detection unit 101 detects road arrow markings (hereinafter simply referred to as arrow markings) indicating the destination of the lane from the image captured by the camera 20. The arrow marking detection unit 101 generates a projection image by projecting the acquired forward image onto a plane (road surface). The arrow marking detection unit 101 generates the projection image and identifies the position and direction of multiple arrow markings just before an intersection, for example, as shown in Figure 2. The position and direction of the arrow markings are identified using image processing such as template matching. Since the method of identifying the position and direction of arrow markings using image processing is a known technology, a detailed description will be omitted. The arrow marking detection unit 101 sets a reference line segment passing through the center position in the lane width direction for each of the multiple arrow markings. For example, the arrow marking detection unit 101 identifies the area of the arrow marking Y from the size of a template T for the arrow marking Y on the projection image, and sets a reference line segment BL as the position of the arrow marking Y. The length of the reference line segment BL is set according to the length of the template T in the lane direction, and matches the length of the arrow marking Y. Furthermore, the "arrow marking direction" here refers to the destination of the lane on which the arrow marking is located, such as "straight ahead," "right turn," or "left turn." Note that while the types of arrow markings are not limited to those shown in Figure 2, because the types of arrow markings Y are finite, it is possible to detect the position and direction of all types of arrow markings.

The arrow sign detection unit 101 outputs data indicating the positions and directions of the multiple arrow signs detected before the intersection to the memory unit 11.

The driving trajectory detection unit 102 detects the driving trajectory of other vehicles using the time-varying positional changes of other vehicles around the vehicle detected by the LiDAR 30 and the position information of the vehicle detected by the GNSS receiver 40. For example, the driving trajectory detection unit 102 verifies (matches) the identity of other vehicles at different times based on the detection results of other vehicles around the vehicle detected by the LiDAR 30, and detects time-varying positional changes of other vehicles based on this matching. The method for detecting the driving trajectory of other vehicles is a known technology, so a detailed description will be omitted.

The driving trajectory detection unit 102 outputs data indicating the detected driving trajectory of the other vehicle to the memory unit 11.

The memory unit 11 stores data indicating the positions and directions of multiple arrow signs before an intersection detected by the arrow sign detection unit 101, and data indicating the driving trajectories of other vehicles detected by the driving trajectory detection unit 102.

The car navigation system 12 is a device that provides guidance on the route the vehicle should take to reach the destination. The car navigation system 12 is connected to the GNSS receiver 40, and when the destination is input, it generates a guided route from the vehicle's current location (or an arbitrarily set starting point) to the destination. Information about the generated guided route is displayed on a display (not shown).

The car navigation system 12 includes a locator 121 and map information 122. The locator 121 acquires latitude and longitude information of the vehicle's location from the GNSS receiver 40 and identifies the vehicle's current location on the map in the map information 122. The car navigation system 12 identifies the road on which the vehicle will travel from among the roads on the map and simultaneously detects the intersection the vehicle will enter. When the car navigation system 12 detects the intersection the vehicle will enter, it transmits information about the intersection (hereinafter referred to as intersection information) to the route estimation unit 13. The intersection information includes, for example, intersection location information. The map information 122 is information stored in the car navigation system 12 and includes various data required for route guidance, such as nodes set at predetermined intervals that indicate the location of roads, links connecting nodes, intersection location information, sign information, traffic light information, lane information (number of lanes, dedicated right and left turn lanes), landmark information, and facility information. The map information 122 may be updated periodically by obtaining the latest map information from a server.

The path estimation unit 13 is connected to the memory unit 11 and the car navigation system 12. When the locator 121 of the car navigation system 12 detects an intersection, it acquires data indicating the position and direction of multiple arrow signs just before the intersection detected by the recognition unit 10, and data indicating the path of another vehicle detected by the path detection unit 102. It then performs a process to determine the destination of the path of the other vehicle using the data indicating the position and direction of the arrow signs and the data indicating the path of the other vehicle (hereinafter referred to as the path destination determination process). It also performs a process to generate data indicating a duplicate path by duplicating the path at the position of the arrow sign indicating the same direction as the destination of the determined path (hereinafter referred to as the path duplication process). It then performs a process to estimate the shape of the path within the intersection based on the data indicating the duplicate path.

Functionally, the roadway estimation unit 13 can be classified into an intersection information acquisition unit 131, an arrow sign acquisition unit 132, a driving trajectory acquisition unit 133, a destination determination unit 134, a driving trajectory duplication unit 135, and a roadway shape estimation unit 136. An overview of each function of the roadway estimation unit 13 is provided below.

