Driving assistance device and driving assistance method

Abstract

To provide a driving support device capable of achieving more suitable parking operation.SOLUTION: A driving assistance device that assists automatic driving of a vehicle based on teaching travel by a user's driving operation, the driving assistance device comprising a storage device that stores a movement route from a travel start position of the teaching travel to a parking position of the teaching travel in the teaching travel, the training travel includes two turns of the traveling direction of the vehicle between the travel start position of the training travel and the parking position of the training travel, and a target route from the travel start position of the automatic travel to the parking position of the automatic travel is generated based on the training travel, the target route in the automatic travel is shorter than the movement route in the training travel, and the automatic travel does not include a turn of the traveling direction of the vehicle between the travel start position of the automatic travel and the parking position of the automatic travel; In the automatic travel, the vehicle travels on the target route at a non-zero vehicle speed.SELECTED DRAWING: Figure 5

Description

[Technical Field]

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

[0002] Generally, many parking spaces in homes are narrow, creating a strong demand for automated parking. Various driving assistance devices are known to realize this type of automated driving. For example, this type of driving assistance device performs automated driving to assist the driver in parking the vehicle from the initial stopping position to a target parking position within the parking space when the driver stops the vehicle at a predetermined initial stopping position. Such a driving assistance device automatically recognizes the parking space and the vehicle's position using various sensors mounted on the vehicle, and automatically performs steering, acceleration, braking, and other operations.

[0003] Patent Document 1 discloses a technology for realizing such automated driving, in which a user (hereinafter referred to as "user") in the vehicle drives the vehicle from a predetermined position outside the parking space to a target parking position within the parking space, and the route traveled during this process is stored as training data, allowing the vehicle to automatically drive along that route in subsequent parking situations. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2017-138664 [Non-patent literature]

[0005] [Non-Patent Document 1] Mikael Persson et al. “Lambda Twist: An Accurate Fast Robust Perspective Three Point (P3P) Solver.”, ECCV 2018, pp 334-349, published in 2018, http: / / openaccess.thecvf.com / content_ECCV_2018 / papers / Mikael_Persson_Lambda_Twist_An_ECCV_2018_paper.pdf Summary of the Invention [Problem to be solved by the invention]

[0006] According to the conventional driving assistance device described in Patent Document 1, the user only needs to perform a single teacher run (meaning a run performed by the user to generate teacher data; the same applies below) once, and thereafter parking operations can be performed automatically, so this technology can be said to be highly convenient.

[0007] However, when the user moves the vehicle into a parking space while turning the vehicle (meaning turning the vehicle's direction of travel from forward to reverse, or from reverse to forward; the same applies below), the vehicle may travel along a path that is not necessarily necessary (hereinafter referred to as an "unnecessary path") in order to change its position (meaning changing the direction the vehicle is facing; the same applies below). In other words, the travel path generated by teacher driving may include an unnecessary path.

[0008] In other words, if the travel route during teacher driving is used as the target route when automatically parking the vehicle in a parking space, as in the driving assistance device described in Patent Document 1, unnecessary driving will be performed every time.

[0009] The present disclosure has been made in consideration of the above-mentioned problems, and aims to provide a driving assistance device and a driving assistance method that are capable of realizing more suitable parking operations. [Means for solving the problem]

[0010] The present disclosure provides: A driving assistance method for assisting automatic driving of a vehicle based on supervised driving by a user's driving operation, comprising: storing a travel route from a travel start position of the teacher travel to a parking position of the teacher travel; generating a target route from a travel start position of the automatic driving to a parking position of the automatic driving based on the teacher travel; In the automatic driving, the vehicle travels along the target route at a non-zero vehicle speed, the teacher travel includes two turns in the travel direction of the vehicle between a travel start position of the teacher travel and a parking position of the teacher travel, the target route in the automatic driving is shorter than the travel route in the teacher driving, The automatic driving does not include a change in the direction of travel of the vehicle from a travel start position of the automatic driving to a parking position of the automatic driving. This is a driving assistance method.

[0011] In other respects, A driving assistance device that assists automatic driving of a vehicle based on a teacher driving operation performed by a user, a storage device that stores a movement route from a travel start position of the teacher travel to a parking position of the teacher travel, the teacher travel includes two turns in the travel direction of the vehicle between a travel start position of the teacher travel and a parking position of the teacher travel, generating a target route from a travel start position of the automatic driving to a parking position of the automatic driving based on the teacher travel; the target route in the automatic driving is shorter than the travel route in the teacher driving, the automatic driving does not include a change in the driving direction of the vehicle from a driving start position of the automatic driving to a parking position of the automatic driving, In the automatic traveling, the vehicle travels along the target route at a non-zero vehicle speed. It is a driving assistance device. [Effects of the Invention]

[0012] According to the driving assistance device according to the present disclosure, it is possible to realize a more suitable parking operation. [Brief explanation of the drawings]

[0013] [Figure 1] A block diagram showing an example of the overall configuration of a vehicle. [Figure 2A] FIG. 10 is a diagram showing an example of a functional block diagram of a driving assistance device (in supervised driving mode) [Figure 2B] A diagram showing an example of a functional block of a driving assistance device (in autonomous driving mode) [Figure 3A] A diagram showing an example of the vehicle's travel path (indicated by a solid arrow) when traveling in teacher driving mode. [Figure 3B] A diagram showing an example of a target route (route indicated by solid arrows) along which a vehicle will automatically travel in autonomous driving mode. [Figure 4] FIG. 10 is a diagram showing an example of teacher data generated by a teacher data generation unit. [Figure 5] A diagram explaining an example of a method for generating training data [Figure 6] FIG. 10 is a diagram showing an example of map data generated by a training data generation unit; [Figure 7] FIG. 10 is a diagram illustrating an example of processing by a position estimation unit. [Figure 8] A flowchart showing an example of an operation when a driving assistance device generates a target route. [Figure 9A] FIG. 10 is a diagram showing an example of a travel path (path indicated by a solid arrow) when a vehicle travels in the instructor travel mode in Modification 1. [Figure 9B] FIG. 10 is a diagram showing an example of a target route (a route indicated by a solid arrow) along which a vehicle will automatically travel in an automatic travel mode in Modification 1. [Figure 10] 10 is a flowchart showing an example of an operation performed by a driving assistance device according to a first modification when generating a target route. [Figure 11] FIG. 10 is a diagram showing an example of a user interface screen displayed on an HMI for a user to perform a selection operation in a driving assistance device according to a fourth modification. [Figure 12] FIG. 10 is a diagram illustrating the process performed by the training data generation unit of the fifth modified example. [Figure 13] 10 is a flowchart showing an example of an operation performed by a driving assistance device according to a fifth modification when generating a target route. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In this specification and drawings, components having substantially the same functions are designated by the same reference numerals, and redundant description will be omitted.

