Walking estimation method and walking estimation device

Abstract

To accurately estimate a shape of a traveling road on which an own vehicle travels in a range including a front area of an intersection where the number of traffic lanes increases or decreases.SOLUTION: A traveling road estimating method estimates a traveling road on which an own vehicle travels, by using map data 2 in which a shape of each road comprising a plurality of traveling roads on which one or a plurality of vehicles can travel side by side is represented by an arrangement of a plurality of nodes. The traveling road estimating method includes: extracting a shape sR2 of a second road along which the vehicles travel in a direction facing the own vehicle along a first road R1 from the map data 2, the second road corresponding to a first predetermined range A including a front area of an intersection of the first road R1 on which the own vehicle is scheduled to travel; and estimating a shape of a traveling road on which the own vehicle travels in the first predetermined range A of the first road R1 based on the shape sR2 of the second road.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to a traveling path estimation method and a traveling path estimation device.

Background Art

[0002] Conventionally, a vehicle lighting device that estimates the shape of a traveling path on which a host vehicle travels using map data (SD-MAP: Standard Map) has been known (Patent Document 1). The vehicle lighting device described in Patent Document 1 acquires position data of a plurality of nodes located before and after the current position of the host vehicle from the map data, and connects the plurality of nodes by interpolation processing to estimate the shape of the traveling path on which the host vehicle travels. The vehicle lighting device controls the irradiation direction of the headlamp of the host vehicle according to the estimated shape of the traveling path on which the host vehicle travels.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, the positions of the nodes registered in the map data are set at the center in the width direction of the road, and the node positions change in the width direction of the road in an area including the vicinity of an intersection where the number of lanes increases or decreases. Therefore, in a range including the vicinity of an intersection of a road, the shape of the road represented by the map data may be different from the actual shape of the traveling path. Therefore, in the traveling path estimation method of connecting a plurality of nodes disclosed in Patent Document 1 by interpolation processing, it may not be possible to accurately estimate the shape of the traveling path in a range including the vicinity of an intersection where the number of lanes increases or decreases.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide a road shape estimation method and a road shape estimation device that can accurately estimate the shape of a road on which a host vehicle travels in a range including the vicinity of an intersection where the number of lanes increases or decreases.

Means for Solving the Problems

[0006] A road shape estimation method according to an aspect of the present invention is a road shape estimation method for estimating a road on which a host vehicle travels, using map data in which the shape of each road consisting of a plurality of roads on which one or a plurality of vehicles can travel side by side is represented by an arrangement of a plurality of nodes. The road shape estimation method extracts, from the map data, the shape of a second road on which traffic proceeds in a direction facing the host vehicle along the first road, corresponding to a first predetermined range including the vicinity of an intersection of a first road on which the host vehicle is scheduled to travel, and estimates the shape of the road on which the host vehicle travels in the first predetermined range of the first road based on the shape of the second road.

Effects of the Invention

[0007] According to the present invention, it is possible to accurately estimate the shape of a road on which a host vehicle travels in a range including the vicinity of an intersection where the number of lanes increases or decreases.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4A

Figure 4B

Figure 5A

Figure 5B

Figure 5C

Figure 6

Mode for Carrying Out the Invention

[0009] With reference to the drawings, embodiments will be described. In the description of the drawings, the same reference numerals are given to the same parts and the description thereof is omitted.

[0010] [Configuration of Travel Path Estimation Device] With reference to FIG. 1, the configuration of the travel path estimation device according to the embodiment will be described. The travel path estimation device is composed of a controller 1 and is mounted on a vehicle (own vehicle). The travel path estimation device extracts the shape of a second road on which traffic travels in a direction facing the own vehicle along the first road, corresponding to a first predetermined range including the area in front of the intersection of the first road on which the own vehicle is scheduled to travel, from the map data 2, and estimates the shape of the travel path on which the own vehicle travels in the first predetermined range of the first road based on the shape of the second road. The first road includes one travel path or a plurality of travel paths on which a plurality of vehicles can travel side by side, but does not include one travel path or a plurality of travel paths on which traffic travels in a direction facing the own vehicle along these plurality of travel paths, that is, the oncoming lanes as seen from the own vehicle. One or a plurality of oncoming lanes as seen from the own vehicle constitute a second road different from the first road. In other words, the "road" in the embodiment is composed of one travel path or a plurality of travel paths on which a plurality of vehicles can travel side by side as a unit.

[0011] The map data 2 (SD-MAP) used by the travel path estimation device according to the embodiment does not include information on a per-lane basis of a road, and each road consisting of one travel path or a plurality of travel paths on which a plurality of vehicles can travel side by side is represented as a single line by a plurality of nodes and links connecting between the plurality of nodes. Therefore, the map data 2 represents each of the first road and the second road as a single line by a plurality of nodes and links. The map data 2 is mainly used for a car navigation system. A node represents a feature point of a road, such as an intersection, a connection point with another road, a turning point, or a dead end. In the map data 2, information related to the node, such as a node number, an absolute position, an elevation, the number of connecting links, a connecting node number, and an intersection name, is stored in association with each node. A link connects between nodes and indicates the shape of the road section. In the map data 2, information related to the link, such as a link number (the numbers of the start and end nodes), the absolute position at each point of the link, a road type, and a route number, is stored in association with each link. The travel path estimation device may acquire the map data 2 from a car navigation system mounted on the host vehicle, or may acquire the map data 2 from a data server via communication.

[0012] The controller 1 is a general-purpose microcomputer including a CPU (Central Processing Unit), a memory (storage unit) such as a RAM and a ROM, and an input / output unit. A computer program for functioning as a travel path estimation device is installed in the microcomputer. By executing the computer program, the microcomputer functions as a plurality of information processing circuits (10, 11, 12, 13, 14) included in the travel path estimation device. In this embodiment, an example is shown in which a plurality of information processing circuits (10, 11, 12, 13, 14) included in the travel path estimation device are realized by software, but it is also possible to prepare dedicated hardware for executing each information process to configure the information processing circuit. Further, the plurality of information processing circuits may be configured by individual hardware. The controller 1 includes, as a plurality of information processing circuits, a host vehicle position estimation unit 10, a travel path detection unit 11, a node selection unit 12 (road shape extraction unit), a node correction unit 13, and a travel path estimation unit 14.

