JP2018083480A - Parking support method and parking support device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To determine an adequate target parking position restraining a self vehicle from protruding from a parking frame when the self vehicle is parked while securing distance from another vehicle parked in an adjacent position.SOLUTION: In a parking support method which supports parking when the self vehicle parks inside the parking frame, it presumes parkable space PS of a self vehicle A except occupied space by the other vehicle B which has been parking from a parking zone S already on the basis of space recognition information that the self vehicle A acquires at a runway R along the parking zone S. It extracts recognition information of a white line SL of the parking frame which the self vehicle A plans to park on the basis of image information that the self vehicle A acquires at the runway R along the parking zone S. It determines a target parking position Pt' on the basis of the parkable space PS and provides limit values L1 and L2 of a parking frame SF on the basis of the recognition information of the white line SL of the parking frame and then determines a target parking position Pt by adding position limit by the limit values L1 and L2 with respect to the determined target parking position Pt'.SELECTED DRAWING: Figure 4

Description

  The present disclosure relates to a parking support method and a parking support device that support parking when the vehicle is parked in a parking frame.

  Conventionally, as a parking support technology, it has a region detection unit for acquiring parking space information and a line detection unit for acquiring parking section information so that parking support according to the actual parking environment can be performed. What was comprised is known (refer patent document 1). The conventional parking assist device calculates the parking target position using the detection result of the parking area when the parking area is detected and the parking lot line is detected and the parking lot line is located outside the parking area. I am doing so.

JP 2014-069722 A

  However, in the conventional apparatus, only when the parking lot line is not detected or the parking lot line is detected outside the parking area, the detection result of the parking area is reflected in the parking target position. Yes. For this reason, when a parking lot line is detected in the parking area, the detection result of the parking lot line is reflected in the parking target position, and the central position of the parking lot line is set as the parking target position of the own vehicle. Therefore, when the parking lot line is detected within the parking area, the parking position of the adjacent vehicle is adjacent to the own vehicle in the parking scene where the parking position of the own vehicle is close to the parking lot line (the parking frame of the own vehicle). There was a problem that the distance to other vehicles parked at the location would be close.

  The present disclosure has been made paying attention to the above-mentioned problem. When the vehicle is parked, an appropriate target parking that suppresses the protrusion from the parking frame while securing the distance from the other vehicle parked at the adjacent position. The purpose is to determine the position.

In order to achieve the above object, the present disclosure supports parking when the vehicle is parked in the parking frame. In this parking support method, based on the space recognition information acquired by the vehicle on the runway along the parking area, the parking space of the own vehicle is estimated by excluding the occupied space by other vehicles already parked from the parking area. To do.
Based on the image information acquired by the own vehicle on the runway along the parking section, the recognition information of the parking frame white line on which the own vehicle is scheduled to be parked is extracted.
The target parking position is determined based on the available parking space, the parking frame limit value is set based on the recognition information of the parking frame white line, and the target parking position is determined by adding the position limit based on the limit value to the calculated target parking position. To do.

  In this way, the target parking position is obtained with reference to the space where parking is possible, and the position is limited by the parking frame as necessary, so when the vehicle is parked, the distance from the other vehicle parked at the adjacent position is secured. However, it is possible to determine an appropriate target parking position that suppresses the protrusion from the parking frame.

1 is a system configuration diagram illustrating a parking assistance system to which a parking assistance method and a parking assistance device of Example 1 are applied. It is a parking assistance control block diagram which shows the control block structure centering on ECU for parking assistance calculation among parking assistance systems. It is explanatory drawing which shows the main words used in parking assistance, and its definition. It is a flowchart which shows the flow of the parking assistance control process performed in the sensor information processing part of Example 1, the image information processing part, ECU for parking assistance calculation, and ECU for vehicle control. In the parking assistance system of Example 1, the target parking position determination effect | action in the parking scene in which the 2nd other vehicle and the 4th other vehicle have parked in the both-sides position adjacent to the parking frame which the own vehicle is planning to park is shown. It is an operation explanatory view. In the parking support system of the first embodiment, the first other vehicle is parked at the jump position on one side and the fourth other vehicle is parked at the adjacent position on the other side of both sides of the parking frame on which the host vehicle is scheduled to park. It is effect | action explanatory drawing which shows the target parking position determination effect | action in the parking scene which exists. It is an operation explanatory view showing the target parking position determination operation in the parking scene where the other vehicle is not parked on both sides of the parking frame in which the own vehicle is scheduled to park in the parking support system of the first embodiment.

  Hereinafter, the best mode for realizing the parking support method and the parking support device of the present disclosure will be described based on the first embodiment shown in the drawings.

First, the configuration will be described.
The parking assistance method and the parking assistance device in the first embodiment are a parking assistance system that performs an autonomous parking operation to a target parking position along a generated parking route when the host vehicle parks in a parking frame of a parking lot. This is applied to the self-driving vehicle installed. Hereinafter, the configuration of the first embodiment will be described by being divided into “entire system configuration”, “parking support control configuration”, “main words and definitions used in parking support”, and “parking support control processing configuration”.

[Overall system configuration]
FIG. 1 is an overall system configuration diagram illustrating a parking support system to which the parking support method and the parking support apparatus according to the first embodiment are applied. The overall system configuration will be described below with reference to FIG.

  As shown in FIG. 1, the parking assist system includes a space recognition sensor 21, a white line recognition sensor 22, a steering angle sensor 23, a wheel speed sensor 24, a vehicle state quantity sensor 25, and a sensor information processing unit 26. And an image information processing unit 27. Furthermore, the ECU 28 includes a parking assist calculation ECU 28, a vehicle control ECU 29, an actuator 30, an image processing unit 31, and a display monitor 32.

  The space recognition sensor 21 is a sensor that is provided in the host vehicle, for example, for acquiring spatial information data around the host vehicle during movement / stop of the runway along the parking section. The space recognition sensor 21 is typified by a commonly used laser range finder. The laser range finder is an apparatus that can irradiate a target with an infrared laser and measure the distance to the target with the degree of reflection thereof, and can acquire distance information of the detected object as point cloud information.

  The white line recognition sensor 22 is provided in the own vehicle, for example, a short-distance camera for acquiring image information data including a parking frame white line in the vicinity of the own vehicle while the runway along the parking section is moving / stopped. It is. The white line recognition sensor 22 is typified by a generally used around view monitor. The around view monitor captures surrounding images in four directions in the front, rear, left, and right directions of the own vehicle with a camera, and obtains an overhead image around the own vehicle by performing viewpoint conversion processing on the captured images.