[0015] [Overall vehicle configuration] An example of the configuration of a vehicle according to an embodiment will be described below with reference to FIG.

[0016] FIG. 1 is a block diagram showing an example of the overall configuration of a vehicle 1. As shown in FIG.

[0017] The vehicle 1 includes a driving assistance device 10, a vehicle drive device 20, an on-board sensor 30, an on-board camera 40, an external storage device 50, and an HMI (Human Machine Interface) 60.

[0018] The vehicle drive device 20 is a means for realizing the driving, braking, and turning motions required for the vehicle 1 to travel, and is configured to include, for example, a drive motor, a power transmission mechanism, a brake device, a steering device, etc., and an ECU (Electronic Control Unit) that controls them. The vehicle drive device 20, for example, generates power using a drive motor, and transmits the power to the wheels via a power transmission mechanism (such as a propeller shaft, a differential gear, and a drive shaft), thereby causing the vehicle 1 to travel. Note that the operation of the vehicle drive device 20 according to this embodiment is controlled by the driving assistance device 10 in the autonomous travel mode.

[0019] The on-vehicle sensors 30 are various sensors mounted on the vehicle 1 and detect the driving state of the vehicle 1. The on-vehicle sensors 30 include, for example, an accelerator opening sensor that detects an accelerator opening, a steering angle sensor that detects a steering angle of a steering device, an acceleration sensor that detects acceleration acting in the longitudinal direction of the vehicle 1, a torque sensor that detects torque acting on a power transmission mechanism between the wheels and the drive motor of the vehicle 1, and a vehicle speed sensor that detects the vehicle speed of the vehicle 1. The on-vehicle sensors 30 output sensor information obtained by detection to the driving assistance device 10.

[0020] The on-board camera 40 is a surrounding sensor mounted on the vehicle 1 and monitors the environment surrounding the vehicle 1. In this embodiment, the on-board camera 40 is used, for example, to detect objects (typically, objects fixed on the ground) present around the vehicle 1 and estimate the current position of the vehicle 1 from the positional relationship between the vehicle 1 and the objects present around the vehicle 1. The on-board camera 40 is configured, for example, with four cameras arranged to capture images in four directions, i.e., the front, rear, left, and right directions of the vehicle 1. The on-board camera 40 outputs camera images generated by itself to the driving assistance device 10. Note that, instead of the on-board camera 40, a LiDAR, a radar, an ultrasonic sensor, or the like may be used as a surrounding sensor for estimating the self-position of the vehicle 1.

[0021] The external storage device 50 is, for example, an auxiliary storage device such as a hard disk drive (HDD), a solid state drive (SSD), or a flash memory. The external storage device 50 stores, for example, teacher data D1 and map data D2 generated by the driving assistance device 10 in the teacher driving mode.

[0022] The HMI 60 is a user interface, such as a touch panel, a commander, a button, or an operation key, that accepts input operations from a user in the vehicle 1. The HMI 60 is configured to be able to accept input operations, such as an execution command for executing the supervised driving mode or an execution command for executing the automatic driving mode. The HMI 60 outputs information related to the input operations accepted from the user to the driving assistance device 10.

[0023] The driving assistance device 10 is an electronic control unit that controls each part of the vehicle 1. In the automatic driving mode, the driving assistance device 10 is configured to control the vehicle drive device 20 while referring to sensor information acquired from the on-board sensor 30, so as to make the vehicle 1 drive automatically (i.e., autonomously).

[0024] The driving assistance device 10 includes, for example, a central processing unit (CPU) 10a, a read-only memory (ROM) 10b, a random access memory (RAM) 10c, an input port (not shown), and an output port (not shown). Each function of the driving assistance device 10, which will be described later, is realized, for example, by the CPU 10a referencing a control program and various data stored in the RAM 10c and the ROM 10b. However, some or all of the functions of the driving assistance device 10 may be realized by processing by a digital signal processor (DSP) or a dedicated hardware circuit (for example, an ASIC or an FPGA) instead of or in addition to processing by the CPU 10a.

[0025] The driving assistance device 10 is interconnected with the vehicle drive unit 20, on-board sensors 30, on-board cameras 40, external storage devices 50, and HMI 60 via an on-board network (e.g., a communication network conforming to the CAN communication protocol), and is capable of transmitting and receiving necessary data and control signals to and from each other.

[0026] [Configuration of driving assistance device] Next, an example of the configuration of the driving support device 10 will be described with reference to FIGS. 2A to 7. FIG.

[0027] The driving assistance device 10 is configured to be able to switch between a supervised driving mode and an automatic driving mode based on an input operation by the user. The supervised driving mode is a mode for registering a target route for automatic driving of the vehicle 1 in the automatic driving mode. In the supervised driving mode, the vehicle 1 is controlled to drive by the driving operation of the user. On the other hand, the automatic driving mode is a mode in which the vehicle 1 is automatically driven according to the target route registered in the supervised driving mode. In the automatic driving mode, the vehicle 1 is automatically controlled to drive by the driving assistance device 10 without the need for driving operation by the user.

[0028] However, the driving assistance device 10 according to this embodiment is configured to cut out unnecessary routes when turning from the travel route of the vehicle 1 during supervised driving, and generate a target route for the vehicle 1 to drive automatically.

[0029] 2A and 2B are diagrams showing examples of functional blocks of the driving assistance device 10. Note that Fig. 2A shows only the functional units that function when the vehicle 1 is in the supervised driving mode, and Fig. 2B shows only the functional units that function when the vehicle 1 is in the automatic driving mode.

[0030] Fig. 3A is a diagram showing an example of a travel route (route indicated by a solid arrow) when vehicle 1 travels in the teacher travel mode. Fig. 3B is a diagram showing an example of a target route (route indicated by a solid arrow) along which vehicle 1 will travel automatically in the automatic travel mode.

[0031] Fig. 3A shows a state in which a parking operation is performed by the user in the supervised driving mode, from a position P1 outside the parking space to a target parking position P2 within the parking space. The target route in Fig. 3B is generated based on the travel route taken by the vehicle 1 when it traveled in the supervised driving mode in Fig. 3A, and most of the target route is the same as the travel route. The target route is generated by changing only the turning positions within the target route from the original travel route.