When the car navigation system 12 detects an intersection that the host vehicle is about to enter, the intersection information acquisition unit 131 acquires intersection information from the car navigation system 12. For example, the intersection information acquisition unit 131 acquires position information of the intersection that the host vehicle is about to enter from the car navigation system 12.

The arrow sign acquisition unit 132 acquires data indicating the position and direction of each of the multiple arrow signs just before the intersection from the memory unit 11 based on the acquired intersection information. For example, the arrow sign acquisition unit 132 acquires data indicating the position and direction of each of the multiple arrow signs just before the intersection from the memory unit 11 based on the acquired intersection position information.

The driving trajectory acquisition unit 133 acquires data indicating the driving trajectories of other vehicles traveling through the intersection from the memory unit 11 based on the acquired intersection information. The driving trajectory acquisition unit 133 acquires data indicating the driving trajectories of other vehicles traveling through the intersection from the memory unit 11 based on, for example, the acquired intersection position information.

The destination determination unit 134 performs processing to determine the destination of the driving trajectory. The destination determination unit 134 calculates the distance in the lane width direction from the driving trajectory of another vehicle to each of the multiple arrow markings. If it is possible to draw a perpendicular line from the driving trajectory to the reference line segment of the arrow markings, the destination determination unit 134 calculates the distance in the lane width direction between the driving trajectory and the reference line segment to which the perpendicular line is shortest as the distance in the lane width direction from the driving trajectory to the arrow marking.

If it is not possible to draw a perpendicular line from the driving trajectory to the reference line segment of the arrow marking, the destination determination unit 134 extends the driving trajectory toward the reference line segment. Then, the destination determination unit 134 calculates the distance in the lane width direction between the extended driving trajectory and the reference line segment, where the perpendicular line drawn from the extended driving trajectory to the reference line segment is the shortest, as the distance in the lane width direction from the driving trajectory to the road arrow marking. In this case, the destination determination unit 134 calculates the angle (first angle) between the extended driving trajectory and the reference line segment, and determines the destination of the driving trajectory if the angle is less than the second threshold value.

The destination determination unit 134 also identifies as a reference arrow marking an arrow marking where the distance in the lane width direction from the travel trajectory of another vehicle to the arrow marking is less than a first threshold. Of the multiple arrow markings, the destination determination unit 134 identifies as a reference arrow marking the road arrow marking where the distance in the lane width direction from the travel trajectory to the arrow marking is less than the first threshold and is the shortest. The destination of the travel trajectory is then determined based on the direction of the reference arrow marking. Note that if there are multiple directions indicated by the reference arrow marking, the destination determination unit 134 calculates the curvature of the travel trajectory and determines the destination of the travel trajectory based on the curvature.

The travel trajectory duplication unit 135 identifies an arrow marking that indicates the same direction as the destination of the travel trajectory of the other vehicle and is different from the reference road arrow marking. It then generates data indicating a duplicated travel trajectory by translating the travel trajectory of the other vehicle to the position of the identified arrow marking. If it is impossible to draw a perpendicular line from the travel trajectory to the reference line segment of the arrow marking, the travel trajectory duplication unit 135 translates the travel trajectory in the lane width direction and determines whether the translated travel trajectory intersects with another travel trajectory. If the translated travel trajectory intersects with another travel trajectory, it calculates the angle between the translated travel trajectory and the other travel trajectory (second angle), and if the angle is less than a third threshold, it translates the travel trajectory to a position where it intersects with the other travel trajectory, generating data indicating a duplicated travel trajectory.

The road shape estimation unit 136 estimates the shape of the road within the intersection based on data indicating the duplicated driving trajectory generated by the driving trajectory duplication unit 135. The road shape estimation unit 136 estimates the shape of the road within the intersection from multiple driving trajectories consisting of the duplicated driving trajectories and driving trajectories within the intersection.

With the above configuration, the roadway estimation device 1 according to this embodiment can estimate the shape of the roadway within an intersection based on the travel trajectories of other vehicles acquired while traveling just before the intersection and the map information 122 of the car navigation system 12. This eliminates the need for complex processing, such as searching for the exit point of the link on the road network corresponding to the entrance point of the intersection, calculating the midpoint, and calculating a spline curve connecting the entrance point, midpoint, and exit point, as was previously required.

Next, the operational flow of each functional unit of the travel path estimation unit 13, as well as the details of the travel path destination determination process and travel path duplication process, will be explained with reference to Figures 3 to 9.