  The steering angle sensor 23 is a sensor that acquires steering angle information for estimating the current tire turning angle amount and the like. The steering angle sensor signal is output to the parking assist calculation ECU 28 and the vehicle control ECU 29.

  The wheel speed sensor 24 is a sensor that acquires wheel speed information for calculating the speed and travel distance of the vehicle. The wheel speed sensor signal is output to the parking assist calculation ECU 28 and the vehicle control ECU 29.

  The vehicle state quantity sensor 25 is a sensor that measures a state quantity of a vehicle represented by yaw rate. The vehicle state quantity sensor signal is output to the parking assist calculation ECU 28.

  The sensor information processing unit 26 inputs point cloud information that is space information from the space recognition sensor 21 and estimates a parking space and a road boundary based on the input information. The parking space and runway boundary estimation information is output to the parking assist calculation ECU 28.

  The image information processing unit 27 receives the image information including the parking frame white line from the white line recognition sensor 22, detects the edge of the parking frame white line based on the input information, and detects the position of the parking frame white line and the direction of the parking frame white line. Is estimated.

  The parking assist calculation ECU 28 calculates a target parking position based on the parking available space from the sensor information processing unit 26. A parking frame limit value is provided based on the parking frame white line recognition information from the image information processing unit 27. Then, a target parking position is determined by adding a position limit based on a limit value to the target parking position. Further, when the target parking position is determined, a backward turning position of forward → backward is searched, and a parking path of the own vehicle from the current own vehicle position to the target parking position is generated. Then, the generated parking route is transmitted to the vehicle control ECU 29.

  The vehicle control ECU 29 inputs the parking route information from the parking assist calculation ECU 28, the steering angle sensor signal from the steering angle sensor 23, and the wheel speed sensor signal from the wheel speed sensor 24. Then, the actuator 30 is driven in accordance with a guidance control command for moving the host vehicle along the generated parking route. As the actuator 30, a drive actuator, a braking actuator, a turning actuator, a range position selection actuator, and the like are provided.

  The image processing unit 31 inputs parking route information from the parking assist calculation ECU 28. The monitor displays a map of the parking lot including the parking lot including the road and the parking frame of the vehicle, and image processing signals based on the current vehicle position / posture, the target parking position / posture, and the parking route on the parking lot map. 32.

  When the image processing signal is input from the image processing unit 31, the display monitor 32 displays the current vehicle position / posture, the target parking position / posture, the parking route, and the like on the parking lot map on the monitor screen. Here, the display monitor 32 is an information display unit for confirming whether or not the host vehicle is automatically guided along the generated parking route, and generally uses a liquid crystal display for navigation provided in the vehicle. can do.

  When the display monitor 32 is switched to a vehicle that does not have the vehicle control ECU 29 or manual driving, the display monitor 32 performs parking by the driver driving operation so that the vehicle reaches the target parking position along the generated parking route. You may use as a driving assistance information display part to assist. In addition, when the driving support information display unit is used, a voice guidance that guides the steering direction or the like so that the vehicle moves along the guidance route using a speaker that is already installed in the vehicle is more finely tuned. Parking guidance is possible.

[Parking assistance control configuration]
FIG. 2 shows a control block configuration centering on the parking assist calculation ECU 28 in the parking assist system of FIG. Hereinafter, a parking assistance control configuration will be described with reference to FIG.

  As shown in FIG. 2, the sensor information processing unit 26 includes a parking space estimation unit 261 and a runway boundary estimation unit 262. The parking space estimation unit 261 inputs the point cloud information from the space recognition sensor 21, and the vehicle is already parked from the parking area based on the space recognition information acquired on the runway along the parking area. Estimate the parking space of your vehicle, excluding the space occupied by other vehicles. The road boundary estimation unit 262 receives the point cloud information from the space recognition sensor 21 and estimates the direction of the road and the road boundary from the line of other parked vehicles.

  As illustrated in FIG. 2, the image information processing unit 27 includes a parking frame white line recognition unit 271 that inputs image information including a parking frame white line from the white line recognition sensor 22. The parking frame white line recognizing unit 271 extracts the parking frame white line on which the vehicle is scheduled to park based on the image information acquired by the own vehicle on the runway along the parking area, and detects the edge of the parking frame white line. To estimate the position of the parking frame white line and the direction of the parking frame white line.

  As shown in FIG. 2, the parking assist calculation ECU 28 includes a target parking position calculation unit 281, a limit value calculation unit 282, a target parking position determination unit 283, a target parking posture calculation unit 284, and the current host vehicle position. A posture information acquisition unit 285, a reverse turn-back position search unit 286, and a parking route generation unit 287.

  The target parking position calculation unit 281 calculates the median value of the parking available space as the target parking position when obtaining the target parking position based on the parking available space.

  When the limit value calculation unit 282 provides the limit value based on the recognition information of the parking frame white line, the limit value calculation unit 282 sets the both end positions of the allowable offset width of the target parking position where the vehicle can park without crossing the parking frame white line. The first limit value and the second limit value are used.

  When determining the target parking position, the target parking position determination unit 283 determines the median value as the target parking position if the median value of the parking space is within the allowable offset width. On the other hand, when the median value of the parking space deviates from the allowable offset width, a limit value close to the median value is determined as the target parking position among the first limit value and the second limit value. When the parking space cannot be estimated, the median value of the first limit value and the second limit value is determined as the target parking position.

  The target parking posture calculation unit 284 indicates the target parking posture when the host vehicle is parked at the target parking position, the estimation information of the road boundary from the road boundary estimation unit 262, and the parking frame white line from the parking frame white line recognition unit 271. Calculation based on the recognition information. At this time, the parking lot structure is estimated from the estimated information of the road boundary and the recognition information of the parking frame white line, and the calculation method is selected according to the column, parallel, and diagonal parking forms obtained from the estimation result of the parking lot structure. To do. Further, when the target parking posture is calculated, the reliability of the road boundary estimation information and the recognition information of the parking frame white line are compared, and the higher reliability is selected. For example, when the parking structure is determined to be an oblique parking form based on the estimation result of the parking lot structure, the recognition information of the parking frame white line is preferentially selected.

  The current vehicle position / posture information acquisition unit 285 inputs sensor information from the steering angle sensor 23, the wheel speed sensor 24, and the vehicle state quantity sensor 25, and acquires detection information of the current vehicle position and vehicle posture. To do.

  The reverse turn-back position search unit 286 includes a road boundary from the road boundary estimation unit 262, a target parking position from the target parking position determination unit 283, a target parking posture from the target parking posture calculation unit 284, and the current vehicle position. Input each information of the current vehicle position / posture from the posture information acquisition unit 285. Then, a suitable position as a reverse turning position from forward to backward is searched in the runway so that the target parking posture is reached when the host vehicle reaches the target parking position.