[0032] Specifically, in the movement route of FIG. 3A, the vehicle turns around (here, turns from forward to reverse) at positions T1 and T2, but in the target route of FIG. 3B, unnecessary paths when vehicle 1 turns around within the movement route are cut, and vehicle 1 turns around at positions T1' and T2'.

[0033] <When running in teacher mode> First, the functional configuration of the driving assistance device 10 that functions when the teacher driving mode is executed will be described.

[0034] When the teacher driving mode is executed, the vehicle information acquisition unit 110, the dead reckoning unit 120, and the teacher data generation unit 130 function in the driving assistance device 10 (see FIG. 2A).

[0035] The vehicle information acquisition unit 110 acquires sensor information indicating the traveling state of the vehicle 1 from the on-board sensor 30. Then, the vehicle information acquisition unit 110 sends the acquired sensor information to the dead reckoning unit 120.

[0036] The dead reckoning unit 120 estimates the current position of the vehicle 1 based on temporal changes in the sensor values ​​of the on-board sensor 30 that detects the driving state of the vehicle 1. For example, the dead reckoning unit 120 calculates the amount of movement of the vehicle 1 from a reference position (for example, the driving start position when the teacher driving mode is started) based on temporal changes in the vehicle speed and yaw rate indicated by the on-board sensor 30, and estimates the current position of the vehicle 1 based on the amount of movement.

[0037] The current position of the vehicle 1 estimated by the dead reckoning unit 120 is a rough current position of the vehicle 1, and the position estimation by the dead reckoning unit 120 has low estimation accuracy. Therefore, in the driving assistance device 10 according to this embodiment, the actual current position of the vehicle 1 is corrected based on a camera image generated by the on-board camera 40. Reasons for the low estimation accuracy of the position estimation by the dead reckoning unit 120 include the difficulty of constructing a complete kinematics model (for example, tire diameter and tread width), the fact that the observation values ​​of the on-board sensor 30 always contain noise, and the fact that external factors that cannot be observed by the on-board sensor 30 (for example, tire slippage) significantly affect the position estimation. Furthermore, errors caused by these factors accumulate in the position estimation by the dead reckoning unit 120.

[0038] The teacher data generation unit 130 stores the travel route during teacher driving and generates teacher data D1 related to a target route (hereinafter simply referred to as the "target route") for automatically driving the vehicle 1 in the automatic driving mode from the travel route. The teacher data generation unit 130 starts storing the travel route, for example, in response to a command from the user to start teacher driving mode, and ends storing the travel route in response to a command from the user to end teacher driving mode. Teacher driving typically starts when the vehicle 1 is stopped at a desired position outside the parking space (P1 in FIG. 3A) and ends when the vehicle 1 is stopped at a parking target position (P2 in FIG. 3A) within the parking space. The teacher data D1 generated by the teacher data generation unit 130 is stored in the external storage device 50.

[0039] Fig. 4 is a diagram showing an example of teacher data D1 generated by the teacher data generation unit 130. Data D1a shown in Fig. 4 is data indicating the route actually traveled by the vehicle 1 and the driving behavior at each position on the route, which is sequentially stored during teacher driving, and teacher data D1 is generated from this data D1a during teacher driving.

[0040] FIG. 5 is a diagram illustrating an example of a method for generating the training data D1.

[0041] 4 and 5 explain the process of cutting out unnecessary routes before and after the turning point T1 in Fig. 3. WP_1, WP_2, ... WP_G in Fig. 4 and 5 represent, in order, each position on the travel route of vehicle 1 stored during training driving, with WP_1 corresponding to the start point of the travel route and WP_G corresponding to the end point of the travel route.

[0042] The training data D1 includes, for example, a target route for automatically driving the vehicle 1, the orientation (i.e., direction) of the vehicle 1 at each target position on the target route, the driving direction (i.e., forward or backward) of the vehicle 1 at each target position on the target route, and reference driving information at each target position on the target route.

[0043] Here, the "target route" of the teacher data D1 is generated by, for example, deleting unnecessary routes before and after a turning point from the data D1a of the travel route during teacher driving, and connecting the routes in between.

[0044] Specifically, the teacher data generation unit 130 searches for a pair of first and second locations on the travel route during teacher driving, where the difference in the direction of travel of the vehicle 1 when the vehicle 1 traveled is less than a first threshold and the difference in distance is less than a second threshold (hereinafter referred to as the "connection condition"). When a pair of first and second locations that satisfies the connection condition is detected, the teacher data generation unit 130 connects the first location or a location adjacent to the first location on the travel route with the second location or a location adjacent to the second location on the travel route, and changes the travel route to a route that is a shortcut at the above-mentioned turning point, thereby generating a target route. In other words, the teacher data generation unit 130 sets the turning point of the vehicle 1 on the target route to the location connecting the first location (or a location adjacent to the first location on the travel route) and the second location (or a location adjacent to the second location on the travel route) instead of the turning point on the travel route during teacher driving.

[0045] The first threshold value for the difference in orientation (the orientation of vehicle 1 when vehicle 1 traveled through the location) that serves as the reference for the connection condition is, for example, a value between 0 degrees and 3 degrees. The second threshold value for the difference in distance that serves as the reference for the connection condition is, for example, a value between 0 m and 1 m.

[0046] The reason why the teacher data generation unit 130 performs such processing is that a pair of first and second locations that satisfy the connection condition can be considered to be locations on the travel route during teacher driving that are in approximately the same position and pass through approximately the same location before and after the turning point of the vehicle 1. In other words, the portion between the pair of first and second locations and the turning point on the travel route can be said to be an unnecessary route that is not effective for changing the attitude of the vehicle 1. In FIG. 5, the locations WP_N and WP_K on the travel route correspond to a pair of first and second locations that satisfy the connection condition.

[0047] Preferably, when generating the target route, the teacher data generation unit 130 selects a point adjacent to at least one of the first and second points on the travel route as a target for connection so that the route between the two connected points is not perpendicular to the routes before and after the point. For example, in FIG. 5, it is preferable to connect points WP_N-1 and WP_K rather than connecting points WP_N and WP_K, which are a pair that satisfies the connection condition. That is, in FIG. 5, it is preferable to set the target route so that the vehicle 1 travels forward from point WP_N-1 to point WP_K, and then turns around at point WP_K (see the connection positions in the teacher data D1 in FIG. 4).

[0048] This prevents the target route from moving the vehicle 1 laterally (i.e., in the vehicle width direction) at the coupling position, and enables the target route to be one that allows the vehicle 1 to travel smoothly. In particular, as shown in Fig. 5, by setting one of the coupling targets to point WP_N-1 instead of point WP_N (i.e., by shifting the point on the starting point side of the movement route of the pair of first and second points one position toward the starting point), it becomes possible to prevent the steering angle from becoming excessive when moving the vehicle 1 along the movement route after point WP_K.

