Travel assistance device, travel route assistance method, and computer program

Abstract

To accurately specify a traveling road end of a traveling road crossing a traveling road on which an own vehicle is traveling.SOLUTION: The driving assistance device 100 includes a first road edge estimating unit 11 that estimates running road edges that are right and left road edges of a running road by using a detection result of a sensor 200 that is mounted on the own vehicle and detects surrounding environments, a first structure estimating unit 13 that estimates positions of first regular structures that are structures provided along the running road by using the detection result, and a second road edge estimating unit 12 that estimates crossing road edges that are right and left road edges of a crossing road that crosses the running road by using the positions of the first regular structures. V0.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a driving assistance device, a driving assistance method, and a computer program. [Background technology]

[0002] A driving assistance device has been proposed that assists a vehicle in driving through a parking lot. The driving assistance device in Patent Document 1 identifies a plurality of parking stall lines lined up in a row in an image captured by a camera, and estimates the straight lines connecting the ends (end points) of the parking stall lines as driving stall lines, i.e., the left and right edges of the lane on which the vehicle is currently traveling. Furthermore, when the vehicle is traveling toward a T-junction and there are parking stall lines lined up in a direction perpendicular to the driving direction at the end of the road, the driving assistance device estimates the driving stall lines of the lane ahead after turning right or left at the T-junction in the same manner as described above. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-66934 Summary of the Invention [Problem to be solved by the invention]

[0004] Generally, lane markings on the inside of a lane when turning right or left are difficult to capture in a captured image. Therefore, the driving assistance device of Patent Document 1 can estimate lane markings on the outside of a lane where the vehicle has turned right or left, but cannot accurately estimate lane markings on the inside of a lane. This poses a problem, for example, when determining a planned driving route after a right or left turn in an automated driving system, where an appropriate route may not be determined. For these reasons, a technology is desired that can accurately estimate the lane edge of a lane that intersects with the lane on which the vehicle is traveling. [Means for solving the problem]

[0005] One aspect of the present disclosure provides a driving assistance device (100) that is mounted on a host vehicle (V0) and assists the host vehicle in driving within a parking lot (Pa). The driving assistance device includes a first road edge estimation unit (11) that estimates in-motion road edges, which are left and right road edges of a current road on which the host vehicle is traveling, using a detection result from a sensor (200) that is mounted on the host vehicle and detects the surrounding environment of the host vehicle, a first structure estimation unit (13) that estimates the position of a first regular structure, which is a structure that is provided in a regular manner along the current road edges, using the detection result, and a second road edge estimation unit (12) that estimates crossing road edges, which are left and right road edges of an intersecting road that is a road that intersects with the current road, using the estimated position of the first regular structure.

[0006] According to this type of driving assistance device, the position of a first regular structure located along the road edge while driving is estimated using the detection results of the sensor, and the estimated position of the first regular structure is used to estimate the intersecting road edges, which are the left and right road edges of the intersecting roads.Therefore, the road edge on the inner side of the intersecting road edges when turning right or left can be accurately estimated. [Brief explanation of the drawings]

[0007] [Figure 1] 1 is a block diagram showing a schematic configuration of a driving assistance device according to an embodiment of the present disclosure. [Figure 2] FIG. 2 is a diagram showing an example of a parking route set in a parking lot. [Figure 3] 10 is a flowchart showing a procedure for processing to estimate an intersecting road edge. [Figure 4] 10 is a flowchart showing the detailed procedure of S10. [Figure 5] FIG. 2 is a diagram showing an example of a first regular structure and a road edge during travel. [Figure 6] FIG. 2 is a diagram showing an example of a first regular structure and a road edge during travel. [Figure 7] 10 is a flowchart showing the detailed procedure of S20. [Figure 8]FIG. 10 is an explanatory diagram showing an example of the determination in S125. [Figure 9] FIG. 10 is an explanatory diagram showing an example of the determination in S125. [Figure 10] FIG. 10 is a diagram showing an example of a first regular structure and a road edge during travel in the second embodiment. [Figure 11] FIG. 10 is an explanatory diagram showing an example of a method for estimating the end of an inside intersecting roadway end in another embodiment. [Figure 12] FIG. 10 is an explanatory diagram showing an example of a method for estimating the end of an inside intersecting roadway end in another embodiment. [Figure 13] FIG. 10 is an explanatory diagram showing an example of a method for estimating the end of an inside intersecting roadway end in another embodiment. [Figure 14] FIG. 10 is a diagram illustrating an example of the configuration of a parking lot according to another embodiment. [Figure 15] FIG. 10 is a diagram illustrating an example of the configuration of a parking lot according to another embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] A. First embodiment: A1.Device configuration: A driving assistance device 100 of the first embodiment shown in FIG. 1 is mounted on a vehicle V0 and assists the vehicle V0 in driving in a parking lot. In this embodiment, the vehicle V0 is configured as a vehicle capable of autonomous driving. In this embodiment, the above-mentioned "driving assistance" means performing a process of recognizing the road on which the vehicle V0 is driving (hereinafter referred to as the "current road") and roads that intersect with the current road (hereinafter referred to as the "intersecting roads"), a process of determining a route for parking (hereinafter referred to as the "parking route") using such recognition results, and a process of automatically driving the vehicle V0 along the route for parking. The vehicle V0 may be configured as any type of vehicle, such as an internal combustion engine vehicle, a hybrid electric vehicle (HEV), a plug-in hybrid electric vehicle (PHEV), an electric vehicle (EV), or a fuel cell electric vehicle (FCV, FCHV).