[0013] The host vehicle position estimation unit 10 estimates the position of the host vehicle on the map data. The host vehicle position estimation unit 10 acquires a GPS signal from a receiver that receives the GPS (Global Positioning System) signal mounted on the vehicle, and measures the position (absolute position) and attitude (absolute attitude) of the host vehicle in the earth coordinates. Further, the host vehicle position estimation unit 10 may perform odometry and dead reckoning. Specifically, the host vehicle position estimation unit 10 measures the relative position, relative attitude, and speed of the host vehicle with respect to a predetermined reference point based on the information acquired from a wheel speed sensor that detects the wheel speed of each wheel of the vehicle and a steering angle sensor that detects the steering angle of the steering wheel. Thereby, even when the reception state of the GPS signal deteriorates, the host vehicle position estimation unit 10 can continuously measure the absolute position and absolute attitude of the host vehicle from the measured relative position and relative attitude of the host vehicle with respect to the predetermined reference point. The host vehicle position estimation unit 10 estimates the position of the host vehicle on the map data from the measured absolute position and absolute attitude of the host vehicle.

[0014] The lane detection unit 11 detects the lane dividing line of the lane on which the host vehicle is traveling, and acquires the absolute position of the lane dividing line of the lane on which the detected host vehicle is traveling. The lane detection unit 11 continuously acquires, at a predetermined cycle, information obtained by a plurality of different types of object detection sensors that detect objects around the host vehicle, such as a camera and a LiDAR (Light Detection and Ranging) mounted on the vehicle. The object detection sensors continuously detect the road on which the host vehicle is traveling and the objects around the road at a predetermined cycle. More specifically, the object detection sensors can detect moving objects including motorcycles and vehicles, and stationary objects such as lane dividing lines, road markings, road signs, curbs, guardrails, walls, and parked vehicles on or around the road. For example, the object detection sensors can detect the position (relative position), attitude, size, speed, acceleration, deceleration, and yaw rate of a moving or stationary object with respect to the vehicle. The lane detection unit 11 extracts information regarding the lane dividing line from the information regarding the road on which the host vehicle is traveling and the objects around the road detected by the plurality of object detection sensors, and determines the lane dividing line closest to the host vehicle as the lane dividing line of the lane on which the host vehicle is traveling. The lane detection unit 11 acquires the position (relative position) of the lane dividing line of the lane on which the host vehicle is traveling from the information regarding the road on which the host vehicle is traveling and the objects around the road. Then, the lane detection unit 11 acquires the absolute position of the host vehicle from the host vehicle position estimation unit 10, and converts the position of the lane dividing line of the lane on which the host vehicle is traveling from the relative position to the absolute position.

[0015] The node selection unit 12 acquires the position of the host vehicle on the map data from the host vehicle position estimation unit, and identifies the first road on which the host vehicle is scheduled to travel. Specifically, the node selection unit 12 identifies, as the first road on which the host vehicle is traveling, the road passing through the position of the host vehicle on the map data or the road closest to the position of the host vehicle on the map data. The node selection unit 12 can more accurately estimate the first road on which the host vehicle is traveling from the position of the host vehicle on the map data by using map matching technology. The node selection unit 12 estimates, as the first road on which the host vehicle is scheduled to travel, the road that is connected in line with the first road on which the host vehicle is traveling. When the planned travel route of the host vehicle is set in the car navigation system, the node selection unit 12 may identify the road on the planned travel route as the first road on which the host vehicle is scheduled to travel. Alternatively, in the case of an autonomous vehicle that presets the planned travel route of the host vehicle and controls the host vehicle to travel along the preset route, the preset planned travel route may be identified as the first road. The node selection unit 12 sets a first predetermined range including the area before the intersection of the first road on which the host vehicle is scheduled to travel, and selects the nodes existing in the first predetermined range of the first road. The node selection unit 12, for example, acquires a node sequence existing in the range from the position of the host vehicle on the first road to 2 km ahead from the map data 2. The range of the node sequence of the first road to be acquired can be appropriately set in advance according to the road type, road shape, etc. The node selection unit 12 acquires information regarding each node included in the node sequence acquired from the map data 2, and identifies the node located at the intersection of the first road from among the node sequence based on the information regarding the acquired node. Thereby, the node selection unit 12 can set the first predetermined range including the area before the intersection of the first road on which the host vehicle is scheduled to travel, and can select the nodes of the first road existing in the first predetermined range. Note that the method for identifying the node indicating the intersection is not limited to this. For example, by acquiring the node sequence of the road intersecting with the first road in addition to the node sequence of the first road from the map data 2, the position where the node sequences intersect can be identified as the intersection. Details regarding the process in which the node selection unit 12 sets the first predetermined range will be described later with reference to FIG. 2.

[0016] The node selection unit 12 extracts the shape of a second road on which traffic travels in a direction facing the host vehicle (i.e., the opposite direction) along a first road corresponding to a first road on which the host vehicle is scheduled to travel, from the map data 2, and calculates the degree of difference in a first predetermined range between the shape of the extracted first road and the shape of the second road. Specifically, the node selection unit 12 extracts a node sequence of the second road corresponding to the first predetermined range of the first road from the map data 2, and calculates the degree of difference between the node sequence of the first predetermined range of the first road and the second node sequence corresponding to the first predetermined range. The node sequence of the second road corresponding to the first predetermined range of the first road is the node sequence of the second road included in the moved first predetermined range when the first predetermined range is moved parallel to the second road so as to include the second road. Here, the node sequence means an array of a plurality of nodes existing in the first predetermined range, or an array of a plurality of nodes and line segments connecting between the plurality of nodes with curves or straight lines. That is, the node sequence is an array of a plurality of nodes, that is, represents the shape of a road on a map, and it suffices to include an array of a plurality of nodes and is not limited to only an array of a plurality of nodes.

[0017] When calculating the degree of difference in the shape of the road (i.e., the node sequence) within the first predetermined range, the node selection unit 12 first sets a second predetermined range including a first road and a second road on the front side in the traveling direction of the host vehicle from the first predetermined range, and calculates a second similarity between the shape of the first road and the shape of the second road within the second predetermined range. The node selection unit 12 determines whether the calculated second similarity is equal to or greater than a second predetermined value. The second predetermined range is, for example, from the intersection exit existing on the front side in the traveling direction of the host vehicle from the first predetermined range to before the next intersection. The node selection unit 12 moves the node sequence of the second road existing within the second predetermined range parallel to the width direction of the road toward the node sequence of the first road so as to match the node sequence of the first road in the road width direction, and calculates the similarity of the second predetermined range. The node selection unit 12 moves parallel to the width direction of the road, for example, so that the position of the node located at the center of the node sequence of the second predetermined range of the second road matches the node located at the center of the node sequence of the second predetermined range of the first road. The second predetermined value may be set, for example, by performing a simulation in advance to calculate the similarity when the shape of the first road and the shape of the second road are actually similar for each road shape, and based on the calculated similarity for each road shape.