[0049] An example of the processing performed by the teacher data generating unit 130 to generate a target route will be described later with reference to FIG.

[0050] The "orientation of vehicle 1 at each target position on the target route" and "driving direction of vehicle 1 at each target position on the target route" in the teacher data D1 directly reflect, for example, the orientation and driving direction of vehicle 1 during teacher driving.

[0051] Furthermore, the "reference driving information at each target position on the target route" in the teacher data D1 is, for example, driving information (e.g., steering angle and vehicle speed) detected at each position on the travel route when the vehicle 1 travels as a teacher. This reference driving information serves as a reference (e.g., referred to as a feedforward element) when the vehicle control unit 160 (described later) controls the vehicle drive device 20 (e.g., steering angle of the steering device) at each position on the travel route when performing automatic driving.

[0052] Preferably, the teacher data generation unit 130 changes the vehicle speed at the junction position (here, point WP_K) on the target route from the vehicle speed during teacher driving to zero [m / sec] in the "reference driving information at each target position on the target route" of the teacher data D1. This makes it possible to prevent a sudden change in vehicle speed when the vehicle 1 turns around at the junction position.

[0053] Here, the teacher data generation unit 130 generates the teacher data D1 while creating map data D2 for estimating the current position of the vehicle 1 from, for example, a camera image captured by the in-vehicle camera 40. This technique is also called SLAM (Simultaneous Localization and Mapping) method.

[0054] Fig. 6 is a diagram showing an example of map data D2 generated by the teacher data generating unit 130. Note that Fig. 6 shows a bird's-eye view of the position in real space of feature point Q in the real scene stored in the map data D2.

[0055] The map data D2 is data that stores, for each of a plurality of feature points in a real scene, the three-dimensional position of the feature point in real space and the feature amount of the feature point obtained from a camera image taken when the map data D2 was created, in association with each other. Feature points stored as the map data D2 are, for example, portions (e.g., corners) from which distinctive image patterns can be obtained from camera images of objects (e.g., trees, walls, pillars, etc.) that can serve as landmarks in the real scene. The plurality of feature points in the map data D2 are stored so as to be individually identifiable, for example, by identification numbers.

[0056] The three-dimensional positions in real space of the feature points stored in the map data D2 are expressed in a three-dimensional Cartesian coordinate system (X, Y, Z).

[0057] The feature amounts of feature points stored in the map data D2 may be brightness or density on the camera image, SIFT (Scale Invariant Feature Transform) feature amounts, SURF (Speeded Up Robust Features) feature amounts, etc. The feature amount data of feature points stored in the map data D2 may be for feature points at the same three-dimensional position, or may be stored separately for each camera shooting position or shooting direction when the feature point was captured. Furthermore, the feature amount data of feature points stored in the map data D2 may be stored in association with an image of an object having the feature point.

[0058] The teacher data generation unit 130 identifies the coordinates of feature points in the actual scene, for example, based on stereo photogrammetry. Specifically, the teacher data generation unit 130 reads multiple camera images generated at different times and associates identical feature points that appear in the multiple camera images. The teacher data generation unit 130 then acquires, for example, information related to the provisional position of the vehicle 1 when the multiple camera images were generated from the dead reckoning unit 120 and identifies the provisional coordinates of the feature points in the actual scene based on the principle of triangulation. The teacher data generation unit 130 then performs bundle adjustment using, for example, the provisional position of the vehicle 1 and the provisional coordinates of the feature points in the actual scene as reference information to calculate the formal position of the vehicle 1 and the formal coordinates of the feature points in the actual scene so as to minimize the reprojection error when each feature point in the actual scene is projected onto all camera images. The teacher data generation unit 130 then stores the formal position of the vehicle 1 as data D1a indicating the position of the vehicle 1 during the teacher driving. Furthermore, the teacher data generating unit 130 stores the formal coordinates of the feature points in the actual scene in the external storage device 50 as map data D2.

[0059] The map data D2 may be generated in advance using a light detection and ranging (LIDAR) or a stereo camera instead of using the SLAM method. However, since the map data D2 often contains distortion depending on the measurement method used when generating the map data D2, it is preferable to use the SLAM method from the viewpoint of suppressing a decrease in the accuracy of position estimation due to distortion in the map data D2 itself.

[0060] <When running in automatic driving mode> Next, the functional configuration of the driving assistance device 10 that functions when the automatic driving mode is executed will be described.

[0061] When the autonomous driving mode is executed, the vehicle information acquisition unit 110, the dead reckoning unit 120, the position estimation unit 140, the target route reading unit 150, and the vehicle control unit 160 function in the driving assistance device 10 (see FIG. 2B).

[0062] The vehicle information acquisition unit 110 and the dead reckoning unit 120 have the same configuration as those described above, and therefore, a description thereof will be omitted here.

[0063] The position estimation unit 140 estimates the current position of the vehicle 1 based on the map data D2 and the camera image of the vehicle-mounted camera 40.

[0064] Fig. 7 is a diagram illustrating an example of processing by the position estimation unit 140. In Fig. 7, points R1, R2, and R3 represent three feature points extracted from the camera image of the in-vehicle camera 40, and points Q1, Q2, and Q3 represent the three-dimensional positions in real space of the feature points R1, R2, and R3 stored in the map data D2. RP1 represents the imaging plane of the in-vehicle camera 40. Point P1 represents the position of the in-vehicle camera 40 (i.e., the position of the vehicle 1) determined from the three feature points R1, R2, and R3 extracted from the camera image of the in-vehicle camera 40 and the points Q1, Q2, and Q3 stored in the map data D2.

[0065] For example, the position estimation unit 140 first compares feature points extracted from the camera image of the in-vehicle camera 40 with feature points stored in the map data D2 using pattern matching, feature search, or the like. Then, the position estimation unit 140 randomly selects several (e.g., three to six) feature points from among the feature points extracted from the camera image of the in-vehicle camera 40 that can be compared with the feature points stored in the map data D2, and estimates the position of the vehicle 1 in real space based on the positions of these several feature points in the camera image and the three-dimensional positions of the several feature points in real space stored in the map data D2. At this time, the position estimation unit 140 calculates the position of the vehicle 1 by solving a PnP problem using, for example, a known method such as Lambda Twist (see, for example, Non-Patent Document 1).