[0009] In addition to the driving assistance device 100, the vehicle V0 is equipped with a sensor 200, a vehicle control device 400, a drive device 410, a steering device 420, and a braking device 430.

[0010] The sensor 200 detects the surrounding environment of the vehicle V0. The sensor 200 transmits the detection results to the driving assistance device 100. In this embodiment, the sensor 200 includes a forward camera 210, a millimeter-wave radar 220, and an ultrasonic sensor 230. The forward camera 210 captures an image of the area ahead of the vehicle V0. The millimeter-wave radar 220 and the ultrasonic sensor 230 emit millimeter waves or ultrasonic waves and receive reflected waves from objects present around the vehicle V0, thereby detecting the position of the object and the distance to the object. Note that the "object" detected by the millimeter-wave radar 220 is a collection of multiple detection points (targets).

[0011] The vehicle control device 400 causes the vehicle V0 to travel along a travel route set by the travel assistance device 100 based on the current travel path and intersecting travel paths recognized by the travel assistance device 100. The travel route includes set target coordinates to be passed through and a vehicle speed when passing through the target coordinates. The vehicle control device 400 controls the drive device 410, the steering device 420, and the braking device 430 so that the vehicle passes through each target coordinate at the target vehicle speed.

[0012] The drive device 410 is a group of devices for driving the vehicle V0. The drive device 410 includes devices that generate drive force, such as an engine and a motor generator, various actuators for driving the engine and the motor generator, and an ECU (Electronic Control Unit) for controlling these actuators. The steering device 420 is a group of devices for steering the vehicle V0. The steering device 420 includes a steering wheel, a steering angle sensor, an assist hydraulic device, an actuator for generating hydraulic pressure, and an ECU for controlling steering. The braking device 430 is a group of devices for generating braking force in the vehicle V0. The braking device 430 includes a disc rotor, brake pads, a brake hydraulic device, an actuator for generating hydraulic pressure, and an ECU for controlling braking. The drive device 410, the steering device 420, and the braking device 430 all communicate with the vehicle control device 400. At this time, the vehicle control device 400 transmits control signals to these devices 410 to 430, and on the other hand, receives signals indicating the operating state of each device from the devices 410 to 430. The above-mentioned "signals indicating the operating state" include signals indicating the vehicle speed of the vehicle V0, whether or not the brakes are operated, whether or not the turn signals are operated, the amount of wheel rotation, the amount of steering, etc.

[0013] In this embodiment, the driving assistance device 100 is configured as an ECU in which a CPU 10, a ROM 20, and a RAM 30 are capable of communicating with each other via an internal bus 90. The ROM 20 includes a non-volatile memory, such as an EEPROM, and stores a control program in advance. The CPU 10 reads the control program stored in the ROM 20, loads it into the RAM 30, and executes it, thereby functioning as a first road edge estimation unit 11, a second road edge estimation unit 12, a first structure estimation unit 13, and a second structure estimation unit 14.

[0014] The first road edge estimation unit 11 estimates the left and right edges of the lane while the vehicle is traveling (hereinafter also referred to as "road edges while the vehicle is traveling") using the detection results of the sensor 200. The second road edge estimation unit 12 estimates the left and right edges of the intersecting lane (hereinafter also referred to as "intersecting road edges"). The second road edge estimation unit 12 estimates the intersecting road edges using the position of a first regular structure. The above-mentioned "first regular structure" refers to a structure that is provided in a regular pattern along the road edge while the vehicle is traveling. Examples of first regular structures include parking spaces painted on the road surface, vehicles parked in parking spaces, pillars, and walls. The first structure estimation unit 13 estimates the position of the first regular structure using the detection results of the sensor 200. The second structure estimation unit 14 estimates the position of the second regular structure using the detection results of the sensor 200. A "second regular structure" is a structure that faces the lane while the vehicle is in motion and is provided in a regular pattern along either the left or right edge of the intersecting lane.

[0015] In the example of FIG. 2, roads R1 and R2 are provided within a parking lot Pa, intersecting with each other. Further, multiple parking stalls PA1-PA7, PA11-PA17, and PA21-PA27 are provided along these roads R1 and R2. Vehicles V1-V7, V13, V14, V16, V17, V23, V24, V26, and V27 are already parked in the parking stalls PA1-PA7, PA13, PA14, PA16, PA17, PA23, PA24, PA26, and PA27. Vehicle V0 (hereinafter also referred to as "host vehicle V0") is traveling on road R1. Therefore, road R1 corresponds to the current traveling lane. Furthermore, road R2 intersects with road R1 and corresponds to an intersecting lane. Furthermore, the left and right ends of road R1, which is the current traveling lane, correspond to current traveling lane edges e1L and e1R. The left and right ends of road R2, which is an intersecting lane, correspond to intersecting lane edges e2L and e2R. Intersecting lane edge e2R corresponds to the outer wheel side when the host vehicle V0 turns to turn left, in other words, the intersecting lane edge on the outer side of the turning direction (hereinafter referred to as the "outer intersecting lane edge"). On the other hand, intersecting lane edge e2L corresponds to the inner wheel side when the host vehicle V0 turns to turn left, in other words, the intersecting lane edge on the inner side of the turning direction (hereinafter referred to as the "inner intersecting lane edge"). Furthermore, parking stalls PA1-PA7 and vehicles V1-V7, as well as parking stalls PA11-PA17 and vehicles V13, V14, V16, and V17, correspond to first regular structures. Furthermore, parking stalls PA21-PA27 and vehicles V23, V24, V26, and V27 correspond to second regular structures.

[0016] The driving assistance device 100 sets a parking path as described above. At this time, the driving assistance device 100 uses the detection results of the sensor 200 to set, for example, the parking path MP1 shown in FIG. 2. The parking path MP1 is a path that goes straight on road R1, turns left at the intersection of two roads R1 and R2, and then goes straight on road R2. Here, there are no obstacles between the position of the host vehicle V0 shown in FIG. 2 and the parking stalls PA23-PA25. Therefore, by detecting these parking stalls PA23-PA25, it is easy to identify the intersecting lane edge e2R, which is the outer intersecting lane edge along these parking stalls PA23-PA25. On the other hand, for example, a vehicle V7 exists as an obstacle between the position of the host vehicle V0 and the intersecting lane edge e2L, which is the inner intersecting lane edge. However, the driving assistance device 100 can accurately estimate the inner intersecting lane edge (intersecting lane edge e2L) by performing the intersecting lane edge estimation process described below, and can set an appropriate route, such as passing through the center of road R2, as the parking route MP1.

[0017] A2. Intersection road edge estimation processing: The intersecting road edge estimation process shown in FIG. 3 is repeatedly executed when the driving assistance device 100 is powered on. The intersecting road edge estimation process includes steps S10 and S20. Hereinafter, "step S" will be simply referred to as "S." In S10, the driving assistance device 100 estimates the position of the first regular structure using the detection result of the sensor 200. In S20, the driving assistance device 100 estimates the intersecting road edge using the position of the first regular structure. The detailed configurations of S10 and S20 will be described below.

[0018] As shown in FIG. 4, S10 includes S105 to S130. In S105, the first structure estimation unit 13 recognizes a parked vehicle using the detection result of the sensor 200. In S110, the first structure estimation unit 13 recognizes a parking space using the detection result of the sensor 200. A "parking space" refers to a frame indicating a partitioned area for parking a vehicle. Specifically, this refers to a rectangle or straight line drawn with white or yellow lines on the road surface of a parking lot. In S105 and S110, the first structure estimation unit 13 may recognize the parked vehicle and the parking space by, for example, performing semantic segmentation using a convolutional neural network on the captured image obtained by the front camera 210. Alternatively, the first structure estimation unit 13 may recognize the parked vehicle and the parking space by, for example, extracting feature points such as edges from the captured image obtained by the front camera 210 or point cloud information and performing pattern matching based on the feature points. Note that the parking space may be recognized as a section defined by the detected white lines by detecting white lines. At this time, the recognized space may be compared with the dimensions of a general parking space to determine the difference, and if the difference is within a predetermined size range, the space may be recognized as a parking space.

[0019] In S115, the first structure-estimating unit 13 recognizes structures other than parked vehicles and parking spaces using the detection results of the sensor 200. In the example of FIG. 2, no structures other than parked vehicles and parking spaces are detected on road R1. However, for example, in the example of FIG. 5, in addition to six parking spaces PA31, PA32, PA33, PA34, PA35, and PA36, three pillars PO1, PO2, and PO3 are provided along road R3, which is the running path. Therefore, in this case, the three pillars PO1, PO2, and PO3 are recognized in S115. Note that in FIG. 5, the squares of each parking space PA31 to PA36 and the small white circles attached to the corners of each pillar PO1 to PO3 facing road R3 indicate detected feature points (corners). Also, for example, in the example of FIG. 6, a wall W1 is provided along road R4, which is the running path. Therefore, in this case, the wall W1 is recognized in S115. In Fig. 6, a group of small white circles on the surface of wall W1 facing road R4 indicates the detected feature point group pg1. Although omitted in Fig. 2, feature points are also detected in each parking space provided along road R1 in Fig. 2 and each vehicle parked in such parking space. The above-mentioned steps S105 to S115 can be executed in any order. Furthermore, at least some of steps S105 to S115 may be executed simultaneously (in parallel).

[0020] As shown in FIG. 4, in S120, the first road edge estimation unit 11 estimates the road edge during travel using the recognition results from S105 to S115. Specifically, first, the corners of each structure recognized in S105 to S115 that are located on the road (road R1) during travel are connected along the direction of travel. For example, in the example of FIG. 2, for the road edge during travel on the left side of road R1, the two corners of each parking stall PA1 to PA7 on the road R1 side and the two corners of each vehicle V1 to V7 on the road R1 side are connected. In this case, whether or not to connect the corners may be determined based on the distance between the points along the width direction of road R1 (hereinafter also referred to as the "road width direction") or the degree of difference (angle) between the direction of the line segment connecting two points and the direction of the line segment connecting other two points. Then, the line segment obtained by extending the line segment connecting the two points closest to the road in the width direction is estimated as the road edge during travel. In the example of Figure 2, road edges e1L and e1R are estimated. In the example of Figure 5, the line segment obtained by extending the line segment connecting the corners of pillars PO1 to PO3 closest to road R3 is estimated as road edge e3L during travel. Note that in the example of Figure 5, the right-hand road edge during travel is omitted. In the example of Figure 6, the line segment obtained by connecting feature point group pg1 is estimated as road edge e4L during travel. Note that in the example of Figure 6, the right-hand road edge during travel is also omitted.

[0021] As shown in FIG. 4, in S125, the first structure-estimating unit 13 determines whether there is a regularity in the components constituting the road edge during travel estimated in S120. "Regularity" means that the components are arranged in a regular pattern. In other words, it means that there is a regularity in the arrangement pattern. For example, in the example of FIG. 2, the components constituting the road edge during travel e1L are parking stalls PA1-PA7 and vehicles V1-V7. These components (structures) are arranged at approximately equal intervals along road R1. Therefore, the first structure-estimating unit 13 determines that there is a regularity in the arrangement of the parking stalls PA1-PA7 and vehicles V1-V7. Also, for example, in the example of FIG. 5, the components constituting the road edge during travel e3L are three pillars PO1-PO3. These pillars PO1-PO3 are positioned approximately equally in the road width direction, and the distance between the pillars along road R3 (the traveling direction of the host vehicle V0) is approximately constant. Therefore, the first structure-estimating unit 13 determines that the three pillars PO1 to PO3 have a regularity. Also, in the example of Fig. 6, for example, the wall W1 is arranged continuously and without interruption along the road R1. Therefore, the first structure-estimating unit 13 determines that the wall W1 has a regularity.

[0022] 4, if it is determined that there is regularity in the components of the road edge during travel (S125: YES), the first structure estimation unit 13 estimates the position of the component as the position of the first regular structure (S130).If it is determined that there is no regularity in the components of the road edge during travel (S125: NO), or after the completion of the above-mentioned S130, the processing proceeds to S20.

[0023] As shown in Figure 7, S20 includes S205 to S235. In S205, the second road edge estimating unit 12 determines whether the position of the first regular structure has been estimated in S10. If the above-mentioned S130 has been executed, the second road edge estimating unit 12 determines that the position of the first regular structure has been estimated. On the other hand, if S130 has not been executed, the second road edge estimating unit 12 determines that the position of the first regular structure has not been estimated.

[0024] If it is determined that the position of the first regular structure has been estimated (S205: YES), the second road edge estimation unit 12 determines whether there is a discontinuity in the first regular structure in S210. The "discontinuity in the first regular structure" means a discontinuity in the direction along the road during travel, in other words, in the traveling direction of the host vehicle V0.

[0025] If it is determined that there is an end to the first regular structure (S210: YES), the second road edge estimation unit 12 determines in S215 whether the foremost first regular structure is located at a distance equal to or greater than a predetermined first threshold distance from the detection range limit of the sensor 200.

[0026] In the example of FIG. 8, the host vehicle V0 is traveling on road R5, which is a traveling lane. In S10, the positions of five parking stalls PA41, PA42, PA43, PA44, and PA45 are estimated as the positions of the first regular structures. In this example, the foremost first regular structure is parking stall PA45. The first regular structures end at this parking stall PA45. Here, the parking stall PA45 is located a distance d1 before (on the side closer to the host vehicle V0) the detection range limit SL1 of the sensor 200. This distance d1 is relatively short and smaller than the first threshold distance. In this configuration, a parking stall PA46 also exists in front of the parking stall PA45. However, since this parking stall PA46 is located farther than the detection range limit SL1, it is not detected as a first regular structure.

[0027] FIG. 9 shows an enlarged view of the vicinity of the parking stall PA7 in FIG. 2. As described above, the positions of the parking stalls PA1 to PA7 and the vehicles V1 to V7 are identified as the positions of the first regular structures. These first regular structures end at the parking stall PA7, which is the foremost first regular structure. The parking stall PA7 is located a distance d2 before the detection range limit SL1 (on the side closer to the host vehicle V0). This distance d2 is relatively large and is greater than the first threshold distance. Therefore, there is no structure that continues to the parking stall PA7 at a position farther away than the detection range limit SL1.

[0028] In the example of FIG. 5, it is determined whether the pillar PO3 is located at least a first threshold distance in front of the detection range limit SL1. In the example of FIG. 6, it is determined whether the front end of the wall W1 is located at least a first threshold distance in front of the detection range limit SL1. S215 is executed to determine whether the end of the first regular structure occurs because the vehicle has reached the front end of a structure actually arranged in a regular pattern, or because the vehicle has reached the detection range limit. Therefore, the first threshold distance in S215 may be set in advance by determining the distance at which the first regular structure can be determined to end in various parking lots. For example, the first threshold distance may be set to a value greater than the width of the road in the parking lot.

[0029] As shown in Fig. 7, when it is determined that the foremost first regular structure is located at a predetermined first threshold distance or more before the detection range limit of the sensor 200 (S215: YES), the second road edge estimation unit 12 estimates the position of the foremost first regular structure as the end of the inner intersecting road edge (S220). For example, in the examples of Figs. 2 and 9, the corner ep1 of the parking stall PA7 is identified as the position of the parking stall PA7, and this corner ep1 is estimated as the end of the inner intersecting road edge (intersecting road edge e2L) of the road R2, which is the intersecting road. The "end of the intersecting road edge" refers to the end of the intersecting road edge (line) located at the intersection (intersection) between the road currently being traveled and the intersecting road.

[0030] As shown in FIG. 7, in S225, the second structure estimation unit 14 estimates the position of the second regular structure using the detection result of the sensor 200. When the host vehicle V0 is located at the position shown in FIG. 2, the second structure estimation unit 14 uses the sensor 200 to detect at least some of the parking stalls PA21-PA27 and the vehicles V23, V24, V26, and V27. At least some of the detected parking stalls PA21-PA27 and the vehicles V23, V24, V26, and V27 face the road R1 and are arranged along the road R2, which is an intersecting lane. In addition, at least some of the detected parking stalls PA21-PA27 and the vehicles V23, V24, V26, and V27 are arranged at approximately equal intervals along the road R2. Therefore, the positions of at least some of the parking stalls PA21 to PA27 and the vehicles V23, V24, V26, and V27 detected by the sensor 200 are estimated as the positions of the second regular structure.

[0031] As shown in Figure 7, in S230, the second road edge estimation unit 12 estimates the outer intersecting road edge using the position of the second regular structure estimated in S225. Specifically, the second road edge estimation unit 12 estimates the outer intersecting road edge using the position of the second regular structure estimated in S225 in a manner similar to the method used to estimate the road edge during travel in S120 described above. In the example of Figure 2, the outer intersecting road edge, intersecting road edge e2R, is estimated in S230.

[0032] As shown in FIG. 7, in S235, the second road edge estimation unit 12 estimates the inner intersecting road edge as a line extending from the end of the inner intersecting road edge estimated in S220 so as to be parallel to the outer intersecting road edge estimated in S230. In the example of FIG. 2, a line extending from corner ep1 so as to be parallel to intersecting road edge e2R is estimated as intersecting road edge e2L, which is the inner intersecting road edge. Completion of S235 completes S20, and as shown in FIG. 3, the process returns to S10. Note that the process returns to S10 if it is determined that the position of the first regular structure has not been estimated (S205: NO), if it is determined that there is no end to the first regular structure (S210: NO), or if it is determined that the foremost first regular structure is not located within the detection range limit of sensor 200 by a predetermined first threshold distance or more (S215: NO).

[0033] According to the driving assistance device 100 of the first embodiment described above, the position of the first regular structure located along the road edge e1L during driving is estimated using the detection results of the sensor 200, and the estimated position of the first regular structure is used to estimate the intersecting road edge e2L, which is the left and right road edges of the intersecting road (road R2), so that the inner intersecting road edge of the intersecting road edges can be estimated with high accuracy.

[0034] Furthermore, since the front end of the first regular structure is estimated to be the end of the intersecting road end e2L, which is the inside intersecting road end, the end of the inside intersecting road end (intersecting road end e2L) can be estimated with high accuracy.

[0035] Furthermore, when the end of the first regular structure is located closer to the sensor 200 than a predetermined first threshold distance from the detection range limit SL1, the end of the first regular structure is estimated to be the end of the inner intersecting road edge, thereby enabling accurate estimation of the end of the inner intersecting road edge (intersecting road edge e2L). When the end of the first regular structure is located further back than the first threshold distance from the detection range limit SL1 of the sensor 200, there is a possibility that a first regular structure exists beyond that end but cannot be detected because it is farther away than the detection range limit. In contrast, when the end of the first regular structure is located closer to the sensor 200 than the first threshold distance from the detection range limit SL1 of the sensor 200, there is a low possibility that a first regular structure exists beyond that end, and such end is likely to be the end of the inner intersecting road edge.

[0036] Furthermore, if the first regular structure is a wall W1, the end of the wall W1 is estimated to be the end of the inner intersecting lane end, so that the end of the inner intersecting lane end can be accurately estimated in an environment where a wall W1 exists along the lane (road R4) during travel.

[0037] Furthermore, when the first regular structure is a plurality of pillars PO1 to PO3 arranged in a predetermined direction, the end of the pillar PO3 that is the foremost among the plurality of pillars PO1 to PO3 in the traveling direction of the vehicle V0 is estimated to be the end of the inner intersecting lane edge, so that the end of the inner intersecting lane edge can be accurately estimated in an environment where a plurality of pillars PO1 to PO3 are arranged along the lane while the vehicle V0 is traveling.

[0038] Furthermore, when the first regular structure is at least one of a parking space and a parked vehicle, the position where the continuous parking space PA1 to PA7 and parked vehicles V1 to V7 end is estimated to be the end of the inner intersecting lane end (intersecting lane end e2L), so that in an environment where at least one of a parking space and a parked vehicle is provided along the lane during travel, the end of the inner intersecting lane end can be estimated with high accuracy.

[0039] Furthermore, the position of the second regular structure that faces the current road (road R1) and is arranged in a regular pattern along the outer intersecting road edge of the intersecting road (road R2) is estimated, and the intersecting road edge e2R, which is the outer intersecting road edge, is estimated using this position, and a line extending from the end of the estimated current road edge e1L so as to be parallel to the estimated outer intersecting road edge (intersecting road edge e2R) is estimated to be the inner intersecting road edge (intersecting road edge e2L), so the inner intersecting road edge can be identified with high accuracy. This is because the left and right road edges of an intersecting road are generally parallel to each other.

[0040] B. Second embodiment: The configuration of the driving support device 100 of the second embodiment is the same as that of the driving support device 100 of the first embodiment, so the same components are given the same reference numerals and detailed descriptions thereof will be omitted. Furthermore, the procedure for estimating an intersection road edge in the second embodiment is the same as that in the first embodiment, so the same steps are given the same reference numerals and detailed descriptions thereof will be omitted. The second embodiment differs from the first embodiment in the method for estimating a first regular structure in S125 and S130.

[0041] In the first embodiment, parked vehicles, parking spaces, and other structures (pillars, walls, etc.) that are arranged in a regular pattern along the road while the vehicle is traveling are all estimated as first regular structures. In the second embodiment, the width direction of the road and the orientation of the structures are also taken into consideration to estimate first regular structures and their positions.

[0042] For example, in the example of FIG. 10, vehicle V0 is traveling on road R6. In this case, road R6 corresponds to the lane during travel. Five structures OB1 to OB5 are located along road R6. The five structures OB1 to OB5 correspond to, for example, parking spaces, parked vehicles, and pillars. In the example of FIG. 10, the distances between adjacent structures along the lane during travel, in other words, along direction D1 parallel to the traveling direction of vehicle V0, vary. Even in this case, if the variation in the distances between structures along direction D1 is within a predetermined range, they are candidates for first regularity structures. On the other hand, if the variation in the distances between structures along direction D1 is outside the predetermined range, the five structures OB1 to OB5 are estimated not to be first regularity structures. Next, the variation in the position of each structure in lane width direction D2 is identified. Specifically, a tentative lane edge during travel is set based on the candidates for first regularity structures. The orientation of the tentative road edge during travel may be determined, for example, by identifying the two corners of each of the structures OB1-OB5 on the road R6 side, identifying the line segments connecting the two corners, and using the average of the line segment orientations as the orientation of the tentative road edge during travel. In this case, if there is a structure with a line segment whose orientation is significantly different from the orientation of the other line segments, the five structures OB1-OB5 are estimated not to be structures with first regularity. In the example of FIG. 10, the line segment connecting the two corners of the structure OB4 on the road R6 side is oriented significantly differently from the line segments obtained based on the other structures OB1-OB3 and OB5. In this case, the five structures OB1-OB5 are estimated not to be structures with first regularity. If the orientation of structure OB4 is the same as that of the other structures OB1-OB3, OB5, and current road edge e6L is set as the tentative current road edge, the distance to current road edge e6L along the road width direction D2 is identified for each of structures OB1-OB5. If all of these distances are equal to or less than a predetermined threshold distance, the five structures OB1-OB5 are estimated to be first regular structures. On the other hand, as shown in Figure 10, if the distance d5 between structure OB5 and current road edge e6L is greater than the threshold distance, the five structures OB1-OB5 are estimated not to be first regular structures.

[0043] The driving assistance device 100 of the second embodiment described above has the same effects as the driving assistance device 100 of the first embodiment. In addition, from the detection results of the sensor 200, a plurality of structures lined up along the traveling direction of the host vehicle V0 are identified, and among the plurality of structures, structures whose distance from the estimated traveling road edge e6L along the road width direction D2 (a direction perpendicular to the traveling direction of the host vehicle V0) is equal to or less than a predetermined second threshold distance are identified as first regularity structures. This makes it possible to prevent structures arranged without regularity from being erroneously identified as first regularity structures. This makes it possible to prevent erroneous estimation of intersecting road edges.

[0044] C. Other Embodiments: (C1) In each of the above embodiments, if it is determined in S215 that "the foremost first regular structure is closer than a predetermined first threshold distance from the detection range limit of the sensor 200" (S215: YES), the second road edge estimation unit 12 estimates the position of the foremost first regular structure as the end of the inner intersecting road edge, but the present disclosure is not limited to this.

[0045] In the example of FIG. 11, the host vehicle V0 is traveling on the road R7. Four parking stalls PA66, PA67, PA68, and PA69 are detected by the sensor 200 along the road R7, which is the traveling path. Here, the distance between the parking stall PA68 and the parking stall PA69 is significantly different from the distance between adjacent parking stalls in the three parking stalls PA66 to PA68 on the near side (the distance along the road R7). In this case, the three parking stalls PA66 to PA68 are determined to be arranged in a regular manner and can be identified as a first regular structure.

[0046] Here, among the three vehicle frames PA66 to PA68, which are the first regularity structures, the parking frame PA68 located at the front is located a distance d3 shorter than the first threshold distance from the detection range limit SL1 of the sensor 200. In this case, in the first embodiment, S220 is not executed, and the position of the parking frame PA68 is not estimated to be the end of the inner intersecting road edge. However, in other embodiments, since the parking frame PA69, which is a structure, is detected in front of the parking frame PA68 at the detection range limit SL1, it is highly likely that the parking frame PA68 is the end of the running road edge and the end of the inner-side intersecting road edge. Therefore, in this case, in other embodiments, the position of the parking frame PA68 is estimated to be the end of the inner intersecting road edge.

[0047] In the example of FIG. 12, the host vehicle V0 is traveling on road R8. The current position of the host vehicle V0 has passed by multiple groups consisting of first regularity structures and is about to pass by a new first regularity structure. Specifically, the host vehicle V0 has passed multiple first regularity structure groups including a first first regularity structure group OG1 and a second first regularity structure group OG2, and is about to approach a third first regularity structure group OG3. By passing by multiple first regularity structure groups, the host vehicle V0 can recognize that the length of each first regularity structure group along road R8 and the distance between adjacent first regularity structure groups are approximately the same. In other words, in the example of FIG. 12, the driving assistance device 100 can estimate that the multiple structures that make up the first regularity structure are arranged in a regular pattern along road R8, and that the first regularity structure groups are also arranged in a regular pattern along road R8. In such a case, even if the distance d4 between the corner ep2 of the forwardmost structure in the third first regular structure group OG3 and the detection range limit SL1 of the sensor 200 is smaller than the first threshold distance, the corner ep2 may be estimated to be the end of the intersecting road edge. In other words, if the length L3 of the third first regular structure group OG3 along the road R8 is approximately equal to the length L1 of the first first regular structure group OG1 and the length L3 of the second first regular structure group OG2, and the distance between the second first regular structure group OG2 and the third first regular structure group OG3 is approximately equal to the distance between the first first regular structure group OG1 and the second first regular structure group OG2, then even if the distance d4 between the corner ep2 of the forwardmost structure of the third first regular structure group OG3 and the detection range limit SL1 of the sensor 200 is smaller than the first threshold distance, the corner ep2 may be estimated to be the end of the intersecting road edge.

[0048] In the example of Fig. 13, the host vehicle V0 is traveling on road R9. Five parking spaces PA74, PA75, PA76, PA77, and PA78 are detected by sensor 200 along road R9, which is the road on which the host vehicle V0 is traveling. In addition, a wall W2 is detected at the end of road R9. Note that a group of small white circles on the surface of wall W2 indicates a detected group of feature points pg2.

[0049] Among the five parking stalls PA74 to PA78, which are the first regularity structures, the parking stall PA78 located furthest forward is located a distance d5 shorter than the first threshold distance from the detection range limit SL1 of the sensor 200. In this case, in the first embodiment, S220 is not executed, and the position of the parking stall PA68 is not estimated to be the end of the inner intersecting road edge. However, in other embodiments, since a wall W2 exists so as to block the road R9, which is the running road, in this case, the position of the parking stall PA78 is estimated to be the end of the inner intersecting road edge. In these configurations of the other embodiments described above, the end of the inner intersecting road edge can also be accurately identified.

[0050] (C2) In each embodiment, the intersecting lane ends estimated by the intersecting lane end estimation process are used to set a parking path, but the present disclosure is not limited to this. The estimated intersecting lane ends may be used to display an area in which the vehicle V0 can travel on a display device mounted on the vehicle V0. For example, the estimated intersecting lane ends may be displayed as lines superimposed on an image captured by the front camera 210 serving as the sensor 200, or the area between the left and right intersecting lane ends may be displayed as bands of a predetermined color superimposed on the image. Furthermore, for example, the estimated intersecting lane ends may be displayed as lines superimposed on a bird's-eye view image of a parking lot, or the area between the left and right intersecting lane ends may be displayed as bands of a predetermined color superimposed on the image.

[0051] (C3) In the parking lot Pa in each embodiment, the intersection of two roads R1 and R2 is a so-called T-shaped intersection, but the present disclosure is not limited to this. For example, as in the parking lot Pa1 in FIG. 14, the intersection of two roads R10 and R11 may be a so-called cross-shaped intersection C1. In the parking lot Pa1, four parking space groups PB1, PB2, PB3, and PB4 are provided, each consisting of a total of eight parking spaces PAx arranged in two rows and four columns. In this configuration, as in each embodiment, the active road edges e7L and e7R and the inner intersecting road edge e8L are identified. Note that the outer intersecting road edge e8R is identified as an intersecting road edge separated into two by the road R10.

[0052] Also, for example, as in parking lot Pa2 in FIG. 15, the intersection of two roads R12 and R13 may be a so-called cross-shaped intersection C2. Parking lot Pa2 has six parking space groups PB11, PB12, PB13, PB14, PB15, and PB16, each consisting of a total of six parking spaces PAx arranged in three rows and two columns. In this configuration, as in the previous embodiments, the current road edges e9L and e9R and the inner intersecting road edge e10L are identified. Note that the outer intersecting road edge e10R is identified as an intersecting road edge separated into two by road R12.

[0053] (C4) Each embodiment is merely an example and can be modified in various ways. For example, S215 may be omitted from the intersecting lane edge estimation process of each embodiment. Furthermore, for example, instead of S235, a line extending from the end of the intersecting lane edge in a direction perpendicular to the lane being traveled may be estimated to be the inner intersecting lane edge. Furthermore, for example, the first and second regular structures are not limited to parking spaces, parked vehicles, pillars, and walls, but may also be any type of structure, such as regularly arranged pylons or guardrails.

[0054] (C5) The driving assistance device 100 and the methods described herein may be implemented by a special-purpose computer configured by configuring a processor and memory programmed to execute one or more functions embodied in a computer program. Alternatively, the driving assistance device 100 and the methods described herein may be implemented by a special-purpose computer configured by configuring a processor with one or more dedicated hardware logic circuits. Alternatively, the driving assistance device 100 and the methods described herein may be implemented by one or more special-purpose computers configured by combining a processor and memory programmed to execute one or more functions with a processor configured with one or more hardware logic circuits. Furthermore, the computer program may be stored in a computer-readable non-transitory tangible recording medium as instructions executed by a computer.

[0055] The present disclosure may be realized in various forms, such as a driving assistance method, a driving assistance device, a computer program for realizing the driving assistance method, a non-transitory recording medium on which such a computer program is recorded, etc.

[0056] The present disclosure is not limited to the above-described embodiments and can be realized in various configurations without departing from the spirit thereof. For example, the technical features in each embodiment corresponding to the technical features in the form described in the Summary of the Invention section can be appropriately replaced or combined to solve some or all of the above-described problems or achieve some or all of the above-described effects. Furthermore, if a technical feature is not described as essential in this specification, it can be appropriately deleted. [Explanation of symbols]

[0057] 11...first driving path estimation unit, 12...second driving path estimation unit, 13...first structure estimation unit, 100...driving assistance device, V0...own vehicle, Pa, Pa1, Pa2...parking lot

Claims

1. A driving assistance device (100) mounted on a vehicle (V0) and assisting the vehicle in driving within a parking lot (Pa), a first road edge estimation unit (11) that estimates road edges on the left and right of a road on which the vehicle is traveling, using a detection result from a sensor (200) mounted on the vehicle that detects the surrounding environment of the vehicle; a first structure estimation unit (13) that estimates the position of a first regular structure, which is a structure that is provided in a regular pattern along the road edge during travel, using the detection result; a second road edge estimation unit (12) that estimates the left and right road edges of an intersecting road that intersects with the road during travel, using the estimated position of the first regular structure; A driving assistance device comprising:

2. The driving support device according to claim 1, A driving assistance device in which the second road edge estimation unit estimates the end of the first regular structure to be the end of an inner intersecting road edge (e2L) among the intersecting road edges, which is the intersecting road edge on the inside of the turning direction when the vehicle turns.

3. The driving support device according to claim 2, A driving assistance device in which the second road edge estimation unit estimates that the end of the first regular structure is the end of the inner intersecting road edge when the end of the first regular structure is located closer to the vehicle than a predetermined first threshold distance from the detection range limit (SL1) of the sensor.

4. The driving support device according to any one of claims 1 to 3, the first regular structure is a wall (W1), A driving assistance device in which the second road edge estimation unit estimates the end of the wall to be the end of the intersecting road edge (e2L) that is on the inside of the turning direction when the vehicle turns, among the intersecting road edges.

5. The driving support device according to any one of claims 1 to 3, the first regular structure is a plurality of pillars (OP1 to OP3) arranged in a predetermined direction, The second roadway edge estimation unit estimates that the end of the pillar among the plurality of pillars that is the foremost in the driving direction of the vehicle is the end of the intersecting road edge (e2L) that is on the inside of the turning direction when the vehicle turns.

6. The driving support device according to any one of claims 1 to 3, the first regular structure is at least one of a parking space (PA1 to PA7) and a parked vehicle (V1 to V7), The second road edge estimation unit estimates a position where the continuous parking spaces or the parked vehicles end as an end of the intersecting road edge.

7. 7. The driving support device according to claim 6, The first structure estimation unit identifies, from the detection results, a plurality of structures lined up along the direction of travel of the vehicle, and identifies, among the plurality of structures, a structure whose distance from the estimated road edge during travel in a direction perpendicular to the direction of travel is less than or equal to a predetermined second threshold distance as the first regularity structure.

8. 7. The driving support device according to claim 6, The first structure estimation unit identifies, from the detection results, a plurality of structures lined up along the direction of travel of the vehicle, and identifies the first regular structure from among the plurality of structures based on the distance to adjacent structures.

9. The driving support device according to any one of claims 1 to 3, a second structure estimation unit (14) that uses the detection results to estimate the position of a second regular structure, which is a structure that faces the running road and is provided in a regular manner along an outer intersecting road edge (e2R), which is an intersecting road edge on the outer side of the turning direction of the vehicle when the vehicle turns, among the intersecting road edges; The second road edge estimation unit uses the estimated position of the second regular structure to estimate the outer intersecting road edge among the intersecting road edges, and estimates that a line extended so as to be parallel to the estimated outer intersecting road edge is the intersecting road edge among the intersecting road edges that is on the inner side of the turning direction when the vehicle is turning.

10. A driving assistance method for assisting a vehicle (V0) in driving in a parking lot (Pa), comprising: A step of estimating road edges on the left and right of a road on which the vehicle is traveling, using a detection result of a sensor (200) mounted on the vehicle that detects the surrounding environment of the vehicle; a step of estimating a position of a first regular structure, which is a structure provided in a regular pattern along the road edge during travel, using the detection result; a step of estimating left and right road ends of an intersecting road that intersects with the road currently being traveled, using the estimated position of the first regular structure; A driving assistance method comprising:

11. A computer program for assisting a vehicle (V0) in driving in a parking lot (Pa), A function of estimating road edges on the left and right of the road on which the vehicle is traveling, using the detection results of a sensor (200) mounted on the vehicle that detects the surrounding environment of the vehicle; a function of estimating the position of a first regular structure, which is a structure provided in a regular pattern along the road edge during travel, using the detection result; a function of estimating the left and right road ends of an intersecting road that is a road that intersects with the road currently being traveled, using the estimated position of the first regular structure; A computer program that enables a computer to realize the above.

Images