  The parking route generation unit 287 adds the backward return position information from the backward return position search unit 286 to the input information to the backward return position search unit 286. Then, a parking route is generated based on a forward route from the current vehicle position and vehicle posture to the reverse return position and a reverse route from the reverse return position to the target parking position and the target parking posture. When generating this parking route, the route is drawn so that the vehicle does not interfere with the route boundary line between the running road and the parking frame of the vehicle.

[Key words and definitions used in parking assistance]
FIG. 3 shows main phrases used in parking assistance and their definitions. Hereinafter, main phrases and their definitions will be described with reference to FIG.

  As a structure of a parking lot, there are a structure in a parallel parking form, a structure in a parallel parking form, a structure in an oblique parking form, and the like. In Example 1, the parking lot by the structure by the most common parallel parking form is taken as an example among these parking forms.

  As shown in FIG. 3, the parking lot PK in the parallel parking mode includes a runway R on which an incoming vehicle and an outgoing vehicle run, and a parking section S provided along the runway R as a parking area for the vehicle. The track R is set by two track boundaries RL and RR having a predetermined track width. The parking section S is set with the runway R via the runway boundary RL, and a plurality of parking frames SF (which become a parking space for one vehicle by drawing a parking frame white line SL (SL1, SL2, SL3, SL4)) SF1, SF2, SF3, SF4). Each parking frame SF (SF1, SF2, SF3, SF4) is configured by a parking frame depth and a parking frame width.

  In this parking lot PK, when the own vehicle A plans to park in one parking frame SF, the own vehicle can be parked in advance in order to keep the distance from the other vehicles B (B2, B4) parked adjacent to each other. It is necessary to recognize a secure parking space PS. This “parkable space PS” is a space excluding the space occupied by other vehicles B (B2, B4) that are already parked from the parking section S including the parking frame SF where the vehicle A is scheduled to park. Define.

  In addition, when the host vehicle A schedules parking in one parking frame SF, it is necessary to provide a limit value in the vehicle width direction in which the host vehicle A does not protrude from the parking frame SF. As shown in FIG. 3, when the allowable offset width ΔW of the target parking position Pt that allows the vehicle A to park without crossing the parking frame white line SL is set as shown in FIG. 3, both end positions of the allowable offset width ΔW are set. Are defined as a first limit value L1 and a second limit value L2, respectively.

  When the parking route Rin that avoids interference with the surrounding obstacle environment is generated by the vehicle A, the current vehicle position Pn and the vehicle attitude Dn, the reverse turn-back position Pc, and the target parking position Pt for completing the parking The target parking posture Dt needs to be determined in advance. It is assumed that the parking route Rin includes a forward route Rinf and a backward route Rinr.

  The “current vehicle position Pn” is defined as the rear axle center position that connects the left and right rear wheels of the vehicle A at the start position for entry on the road. “Current vehicle posture Dn” is defined by the angle θ formed by the center axis in the front-rear direction passing through the center of gravity of the vehicle A at the start position for warehousing on the road with respect to the y-axis (axis in the parking frame width direction). . For example, when the host vehicle A is in a direction parallel to the road R, the angle θ representing the “current host vehicle posture Dn” is θ = 0 °.

The “reverse turning position Pc” is defined as a rear axle center position that connects the left and right rear wheels of the own vehicle A at a position that is reached by moving forward along the forward path Rinf from the current own vehicle position Pn.
The “reverse turning position Pc” also has the meaning of the movement start position when the turning direction is turned back and the vehicle moves backward along the backward path Rinr.

  The “target parking position Pt” is defined as the center position of the rear axle connecting the left and right rear wheels of the own vehicle A at the warehousing completion position when the own vehicle A completes the warehousing movement along the backward route Rinr. The “target parking posture Dt” indicates that when the vehicle A completes warehousing movement along the reverse route Rinr, the longitudinal center axis passing through the center of gravity of the vehicle A is relative to the y axis (axis in the parking frame width direction). This is defined by the angle θ to be formed. For example, when the host vehicle A stops in a direction parallel to the parking frame depth direction at the warehousing completion position, the angle θ representing the “target parking posture Dt” is θ = 90 °.

[Parking assistance control processing configuration]
FIG. 4 is a flowchart illustrating a flow of a parking support control process executed by the sensor information processing unit 26, the image information processing unit 27, the parking support calculation ECU 28, and the vehicle control ECU 29 according to the first embodiment. Hereinafter, each step of FIG. 4 showing a parking assistance control process structure is demonstrated.

In step S1, when the own vehicle A that has traveled on the runway R of the parking lot PK has moved to a position near the parking frame SF that is scheduled to be parked, the space recognition sensor 21 mounted on the own vehicle A uses the space recognition sensor 21 Is acquired, and the process proceeds to step S2.
Here, the spatial information data acquired by the spatial recognition sensor 21 is input to the sensor information processing unit 26.

In step S2, following the acquisition of the spatial information data in step S1, the direction (running road boundary RL, RR) of the road R on which the vehicle A can travel is estimated based on the acquired spatial information data, Proceed to S3.
Here, “estimation of the road boundary” is performed by the road boundary estimation unit 262 of the sensor information processing unit 26.

In step S3, following the estimation of the direction of the road (running road boundary) in step S2, a parking available space PS, which is a space in which the vehicle A can park, is estimated based on the acquired spatial information data. Proceed to S4.
Here, “estimation of the parking space PS” is performed by the parking space estimation unit 261 of the sensor information processing unit 26. For example, the space excluding the occupied space by the already parked other vehicle B is estimated from the parking section S including the parking frame SF where the host vehicle A is scheduled to park.

In step S4, following the estimation of the parking space PS in step S3, the target parking position Pt ′ is calculated based on the parking space PS, and the process proceeds to step S5.
Here, “target parking position Pt ′” represents a target parking position calculated value based on the estimation of the parking available space PS with respect to the finally determined target parking position Pt. “Calculation of the target parking position Pt ′” is performed by the target parking position calculation unit 281 of the parking assist calculation ECU 28, and the median value of the parking space PS is calculated as the target parking position Pt ′.

In step S5, following the calculation of the target parking position Pt in step S4, image information from the white line recognition sensor 22 is acquired, and the process proceeds to step S6.
Here, the image information acquired by the white line recognition sensor 22 is input to the image information processing unit 27.

In step S6, following the white line recognition sensor information acquisition in step S5, the parking frame white line information (edge detection) that is the parking frame white line SL surrounding the parking frame SF in which the vehicle A is scheduled to be parked is extracted. Proceed to step S7.
Here, “extraction of parking frame white line information” is performed by the parking frame white line recognition unit 271 of the image information processing unit 27.