[0018] Next, the node selection unit 12 sets a third predetermined range including the first road and the second road on the rear side of the traveling direction of the host vehicle from the first predetermined range, and calculates a third similarity between the shape of the first road and the shape of the second road in the third predetermined range. The node selection unit 12 determines whether the calculated third similarity is equal to or greater than a third predetermined value. The third predetermined range is, for example, a range from in front of an intersection existing on the rear side of the traveling direction of the host vehicle from the first predetermined range to the exit of an intersection existing further on the rear side of the traveling direction of the host vehicle. The node selection unit 12 moves the node sequence of the second road existing in the third predetermined range parallel to the width direction of the road toward the node sequence of the first road so as to match the node sequence of the first road, and calculates the similarity of the third predetermined range. The node selection unit 12 moves, for example, the node located at the center of the node sequence of the third predetermined range of the second road so as to match the node located at the center of the node sequence of the third predetermined range of the first road. The third predetermined value may be set, for example, by performing a simulation in advance, calculating the similarity when the shape of the first road and the shape of the second road are actually similar for each road shape, and based on the calculated similarity for each road shape.

[0019] The second similarity and the third similarity are calculated by the following formula (1). N is the number of nodes of the first road or the second road within the second predetermined range or the third predetermined range. In the case where the number of nodes of the first road and the number of nodes of the second road are different within the second predetermined range or the third predetermined range, N may be the number of nodes of the road with fewer nodes. di is the distance between each node of the first road and the corresponding node of the second road when the node located at the center of the node sequence of the second road is made to coincide with the node located at the center of the node sequence of the first road within the second predetermined range or the third predetermined range. B is a value equal to half of the width (lane width) of the road surface of the first road on which the host vehicle is scheduled to travel within the second predetermined range or the third predetermined range. The node selection unit 12 can estimate the lane width of the road by acquiring the road type associated with the link connecting the nodes. The lane width of the road is defined for each road type according to the Road Structure Order. Therefore, the node selection unit 12 can estimate the lane width of the road by acquiring the lane width for each road type stored in advance in the memory of the controller 1.

[0020]

Equation

[0021] When the node selection unit 12 determines that the similarity of the second predetermined range is equal to or greater than the second predetermined value and the similarity of the third predetermined range is equal to or greater than the third predetermined value, the dissimilarity of the first predetermined range is calculated. That is, the node selection unit 12 selects the nodes existing in the first predetermined range as the objects to be corrected. Note that the condition for calculating the dissimilarity of the first predetermined range is not limited to this. For example, when it is determined that the similarity of the second predetermined range is equal to or greater than the second predetermined value, or when it is determined that the similarity of the third predetermined range is equal to or greater than the third predetermined value, the dissimilarity of the first predetermined range may be calculated.

[0022] The node selection unit 12 sets a local range including a first road smaller than a first predetermined range, moves the local range in the traveling direction of the host vehicle at a position overlapping with the first predetermined range, and calculates a first similarity between the shape of the first road in each moved local range and the shape of a second road corresponding to each local range. Note that the shape (node sequence) of the second road corresponding to each moved local range is the shape (node sequence) of the second road included in the moved local range when the local range is moved parallel to the second road so as to include the second road. The first similarity is calculated by the above formula (1).

[0023] The node selection unit 12 matches the shape of the second road corresponding to the local range with the highest first similarity to the shape of the first road in the local range, and calculates the degree of difference in the first predetermined range. The degree of difference in the first predetermined range is calculated by the following formula (2). N is the number of nodes of the first road or the second road in the first predetermined range. Note that when the number of nodes of the first road and the number of nodes of the second road are different in the first predetermined range, N may be the number of nodes of the road with fewer nodes. di is the distance between each node of the first road in the first predetermined range and the node of the second road corresponding to each node of the first road when the shape of the second road corresponding to the local range with the highest first similarity is matched to the shape of the first road in the local range. When the node selection unit 12 determines that the degree of difference in the first predetermined range is equal to or greater than a first predetermined value, the node selection unit 12 selects the nodes existing in the first predetermined range of the first road as the objects to be corrected. In the present embodiment, the degree of difference for all the nodes existing in the first predetermined range is calculated, but the degree of difference may be calculated for each node, and it may be determined whether to select each node as an object to be corrected. The first predetermined value may be set, for example, by performing a simulation in advance to calculate the degree of difference for each road shape when the number of lanes increases or decreases in a predetermined range including in front of the intersection of the first road, and based on the calculated degree of difference for each road shape. Thereby, when the number of lanes increases or decreases in the predetermined range of the first road, the degree of difference in the first predetermined range can be set to be equal to or greater than the first predetermined value.

[0024]

Number

[0025] When it is determined that the degree of difference within the first predetermined range is equal to or greater than the first predetermined value, the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range of the first road. When the range up to before the intersection is set as the first predetermined range, the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range of the first road by the following method. The node correction unit 13 moves the node sequence corresponding to the first predetermined range in the direction of the node sequence of the second road so that the positions in the lateral direction (road width direction) with respect to the traveling direction of the host vehicle of the node located behind the traveling direction of the host vehicle in the first predetermined range of the first road match the nodes of the second road corresponding to the node located behind. Then, the node correction unit 13 corrects the shape of the node sequence existing in the first predetermined range of the first road to a shape along the shape of the second road corresponding to the first predetermined range by moving the positions of the plurality of nodes existing in the first predetermined range of the first road. Alternatively, the node correction unit 13 sets a composition ratio that increases as it approaches before the intersection for the plurality of nodes of the second road corresponding to the first predetermined range of the first road, and corrects the positions of the plurality of nodes existing in the first predetermined range of the first road to positions along the shape of the second road corresponding to the first predetermined range based on the set composition ratio. The node correction unit 13 sets a composition ratio that increases as it approaches the intersection exit not only for the plurality of nodes existing in the first predetermined range before the intersection of the first road, but also when the first predetermined range including within the intersection and up to the intersection exit is set. Further, the node correction unit 13 sets a composition ratio that decreases as it approaches the intersection exit for the plurality of nodes corresponding to the first predetermined range of the second road. Then, the node correction unit 13 synthesizes the shape of the first road and the shape of the second road based on the respective composition ratios set for the nodes of the first road and the nodes of the second road, and corrects the positions of the plurality of nodes existing in the first predetermined range of the first road to the positions where the shape of the first road and the shape of the second road are synthesized. Here, the composition ratio means the movement ratio of the plurality of nodes existing in the first predetermined range of the first road to the plurality of nodes of the second road corresponding to the first predetermined range of the first road.That is, for each of the plurality of nodes existing in the first predetermined range of the first road, the lateral (road width direction) distance from the nodes of the second road corresponding to the first predetermined range of the first road is calculated, and each node existing in the first predetermined range of the first road is moved by the distance obtained by multiplying the calculated distance by the synthesis ratio, thereby correcting the shape of the node sequence existing in the first predetermined range of the first road.