[0066] When comparing the feature points extracted from the camera image of the in-vehicle camera 40 with the feature points stored in the map data D2, the position estimation unit 140 may, for example, use the current position of the vehicle 1 estimated by the dead reckoning unit 120 as a reference to narrow down the feature points stored in the map data D2 to be compared with the feature points extracted from the camera image of the in-vehicle camera 40.

[0067] Information relating to the current position of the vehicle 1 estimated by the position estimation unit 140 is sent to the vehicle control unit 160. Note that the information relating to the current position of the vehicle 1 estimated by the position estimation unit 140 includes, for example, information relating to the two-dimensional position of the vehicle 1 in real space (X coordinate, Y coordinate) and information relating to the orientation (i.e., direction) of the vehicle 1.

[0068] The target route reading unit 150 reads out the training data D1 for moving the vehicle 1 from the external storage device 50 and sends it to the vehicle control unit 160.

[0069] When receiving a command to execute the automatic driving mode from the user, the vehicle control unit 160 moves the vehicle 1 along the target route indicated by the teacher data D1 acquired from the target route reading unit 150 while recognizing the current position of the vehicle 1. Then, the vehicle control unit 160 stops the vehicle 1 at the end point of the target route (i.e., the position corresponding to the end point of the travel route during teacher driving).

[0070] The vehicle control unit 160 typically controls the traveling of the vehicle 1 (e.g., the steering angle of the steering device of the vehicle 1) by feedback control based on the current position of the vehicle 1 estimated by the position estimation unit 140 and each target position on the target route indicated by the teacher data D1. The reason for performing such control is that the initial position of the vehicle 1 when the autonomous traveling mode is started may not exactly match the initial position of the target route indicated by the teacher data D1. Furthermore, a method of reproducing a travel route using only dead reckoning is likely to result in large errors, especially in low-speed areas or travel routes with turns. In addition, since the target route during autonomous traveling is a modified version of the travel route during teacher traveling, if the vehicle speed and steering angle during teacher traveling are used as is, there is a risk that the actual traveling route will deviate from the target route.

[0071] At this time, the vehicle control unit 160 may use the reference driving information (vehicle speed, steering angle, etc.) set in the teacher data D1 as a feedforward element.

[0072] Furthermore, when causing the vehicle 1 to automatically travel, the vehicle control unit 160 does not necessarily have to start the automatic travel from the starting point of the target route indicated by the teacher data D1, but may also cause the vehicle 1 to automatically travel from a position between the starting point and the end point of the target route indicated by the teacher data D1. In this case, the vehicle control unit 160 may use, for example, a predetermined evaluation function to determine the easiest target position to enter from among the target positions between the starting point and the end point of the target route indicated by the teacher data D1, based on the initial position and orientation of the vehicle 1 at the start of the automatic travel mode. This allows the vehicle 1 to reach the target route more smoothly, even if the initial position of the vehicle 1 at the start of the automatic travel mode is deviated from the starting point of the target route.

[0073] [Operation flow of driving assistance device] Fig. 8 is a flowchart showing an example of the operation of the driving assistance device 10 (teacher data generation unit 130) when generating a target route. The flowchart shown in Fig. 8 is a process executed by the driving assistance device 10, for example, after the end of a teacher driving. In the flowchart of Fig. 8, WP_N is used as a variable representing one of a pair of two points, among the points WP_1, WP_2, ...WP_G on the travel route.

[0074] In step S1, first, the driving assistance device 10 sets N=1, and sets one candidate point of the two-point pair as the starting point within the travel route.

[0075] In step S2, the driving assistance device 10 searches for a point that satisfies the connection condition from each point on the travel route between point WP_G and point WP_N+1, based on the orientation and position of the vehicle 1 when it traveled through point WP_N. That is, here, the driving assistance device 10 sequentially searches for a point that pairs with point WP_N that satisfies the connection condition within the travel route, starting from the end point of the travel route. As a result, if there is a point within the travel route that pairs with point WP_N that satisfies the connection condition, the point closest to the end point of the travel route will be detected first.

[0076] In step S3, the driving assistance device 10 determines whether a point satisfying the connection condition has been detected at any point from point WP_G to point WP_N+1 on the travel route. If no point satisfying the connection condition has been detected at any point from point WP_G to point WP_N+1 on the travel route (S3: NO), the driving assistance device 10 proceeds to step S4. If a point satisfying the connection condition has been detected at any point from point WP_G to point WP_N+1 on the travel route (here, point WP_K) (S3: YES), the driving assistance device 10 proceeds to step S5.

[0077] In step S4, the driving assistance device 10 increments one of the candidate points of the two point pair to N=N+1 (i.e., shifts one of the candidate points of the two point pair by one point toward the end point of the travel route), and proceeds to step S8.

[0078] In step S5, the driving assistance device 10 removes WP_N to WP_K-1 from the WP column of the data D1a related to the travel route generated during teacher driving, and connects WP_N-1 and WP_K. This generates teacher data D1 related to the target route from which unnecessary routes have been deleted from the travel route during teacher driving (see FIGS. 4 and 5).

[0079] In step S6, the driving assistance device 10 sets the point WP_K in the teacher data D1 to be a turning point. Specifically, since the vehicle speed at the point WP_K during teacher driving is non-zero, the driving assistance device 10 changes the vehicle speed at the point WP_K in the data D1a related to the travel route generated during teacher driving to zero in the teacher data D1. Note that at this time, the driving assistance device 10 may also change the vehicle speeds at points before and after the point WP_K (WP_K+1, WP_N-1, etc.) to values ​​close to zero, in order to further suppress changes in vehicle speed during turning.

[0080] In step S7, the driving assistance device 10 increments one of the candidate locations of the two location pairs to N=K+1, and proceeds to step S8. Note that the reason for incrementing N=K+1 here is because WP_N to WP_K-1 have been linked by the processing in step S5.

[0081] In step S8, the driving assistance device 10 determines whether the variable related to one of the candidate locations of the two location pairs has been incremented to N=G-1. If the variable related to the candidate location has not been incremented to N=G-1 (S8: NO), the driving assistance device 10 returns to step S2 and executes the same process, and if the variable related to the candidate location has been incremented to N=G-1 (S8: YES), the driving assistance device 10 proceeds to step S9.

[0082] In step S9, the driving assistance device 10 stores in the external storage device 50 the teacher data D1 generated in steps S2 to S8.

[0083] Through the above-described processing, a target route is generated in which unnecessary routes are removed from the travel route during teacher travel.

[0084] [effect] As described above, the driving assistance device 10 of this embodiment searches for a pair of first and second points on the travel route, on either side of a turning point in the travel direction, where the difference in direction when the vehicle 1 traveled is less than a first threshold and the difference in distance is less than a second threshold.If such a pair is detected, the driving assistance device 10 connects the first point or a point adjacent to the first point on the travel route with the second point or a point adjacent to the second point on the travel route, and changes the travel route to a route that is a shortcut at the turning point, thereby generating a target route for automatic driving.

[0085] As a result, when the vehicle 1 is automatically parked at a target parking position within a parking space during autonomous driving, the vehicle 1 can be moved to the target parking position without following an unnecessary route, thereby shortening the driving time during autonomous driving and optimizing vehicle driving.

[0086] As a method for generating a target route according to the prior art, for example, a method has been considered in which only the start point, the turning point, and the end point of the travel route are stored, and a route is automatically generated by calculation using an evaluation function to smoothly connect them. However, with such a method, the route between intermediate points is optimized by the evaluation score, so even if the user has performed supervised driving to avoid places that the user intentionally does not want to pass through (e.g., grass) or places that are difficult for the sensor to detect as obstacles (e.g., thin poles, chains, ponds, etc.), the robot will still try to pass through these places during automatic driving.

[0087] In this regard, the driving assistance device 10 according to this embodiment is useful in that it basically controls the driving of the vehicle 1 so that it follows the route intended by the user, thereby preventing the occurrence of problems that occur in the prior art.

[0088] (Variation 1) In the above embodiment, when generating a target route, the teacher data generation unit 130 connects pairs of two points that have different driving directions and are separated by a turning point in the driving direction. However, when generating a target route, the teacher data generation unit 130 may also connect pairs of two points that have the same driving direction and are separated by a turning point in the driving direction, as long as the pair satisfies a connecting condition.

[0089] 9A is a diagram showing an example of a travel route (route indicated by a solid arrow) when vehicle 1 travels in the teacher travel mode in this modified example 1. FIG. 9B is a diagram showing an example of a target route (route indicated by a solid arrow) along which vehicle 1 will travel automatically in the automatic travel mode in this modified example 1.

[0090] FIG. 9A shows a state in which the vehicle 1, driven by the user's driving operation, changes direction twice as it travels from a position P1 outside the parking space to a target parking position P2 within the parking space. Here, the training data generation unit 130 determines that the pair of points WP_L and WP_M within the travel route satisfy the conditions that the difference in direction is less than a first threshold and the difference in distance is less than a second threshold during the search. Therefore, in this first modification, the training data generation unit 130 generates a target route by connecting point WP_L-1 (a point adjacent to point WP_L on the start side of the travel route) with point WP_M, as shown in FIG. 9B. Note that, in this case, the travel direction at point WP_L and the travel direction at point WP_M are the same (here, reverse travel).

[0091] Even with this configuration, it is possible to generate a target route in which unnecessary routes within the travel route during teacher driving are cut out. However, in this case, the vehicle 1 will travel without making any changes in direction at the position where two points are connected. In such a situation, as in the above embodiment, if the vehicle 1 is stopped once at the position where two points are connected during autonomous driving, it will cause the user to feel uncomfortable. Therefore, in this modified example 1, when connecting two points with the same traveling direction, it is preferable that the teacher data generation unit 130 sets the vehicle speed at the connecting position to a non-zero value (for example, the vehicle speed during teacher driving remains the same).

[0092] Fig. 10 is a flowchart showing an example of the operation of the driving assistance device 10 (teacher data generation unit 130) according to Modification 1 when generating a target route. The processing of steps S1 to S9 is the same as the processing described in the flowchart of Fig. 8, but this flowchart differs from the flowchart of Fig. 8 in that step Sa is inserted between step S4 and step S5.

[0093] Step Sa is a process that is carried out after a pair of two points WP_N and WP_K that satisfies the connection condition is detected and points WP_N-1 and WP_K corresponding to this pair of two points are connected.

[0094] In step Sa, the driving assistance device 10, for example, references the shift information at point WP_N and the shift information at point WP_K during teacher driving to determine whether the driving direction at point WP_N and the driving direction at point WP_K are the same or opposite. If the driving direction at point WP_N and the driving direction at point WP_K are opposite (step Sa: No), the driving assistance device 10 sets the joining position (here, point WP_K) as a turning point in the teacher data D1 (step S6) and proceeds to step S7. If the driving direction at point WP_N and the driving direction at point WP_K are the same (step Sa: Yes), the driving assistance device 10 proceeds to step S7 without setting the joining position (here, point WP_K) as a turning point in the teacher data D1.

[0095] Here, when the driving assistance device 10 sets the connection position as a turning point in the teacher data D1 (step S6), for example, as in the above embodiment, it sets the target vehicle speed at the connection position set in the teacher data D1 (here, point WP_K) to zero. On the other hand, when the driving assistance device 10 does not set the connection position as a turning point in the teacher data D1, for example, it leaves the target vehicle speed at the connection position set in the teacher data D1 (here, point WP_K) as the vehicle speed during teacher driving.

[0096] As described above, according to the target route setting method of the present modified example 1, it may be possible to set a more optimal target route compared to the target route setting method according to the above embodiment.

[0097] (Variation 2) In the above embodiment, when searching for a pair of two locations, the teacher data generation unit 130 applies the same connection conditions (the first threshold value related to the direction and the second threshold value related to the distance) to each candidate location. However, when searching for a pair of two locations, the teacher data generation unit 130 may set the first threshold value and / or the second threshold value related to the connection conditions to be larger as the distance from the candidate location to the end point of the travel path during teacher driving increases. In other words, the teacher data generation unit 130 may relax the connection conditions for determining whether a candidate location that is far from the target parking position corresponds to a pair of two locations.

[0098] Generally, there are often only narrow spaces near the target parking position, but there are often relatively more spacious spaces at positions away from the target parking position. In other words, when vehicle 1 is driven automatically, if vehicle 1 deviates even slightly from the travel path during teacher driving near the target parking position, there is a risk that vehicle 1 will collide with an obstacle, but at positions away from the target parking position, vehicle 1 can be guided to the target parking position without colliding with an obstacle even if it deviates slightly from the travel path during teacher driving.

[0099] In this regard, as in this variant example 2, when searching for a pair of two locations, the greater the distance from the candidate location to the end point of the travel route during teacher driving, the larger the first threshold and / or second threshold related to the above-mentioned connection condition can be set, making it possible to more actively eliminate unnecessary routes.

[0100] (Variation 3) In the above embodiment, when the teacher data generation unit 130 searches for two-point pairs, it searches all points from the start point to the end point of the travel route during teacher driving. However, when searching for two-point pairs, the teacher data generation unit 130 may exclude from the search targets points on the travel route during teacher driving whose curvature is equal to or greater than a third threshold. Here, the third threshold is, for example, a curvature converted from the general steering angle limit of the vehicle 1.

[0101] Generally, due to the limitations of the steering angle of the vehicle 1, when the vehicle 1 is driven along a certain route, the curvature of the route must be equal to or less than a predetermined value. In this regard, after changing direction of travel, the vehicle 1 is usually driven along a route with a certain degree of curvature. Therefore, if a point with a curvature equal to or greater than the predetermined value is connected as one of a pair of two points, the curvature of the subsequent route may become excessively large. As a result, the vehicle 1 may not be able to drive along the target route and may deviate from the target route. If such a situation occurs near the target parking position, the vehicle 1 may collide with an obstacle. This is because there is usually only a narrow space near the target parking position.

[0102] In this regard, by limiting the search targets as in the third modification, it is possible to prevent the curvature of the target route from becoming too large.

[0103] The curvature at each point determined in this modification 3 may be the curvature of the curve connecting the point and its adjacent points, or may be the curvature of the curve connecting the point and the target parking position. In this case, only points near the target parking position can be excluded from the search target.

[0104] (Variation 4) When a pair of two locations that satisfies the connection condition is detected, the teacher data generating unit 130 may determine whether or not to connect the pair of two locations based on a selection operation by the user.

[0105] Fig. 11 is a diagram showing an example of a user interface screen displayed on the HMI 60 for the user to perform selection operations in the driving assistance device 10 according to Modification 4. The left diagram in Fig. 11 shows the travel route during supervised driving, and the right diagram in Fig. 11 shows the target route proposed by the supervised driving data generation unit 130.

[0106] In Fig. 11, the user can select whether or not to change to a shortcut route for each of the turning points T1 and T2 while viewing the user interface screen. Note that Rb1, Rb2, and Rb3 in Fig. 11 are icons for the user to input.

[0107] When the user drives the vehicle 1 in the supervised driving mode, the user may drive the vehicle 1 along the route with some intention. For example, the user may want to activate a monitoring sensor installed in a parking space by passing a predetermined position. In such a case, changing the target route to a shortcut from the route driven in the supervised driving mode would be contrary to the user's intention.

[0108] In this regard, as in this variant example 4, the user can select whether to change the target route to a route that is a shortcut at the location of a pair of two points that satisfies the connection conditions from the route traveled during teacher driving, or to use the same route as the route traveled during teacher driving, thereby making it possible to set a target route that is in line with the user's intentions.

[0109] (Variation 5) In the above embodiment, the teacher data generation unit 130 generates a target route for autonomous driving by referencing only the data of the movement route from a single supervised driving session and optimizing that movement route. However, when generating a target route for autonomous driving, the teacher data generation unit 130 may also refer to data of the movement routes from previous supervised driving sessions in addition to data of the movement route from the current supervised driving session and combine these to generate the target route.

[0110] When parking for teacher driving, unless the user is an experienced driver, a single teacher driving may include unnecessary routes in the travel route. In this regard, it would be very convenient if the optimal travel route could be generated by performing teacher driving several times and combining the travel routes taken during those teacher driving. The driving assistance device 10 according to the fifth modification responds to such a demand.

[0111] Figure 12 is a diagram illustrating the processing performed by the teacher data generation unit of Variation 5. The left diagram of Figure 12 shows the travel route when teacher driving is performed from a position P1 outside the parking space to a target parking position P2 within the parking space, with the solid line indicating the travel route during the current teacher driving and the dashed line indicating the travel route during previous teacher driving (i.e., saved data). The right diagram of Figure 12 shows a target route generated by combining the travel route during the current teacher driving and the travel route during previous teacher driving.

[0112] For example, the teacher data generation unit 130 compares the vehicle attitude at each position on the travel route during this teacher driving with the vehicle attitude at each position on the travel route during previous teacher driving, and searches for a point where the difference in orientation is less than a first threshold and the difference in distance is less than a second threshold. If there is a point where the difference in orientation is less than the first threshold and the difference in distance is less than the second threshold, the teacher data generation unit 130 determines that this is a candidate point (hereinafter also referred to as a "point of similar orientation") for combining the travel route during this teacher driving with the travel route during previous teacher driving.

[0113] The teacher data generation unit 130 generates a target route that combines the route taken during the current teacher driving and the route taken during the previous teacher driving by selecting either the route taken during the current teacher driving or the route taken during the previous teacher driving for each of the routes, such as the route from the start point to a candidate point, the route from that candidate point to the next candidate point, and the route from the next candidate point to the goal point. Note that, at this time, the teacher data generation unit 130 determines which of the route taken during the current teacher driving or the route taken during the previous teacher driving is the most optimal route using an evaluation function based on criteria such as the number of turns and the length of the distance for the routes connecting each point, and then selects either the route taken during the current teacher driving or the route taken during the previous teacher driving.

[0114] In Figure 12, points Ts11 and Ts12 correspond to candidate points where the difference in direction is less than the first threshold and the difference in distance is less than the second threshold. The right diagram in Figure 12 shows an example in which the route taken during a previous supervised run is selected as the route from the start point to candidate point Ts11, the route taken during the current supervised run is selected as the route from candidate point Ts11 to candidate point Ts12, and the route taken during a previous supervised run is selected as the route from candidate point Ts12 to the finish point.

[0115] FIG. 13 is a flowchart showing an example of the operation of the driving assistance device 10 (teacher data generating unit 130) according to the fifth modification when generating a target route.

[0116] In step S11, the driving assistance device 10 compares the saved route A (representing the route traveled during a previous instructor run; the same applies below) with the current route B to search for the presence of an attitude similarity point Ts (i.e., a point where the difference in orientation is less than the first threshold and the difference in distance is less than the second threshold).

[0117] In step S12, the driving assistance device 10 determines whether or not a posture similar point Ts exists. If a posture similar point Ts exists, the driving assistance device 10 proceeds to step S13, and if a posture similar point Ts does not exist, the driving assistance device 10 proceeds to step S15.

[0118] In step S13, the driving assistance device 10 selects either route A or route B as the optimal route from the start point to the posture similar point Ts. Then, the driving assistance device 10 selects either route A or route B as the optimal route from the previous posture similar point Ts to the next posture similar point Ts (or the goal point if there is no next posture similar point Ts). Then, the driving assistance device 10 connects the respective optimal routes at the posture similar point Ts. The driving assistance device 10 performs this process for all of the detected posture similar points Ts.