In step S7, following the extraction of the parking frame white line information in step S6, the limit values (the first limit value L1 and the second limit value L2) of the parking frame SF that the vehicle A is planning to park are calculated. Proceed to step S8.
Here, the “calculation of limit value” is performed by the limit value calculation unit 282 of the parking assist calculation ECU 28. For example, the calculation of the limit value of the parking frame SF is based on the fact that the both end positions of the allowable offset width ΔW of the target parking position Pt that the vehicle A can park without crossing the parking frame white line SL of the parking frame SF is the first limit. The value L1 and the second limit value L2 are calculated.

  In step S8, following the calculation of the limit value of the parking frame in step S7, it is determined whether the parking space PS can be estimated or whether the parking is not oblique due to the direction of another vehicle already parked. If YES (parkable space PS can be estimated and not diagonally parked), the process proceeds to step S9. If NO (parkable space PS cannot be estimated or diagonally parked), the process proceeds to step S11.

  In step S9, following the determination that the parking space PS can be estimated in step S8 and that the parking is not oblique, the target parking position Pt ′ calculated in step S4 is the limit values L1, L2 of the parking frame SF. It is judged whether it is within. If YES (target parking position Pt ′ ≦ allowable offset width ΔW), the target parking position Pt ′ is set as the target parking position Pt as a determined value as it is, and the process proceeds to step S12. If NO (target parking position Pt ′> allowable offset width ΔW), the process proceeds to step S10.

  In step S10, following the determination that the target parking position Pt ′> the allowable offset width ΔW in step S9, the target parking position Pt as the determined value is set to the target value of the first limit value L1 and the second limit value L2. The limit value on the side closer to the parking position Pt ′ is changed, and the process proceeds to step S12.

In step S11, following the determination that the available parking space PS cannot be estimated in step S8 or that the parking is oblique, the target parking position Pt as the determined value is set to the first limit value L1 and the second limit value L2. And the process proceeds to step S12.
Here, the calculation of the target parking position Pt may be performed using the median value of the parking frame, which is the basic parking position in the parking frame SF, instead of using the median value of the first limit value L1 and the second limit value L2.
In addition, the process which determines the target parking position Pt from step S8 to step S11 is performed in the target parking position determination part 283 of ECU28 for parking assistance calculation.

In step S12, following the determination of the target parking position Pt in step S9, step S10, or step S11, estimation of reliability between the estimation information of the road boundary RL, RR and the recognition information of the parking frame white line SL. Then, the comparison is performed, and the process proceeds to step S13.
Here, “estimating the reliability” means that when calculating the target parking posture Dt, for example, the parking lot structure is estimated from the estimation information of the road boundary RL, RR and the recognition information of the parking frame white line SL, It means estimating the superiority or inferiority of the reliability of two pieces of information according to the parking form obtained from the estimation result of the parking lot structure. "Comparison of reliability" means that when calculating the target parking posture Dt, which information is more reliable among the estimated information of the road boundary RL, RR and the recognition information of the parking frame white line SL To compare.

  For example, when the parking structure is determined to be an oblique parking form based on the estimation result of the parking lot structure, the recognition information of the parking frame white line SL is preferentially selected when calculating the target parking posture Dt. In addition, when the number of other vehicles B parked as the spatial information is small and the variation in the direction detection of the runway R is large, the recognition information of the parking frame white line SL is preferentially selected when calculating the target parking posture Dt. . When the accuracy of image detection cannot be ensured due to the light environment or the like, when calculating the target parking posture Dt, the estimated information of the road boundaries RL and RR is preferentially selected.

In step S13, following the estimation and comparison of the reliability in step S12, based on the estimation and comparison of the reliability between the estimation information of the road boundary RL and RR and the recognition information of the parking frame white line SL, the target parking posture Dt Is calculated, and the process proceeds to step S14.
Here, when calculating the target parking posture Dt, if the reliability of the estimation information of the road boundary RL, RR and the reliability of the recognition information of the parking frame white line SL are given superiority, use the priority information It will be calculated. For example, when the parking form is obtained from the estimation result of the parking lot structure, the calculation method of the target parking posture Dt is selected depending on whether it is a parallel parking form, a parallel parking form, or an oblique parking form. Will be.
In addition, the calculation process of the target parking posture Dt in step S12 to step S13 is performed by the target parking posture calculation unit 284 of the parking assist calculation ECU 28.

In step S14, following the calculation of the target parking posture in step S13, the backward turn-back position Pc is searched, and the process proceeds to step S15.
Here, the “reverse turning position Pc” is, for example, a position on the route when the host vehicle A is moved forward by turning right, and the target parking position Pt / target parking posture Dt is obtained by moving backward by turning left. Decide by searching for a location that can reach.
Note that the search for the reverse turning position Pc in step S14 is performed by the reverse turning position searching unit 286 of the parking assist calculation ECU 28.

In step S15, following the search for the reverse turning position in step S14, a parking route Rin is generated, and the process proceeds to step S16.
Here, the “parking route Rin” is a forward route Rinf that connects the current vehicle position Pn / own vehicle posture Dn and the reverse turning position Pc with a smooth arc, a reverse turning position Pc, the target parking position Pt / target parking. The backward path Rinr that connects the posture Dt with a smooth arc is generated. The generation of the parking route Rin in step S15 is performed by the parking route generation unit 287 of the parking assist calculation ECU 28.

In step S16, following the generation of the parking route in step S15 or the determination that the target parking position has not been reached in step S18, the vehicle control target value is calculated, and the process proceeds to step S17.
Here, the “vehicle control target value” refers to a vehicle speed target value, a steering angle target value, a range position target value, and the like in vehicle guidance control for guiding and moving the host vehicle A along the generated parking route Rin.

  In step S17, following the calculation of the vehicle control target value in step S16, vehicle control is performed according to the calculated target value, and the process proceeds to step S18.

In step S18, following the vehicle control in step S17, it is determined whether or not the target parking position Pt has been reached. If YES (target parking position reached), the process proceeds to the end. If NO (target parking position not reached), the process returns to step S16.
The vehicle guidance control in steps S16 to S18 is performed by the vehicle control ECU 29 that receives the parking route information from the parking assist calculation ECU 28.

Next, the operation will be described.
The operation of the first embodiment will be described by being divided into “parking support control processing operation”, “target parking position determination operation in each parking scene”, and “parking support control characteristic operation”.