Figure 3 is a flowchart showing an example of the flow of operations performed by each functional unit of the roadway estimation unit 13. First, in step S10 of Figure 3, the intersection information acquisition unit 131 acquires intersection information from the car navigation system 12.

The process proceeds to step S11, where the arrow sign acquisition unit 132 acquires data indicating the positions and directions of multiple arrow signs just before the intersection from the memory unit 11 based on the acquired intersection information.

The process proceeds to step S12, where the driving trajectory acquisition unit 133 acquires data indicating the driving trajectories of other vehicles traveling through the intersection from the memory unit 11 based on the acquired intersection information.

The process proceeds to step S13, where the destination determination unit 134 performs a process to determine the destination of the acquired driving trajectory (driving trajectory destination determination process). The detailed method for the driving trajectory destination determination process will now be described with reference to Figures 4 to 6.

[Driving trajectory destination determination process]
4 is a flowchart showing the processing procedure for determining the destination of a travel locus by the destination determination unit 134. For example, in FIG. 5, when arrows Y1 to Y3 are present before an intersection and other vehicles V1 and V2 are traveling through the intersection, the positions and directions of each of the arrows Y1 to Y3 are assumed to have been acquired. As the positions of the arrows Y1 to Y3, reference line segments BL1 to BL3 of each of the arrows Y1 to Y3 are assumed to have been acquired. Furthermore, a travel locus 501 of the other vehicle V1 and a travel locus 502 of the other vehicle V2 are assumed to have been acquired.

In step S1301 of FIG. 4, the destination determination unit 134 calculates the distance from the travel trajectory of another vehicle to each of the multiple arrow markings just before the intersection. For example, in FIG. 5, the destination determination unit 134 draws perpendicular lines from the travel trajectory 501 of another vehicle V1 to each of the reference line segments BL1 to BL3 of the arrow markings Y1 to Y3, and calculates the distances D1 to D3 in the lane width direction between the travel trajectory 501 and the reference line segments BL1 to BL3 where the perpendicular lines are shortest. On the other hand, because the travel trajectory 502 of another vehicle V2 is short, it is not possible to draw perpendicular lines from the travel trajectory 502 to the reference line segments BL1 to BL3 of the arrow markings Y1 to Y3. In this case, the destination determination unit 134 does not calculate the distance from the travel trajectory 502 to the arrow markings.

Processing proceeds to step S1302, where the destination determination unit 134 performs distance determination 1. In distance determination 1, it is determined whether the distance from the driving trajectory to the arrow marking calculated in step S1301 is less than a first threshold. The first threshold may be set to a value approximately half the lane width; for example, if the lane width is 3 m, it is set to 1.5 m. In distance determination 1, if the distance from the driving trajectory to the arrow marking is less than the first threshold, the destination determination unit 134 identifies the arrow marking whose distance from the driving trajectory to the arrow marking is less than the first threshold as the reference arrow marking. The reference road arrow marking is the arrow marking among the multiple arrow markings whose distance from the driving trajectory to the arrow marking is less than the first threshold and is the shortest. Processing then proceeds to step S1308.

For example, in FIG. 5, the destination determination unit 134 determines whether distances D1 to D3 are less than a first threshold. If distance D1 is less than the first threshold, the destination determination unit 134 identifies arrow mark Y1 as the reference arrow mark.

On the other hand, if the distance from the driving trajectory to the arrow marking is equal to or greater than the first threshold in distance determination 1, the destination determination unit 134 ends the processing in FIG. 4, and the path estimation unit 13 ends the processing in FIG. 3. Also, if the distance from the driving trajectory to the arrow marking is not calculated in step S1301 in distance determination 1, the destination determination unit 134 determines that distance determination is impossible, and the processing proceeds to step S1303.

In step S1303, the destination determination unit 134 extends the travel trajectory toward the reference line segment. That is, if it is not possible to draw a perpendicular line from the travel trajectory toward the reference line segment indicated by the arrow marking, the destination determination unit 134 extends the travel trajectory toward the reference line segment. Specifically, the destination determination unit 134 extends the travel trajectory toward the reference line segment using a straight line. For example, as shown in FIG. 6, the destination determination unit 134 extends the travel trajectory 502 of the other vehicle V2 in FIG. 5 toward the reference line segment BL3 using a straight line 70.