[0026] When the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range of the first road, the corrected node sequence (road shape represented by the node sequence) is set as the first road shape. Alternatively, when the node correction unit 13 does not correct the positions of the nodes existing in the first predetermined range of the first road, the node sequence (road shape represented by the node sequence) extracted from the map data 2 is set as the shape of the first road. Therefore, the shape of the first road extracted from the map data 2 is set as it is.

[0027] Based on the lane dividing line on which the host vehicle detected by the lane detection unit 11 is traveling and the shape of the set first road, the travel path estimation unit 14 estimates the shape of the travel path of the host vehicle. Specifically, the travel path estimation unit 14 identifies the shape of the detected lane dividing line, combines the identified shape of the lane dividing line with the shape of the set first road, and extends the lane boundary line along the shape of the first road, thereby estimating the shape of the travel path of the host vehicle.

[0028] Next, with reference to FIG. 2, the process in which the node selection unit 12 sets a first predetermined range will be schematically described. FIG. 2 shows a first road R1 along which the host vehicle is scheduled to travel and a second road R2 along which traffic travels in a direction facing the host vehicle along the first road R1. In FIG. 2, the current position of the host vehicle is indicated by P, and the traveling direction of the host vehicle is the x direction. The traveling direction of the vehicles on the second road R2 is the direction opposite to the x direction. FIG. 2 shows an intersection C2 existing on the front side in the traveling direction x of the host vehicle and an intersection C1 existing at the current position P of the host vehicle, and a first predetermined range A including the vicinity of the intersection C2 of the first road R1 set by the node selection unit 12 is shown. The predetermined range A may also include the intersection C2.

[0029] The node selection unit 12 calculates the difference between the shapes sR1 and sR2 of the first road and the second road for each node at the intersection (between C1 and C2) extracted from the map data 2. The node selection unit 12 repeatedly calculates the difference while moving the node at the end on the C1 side of the intersection within the range for which the difference is calculated toward the C2 side of the intersection. As the node at the end on the C1 side of the intersection moves, the node selection unit 12 sets the range from the position of the node n7R1, where the difference for each node changes from decreasing to increasing in the traveling direction x of the host vehicle, to before intersection C1 or the exit of intersection C1 as the first predetermined range A. The second road R2 has an increasing number of lanes toward before intersection C1 in the traveling direction of the second road R2 (the direction opposite to the x direction). Therefore, the difference for each node decreases from intersection C1 in the traveling direction x of the host vehicle. However, the number of lanes of the first road R1 increases toward before intersection C2 where the lanes of the first road R1 exist in the traveling direction x of the host vehicle. Therefore, the position of the node of the first road R1 changes from the position of the node n7R1 toward intersection C2. That is, a difference occurs between the shape sR1 of the first road and the shape sR2 of the second road. Therefore, the difference between the node n7R1 of the first road R1 and the node n7R2 of the second road corresponding to the node n7R1 changes from decreasing to increasing in the traveling direction x of the host vehicle. Accordingly, the node selection unit 12 selects the node n7R1 where the difference for each node changes from decreasing to increasing in the traveling direction x of the host vehicle, and sets the range from the position of the node n7R1 to before intersection C2 or the exit of intersection C2 as the first predetermined range A.

[0030] Next, with reference to FIG. 3, the second predetermined range C and the third predetermined range D set by the node selection unit 12 will be described. The second predetermined range C is set to the range from the exit of intersection C2, which is on the front side in the traveling direction x of the host vehicle from the first predetermined range A, to before the next intersection. The third predetermined range D is set to the range from before intersection C1, which is on the rear side in the traveling direction x of the host vehicle from the first predetermined range A, to the exit of intersection C0, which is further on the rear side in the traveling direction x of the host vehicle.

[0031] Next, with reference to FIGS. 4A and 4B, a process in which the node selection unit 12 calculates the similarity of the local range B and the difference of the first predetermined range A will be schematically described. FIG. 4A shows the first predetermined range A and the local range B. The node selection unit 12 moves the local range B including the first road R1 smaller than the first predetermined range A in the traveling direction x of the host vehicle at a position overlapping the first predetermined range A. The node selection unit 12 moves the local range B in units of nodes. The node selection unit 12 calculates a first similarity between the shape sR1 of the first road in each moved local range B and the shape sR2 of the second road corresponding to each local range B. As shown in FIG. 4A, the node selection unit 12 sets the local range B at the position where the first similarity is the highest. The node selection unit 12 parallel-translates the shape sR2 of the second road corresponding to the local range B with the highest first similarity in the road width direction toward the shape sR1 of the first road to match the shape sR1 of the first road in the local range B, and calculates the difference in the first predetermined range A. The shape sR2 of the second road shown in the first road R1 is the shape sR2 of the second road corresponding to the local range B with the highest first similarity made to match the shape sR1 of the first road in the local range B. FIG. 4B is a diagram showing each node of the first road R1 and the second road R2 superimposed in FIG. 4A, and is a diagram for explaining a process in which the node selection unit 12 calculates the difference of the first predetermined range A. The node selection unit 12 superimposes the nodes (n8R1, n8R2) on the rearmost side in the traveling direction of the local range B, and calculates the difference between the nodes (n1R1 to n7R1) of the first road in the first predetermined range A and the nodes (n1R2 to n7R2) of the second road corresponding to the first road. In FIG. 4B, the nodes n1R2 to n7R2 of the second road respectively correspond to the nodes n1R1 to n7R1 existing in the first predetermined range A of the first road. The node selection unit 12 can calculate the difference in the first predetermined range A by calculating the distance between the corresponding nodes (d1 in the case of node n1R1 and node n1R2).