[Parking assistance control processing action]
Hereinafter, the parking assistance control processing operation will be described based on the flowchart of FIG.
First, when the own vehicle A autonomously travels on the road S toward the parking frame SF and stops at the parking support control start position close to the parking frame SF, in the flowchart of FIG. 4, step S1, step S2, step S3, step S4. → Proceed to step S5 → step S6 → step S7. That is, in step S1, when the vehicle A is parked, space information data around the vehicle is acquired by the space recognition sensor 21 mounted on the vehicle A. In step S2, the direction of the road (running road boundaries RL, RR) is estimated based on the acquired spatial information data. In step S3, a parking space PS, which is a space in which the vehicle A can park, is estimated based on the acquired road boundary RL, RR and position information of the other vehicle B. In step S4, the target parking position Pt ′ is calculated based on the parking space PS. In step S5, image information from the white line recognition sensor 22 is acquired. In step S6, parking frame white line information (edge detection) of the parking frame SF in which the vehicle A is scheduled to be parked is extracted. In step S7, the limit values (first limit value L1 and second limit value L2) of the parking frame SF in which the vehicle A is scheduled to be parked are calculated.

  When the parking space PS can be estimated, the parking is not oblique, and the target parking position Pt ′ is within the limit values L1 and L2 of the parking frame SF, steps S7 to S8 → step S9 → step. Proceed to S12. That is, by proceeding directly from step S9 to step S12, the target parking position Pt as the determined value is set to the value as the target parking position Pt ′ obtained by the calculation.

  On the other hand, when the available parking space PS can be estimated and the target parking position Pt ′ is out of the limit values L1 and L2 of the parking frame SF, although it is not oblique parking, from step S7 to step S8 → step S9 → step. The process proceeds from S10 to step S12. That is, in step S10, the target parking position Pt as the determined value is changed to a limit value closer to the target parking position Pt 'between the first limit value L1 and the second limit value L2.

  When the parking space PS cannot be estimated or when the parking is oblique, the process proceeds from step S7 to step S8 → step S11 → step S12. That is, in step S11, the target parking position Pt as the determined value is calculated from the median value of the first limit value L1 and the second limit value L2.

  When the target parking position Pt is determined, the process proceeds from step S12 to step S13 → step S14 → step S15. That is, in step S12, the estimation and comparison of the reliability between the estimated information of the road boundaries RL and RR and the recognition information of the parking frame white line SL are performed. In step S13, the target parking posture Dt is calculated based on the reliability estimation and comparison between the estimation information of the road boundary RL and RR and the recognition information of the parking frame white line SL. In step S14, the reverse turning position Pc is searched. In step S15, a parking route Rin is generated.

  When the parking route Rin is generated, the process proceeds from step S15 to step S16 → step S17 → step S18, and the process of step S16 → step S17 → step S18 is repeated until the target parking position Pt is reached. That is, in step S16, a vehicle control target value is calculated. In step S17, vehicle control is performed according to the calculated target value. In step S18, it is determined whether the target parking position Pt has been reached. Then, when the target parking position Pt is reached, the process proceeds from step S18 to the end, and the parking support control ends.

[Determination of target parking position in each parking scene]
First, the determination operation of the target parking position Pt in the parking scene where the second other vehicle B2 and the fourth other vehicle B4 are parked at both side positions adjacent to the parking frame SF3 where the own vehicle A is scheduled to park, This will be described with reference to FIG.

  In the parking scene where the second other vehicle B2 and the fourth other vehicle B4 are parked on both sides adjacent to the parking frame SF3 where the host vehicle A is scheduled to park, the target parking position Pt 'is the limit of the parking frame SF3. The condition that it is within the value is satisfied. For this reason, when the median value CL of the available parking space PS is calculated as the target parking position Pt ′, the finally determined target parking position Pt is a value that remains as the target parking position Pt ′ obtained by the calculation. The

  Thus, in the parking scene in which the second other vehicle B2 and the fourth other vehicle B4 are parked at both side positions adjacent to the parking frame SF3, the distance between the adjacent other vehicles B2 and B4 in the parking space PS is the largest. A parking position that can be secured is determined as the target parking position Pt. For this reason, when generating the parking route Rin to the parking frame SF3, it is possible to easily generate a route that avoids interference with the other vehicles B2 and B4. Moreover, when leaving the own vehicle A from the parking frame SF3 while the other vehicles B2 and B4 remain parked, it is possible to easily generate a delivery route that avoids interference with the other vehicles B2 and B4.

By the way, in the parking scene where other vehicles are parked in each side position adjacent to the parking frame of the own vehicle, as shown by the phantom line in FIG. In the case of the comparative example with the target parking position, the own vehicle A ′ and the fourth other vehicle B4 are at the interval W1.
On the other hand, in the same parking scene, in the case of Example 1, the own vehicle A and the 4th other vehicle B4 become the space | interval W2 (> W1). In other words, since the target parking position Pt is determined based on the parking space PS, the parking frame width of the parking frame SF3 of the own vehicle A is utilized to the maximum extent, and sufficient between the own vehicle A and the other vehicle B4. A distance can be secured.

  Next, the first other vehicle B1 is parked at the jump position on one side of both sides of the parking frame SF3 where the own vehicle A is scheduled to park, and the fourth other vehicle B4 is parked at the adjacent position on the other side. The target parking position determination operation in the parking scene will be described with reference to FIG.

  In a parking scene in which the first other vehicle B1 is parked at the jump position on one side of the parking frame SF3 on which the own vehicle A is scheduled to park and the fourth other vehicle B4 is parked at the adjacent position on the other side. The condition that the target parking position Pt ′ is within the limit value of the parking frame SF3 is not satisfied. For this reason, even if the median CL of the parking space PS is calculated as the target parking position Pt ′, the target parking position Pt finally determined is the target parking position among the first limit value L1 and the second limit value L2. The first limit value L1 closer to the position Pt ′ is changed.

  That is, as described above, based on the parking space PS as a reference, the median value CL in the parking space PS (target value that secures the distance from the other adjacent vehicles most) is determined as the target parking position Pt in principle. . However, if the target parking position Pt is determined only by the parking space PS, there is a possibility that the target parking position Pt is set such that the parking frame SF3 of the own vehicle A protrudes and stops depending on the parking position of another adjacent vehicle. There is. For this reason, the edge position information of the parking frame white line is used as the limit value of the target parking position Pt, and the first limit value L1 and the second limit value L2 are set as threshold values of the target parking position from the edge position information. deep. Thereby, when the median value CL in the parking space PS deviates from the allowable offset width ΔW by the first limit value L1 and the second limit value L2, the own vehicle A stops regardless of the parking position of the adjacent other vehicle B. It is possible to park in the parking frame SF3. Then, by changing the final target parking position Pt to a limit value closer to the target parking position Pt ′, while the vehicle A is parked in the parking frame SF3, the adjacent fourth other vehicle The maximum distance from B4 can be secured.