The process proceeds to step S1304, where the destination determination unit 134 calculates the distance in the lane width direction from the extended driving trajectory to each of the multiple arrow markings just before the intersection. More specifically, the destination determination unit 134 draws a perpendicular line from the straight line extending the driving trajectory to the reference line segment of the arrow markings, and calculates the distance in the lane width direction from the driving trajectory to the arrow markings as the distance in the lane width direction between the extended driving trajectory and the reference line segment where the perpendicular line is shortest. Note that if the extended driving trajectory intersects with the reference line segment, the destination determination unit 134 calculates the distance in the lane width direction from the driving trajectory to the arrow markings as 0. For example, as shown in FIG. 6, if a straight line 70 extending the driving trajectory 502 of another vehicle V2 intersects with the reference line segment BL3 of the arrow marking Y3, the destination determination unit 134 calculates the distance in the lane width direction from the driving trajectory 502 to the arrow marking Y3 as 0.

The process proceeds to step S1305, where the destination determination unit 134 performs distance determination 2. In distance determination 2, it determines whether the distance from the extended travel path to the arrow marking calculated in step S1304 is less than a first threshold. The destination determination unit 134 also identifies the arrow marking whose distance from the extended travel path to the arrow marking is less than the first threshold as the reference arrow marking. The process then proceeds to step S1306.

For example, in FIG. 6 , in distance determination 2, the destination determination unit 134 determines that the distance from the extended travel path 502 to the arrow marking Y3 is less than the first threshold value because the distance from the straight line 70 extended from the travel path 502 to the arrow marking Y3 is 0. The destination determination unit 134 also identifies the arrow marking Y3 as the reference arrow marking.

On the other hand, if, in distance determination 2, the distance from the extended driving trajectory to the arrow marking is equal to or greater than the first threshold, the destination determination unit 134 ends the processing in Figure 4, and the path estimation unit 13 ends the processing in Figure 3.

In step S1306, the destination determination unit 134 calculates the angle (first angle) between the extended travel path and the reference line segment. For example, in FIG. 6, the destination determination unit 134 calculates the angle 60 between the line 70, which is an extension of the travel path 502 of the other vehicle V2, and the reference line segment BL3.

The process proceeds to step S1307, where the destination determination unit 134 performs angle determination. In the angle determination, it is determined whether the angle between the extended driving trajectory and the reference line segment calculated in step S1306 is less than a second threshold. The second threshold is set, for example, to an absolute value of 10 degrees. If the angle between the extended driving trajectory and the reference line segment is less than the second threshold, the process proceeds to step S1308. On the other hand, if the angle between the extended driving trajectory and the reference line segment is equal to or greater than the second threshold, the destination determination unit 134 ends the process in FIG. 4, and the path estimation unit 13 ends the process in FIG. 3.

In step S1308, the destination determination unit 134 determines whether the direction indicated by the reference arrow marking is unique. A unique direction of the reference arrow marking means that there is only one direction indicated by the reference arrow marking. A non-unique direction of the reference arrow marking means that there are multiple directions indicated by the reference arrow marking. If the direction indicated by the reference arrow marking is unique, processing proceeds to step S1309. On the other hand, if the direction indicated by the reference arrow marking is not unique, processing proceeds to step S1310.

In step S1309, the destination determination unit 134 determines that the destination of the travel trajectory is in the direction of the reference arrow marking, and the processing in Figure 4 ends. For example, in Figure 5, the destination determination unit 134 determines that the destination of travel trajectory 501 is in the direction of arrow marking Y1, which has been identified as the reference arrow marking. In other words, the destination of travel trajectory 501 is determined to be "turn right," and the processing ends.

In step S1310, the destination determination unit 134 calculates the curvature ρ of the travel trajectory. Processing proceeds to step S1311, where the destination determination unit 134 determines whether the curvature ρ of the travel trajectory is greater than a positive threshold θ, greater than or equal to a negative threshold -θ and less than a positive threshold θ, or less than a negative threshold -θ. Here, the threshold θ is set to 1/100, for example. By setting the threshold θ to 1/100, it is possible to determine whether the travel trajectory is a right turn, a left turn, or going straight, with a turning radius of 100 m. If the curvature ρ is greater than the positive threshold θ, processing proceeds to step S1312, where the destination determination unit 134 determines that the destination of the travel trajectory is a "left turn," and the processing of Figure 4 ends. If the curvature ρ is greater than or equal to the negative threshold -θ and less than or equal to the positive threshold θ, processing proceeds to step S1313, where the destination determination unit 134 determines the destination of the travel trajectory to be "straight ahead," and the processing in Figure 4 ends. If the curvature ρ is less than the negative threshold -θ, processing proceeds to step S1314, where the destination determination unit 134 determines the destination of the travel trajectory to be "right turn," and the processing in Figure 4 ends.