[0032] Next, with reference to FIGS. 5A to 5C, a process in which the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range A of the first road will be schematically described.

[0033] Referring to FIG. 5A, when the area before the intersection is set as the first predetermined range, an embodiment of the process in which the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range A of the first road will be described. The node correction unit 13 moves the node sequence of the second road corresponding to the first predetermined range A so that the lateral positions with respect to the traveling direction x of the host vehicle of the node n1R1 located behind the traveling direction x of the host vehicle in the first predetermined range A of the first road and the node n1R2 of the second road corresponding to the node n1R1 coincide. Then, the node correction unit 13 corrects the positions of the plurality of nodes nR1 existing in the first predetermined range A of the first road to positions ns along the shape sR2 of the second road corresponding to the first predetermined range A after the positions of the plurality of nodes nR1 are moved. Thereby, the shape of the node sequence formed by the plurality of nodes nR1 existing in the first predetermined range A of the first road is corrected to a shape along the node sequence formed by the plurality of nodes n1R2 of the corresponding second road.

[0034] Referring to FIG. 5B, when the area before the intersection is set as the first predetermined range, a different embodiment from the embodiment shown in FIG. 5A of the process in which the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range A of the first road will be described. The node correction unit 13 sets a composite rate cR2 that increases as it approaches the intersection (the traveling direction x of the host vehicle) for a plurality of nodes of the second road corresponding to the first predetermined range A of the first road. Then, the node correction unit 13 corrects the positions of the plurality of nodes nR1 existing in the first predetermined range A of the first road to positions ns along the shape sR2 of the second road corresponding to the first predetermined range A based on the composite rate cR2. Thereby, the shape of the node sequence formed by the plurality of nodes nR1 existing in the first predetermined range A of the first road is corrected to approach the node sequence formed by the plurality of nodes n1R2 of the corresponding second road. That is, for the distance in the road width direction between each node nR1 existing in the first predetermined range A of the first road and the node nR2 of the corresponding second road, each node nR1 existing in the first predetermined range A of the first road is moved in the direction of the node nR2 of the second road by a distance obtained by multiplying the distance by the composite rate cR2, and the positions of the nodes nR1 existing in the first predetermined range A of the first road are corrected. Here, it can be said that the embodiment shown in FIG. 5A is an embodiment in which the composite rate cR2 is 100% regardless of the distance from the intersection.

[0035] Referring to FIG. 5C, when a first predetermined range A is set that includes not only in front of the intersection but also within the intersection and up to the intersection exit, the node correction unit 13 corrects the positions of the nodes nR1 existing in the first predetermined range A of the first road. The node correction unit 13 sets a composite rate cR1 that increases as it approaches the intersection exit (the traveling direction x of the own vehicle) for a plurality of nodes nR1 existing in the first predetermined range A of the first road. Further, the node correction unit 13 sets a composite rate cR2 that decreases as it approaches the intersection exit (the traveling direction x of the own vehicle) for a plurality of nodes nR2 corresponding to the first predetermined range A of the second road. Then, the node correction unit 13 synthesizes the shape sR1 of the first road and the shape sR2 of the second road based on the respective composite rates set for the nodes nR1 of the first road and the nodes nR2 of the second road. The node correction unit 13 corrects the positions of the plurality of nodes nR1 existing in the first predetermined range A of the first road to the position ss where the shape sR1 of the first road and the shape sR2 of the second road are synthesized. Thereby, the shape of the node sequence formed by the plurality of nodes nR1 existing in the first predetermined range A of the first road is corrected to approach the node sequence formed by the plurality of nodes nR2 of the corresponding second road. Note that the content of the arithmetic processing in the position correction of the plurality of nodes nR1 using the composite rate is the same as the example shown in FIG. 5B.

[0036] [Lane Estimation Method] Next, with reference to FIG. 6, an example of the processing of the lane estimation device shown in FIG. 1 will be described. The operation of the lane estimation device shown in the flowchart of FIG. 6 starts simultaneously when the ignition switch or power switch of the vehicle is turned ON, and the processing ends when the ignition switch or power switch is turned OFF.

[0037] In step S10, the host vehicle position estimation unit 10 estimates the position of the host vehicle on the map data based on the acquired GPS signal. Proceed to step S20, where the lane detection unit 11 continuously acquires, at a predetermined period, information obtained by a plurality of different types of object detection sensors that detect objects around the host vehicle, such as a camera and a lidar mounted on the vehicle, and acquires the absolute position of the lane dividing line of the lane on which the host vehicle is traveling. Proceed to step S30, where the node selection unit 12 estimates the first road on which the host vehicle is scheduled to travel and acquires the node sequence of the first road on which the host vehicle is scheduled to travel from the map data. The node selection unit 12 identifies the road passing through the position of the host vehicle on the map data or the road closest to the position of the host vehicle on the map data as the first road on which the host vehicle is traveling, and estimates the road connected in line with the identified first road as the first road on which the host vehicle is scheduled to travel.

[0038] Proceed to step S40, where the node selection unit 12 acquires information about each node of the node sequence of the first road from the map data. The node selection unit 12 identifies the node located at the intersection of the first road from among the node sequence based on the information about each node, and sets a first predetermined range including the area before the intersection of the first road on which the host vehicle is scheduled to travel. Proceed to step S50, where the node selection unit 12 selects the nodes of the first road existing within the first predetermined range. Proceed to step S60, where the node selection unit 12 extracts the shape of the second road on which traffic is flowing in the direction facing the host vehicle along the first road corresponding to the first road from the map data.

[0039] Proceed to step S70. The node selection unit 12 sets a second predetermined range including a first road and a second road on the front side of the traveling direction of the host vehicle from the first predetermined range, and calculates a second similarity between the shape of the first road and the shape of the second road in the second predetermined range. Proceed to step S80. The node selection unit 12 determines whether the second similarity is equal to or greater than a second predetermined value. If the node selection unit 12 determines that the second similarity is equal to or greater than the second predetermined value (YES in step S80), the process proceeds to step S90. If the node selection unit 12 determines that the second similarity is less than the second predetermined value (NO in step S80), the process ends.

[0040] In step S90, the node selection unit 12 sets a third predetermined range including a first road and a second road on the rear side of the traveling direction of the host vehicle from the first predetermined range, and calculates a third similarity between the shape of the first road and the shape of the second road in the third predetermined range. Proceed to step S100. The node selection unit 12 determines whether the third similarity is equal to or greater than a third predetermined value. If the node selection unit 12 determines that the third similarity is equal to or greater than the third predetermined value (YES in step S100), the process proceeds to step S110. If the node selection unit 12 determines that the third similarity is less than the third predetermined value (NO in step S100), the process ends.

[0041] In step S110, the node selection unit 12 calculates the degree of difference in the first predetermined range. Specifically, the node selection unit 12 sets a local range including a first road smaller than the first predetermined range, and while moving the local range in the traveling direction of the host vehicle at a position overlapping the first predetermined range, calculates a first similarity between the shape of the first road in each moved local range and the shape of the second road corresponding to each local range. The node selection unit 12 makes the shape of the second road corresponding to the local range with the highest first similarity coincide with the shape of the first road in the local range (moves the shape of the second road in the road width direction and overlaps it with the shape of the first road), and calculates the degree of difference in the first predetermined range.

[0042] Proceed to step S120, and the node selection unit 12 determines whether the dissimilarity within the first predetermined range is equal to or greater than the first predetermined value. If the node selection unit 12 determines that the dissimilarity within the first predetermined range is equal to or greater than the first predetermined value (YES in step S120), the node selection unit 12 selects the nodes existing in the first predetermined range of the first road as the objects to be corrected, and the process proceeds to step S130. If the node selection unit 12 determines that the dissimilarity within the first predetermined range is less than the first predetermined value (NO in step S120), the node selection unit 12 does not select the nodes existing in the first predetermined range as the objects to be corrected, and the process proceeds to step S140.

[0043] In step S130, the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range of the first road. When the range up to before the intersection is set as the first predetermined range, the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range of the first road by the following method. The node correction unit 13 moves the node sequence of the second road corresponding to the first predetermined range so that the lateral positions of the nodes located behind the traveling direction of the host vehicle within the first predetermined range of the first road and the nodes of the second road corresponding to the nodes located behind match. Then, the node correction unit 13 corrects the positions of the plurality of nodes existing in the first predetermined range of the first road to the positions along the shape of the second road corresponding to the first predetermined range after the movement. Alternatively, the node correction unit 13 sets a synthesis rate that increases as it approaches the intersection for the plurality of nodes of the second road corresponding to the first predetermined range of the first road, and corrects the positions of the plurality of nodes existing in the first predetermined range of the first road so as to approach the shape of the second road corresponding to the first predetermined range based on the set synthesis rate.

[0044] In step S130, when not only in front of the intersection of the first road but also a first predetermined range including within the intersection and up to the intersection exit is set, the node correction unit 13 sets a composition ratio that becomes higher as it goes toward the intersection exit for a plurality of nodes existing in the first predetermined range of the first road. Further, the node correction unit 13 sets a composition ratio that becomes lower as it goes toward the intersection exit for a plurality of nodes corresponding to the first predetermined range of the second road. Then, the node correction unit 13 synthesizes the shape of the first road and the shape of the second road based on the respective composition ratios set for the nodes of the first road and the nodes of the second road, and corrects the positions of the plurality of nodes existing in the first predetermined range of the first road at the position where the shape of the first road and the shape of the second road are synthesized.

[0045] Proceed to step S140. When the node correction unit 13 corrects the positions of the nodes existing in the first predetermined range of the first road, the node correction unit 13 sets the corrected node sequence as the shape of the first road. Or, when the node correction unit 13 has not corrected the positions of the nodes existing in the first predetermined range of the first road, the node correction unit 13 sets the node sequence extracted from the map data 2 as the shape of the first road.

[0046] In step S150, based on the lane division line on which the host vehicle detected by the lane detection unit 11 is traveling and the shape of the set first road, the lane estimation unit 14 estimates the shape of the lane on which the host vehicle is traveling. Specifically, the lane estimation unit 14 identifies the shape of the detected lane division line, synthesizes the identified shape of the lane division line and the shape of the set first road, and extends the lane boundary line along the shape of the first road, thereby estimating the shape of the lane on which the host vehicle is traveling.

[0047] [Operational Effects] As described above, according to this embodiment, the following operational effects can be obtained.

[0048] The travel path estimation device extracts, from map data, the shape of a second road on which traffic proceeds in a direction facing the host vehicle along a first road, the second road corresponding to a first predetermined range including the area in front of an intersection on the first road on which the host vehicle is scheduled to travel, and estimates the shape of the travel path on which the host vehicle travels in the first predetermined range of the first road based on the shape of the second road. The positions of the nodes registered in the map data are often set at the center positions in the vehicle width direction of a road composed of a plurality of travel paths on which one or more vehicles can travel side by side. Therefore, the shape of the road in front of the intersection on the first road on which the host vehicle is scheduled to travel, represented by the arrangement of a plurality of nodes registered in the map data, may not conform to the shape of the actual travel path due to the addition of right / left turn lanes or the reduction of lanes. On the other hand, in the second road on which traffic proceeds in a direction facing the host vehicle along the first road, the position corresponding to the area in front of the intersection on the first road becomes the intersection exit. Therefore, there are no right / left turn lanes at the intersection exit of the second road, and the shape of the second road corresponding to the area in front of the intersection on the first road registered in the map data is likely to conform to the shape of the travel path on the actual first road. Therefore, the travel path estimation device can estimate the shape of the travel path on which the host vehicle travels in the first predetermined range including the area in front of the intersection on the first road based on the shape of the second road, and can estimate the shape of the travel path on which the host vehicle travels with higher accuracy.

[0049] The travel path estimation device estimates the shape of the travel path on which the host vehicle travels based on the travel path dividing line on which the host vehicle is traveling and the shape of the first road. Thereby, based on the shape of the travel path on which the host vehicle is actually traveling and the shape of the road on which the host vehicle travels extracted from the map data, the shape of the travel path on which the host vehicle travels can be estimated, and the shape of the travel path on which the host vehicle travels can be estimated with higher accuracy.

[0050] The travel path estimation device calculates the degree of difference in a first predetermined range between the shape of a first road extracted from map data and the shape of a second road. When the degree of difference in the first predetermined range is equal to or greater than a first predetermined value, the travel path estimation device selects, as a target for correction, the nodes existing in the first predetermined range of the first road. Thereby, the travel path estimation device can determine whether there is an increase or decrease in the number of lanes on the first road based on the degree of difference in the first predetermined range between the shape of the first road extracted from the map data and the shape of the second road. Then, when the degree of difference is equal to or greater than the first predetermined value, the travel path estimation device can determine that there is an increase or decrease in the number of lanes on the first road in the first predetermined range and select, as a target for correction, the nodes existing in the first predetermined range. That is, the travel path estimation device can appropriately estimate the shape of the travel path on which the host vehicle travels in the first predetermined range of the first road as needed.

[0051] The travel path estimation device moves a local range including a first road smaller than the first predetermined range in the traveling direction of the host vehicle at a position overlapping the first predetermined range, and calculates a first similarity between the shape of the first road in each moved local range and the shape of the second road corresponding to each local range. Thereby, the travel path estimation device can identify the local range having the highest first similarity among the respective local ranges overlapping the first predetermined range. Then, the travel path estimation device moves the shape of the second road in the road width direction so that the shape of the second road corresponding to the local range having the highest first similarity matches the shape of the first road in the local range, and calculates the degree of difference in the first predetermined range by overlapping it with the shape of the first road. Thereby, it is possible to overlap the location where the shape of the first road and the shape of the second road match most in each local range overlapping the first predetermined range, and it is possible to calculate the degree of difference in the first predetermined range with higher accuracy.

[0052] The travel path estimation device calculates the difference degree for each node between the shape of a first road and the shape of a second road between intersections extracted from map data, and sets, as a first predetermined range, the distance from the position of the node where the difference degree for each node changes from decreasing to increasing in the traveling direction of the host vehicle to before the intersection or to the intersection exit. Thereby, the travel path estimation device can set the first predetermined range so as not to include a range where the difference degree on the front side in the traveling direction of the host vehicle on the first road is high. That is, when the second road is before the intersection and the shape of the second road registered in the map data differs from the actual shape of the second road due to an increase or decrease in the number of lanes of the second road, the range where they differ can be excluded. Therefore, the travel path estimation device can estimate the travel path on which the host vehicle travels in the first predetermined range of the first road using only a range where there is a high possibility that the shape of the second road registered in the map data is the same as the actual shape of the second road. Thus, the travel path estimation device can estimate the shape of the travel path on which the host vehicle travels with higher accuracy.

[0053] The travel path estimation device determines whether the shape of the first road and the shape of the second road in a second predetermined range on the front side in the traveling direction of the host vehicle from the first predetermined range are similar. When the similarity degree between the shape of the first road and the shape of the second road in the second predetermined range is equal to or greater than a second predetermined value, the travel path estimation device can determine that the shape of the first road and the shape of the second road are overall similar and that there is a high possibility that only the first predetermined range is different. Therefore, the travel path estimation device can select the nodes existing in the first predetermined range as objects to be corrected after confirming that the shape of the first road and the shape of the second road are overall in agreement, and can estimate the travel path on which the host vehicle travels with higher accuracy.

[0054] The lane estimation device determines whether the shape of the first road and the shape of the second road in a third predetermined range in front of the traveling direction of the host vehicle from a first predetermined range are similar. When the similarity between the shape of the first road and the shape of the second road in the third predetermined range is equal to or greater than a third predetermined value, the lane estimation device can determine that the shape of the first road and the shape of the second road are generally similar, and it is highly likely that only the first predetermined range is different. Therefore, after confirming that the shape of the first road and the shape of the second road are generally the same, the lane estimation device can select the nodes existing in the first predetermined range as the objects to be corrected, and can estimate the lane on which the host vehicle travels with higher accuracy.

[0055] When the lane estimation device sets the range up to before the intersection as the first predetermined range, the second road corresponding to the front of the intersection in the first predetermined range of the first road becomes the intersection exit. Therefore, the second road corresponding to the first predetermined range does not include the front of the intersection. In other words, it can be determined that the shape of the second road corresponding to the first predetermined range of the first road extracted from the map data is highly likely to match the actual shape of the second road. Therefore, the lane estimation device can estimate the shape of the lane on which the host vehicle travels in the first predetermined range of the first road by correcting the positions of a plurality of nodes existing in the first predetermined range of the first road to the positions along the shape of the corresponding second road. Thus, the lane estimation device can estimate the shape of the lane on which the host vehicle travels with higher accuracy.

[0056] When the walking path estimation device sets the area up to before the intersection within a first predetermined range, the second road corresponding to the area before the intersection in the first predetermined range of the first road becomes the intersection exit. Therefore, the second road corresponding to the first predetermined range does not include the area before the intersection. In other words, it can be determined that the shape of the second road corresponding to the first predetermined range of the first road extracted from the map data is likely to match the actual shape of the second road. Also, the width of the first road gradually changes at the area before the intersection in the first predetermined range as the number of lanes increases or decreases. Therefore, the shape of the first road in the first predetermined range extracted from the map data is likely to gradually deviate from the actual shape of the first road as the host vehicle proceeds in the traveling direction. Therefore, the driving support device sets a higher synthesis rate for the plurality of nodes of the second road corresponding to the first predetermined range of the first road as it approaches the area before the intersection, and corrects the positions of the plurality of nodes existing in the first predetermined range of the first road to positions along the shape of the second road corresponding to the first predetermined range based on the set synthesis rate. As a result, the walking path estimation device can gradually approximate the shape of the walking path on which the host vehicle travels to the shape of the second road, and can prevent the estimated walking path from changing abruptly.