[0119] In step S14, the driving assistance device 10 presents the route optimized in step S13 to the user along with route A and route B, prompts the user to select which route to adopt, and proceeds to step S16.

[0120] In step S15, the driving assistance device 10 presents both route A and route B to the user, prompts the user to select which route to adopt, and then proceeds to step S16.

[0121] In step S16, the driving assistance device 10 stores the route selected by the user in the external storage device 50 as training data (ie, a target route for automatic driving).

[0122] As described above, according to the target route setting method of the fifth modification, it may be possible to set a more optimal target route compared to the target route setting method according to the above embodiment.

[0123] Although specific examples of the present invention have been described above in detail, these are merely examples and do not limit the scope of the claims. The technology described in the claims includes various modifications and alterations of the specific examples exemplified above. [Industrial Applicability]

[0124] According to the driving assistance device according to the present disclosure, it is possible to realize a more suitable parking operation. [Explanation of symbols]

[0125] 1 vehicle 10 Driving assistance devices 20 Vehicle drive unit 30 In-vehicle sensors 40 In-vehicle camera 50 External storage device 110 Vehicle information acquisition unit 120 Dead Reckoning 130 Teacher data generation unit 140 Position estimation part 150 Target path readout unit 160 Vehicle control unit D1 Training data D2 map data

Claims

1. A driving assistance method for assisting automatic driving of a vehicle based on supervised driving by a user's driving operation, comprising: storing a travel route from a travel start position of the teacher travel to a parking position of the teacher travel; generating a target route from a travel start position of the automatic driving to a parking position of the automatic driving based on the teacher travel; In the automatic driving, the vehicle travels along the target route at a non-zero vehicle speed, the teacher travel includes two turns in the travel direction of the vehicle between a travel start position of the teacher travel and a parking position of the teacher travel, the target route in the automatic driving is shorter than the travel route in the teacher driving, The automatic driving does not include a change in the direction of travel of the vehicle from a travel start position of the automatic driving to a parking position of the automatic driving. Driving assistance methods.

2. The change in the traveling direction of the vehicle is either a change in the traveling direction of the vehicle from forward to reverse, or a change in the traveling direction of the vehicle from reverse to forward. The driving assistance method according to claim 1 .

3. the travel path includes a first turning position and a second turning position in a turning of the traveling direction of the vehicle, the first turning point is a position on the travel path that is closer to a travel start position of the instructor travel than the second turning point, the second turning position is a position on the travel path that is closer to a parking position for the teacher travel than the first turning position, the target route does not include the first turning position and the second turning position on the target route; The driving assistance method according to claim 1 .

4. the travel path includes a first turning position and a second turning position in a turning of the traveling direction of the vehicle, the first turning point is a position on the travel path that is closer to a travel start position of the instructor travel than the second turning point, the second turning position is a position on the travel path that is closer to a parking position for the teacher travel than the first turning position, the target route includes, on the target route, a first position between a travel start position of the teacher travel and the first turning position, and a second position between a parking position of the teacher travel and the second turning position, the target route includes a route connecting the first position and the second position; The driving assistance method according to claim 1 .

5. the travel path includes a first turning position and a second turning position in a turning of the traveling direction of the vehicle, the first turning point is a position on the travel path that is closer to a travel start position of the instructor travel than the second turning point, the second turning position is a position on the travel path that is closer to a parking position for the teacher travel than the first turning position, the target route includes, on the target route, a first position between a travel start position of the teacher travel and the first turning position, and a second position between a parking position of the teacher travel and the second turning position, the target route includes a route connecting the first position and the second position; the target route includes a position on a route connecting the first position and the second position that has not been traveled during the teacher travel; The driving assistance method according to claim 1 .

6. The movement route during the teacher's travel is displayed on a display device. The driving assistance method according to claim 1 .

7. displaying the target route for the automatic driving on a display device; The driving assistance method according to claim 1 .

8. A driving assistance device that assists automatic driving of a vehicle based on a teacher driving operation performed by a user, a storage device that stores a movement route from a travel start position of the teacher travel to a parking position of the teacher travel, the teacher travel includes two turns in the travel direction of the vehicle between a travel start position of the teacher travel and a parking position of the teacher travel, generating a target route from a travel start position of the automatic driving to a parking position of the automatic driving based on the teacher travel; the target route in the automatic driving is shorter than the travel route in the teacher driving, the automatic driving does not include a change in the driving direction of the vehicle from a driving start position of the automatic driving to a parking position of the automatic driving, In the automatic traveling, the vehicle travels along the target route at a non-zero vehicle speed. Driving assistance device.

9. The change in the traveling direction of the vehicle is either a change in the traveling direction of the vehicle from forward to reverse, or a change in the traveling direction of the vehicle from reverse to forward. The driving assistance device according to claim 8.

10. the travel path includes a first turning position and a second turning position in a turning of the traveling direction of the vehicle, the first turning point is a position on the travel path that is closer to a travel start position of the instructor travel than the second turning point, the second turning position is a position on the travel path that is closer to a parking position for the teacher travel than the first turning position, the target route does not include the first turning position and the second turning position on the target route; The driving assistance device according to claim 8.

11. the travel path includes a first turning position and a second turning position in a turning of the traveling direction of the vehicle, the first turning point is a position on the travel path that is closer to a travel start position of the instructor travel than the second turning point, the second turning position is a position on the travel path that is closer to a parking position for the teacher travel than the first turning position, the target route includes, on the target route, a first position between a travel start position of the teacher travel and the first turning position, and a second position between a parking position of the teacher travel and the second turning position, the target route includes a route connecting the first position and the second position; The driving assistance device according to claim 8.

12. the travel path includes a first turning position and a second turning position in a turning of the traveling direction of the vehicle, the first turning point is a position on the travel path that is closer to a travel start position of the instructor travel than the second turning point, the second turning position is a position on the travel path that is closer to a parking position for the teacher travel than the first turning position, the target route includes, on the target route, a first position between a travel start position of the teacher travel and the first turning position, and a second position between a parking position of the teacher travel and the second turning position, the target route includes a route connecting the first position and the second position; the target route includes a position on a route connecting the first position and the second position that has not been traveled during the teacher travel; The driving assistance device according to claim 8.

13. the driving assistance device displays the travel route in the instructor driving on a display device. The driving assistance device according to claim 8.

14. the driving assistance device displays the target route for the automated driving on a display device. The driving assistance device according to claim 8.

Images