Returning to Figure 3, if the destination of the travel trajectory is determined in step S13, processing proceeds to step S14, where the travel trajectory duplication unit 135 performs processing to duplicate the travel trajectory (travel trajectory duplication processing). The travel trajectory duplication unit 135 identifies an arrow marking that indicates the same direction as the destination of the travel trajectory of the other vehicle and that is different from the reference road arrow marking. Then, the travel trajectory of the other vehicle is translated in parallel to the position of the identified arrow marking to generate data indicating a duplicate travel trajectory. Details of the travel trajectory duplication processing will now be described with reference to Figures 7 to 9.

[Duplication process of driving locus]
7 is a flowchart showing the processing steps of the travel locus duplication process by the travel locus duplication unit 135. The processing in FIG. 7 is executed when the destination determination unit 134 determines the destination of the travel locus.

In step S1401 of FIG. 7, the driving trajectory duplication unit 135 determines whether the distance between the driving trajectory and the arrow markings can be calculated. The driving trajectory duplication unit 135 acquires information from the destination determination unit 134 regarding whether the distance between the driving trajectory of another vehicle and each of the multiple arrow markings just before the intersection could be calculated, and determines whether the distance between the driving trajectory and the arrow markings can be calculated. If the distance between the driving trajectory and the arrow markings can be calculated (YES in step S1401), the process proceeds to step S1402.

For example, in Figure 8, when there are multiple arrow markings Y5 to Y7 just before an intersection and another vehicle V is traveling through the intersection, arrow marking Y5 is identified as the reference arrow marking and the destination of the travel trajectory 504 of the other vehicle V is determined to be "straight ahead." As shown in Figure 8, if the length of the travel trajectory 504 of the other vehicle V is sufficiently long and it is possible to draw a perpendicular line from the travel trajectory 504 to the reference line segment (not shown) of the arrow marking Y5, the destination determination unit 134 can calculate the distance between the travel trajectory 504 and the arrow marking. Therefore, in Figure 8, the travel trajectory duplication unit 135 determines that the distance between the travel trajectory 504 and the arrow marking can be calculated.

On the other hand, if the distance between the travel trajectory and the arrow marking cannot be calculated (NO in step S1401), processing proceeds to step S1403. For example, in FIG. 9, when there are multiple arrow markings Y8-Y10 before an intersection and another vehicle V is traveling through the intersection, arrow marking Y8 is identified as the reference arrow marking and the destination of the travel trajectory 505 of the other vehicle V is determined to be "straight ahead." As shown in FIG. 9, if the length of the travel trajectory 505 of the other vehicle V is short and it is not possible to draw a perpendicular line from the travel trajectory 504 to the reference line segment (not shown) of arrow marking Y8, the destination determination unit 134 cannot calculate the distance between the travel trajectory 505 and the arrow marking. Therefore, in FIG. 9, the travel trajectory duplication unit 135 determines that the distance between the travel trajectory 505 and the arrow marking cannot be calculated. In Figure 9, it is assumed that there are other vehicle travel trajectories (hereinafter referred to as "other travel trajectories") 506, 607 that are different from the travel trajectory 505 of the other vehicle V. Here, it is assumed that processing is being performed on travel trajectory 505 in Figure 9.

In step S1402, the travel trajectory duplication unit 135 identifies an arrow marking that indicates the same direction as the destination of the travel trajectory and is different from the reference arrow marking. The travel trajectory is then translated to the position of the identified arrow marking, and processing proceeds to step S1407. For example, in FIG. 8, the travel trajectory duplication unit 135 identifies arrow marks Y6 and Y7 that indicate the same direction as the destination "straight ahead" of the travel trajectory 504 and are different from the reference arrow marking Y5. The travel trajectory 504 is then translated to the lane width direction positions of the identified arrow marks Y6 and Y7.

In step S1403, the travel trajectory duplication unit 135 translates the travel trajectory in the lane width direction. Processing proceeds to step S1404, where the travel trajectory duplication unit 135 determines whether the translated travel trajectory intersects with other travel trajectories. If the translated travel trajectory intersects with other travel trajectories (YES in step S1404), processing proceeds to step S1405. If the travel trajectory does not intersect with other travel trajectories (NO in step S1404), the travel trajectory duplication unit 135 ends the processing in Figure 7, and the roadway estimation unit 13 ends the processing in Figure 3.

For example, in Figure 9, the travel trajectory duplication unit 135 translates the travel trajectory 505 in the lane width direction and determines whether the translated travel trajectory 505 touches other travel trajectories 506 and 507. In Figure 9, the travel trajectory duplication unit 135 determines that the translated travel trajectory 505 touches other travel trajectories 506 and 507.