  Next, the target parking position determination operation in a parking scene in which the other vehicle B is not parked on both sides of the parking frame SF3 where the host vehicle A is scheduled to park will be described with reference to FIG.

  In the parking scene where the other vehicle B is not parked on both sides of the parking frame SF3 where the own vehicle A is scheduled to park, the parking space is parked because there is no occupied space information of the other vehicle B parked in the parking section S. The possible space PS itself cannot be estimated. For this reason, the target parking position Pt finally determined is calculated by the median value of the first limit value L1 and the second limit value L2. Note that, for example, if the edges at both ends of the parking frame white line SL3 surrounding the parking frame SF3 are detected, the target parking position Pt may be determined by the median value at both ends. Furthermore, as long as an edge on only one side of the parking frame white line SL3 is detected, the target parking position Pt may be determined from a general parking frame standard (for example, 2.5 m).

Thus, in the parking scene where the other vehicle B is not parked on both sides of the parking frame SF3 where the host vehicle A is scheduled to park, the center of the parking frame SF3 is set according to the normal target parking position setting method. The value can be determined as the target parking position Pt.
In addition, when the form of the parking lot PK is an oblique parking form, the parking space PS can be estimated, but the space estimation accuracy is lowered. For this reason, the median value of the parking frame SF is determined as the target parking position Pt, as in the case where the parking space PS cannot be estimated.

[Characteristics of parking support control]
In Example 1, target parking position Pt 'is calculated | required based on the parking possible space PS of the own vehicle A estimated by remove | excluding the occupied space by the other vehicle B which has already parked from the parking area S. FIG. The limit values L1 and L2 of the parking frame SF are provided based on the recognition information of the parking frame white line SL that the host vehicle A is planning to park. Then, the target parking position Pt is determined by adding position restrictions based on the limit values L1 and L2 to the obtained target parking position Pt ′.
That is, by obtaining the target parking position Pt ′ on the basis of the parking space PS, the distance from the other vehicle B parked at the adjacent position is not too close. The distance from the other vehicle B parked in the vehicle is secured. Then, by restricting the position of the target parking position Pt ′ obtained based on the parking space PS with the parking frame SF as necessary, the vehicle A does not protrude beyond the parking frame SF. As a result, when the vehicle A is parked, an appropriate target parking position Pt corresponding to the parking environment can be determined.

In Example 1, when calculating | requiring target parking position Pt 'based on parking possible space PS, the median value CL of parking possible space PS is made into target parking position Pt'.
That is, the median value CL of the parking space PS is the position where the possibility of interference with the other vehicle B parked at the adjacent position is the lowest. Accordingly, when the host vehicle A is parked, the host vehicle A can be parked at a position where the possibility of interference with the other vehicle B parked at the adjacent position is the lowest.

In the first embodiment, when the limit value is set based on the recognition information of the parking frame white line SL, both end positions of the allowable offset width ΔW of the target parking position Pt that can be parked without the host vehicle A straddling the parking frame white line SL. Are a first limit value L1 and a second limit value L2, respectively.
That is, the first limit value L1 and the second limit value L2 are set to limit values at which the own vehicle A does not protrude from the parking frame SF. Therefore, when the host vehicle A is parked, the host vehicle A can be parked within the frame of the parking frame SF while securing a distance from the other vehicle B parked at the adjacent position.

In Example 1, when determining the target parking position Pt, if the median value CL of the parking space PS is within the allowable offset width ΔW, the median value CL is determined as the target parking position Pt. When the median value CL of the parking space PS deviates from the allowable offset width ΔW, a limit value close to the median value CL is determined as the target parking position Pt among the first limit L1 and the second limit value L2.
In other words, the offset width for determining the median value CL of the parking space PS as the target parking position Pt, and the allowable offset width ΔW for maximizing the effective use of the parking frame width of the parking frame SF that the vehicle A is planning to park. It is said. Therefore, when the own vehicle A is parked, the distance between the own vehicle A and the other vehicle B can be ensured to the maximum by effectively utilizing the parking frame width of the parking frame SF and determining the target parking position Pt. It becomes possible.

In the first embodiment, when the parking available space PS cannot be estimated, the median value of the parking frame SF where the vehicle A is scheduled to park is determined as the target parking position Pt.
That is, even when the parking space PS cannot be estimated, when the recognition information of the parking frame white line SL is acquired, the median value of the parking frame SF is set as the target parking position Pt. Therefore, even when there is no other vehicle B parked at the adjacent position and information on the parking space PS cannot be acquired, the own vehicle A can be parked within the parking frame SF.

In the first embodiment, the target parking posture Dt when the host vehicle A is parked at the target parking position Pt, the estimated information of the runway boundaries RL and RR estimated from the arrangement of other parked vehicles B, and the parking frame white line Calculation based on SL recognition information.
That is, the target parking posture Dt can be calculated from both the estimated information of the direction of the road (running road boundaries RL, RR) from the parked other vehicle B and the recognition information of the parking white line SL that is a parking lot line. Therefore, robustness is ensured. Therefore, when the parking route Rin is generated, the generation accuracy of the parking route Rin can be improved by adding the target parking posture Dt in which robustness is ensured to the target parking position Pt.

In the first embodiment, when calculating the target parking posture Dt, the reliability of the estimated information of the road boundary RL, RR and the recognition information of the parking frame white line SL is compared, and the higher reliability is selected.
That is, according to various parking environments, the target parking posture Dt can be calculated using the most reliable information in the situation. Therefore, when calculating the target parking posture Dt, it is possible to obtain the target parking posture Dt with high accuracy.

In the first embodiment, when the target parking posture Dt is calculated, the parking lot structure is estimated from the estimation information of the runway boundaries RL and RR and the recognition information of the parking frame white line SL, and obtained from the estimation result of the parking lot structure. The calculation method is selected according to the vertical, parallel, and oblique parking modes.
That is, when calculating the target parking posture Dt, information that should be prioritized is selected according to various parking modes (column, parallel, diagonal) from the estimated information of the road boundaries RL and RR and the recognition information of the parking frame white line SL. I can do it.

In the first embodiment, when the target parking posture Dt is calculated, if the parking structure is determined to be an oblique parking form based on the estimation result of the parking lot structure, the recognition information of the parking frame white line SL is preferentially selected.
That is, in the case of diagonal parking, the target parking posture Dt can be calculated by giving priority to the recognition information of the parking frame white line SL. Therefore, the parking posture accuracy of the host vehicle A can be improved in the oblique parking mode.

Next, the effect will be described.
In the parking support method and the parking support apparatus according to the first embodiment, the effects listed below can be obtained.