[0057] When the walking estimation device sets the area up to the intersection exit of the first road within a first predetermined range, the second road corresponding to the intersection exit within the first predetermined range of the first road is in front of the intersection. Therefore, within the first predetermined range, it can be determined that the shapes of the first road and the second road extracted from the map data deviate from the actual road shape due to the increase or decrease of lanes from in front of the intersection to the center of the intersection for each road. For this reason, the walking estimation device sets a higher synthesis rate for a plurality of nodes existing within the first predetermined range of the first road, such that the rate increases as it approaches the intersection exit, and sets a lower synthesis rate for a plurality of nodes corresponding to the first predetermined range of the second road, such that the rate decreases as it approaches the intersection exit of the first road. Then, the walking estimation device synthesizes the shape of the first road and the shape of the second road based on the respective synthesis rates set for the nodes of the first road and the nodes of the second road, and corrects the positions of the plurality of nodes existing within the predetermined range of the first road to the position where the shape of the first road and the shape of the second road are synthesized. Therefore, at the intersection approach of the first road, the walking estimation device sets a high synthesis rate for the nodes of the second road corresponding to the intersection approach of the first road, and sets a low synthesis rate for the nodes of the first road. Thereby, the walking estimation device can correct the positions of the nodes at the intersection approach of the first road to match the shape of the second road, which is highly likely to match the actual road shape. Also, at the intersection exit of the first road, the walking estimation device sets a low synthesis rate for the nodes of the second road corresponding to the intersection exit of the first road, and sets a high synthesis rate for the nodes of the first road. Thereby, the walking estimation device can match the positions of the nodes at the intersection exit of the first road to the shape of the first road, which is highly likely to match the actual road shape. Therefore, the walking estimation device can more accurately estimate the driving path of the own vehicle up to the intersection exit of the first road.

Explanation of Signs

[0058] 1 Controller 2 Map Data 10 Own Vehicle Position Estimation Unit 11 Travel Path Detection Unit 12 Node Selection Unit 13 Node Correction Unit 14 Path Estimation Unit A First Predetermined Range B Local Range C Second Predetermined Range D Third Predetermined Range R1 First Road R2 Second Road cR1 Composite Ratio of the First Road cR2 Composite Ratio of the Second Road nR1 Nodes of the First Road nR2 Nodes of the Second Road sR1 Shape of the First Road sR2 Shape of the Second Road

Claims

1. A lane estimation method for estimating a lane on which a host vehicle travels, using map data in which the shape of each of a plurality of roads on which one or more vehicles can travel side by side is represented by an arrangement of a plurality of nodes, extracting, from the map data, the shape of a second road on which traffic travels in a direction facing the host vehicle along the first road, corresponding to a first predetermined range including a point before an intersection of the first road on which the host vehicle is scheduled to travel, estimating the shape of the lane on which the host vehicle travels in the first predetermined range of the first road based on the shape of the second road Lane estimation method.

2. estimating the shape of the lane on which the host vehicle travels in the first predetermined range of the first road by correcting the shape of the lane on which the host vehicle travels in the first predetermined range of the first road extracted from the map data based on the shape of the second road The lane estimation method according to claim 1.

3. detecting a lane dividing line of the lane on which the host vehicle is traveling, estimating the shape of the lane on which the host vehicle travels based on the detected lane dividing line and the shape of the first road extracted from the map data The lane estimation method according to claim 1.

4. calculating a degree of difference between the shape of the first road and the shape of the second road in the first predetermined range extracted from the map data, when the degree of difference in the first predetermined range is equal to or greater than a first predetermined value, selecting nodes existing in the first predetermined range of the first road as objects to be corrected The lane estimation method according to claim 1.

5. moving a local range including the first road smaller than the first predetermined range in the traveling direction of the host vehicle at a position overlapping the first predetermined range, and calculating a first similarity between the shape of the first road in each of the moved local ranges and the shape of the second road corresponding to each of the local ranges, calculating the degree of difference in the first predetermined range by overlapping the shape of the second road with the shape of the first road by moving the shape of the second road in the road width direction so as to match the shape of the second road corresponding to the local range having the highest first similarity with the shape of the first road in the local range The lane estimation method according to claim 4.

6. calculating the degree of difference for each node between the shape of the first road and the shape of the second road between intersections extracted from the map data Set the position from the node where the difference degree for each node changes from decreasing to increasing in the traveling direction of the host vehicle to before the intersection or the intersection exit as the first predetermined range. The lane estimation method according to claim 1.

7. Set a second predetermined range including the first road and the second road on the front side of the traveling direction of the host vehicle with respect to the first predetermined range. When the second similarity between the shape of the first road and the shape of the second road in the second predetermined range is equal to or greater than a second predetermined value, select the nodes existing in the first predetermined range of the first road as objects to be corrected. The lane estimation method according to claim 1.

8. Set a third predetermined range including the first road and the second road on the rear side of the traveling direction of the host vehicle with respect to the first predetermined range. When the third similarity between the shape of the first road and the shape of the second road in the third predetermined range is equal to or greater than a third predetermined value, select the nodes existing in the first predetermined range of the first road as objects to be corrected. The lane estimation method according to claim 1.

9. When setting the range up to before the intersection as the first predetermined range, correct the positions of a plurality of nodes existing in the first predetermined range of the first road to positions along the shape of the second road corresponding to the first predetermined range. The lane estimation method according to claim 1.

10. When setting the range up to before the intersection as the first predetermined range, set a synthesis rate that becomes higher as it goes toward before the intersection for a plurality of nodes in the first predetermined range of the first road corresponding to the first predetermined range, Based on the synthesis rate, correct the positions of a plurality of nodes existing in the first predetermined range of the first road so as to approach the shape of the second road corresponding to the first predetermined range. The lane estimation method according to claim 1.

11. When setting the range up to the intersection exit of the first road as the first predetermined range, set a synthesis rate that becomes higher as it goes toward the intersection exit for a plurality of nodes existing in the first predetermined range of the first road, and set a synthesis rate that becomes lower as it goes toward the intersection exit for a plurality of nodes corresponding to the first predetermined range of the second road. Based on the respective synthesis rates set for the nodes of the first road and the nodes of the second road, synthesize the shape of the first road and the shape of the second road. Modify the positions of a plurality of nodes existing in the first predetermined range of the first road at a position where the shape of the first road and the shape of the second road are combined. The lane estimation method according to claim 1.

12. A lane estimation device that estimates a lane on which a host vehicle travels using map data in which the shape of each road consisting of a plurality of lanes on which one or more vehicles can travel side by side is represented by an array of a plurality of nodes, A road shape extraction unit that extracts, from the map data, the shape of a second road on which traffic travels in a direction facing the host vehicle along the first road, corresponding to a first predetermined range including a position in front of an intersection of the first road on which the host vehicle is scheduled to travel; A lane estimation unit that estimates the shape of a lane on which the host vehicle travels in the first predetermined range of the first road based on the shape of the second road. Lane estimation device.

Images