In step S1405, the travel trajectory duplication unit 135 calculates the angle (second angle) between the translated travel trajectory and the other travel trajectory. The process proceeds to step S1406, where the travel trajectory duplication unit 135 determines whether the angle between the translated travel trajectory and the other travel trajectory is less than a third threshold. The third threshold is set, for example, to an absolute value of 10 degrees. If the angle between the translated travel trajectory and the other travel trajectory is less than the third threshold (YES in step S1406), the process proceeds to step S1407. On the other hand, if the angle between the translated travel trajectory and the other travel trajectory is equal to or greater than the third threshold (NO in step S1406), the travel trajectory duplication unit 135 ends the process of FIG. 7, and the travel path estimation unit 13 ends the process of FIG. 3.

In step S1407, the travel trajectory duplication unit 135 generates a duplicate travel trajectory by duplicating the travel trajectory, and then ends the processing in Figure 7. Then, the travel path shape estimation unit 136 estimates the shape of the road within the intersection based on the travel trajectory and the duplicate travel trajectory.

For example, in Figure 8, the travel trajectory duplication unit 135 generates duplicate travel trajectories 701 and 702 by translating travel trajectory 504 to the positions of arrow markers Y6 and Y7. Also, for example, in Figure 9, the travel trajectory duplication unit 135 generates duplicate travel trajectories 703 and 704 by translating travel trajectory 505 to positions where it meets travel trajectories 506 and 507.

Returning to Figure 3, if data indicating a duplicated driving trajectory is generated in step S14, processing proceeds to step S15, where the road shape estimation unit 136 estimates the shape of the road within the intersection based on the data indicating the duplicated driving trajectory generated by the driving trajectory duplication unit 135. The road shape estimation unit 136 estimates the shape of the road within the intersection from multiple driving trajectories consisting of the duplicated driving trajectory and the driving trajectory within the intersection.

For example, in Figure 8, the road shape estimation unit 136 estimates the shape of the road within the intersection from duplicate driving trajectories 701, 702 and driving trajectory 504 within the intersection. Also, for example, in Figure 9, the road shape estimation unit 136 estimates the shape of the road within the intersection from duplicate driving trajectories 703, 704 and driving trajectories 505, 506, and 507 within the intersection.

[Effects of this embodiment]
As described above, the lane estimation method and device according to this embodiment acquire data indicating the position and direction of each of multiple road arrow markings before an intersection. Data indicating the travel trajectory of a vehicle traveling through the intersection is acquired. The distance in the lane width direction from the travel trajectory to each of the multiple road arrow markings is calculated. The destination of the travel trajectory is determined based on the direction of a reference road arrow marking, which is a road arrow marking whose distance from the travel trajectory to the road arrow marking is less than a first threshold. Data indicating a duplicate travel trajectory is generated by translating the travel trajectory to a position of a road arrow marking that indicates the same direction as the destination of the travel trajectory and is different from the reference road arrow marking. The shape of the lane within the intersection is estimated based on the data indicating the duplicate travel trajectory.

The destination of a road arrow marking whose lane width distance from the vehicle's driving trajectory is less than a threshold value can be determined to be the destination of the driving trajectory. A duplicate driving trajectory is generated by duplicating the driving trajectory at a position of a road arrow marking that indicates the same direction as the driving trajectory destination and is different from the reference road arrow marking. The shape of the road within the intersection can be estimated based on the duplicate driving trajectory. This makes it possible to estimate the shape of the road even in areas of the intersection where no driving trajectory has been acquired.

In this embodiment, the reference road arrow marking is the road arrow marking among the multiple road arrow markings whose distance from the travel path to the road arrow marking is less than the first threshold and is the shortest.

The destination of the road arrow marking whose distance in the lane width direction from the vehicle's driving trajectory is less than the threshold and is the shortest is determined to be the destination of the driving trajectory. This allows the destination of the driving trajectory to be accurately determined regardless of the shape of the driving trajectory or the shape of the intersection.

In this embodiment, a reference line segment passing through the center position in the lane width direction is set for each of the multiple road arrow markings. If it is possible to draw a perpendicular line from the driving trajectory to the reference line segment, the distance in the lane width direction between the driving trajectory and the reference line segment where the perpendicular line is shortest is calculated as the distance in the lane width direction from the driving trajectory to the road arrow marking.

The direction of the road arrow marking where the distance in the lane width direction between the vehicle's driving trajectory and the reference line segment, which is the shortest perpendicular line drawn from the vehicle's driving trajectory to the reference line segment passing through the lane width center position of the road arrow marking, is less than a threshold, can be determined to be the destination of the driving trajectory. This allows the destination of the driving trajectory to be accurately determined regardless of the shape of the driving trajectory or the shape of the intersection.

In this embodiment, a reference line segment passing through the center position in the lane width direction is set for each of the multiple road arrow markings. If it is not possible to draw a perpendicular line from the driving trajectory to the reference line segment, the driving trajectory is extended toward the reference line segment. The distance in the lane width direction between the extended driving trajectory and the reference line segment, where the perpendicular line drawn from the extended driving trajectory to the reference line segment is the shortest, is calculated as the distance in the lane width direction from the driving trajectory to the road arrow marking.

The destination of the road arrow marking can be determined as the destination of the driving trajectory when the distance in the lane width direction between the extended vehicle driving trajectory and the reference line segment is less than a threshold value at the point where the perpendicular line drawn from the extended vehicle driving trajectory to the reference line segment passing through the lane width center position of the road arrow marking is the shortest. This makes it possible to accurately determine the destination of a driving trajectory, even for short driving trajectories detected partway through an intersection, regardless of the shape of the driving trajectory or the shape of the intersection.

In this embodiment, a first angle formed between the extended driving trajectory and the reference line segment is calculated. If the first angle is less than a second threshold, the destination of the driving trajectory is determined based on the direction of a reference road arrow marking, which is a road arrow marking whose lane width direction distance from the driving trajectory to the road arrow marking is less than the first threshold.

The direction of the road arrow marking where the angle between the extended driving trajectory and the reference line segment passing through the lane width center position of the road arrow marking is less than a threshold value can be determined as the destination of the driving trajectory. This allows for more accurate determination of the destination of a driving trajectory, even for short driving trajectories detected partway through an intersection, regardless of the shape of the driving trajectory or the shape of the intersection.

In this embodiment, if there are multiple directions indicated by the reference road arrow marking, the curvature of the travel path is calculated. The destination of the travel path is determined based on the curvature.

When the reference road arrow marking indicates multiple directions, the destination of the driving path is determined based on the curvature of the driving path. This makes it possible to identify a single destination for the driving path.

In this embodiment, a reference line segment passing through the center position in the lane width direction is set for each of the multiple road arrow markings. If it is not possible to draw a perpendicular line from the driving trajectory to the reference line segment, the driving trajectory is translated in the lane width direction, and a determination is made as to whether the translated driving trajectory intersects with another driving trajectory. If the translated driving trajectory intersects with another driving trajectory, a second angle formed between the translated driving trajectory and the other driving trajectory is calculated. If the second angle is less than a third threshold, the driving trajectory is translated to a position where it intersects with the other driving trajectory, and data indicating a duplicate driving trajectory is generated by duplicating the driving trajectory.

Travel trajectories that do not touch the arrow markings are duplicated by moving them in parallel to the point where they intersect with other travel trajectories, so even short travel trajectories can be duplicated. Based on the duplicated travel trajectories, the route within the intersection can be estimated.

When generating a duplicated traveling trajectory, the traveling trajectory duplication unit 135 may adjust the turning radius of the duplicated traveling trajectory by the amount of translation of the traveling trajectory. For example, in FIG. 10A , arrow markings Y11 and Y12 are present before an intersection, and another vehicle V is traveling through the intersection. Arrow marking Y11 is identified as the reference arrow marking, and the destination of the traveling trajectory 508 of the other vehicle V is determined to be "left turn." Then, suppose that a duplicated traveling trajectory 705 is generated by translating the traveling trajectory 508 to the position of arrow marking Y12. In this case, the traveling trajectory duplication unit 135 may adjust the turning radius of the duplicated traveling trajectory 705 by the amount of translation of the traveling trajectory 508 to the position of arrow marking Y12, thereby increasing the curvature of the traveling trajectory 508 and decreasing the radius of the traveling trajectory 508.

Also, for example, in Figure 10B, when arrow markings Y13 and Y14 are present just before the intersection and another vehicle V is traveling through the intersection, arrow marking Y13 is identified as the reference arrow marking and the destination of the other vehicle V's traveling trajectory 509 is determined to be "right turn." Then, suppose that a duplicated traveling trajectory 706 is generated by translating traveling trajectory 509 to the position of arrow marking Y14. In this case, the traveling trajectory duplicating unit 135 adjusts the turning radius of duplicated traveling trajectory 706 by the amount of translation of traveling trajectory 509 to the position of arrow marking Y14, thereby reducing the curvature of traveling trajectory 509 and increasing the radius of traveling trajectory 509.