(1) Assist parking when the vehicle A parks in the parking frame SF.
In this parking support method, the own vehicle A excluding the occupied space by the other vehicle B already parked from the parking section S based on the space recognition information acquired by the own vehicle A on the runway R along the parking section S. Estimated parking space PS.
Based on the image information acquired by the own vehicle A on the runway R along the parking section S, the recognition information of the parking frame white line SL that the own vehicle A is planning to park is extracted.
The target parking position Pt ′ is calculated based on the parking space PS, the limit values L1 and L2 of the parking frame SF are provided based on the recognition information of the parking frame white line SL, and the limit value L1 for the calculated target parking position Pt ′ , L2 is added to determine the target parking position Pt (S1 to S11 in FIG. 4).
For this reason, when the host vehicle A parks, the parking assistance method which determines the suitable target parking position Pt which suppresses the protrusion from the parking frame SF, ensuring the distance with the other vehicle B parked in the adjacent position. Can be provided.

(2) When the target parking position Pt ′ is obtained based on the parking space PS, the median CL of the parking space PS is set as the target parking position Pt ′ (S4 in FIG. 4).
For this reason, in addition to the effect of (1), when the vehicle A is parked, the vehicle A can be parked at a position where the possibility of interference with the other vehicle B parked at the adjacent position is the lowest.

(3) When the limit value is set based on the recognition information of the parking frame white line SL, both end positions of the allowable offset width ΔW of the target parking position Pt that the vehicle A can park without crossing the parking frame white line SL are respectively set. The first limit value L1 and the second limit value L2 are set (S7 in FIG. 4).
For this reason, in addition to the effect of (2), when the own vehicle A is parked, the own vehicle A is parked within the frame of the parking frame SF while securing a distance from the other vehicle B parked at the adjacent position. be able to.

(4) When the target parking position Pt is determined, if the median value CL of the parking space PS is within the allowable offset width ΔW, the median value CL is determined as the target parking position Pt. When the median value CL of the parking space PS deviates from the allowable offset width ΔW, a limit value close to the median value CL is determined as the target parking position Pt among the first limit L1 and the second limit value L2 (S9 in FIG. 4). , S10).
For this reason, in addition to the effect of (2) or (3), when the vehicle A is parked, the target parking position Pt is determined by effectively utilizing the parking frame width of the parking frame SF. The distance from the car B can be secured to the maximum.

(5) When the available parking space PS cannot be estimated, the median value of the parking frame SF that the vehicle A is planning to park is determined as the target parking position Pt (S8 → S11 in FIG. 4).
For this reason, in addition to the effects of (1) to (4), even when there is no other vehicle B parked in the adjacent position and information on the parking space PS cannot be acquired, the vehicle is within the parking frame SF. A can be parked.

(6) Estimated information of the road boundary RL, RR estimated from the alignment of other parked vehicles B and the parking frame white line SL of the target parking posture Dt when the vehicle A is parked at the target parking position Pt Calculation is performed based on the recognition information (S13 in FIG. 4).
For this reason, in addition to the effects of (1) to (5), when generating the parking route Rin, the target parking position Pt is added with the target parking posture Dt in which robustness is ensured, thereby improving the generation accuracy of the parking route Rin. Can be improved.

(7) When calculating the target parking posture Dt, the reliability of the estimated information of the road boundaries RL and RR and the recognition information of the parking frame white line SL is compared, and the one with the higher reliability is selected (S12 in FIG. 4). ).
For this reason, in addition to the effect of (6), when calculating the target parking posture Dt, it is possible to obtain the target parking posture Dt with high accuracy by using the most reliable information in various parking environments.

(8) When calculating the target parking posture Dt, the parking lot structure is estimated from the estimated information of the runway boundaries RL and RR and the recognition information of the parking frame white line SL, and the column obtained from the estimation result of the parking lot structure, The calculation method is selected according to the parallel and oblique parking mode (S12 in FIG. 4).
For this reason, in addition to the effect of (6) or (7), when calculating the target parking posture Dt, it is possible to select information to be prioritized according to various parking modes.

(9) When calculating the target parking posture Dt, if the parking structure is determined to be an oblique parking form based on the estimation result of the parking lot structure, the recognition information of the parking frame white line SL is preferentially selected (S12 in FIG. 4).
For this reason, in addition to the effect of (8), the parking posture accuracy of the vehicle A can be improved by deciding the target parking posture Dt by giving priority to the recognition information of the parking frame white line SL in the case of the oblique parking mode. .

(10) A parking assistance controller (a sensor information processing unit 26, an image information processing unit 27, and a parking assistance calculation ECU 28) that assists parking when the host vehicle A parks in the parking frame SF is provided.
In this parking assistance device, the parking assistance controller (the sensor information processing unit 26, the image information processing unit 27, and the parking assistance calculation ECU 28) includes a parking space estimation unit 261, a parking frame white line recognition unit 271, and a target parking position determination. Part 283.
Based on the space recognition information acquired by the own vehicle A on the runway R along the parking section S, the parking space estimation unit 261 automatically excludes the space occupied by the other vehicle B already parked from the parking section S. The parking space PS for car A is estimated.
The parking frame white line recognition unit 271 extracts the recognition information of the parking frame white line SL that the vehicle A plans to park based on the image information acquired by the vehicle A on the runway R along the parking section S. .
The target parking position determination unit 283 obtains the target parking position Pt ′ based on the parking space PS, and provides the limit values (first limit value L1, second limit value L2) based on the recognition information of the parking frame white line SL. Then, a target parking position Pt is determined by adding a position restriction based on a restriction value to the obtained target parking position Pt ′ (FIG. 2).
For this reason, when the own vehicle A parks, the parking assistance apparatus which determines the suitable target parking position Pt which suppresses the protrusion from the parking frame SF, ensuring the distance with the other vehicle B parked in the adjacent position. Can be provided.

  As mentioned above, although the parking assistance method and parking assistance device of this indication were explained based on Example 1, it is not restricted to this Example 1 about a concrete composition, and it is in each claim of a claim Design changes and additions are allowed without departing from the gist of the invention.

  In the first embodiment, as the target parking position calculation unit 281, when the target parking position Pt ′ is obtained based on the parking available space PS, the median CL of the parking available space PS is set as the target parking position Pt ′. However, as the target parking position calculation unit, when the target parking position is obtained based on the parking space, any direction of another vehicle adjacent from the median value of the parking space (for example, a direction with higher spatial information accuracy) It is good also as an example which makes the position slightly offset to the target parking position.