By increasing or decreasing the turning radius of the driving trajectory by the amount of movement during duplication, it is possible to duplicate a driving trajectory that does not intersect with adjacent lanes.

As described above, an embodiment of the present invention has been described. However, the descriptions and drawings that form part of this disclosure should not be understood as limiting this invention. Various alternative embodiments, examples, and operating techniques will become apparent to those skilled in the art from this disclosure.

1 Travel path estimation device 501-509 Travel trajectory 701-706 Duplicate travel trajectory D1-D3 Distance Y, Y1-Y14 Road arrow marking (arrow marking)
V, V1-V2 vehicles (other vehicles)

Claims (9)

A method for estimating a path within an intersection, comprising: acquiring data indicating the position and direction of each of a plurality of road arrow markings in front of the intersection;
acquiring data indicating a travel path of a vehicle traveling through the intersection;
Calculating a distance in a lane width direction from the travel path to each of the plurality of road arrow markings;
determining a destination of the travel path based on a direction of a reference road arrow marking, which is the road arrow marking whose distance is less than a first threshold value;
generating data indicating a duplicated travel locus obtained by translating the travel locus to a position of the road arrow marking that indicates the same direction as the destination of the travel locus and is different from the reference road arrow marking;
A travel path estimation method comprising estimating the shape of a travel path within the intersection based on data indicating the replicated travel path. 2. The lane estimating method according to claim 1, wherein the reference road arrow marking is a road arrow marking, among the plurality of road arrow markings, for which the distance is less than the first threshold and is the shortest. A reference line segment passing through a center position in a lane width direction is set for each of the plurality of road arrow markings;
3. The lane estimating method according to claim 1, wherein, when it is possible to draw a perpendicular line from the driving trajectory to the reference line segment, the distance in the lane width direction between the driving trajectory and the reference line segment at which the perpendicular line is shortest is calculated as the distance in the lane width direction from the driving trajectory to the road arrow marking. A reference line segment passing through a center position in a lane width direction is set for each of the plurality of road arrow markings;
If it is impossible to draw a perpendicular line from the travel path to the reference line segment, the travel path is extended toward the reference line segment;
The lane width direction distance between the extended driving trajectory and the reference line segment, where the perpendicular line from the extended driving trajectory to the reference line segment is the shortest, is calculated as the lane width direction distance from the driving trajectory to the road arrow marking. Calculating a first angle between the extended travel path and the reference line segment;
The lane estimation method according to claim 4, characterized in that, when the first angle is less than a second threshold, the destination of the driving trajectory is determined based on the direction of the reference road arrow marking, which is the road arrow marking whose distance is less than the first threshold. If there are a plurality of directions indicated by the reference road arrow markings, a curvature of the travel path is calculated;
3. The method for estimating a travel path according to claim 1, wherein a destination of the travel path is determined based on the curvature. 3. The method for estimating a travel path according to claim 1, further comprising adjusting a turning radius of the replicated travel path by an amount corresponding to a translational movement of the travel path. A reference line segment passing through a center position in a lane width direction is set for each of the plurality of road arrow markings;
If it is impossible to draw a perpendicular line from the travel path to the reference line segment, the travel path is translated in the lane width direction;
determining whether the translated travel path touches another travel path;
When the translated travel trajectory is in contact with the other travel trajectory, a second angle formed between the translated travel trajectory and the other travel trajectory is calculated;
A path estimation method as described in either 1 or 2, characterized in that if the second angle is less than a third threshold, the driving trajectory is translated in parallel to a position where it is tangent to the other driving trajectory, and data indicating the duplicated driving trajectory is generated by duplicating the driving trajectory. A travel path estimation device including a control unit that estimates a travel path within an intersection,
The control unit
acquire data indicating the position and direction of each of the plurality of road arrow markings before the intersection;
acquiring data indicating a travel path of a vehicle traveling through the intersection;
Calculating a distance in a lane width direction from the travel path to each of the plurality of road arrow markings;
determining a destination of the travel path based on a direction of a reference road arrow marking, which is the road arrow marking whose distance is less than a first threshold value;
generating data indicating a duplicated travel locus obtained by translating the travel locus to a position of the road arrow marking that indicates the same direction as the destination of the travel locus and is different from the reference road arrow marking;
A travel path estimation device that estimates the shape of a travel path within the intersection based on data indicating the replicated travel path.