  In the first embodiment, when the limit value is provided based on the recognition information of the parking frame white line SL as the limit value calculation unit 282, the target parking position Pt that can be parked without the host vehicle A straddling the parking frame white line SL. An example is shown in which both end positions of the allowable offset width ΔW are the first limit value L1 and the second limit value L2, respectively. However, the limit value calculation unit draws a reference line at a position a predetermined amount from the parking frame white line, determines an allowable offset width based on the reference line, and sets both end positions of the allowable offset width as the first limit value and the first limit value, respectively. An example of 2 limit values may be used.

  In the first embodiment, as the target parking position determination unit 283, when the median value CL of the parking space PS is within the allowable offset width ΔW, the median value CL is determined as the target parking position Pt. When the median value CL of the parking space PS deviates from the allowable offset width ΔW, an example is shown in which the limit value close to the median value CL is determined as the target parking position Pt among the first limit L1 and the second limit value L2. However, the target parking position determination unit may be an example in which the median value of the parking frame is determined as the target parking position when the median value of the parking space deviates from the allowable offset width.

  In the first embodiment, as the target parking position determination unit 283, when the parking available space PS cannot be estimated, an example in which the median value of the parking frame SF in which the vehicle A is scheduled to park is determined as the target parking position Pt is shown. . However, the target parking position determination unit may be an example in which a value within the allowable offset width of the parking frame in which the host vehicle is scheduled to park is determined as the target parking position when the parking available space cannot be estimated.

  In the first embodiment, as the target parking posture calculation unit 284, the target parking posture Dt when the host vehicle A is parked at the target parking position Pt is determined from the road boundaries RL, RR estimated from the alignment of other parked vehicles B. The example which calculates based on the estimation information of this and the recognition information of the parking frame white line SL was shown. However, the target parking posture calculation unit may be an example of calculating the target parking posture based only on the recognition information of the parking frame white line. Furthermore, for example, when the parking form of the parking lot is clear in advance, an example in which the calculation of the target parking posture may be omitted.

  In Example 1, the example which applies the parking assistance method and parking assistance apparatus of this indication to the automatic driving vehicle carrying the parking assistance system by autonomous parking control was shown. However, the parking support method and the parking support device of the present disclosure are also applied to a vehicle equipped with a parking support system that only displays and guides the target parking position and the parking route in order to support the parking operation performed by the driver. Can do. Moreover, it is applicable also to the vehicle carrying the automatic parking assistance system which implements only the parking guidance control along the parking path produced | generated according to the instruction | indication when stopping near a parking frame. In short, the present invention can be applied to any vehicle having a parking support function that supports the driver's parking operation in some form by determining the target parking position when the vehicle is parked in the parking frame.

DESCRIPTION OF SYMBOLS 21 Space recognition sensor 22 White line recognition sensor 23 Steering angle sensor 24 Wheel speed sensor 25 Vehicle state quantity sensor 26 Sensor information processing part 261 Parking space estimation part 262 Runway boundary estimation part 27 Image information processing part 271 Parking frame white line recognition part 28 ECU for parking assist calculation
281 Target parking position calculation unit 282 Limit value calculation unit 283 Target parking position determination unit 284 Target parking posture calculation unit 285 Current own vehicle position / posture information acquisition unit 286 Reverse turn-back search unit 287 Parking path generation unit 29 Vehicle control ECU
30 Actuator 31 Image processing unit 32 Display monitor A Own vehicle B Other vehicle R Runway
RL, RR Track boundary S Parking lot
SF parking space
SL parking frame white line
L1 First limit value
L2 Second limit value
Rin parking route
Pn Current vehicle position
Dn Current vehicle attitude
Pc Reverse turning position
Pt Target parking position
Dt Target parking posture

Claims (10)

In the parking support method for assisting parking when the vehicle parks in the parking frame,
Based on the space recognition information acquired by the vehicle on the runway along the parking area, the parking space of the own vehicle is estimated except for the occupied space by other vehicles already parked from the parking area,
Based on the image information acquired by the vehicle on the runway along the parking area, it extracts the recognition information of the parking frame white line on which the vehicle is scheduled to park,
A target parking position is obtained based on the parking space, a limit value for the parking frame is provided based on the recognition information of the parking frame white line, and a position restriction based on the restriction value is added to the obtained target parking position. A parking support method characterized by determining a target parking position. In the parking assistance method according to claim 1,
A parking support method, wherein when a target parking position is obtained based on the parking space, a median value of the parking space is set as the target parking position. In the parking assistance method according to claim 2,
When the limit value is set based on the recognition information of the parking frame white line, both end positions of the allowable offset width of the target parking position that can be parked without the vehicle straddling the parking frame white line are respectively set as the first limit value. A parking support method characterized in that the second limit value is used. In the parking assistance method according to claim 3,
When determining the target parking position, if the median value of the parking space is within the allowable offset width, the median value is determined as the target parking position, and the median value of the parking space is the allowable offset width. When deviating from the above, a limit value close to the median value among the first limit value and the second limit value is determined as the target parking position. In the parking assistance method according to claim 3 or claim 4,
When the parking available space cannot be estimated, a median value of the parking frame in which the vehicle is scheduled to park is determined as the target parking position. In the parking assistance method according to any one of claims 1 to 5,
The target parking posture when the host vehicle is parked at the target parking position is calculated based on the estimation information of the road boundary estimated from the arrangement of other parked vehicles and the recognition information of the parking frame white line. A parking assistance method characterized by that. In the parking assistance method according to claim 6,
When calculating the target parking posture, the reliability of the road boundary estimation information and the recognition information of the parking frame white line are compared, and the one with higher reliability is selected. In the parking assistance method according to claim 6 or 7,
When calculating the target parking posture, a parking lot structure is estimated from the estimation information of the road boundary and the recognition information of the parking frame white line, and a tandem, parallel, and diagonal obtained from the estimation result of the parking lot structure A parking support method, wherein a calculation method is selected according to a parking mode. In the parking assistance method according to claim 8,
When calculating the target parking posture, the parking support white line recognition information is preferentially selected when it is determined as an oblique parking form based on the estimation result of the parking lot structure. In a parking support device including a parking support controller that supports parking when the vehicle is parked in the parking frame,
The parking assistance controller is:
Based on the space recognition information acquired by the own vehicle on the runway along the parking area, the parking space estimation for estimating the parking space of the own vehicle excluding the occupied space by other vehicles already parked from the parking area And
A parking frame white line recognition unit that extracts recognition information of a parking frame white line on which the vehicle is scheduled to park based on image information acquired by the own vehicle on a runway along the parking section;
A target parking position is obtained based on the parking space, a limit value is set based on the recognition information of the parking frame white line, and a position restriction based on the limit value is added to the obtained target parking position to obtain a target parking position. A target parking position determination unit to be determined;
A parking assistance device comprising: