JP6976050B2 - Posture estimation method of parking control device and posture estimation device - Google Patents

Description

The present invention relates to a posture estimation method for a parking control device and a posture estimation device.

Conventionally, as described in Patent Document 1, based on the self-position and posture estimated based on the rotation speed of the wheel and the steering angle of the vehicle, and the self-position and posture estimated based on the environment map. A method of determining the final self-position and posture is known.

The environment map (for example, a three-dimensional map) is created from data acquired by a vehicle that has traveled in the past, and a plurality of feature points and features such as road signs and structures are stored. Then, by collating the environmental information such as the shape and texture of the object acquired by the surrounding detection sensor such as the camera or laser radar mounted on the vehicle with the feature points and features of the environmental map, the position of the vehicle and the position of the vehicle on the environmental map and the features are collated. Estimate the posture.

Japanese Unexamined Patent Publication No. 2009-149194

However, in the conventional example disclosed in Patent Document 1 described above, since the feature points are stored without considering the distance from the vehicle, the posture of the vehicle on the environment map may not be estimated with high accuracy.

The present invention has been made to solve such a conventional problem, and an object thereof is a posture estimation method of a parking control device capable of estimating the posture of a vehicle with high accuracy, and a posture. The purpose is to provide an estimation device.

In order to achieve the above object, the present invention determines whether or not the distance from the own vehicle to the parking target position has reached a nearby area which is a region less than a preset threshold distance, and the parking target position from the own vehicle is determined. If the distance to is judged to have reached the vicinity area of the feature points included in the environment map, the target parking position or we choose a predetermined distance or more away feature points, and its feature points, the area around the vehicle collating the feature points definitive to be, we estimate the posture of the vehicle.

According to the present invention, the posture of the vehicle can be estimated with high accuracy, and the vehicle can be parked at the correct position and posture with respect to the parking target position.

FIG. 1 is a block diagram showing a configuration of a posture estimation device and its peripheral devices according to an embodiment of the present invention. FIG. 2 is a flowchart showing a processing procedure of the posture estimation device according to the first embodiment of the present invention. FIG. 3 is an explanatory diagram showing a target route set by the posture estimation device according to the first embodiment of the present invention. FIG. 4 is an explanatory diagram showing a state of selecting a feature point away from a parking target position according to the first embodiment of the present invention. FIG. 5 is an explanatory diagram showing a change in an image captured by a camera when the posture of the vehicle changes. FIG. 6 is a flowchart showing a processing procedure of the posture estimation device according to the second embodiment of the present invention. FIG. 7 is an explanatory diagram showing how to select a feature point far away from the parking target position and a feature point near the parking target position according to the second embodiment of the present invention. FIG. 8 is a flowchart showing a processing procedure of the posture estimation device according to the third embodiment of the present invention. FIG. 9 is an explanatory diagram showing how a feature point away from the parking target position and a planar or linear feature are selected according to the third embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[Explanation of the first embodiment]
FIG. 1 is a block diagram showing a configuration of a posture estimation device and a peripheral device thereof according to the first embodiment of the present invention. As shown in FIG. 1, the posture estimation device 100 according to the present embodiment includes a camera 11 (peripheral detection unit) that captures an image of the surrounding environment of the vehicle, and three-dimensional data around the vehicle and feature points included in the three-dimensional data. It includes an environment map 12 for storing features, an odometry sensor 13 for detecting the odometry of the vehicle, and a control unit 15 for outputting a control signal according to the posture of the vehicle. The control unit 15 is connected to an actuator 16 that controls steering, accelerator, brake, shift, parking, etc. mounted on the vehicle, and outputs a control signal to the actuator 16.

The environment map 12 may store two-dimensional data around the vehicle. In addition to the case where it is mounted on a vehicle, it may be configured to be connected via a network. Further, although the case where the camera 11 is used as a sensor for detecting the surrounding condition of the vehicle will be described as an example, it is also possible to detect the surrounding condition of the vehicle by using a laser radar, sonar, a beacon, and WiFi communication. ..

The posture of the own vehicle on the environment map 12 is estimated by detecting the distance and the angle from the feature points. In general, the farther the vehicle is from the feature point, the easier it is to estimate the posture of the vehicle. For example, when a feature point exists in front of the vehicle, if the angle of the posture of the vehicle changes by one degree, the larger the distance between the vehicle and the feature point, the larger the position of the feature point will be detected. .. On the contrary, when the distance between the vehicle and the feature point is small, even if the angle of the posture of the vehicle changes by 1 degree, the deviation of the position of the feature point is small and it is difficult to detect.

Therefore, in the present embodiment, when the vehicle V1 approaches the parking target position 21 (when the vehicle V1 enters the vicinity region R1 in FIG. 3, which will be described later), the posture is determined by using the feature points farther from the parking target position 21. By estimating, the estimation accuracy of the posture is improved. The reason why the accuracy of the posture of the vehicle V1 is required as the distance between the vehicle V1 and the parking target position 21 is smaller is that the closer the vehicle V1 is to the parking target position 21, the more sensitive the occupant is to the posture of the vehicle V1. Because there is. For example, when the vehicle V1 is parked at the parking target position 21, if the posture of the vehicle V1 with respect to the parking target position 21 is deviated, the occupant may feel uncomfortable, and further, it is conceivable to re-park. Therefore, the smaller the distance between the vehicle V1 and the parking target position 21, the more accurate the posture of the vehicle V1 is required.

On the other hand, the estimation accuracy of the position of the vehicle V1 on the environment map 12 is improved as the distance from the vehicle V1 is shorter. For example, when a feature point exists in front of the vehicle V1, if the vehicle V1 moves to the side by 10 cm, the feature point that is closer to the vehicle V1 will be detected with a large deviation in the position of the feature point. .. On the contrary, when the distance between the vehicle V1 and the feature point is large, even if the vehicle V1 moves to the side by 10 cm, the deviation of the position of the feature point is small, and the deviation of the position of the feature point can be detected. difficult.

The control unit 15 includes a feature point / feature selection unit 151 that selects a feature point or a feature included in the environment map 12 based on the conditions described later with reference to the environment map 12. Further, the feature point / feature object detection unit 152 (feature point detection circuit) for detecting the feature points around the vehicle or the feature objects from the image captured by the camera 11 is provided.

Further, the posture of the vehicle is estimated by collating the feature point or feature selected by the feature point / feature selection unit 151 with the feature point or feature detected by the corresponding feature point / feature detection unit 152. It is provided with a posture estimation unit 153 (posture estimation circuit).

Each function shown in this embodiment may be implemented by one or more processing circuits. The processing circuit includes a processing device including an electric circuit. Processing devices also include devices such as application specific integrated circuits (ASICs) and conventional circuit components arranged to perform the functions described in the embodiments. Then, the posture estimation device 100 according to the present embodiment detects the posture of the vehicle with high accuracy when the vehicle approaches the parking target position (for example, the parking space of the parking lot), and the posture is correct with respect to the parking target position. Control the vehicle so that it stops at.

Hereinafter, the processing procedure of the posture estimation device 100 according to the present embodiment will be described with reference to the flowchart shown in FIG. This control is executed by the control unit 15 shown in FIG.

First, in step S11, the control unit 15 determines whether or not the vehicle has approached the parking target position 21. In this process, for example, when the occupant of the vehicle determines that the vehicle has approached the parking target position 21, and the operation switch (not shown) is pressed, it can be determined that the vehicle has approached. Alternatively, in a vehicle equipped with a navigation device (not shown), a parking target position 21 is registered in advance in the navigation device, and the vehicle has a first threshold value with respect to the parking target position 21 based on the GPS vehicle position. It is also possible to determine that the vehicle has approached within a distance. The reference point of the parking target position 21 is, for example, the center position (for example, the position of reference numeral Q1 in FIG. 3) when the vehicle is correctly parked in the parking target position 21. Further, the first threshold distance can be set to, for example, 20 m. That is, when the vehicle reaches a radius of 20 m from the point of reference numeral Q1, it can be determined that the vehicle has approached the parking target position 21. Further, when the vehicle reaches a radius of 20 m, the occupant may be notified by emitting a sound or light to encourage the occupant to operate the switch. It is also possible to set the position of the camera 11 as the reference position when the vehicle is correctly stopped within the parking target position 21.

In step S12, the control unit 15 collates the feature points, features, and obstacles included in the image captured by the camera 11 with the feature points, features, and obstacles of the environment map 12, and the parking target position. Estimate the position and attitude of the vehicle with respect to. The feature points are arbitrary points such as road signs and buildings. Features are, for example, white lines laid on the road surface, walls of buildings, and the like. Obstacles are objects that are installed on the road surface and hinder the running of the vehicle.

In this embodiment, an example in which a camera 11 is used as a peripheral detection unit and feature points, features, and obstacles around the vehicle are detected using an image captured by the camera 11 will be described. As another example, it is also possible to use a laser radar or the like. At the same time, it is better to acquire the odometry information such as the wheel speed pulse and the yaw rate sensor by the odometry sensor 13.

In step S13, the control unit 15 refers to the vehicle position, posture information, and obstacle information included in the image captured by the camera 11, and the vehicle reaches the parking target position 21 from the current position. Generate a target route. For example, as shown in FIG. 3, when the vehicle V1 approaches the parking target position 21, a suitable target route L1 for parking the vehicle V1 at the parking target position 21 is generated. At this time, the vehicle V1 recognizes the position and the posture with reference to, for example, the roadside structure 31.

In step S14, the control unit 15 outputs a control signal to the actuator 16 so that the vehicle V1 travels along the target path L1 and executes parking control.

In step S15, the control unit 15 determines whether or not the distance from the vehicle V1 to the parking target position 21 has reached a region equal to or less than a preset second threshold distance (this is referred to as a “nearby region R1”). to decide. The second threshold distance is set to, for example, 10 m. For example, a range having a radius of 10 m is set in the vicinity region R1 with reference to the center point Q1 of the parking target position 21 shown in FIG. When the vehicle V1 enters the vicinity region R1, the process proceeds to step S16.

In step S16, the control unit 15 detects a feature point from the environment map 12. Further, an image captured by the camera 11 is acquired, and the image is analyzed to detect feature points included in an object or a building existing around the vehicle.

In step S17, the control unit 15 selects a feature point included in the environment map 12 that is separated from the center point Q1 by a predetermined distance or more. Specifically, as shown in FIG. 4, a region R2 less than a predetermined distance from the center point Q1 is set, and feature points existing outside the region R2 are selected. For example, the feature point p1 shown in FIG. 4 is selected. By selecting a feature point that is separated from the center point Q1 by a predetermined distance or more, it is possible to improve the estimation accuracy of the posture of the vehicle on the environment map 12. Further, by selecting the feature point located at the position farthest from the center point Q1, it is possible to further improve the estimation accuracy of the posture of the own vehicle.

In step S18, the control unit 15 stores the information of the feature points selected in the process of step S17 in the environment map 12 as the information of the feature points selected at the time of parking control. The information of the feature points stored in the environment map 12 can be used when the vehicle V1 is parked at the parking target position 21 from the next time onward.

In step S19, the control unit 15 collates the feature points selected on the environment map 12 with the feature points on the image captured by the corresponding camera 11, and estimates the posture of the vehicle V1 on the environment map 12. .. Specifically, as shown in FIG. 4, when the feature point p1 is selected, the feature point p1 is collated with the feature point on the image to estimate the posture of the vehicle V1 on the environment map 12. do. By doing so, when the vehicle V1 approaches the parking target position 21, the estimation accuracy of the posture of the vehicle V1 with respect to the parking target position 21 can be improved.

Hereinafter, the reason for selecting a feature point far from the parking target position 21 when estimating the posture of the vehicle will be described in detail.

FIG. 5 is an explanatory diagram schematically showing a surrounding image captured by a camera 11 mounted on a vehicle. The image D1 shows an image taken when the vehicle is in the posture F1, and the image D2 shows an image taken when the posture angle of the vehicle is slightly changed with respect to the posture F1. In the posture F1, the imaging region is f1, and in the posture F2, the imaging region is f2. The image D1 captured by the camera 11 in the posture F1 includes an object B1 (for example, a road sign) existing in the vicinity of the vehicle and an object B2 (for example, a mountain) existing in the distance. Further, the images D2 captured by the camera 11 in the posture F2 also include the objects B1 and B2.

Then, when the images D1 and D2 are compared, it is understood that the farther the object is from the vehicle, the larger it moves and is displayed in response to the change in posture. Specifically, the change of the object B2 of the image D1 and the change of the object B2 of the image D2 is larger than the change of the object B1 of the image D1 and the object B1 of the image D2. That is, the farther the object is from the vehicle, the greater the influence of the posture change. Further, as described above, the closer the object is to the vehicle, the greater the influence of the position change.

From the above, it can be seen that the posture of the vehicle can be estimated with higher accuracy by using the feature points existing farther from the vehicle within the range that can be captured by the camera 11. Therefore, in the process of step S17, the feature point p1 existing farthest from the parking target position 21 is selected and the posture of the vehicle is estimated.

Next, in step S20 of FIG. 2, the control unit 15 controls the vehicle V1 to travel along the target route L1 and moves the vehicle V1 to the parking target position 21.

In step S21, the control unit 15 determines whether or not the vehicle V1 has reached the parking target position 21 and the parking operation is completed, and if it is completed, the control unit 15 ends this process. When parking is completed, it is possible to detect that the occupant has put the shift position into parking or operated the parking brake and recognize that the vehicle V1 has been parked at the parking target position 21.

In this way, in the posture estimation device 100 according to the first embodiment, when the vehicle V1 is parked at the parking target position 21, the feature points of the environment map 12 and the feature points of the image captured by the camera 11 are collated. The posture of the vehicle V1 with respect to the parking target position 21 on the environment map 12 is estimated. At this time, a feature point separated from the parking target position 21 by a predetermined distance or more (a feature point outside the region R2 in FIG. 4) or a feature point farthest from the center point Q1 is selected and stored. As a result, when the vehicle V1 parks at the parking target position 21 using the environment map 12, the feature points existing at a position far from the parking target position 21 can be used, so that the posture of the vehicle V1 can be accurately determined. It can be estimated that the vehicle can be parked in the correct position with respect to the parking target position 21.

Further, since the information of the selected feature points is stored in the environment map 12, when the vehicle V1 is parked at the same parking target position 21 from the next time onward, the feature points stored in the environment map 12 are selected. be able to. Therefore, there is no need for a new process of selecting a feature point, and the calculation load can be reduced.

Further, instead of the center point Q1 described above, the distance to the feature point can be calculated more accurately by using the mounting position of the camera 11 as a reference point when the vehicle V1 is parked at the parking target position 21. The distance to the feature point can be determined, and the posture estimation accuracy can be improved.

[Explanation of the second embodiment]
Next, a second embodiment of the present invention will be described. Since the device configuration is the same as the block diagram shown in FIG. 1, the description thereof will be omitted.

In the second embodiment, among the feature points of the environment map 12, in addition to the feature points separated from the parking target position 21 by a predetermined distance or more, the feature points existing below the predetermined distance are selected, and both feature points and the camera 11 are used. The position and posture of the vehicle V1 are estimated with high accuracy by collating the feature points included in the captured image. Hereinafter, the processing procedure of the posture estimation device according to the second embodiment will be described with reference to the flowchart shown in FIG.

Since the processing of steps S51 to 56 shown in FIG. 6 is the same as the processing of steps S11 to S16 shown in FIG. 2, the description thereof will be omitted.

In step S57, among the feature points of the environment map 12, the control unit 15 is less than a predetermined distance in addition to the feature points separated from the parking target position 21 by a predetermined distance or more (the feature points outside the region R2 in FIG. 4). Select a feature point (feature point inside region R2). Specifically, as shown in FIG. 7, a feature point p1 far from the parking target position 21 and a feature point p2 near the parking target position 21 (points at the corners of the white line) are selected.

In step S58, the control unit 15 stores the selected feature points in the environment map 12 as feature points used for parking control. After that, in step S59, the control unit 15 collates the selected feature points p1 and p2 with the feature points included in the image captured by the camera 11 and estimates the position and posture of the vehicle V1. By selecting the feature point p2 that exists near the parking target position 21, the position of the vehicle V1 with respect to the parking target position 21 can be estimated with high accuracy, and as a result, the position and posture of the vehicle V1 can be estimated with high accuracy. become.

In step S60, the control unit 15 controls the vehicle V1 to travel along the target path, and moves the vehicle V1 to the parking target position 21.

In step S61, the control unit 15 determines whether or not the vehicle V1 has reached the parking target position 21 and the parking operation is completed, and if it is completed, the control unit 15 ends this process. When parking is completed, it is possible to detect that the occupant has put the shift position into parking or operated the parking brake and recognize that the vehicle V1 has been parked at the parking target position 21.

In this way, in the posture estimation device according to the second embodiment, both the feature points separated from the parking target position 21 by a predetermined distance or more and the feature points less than a predetermined distance are selected from the feature points of the environment map 12. The position and posture of the vehicle V1 are estimated. By using the feature points separated by a predetermined distance or more, the estimation accuracy of the posture of the vehicle V1 on the environment map 12 can be improved. By using feature points less than a predetermined distance, it is possible to improve the estimation accuracy of the position of the vehicle V1 on the environment map 12. Therefore, the position and posture when the vehicle V1 is parked in the parking target position 21 can be estimated with high accuracy, and the vehicle V1 can be parked at the correct position of the parking target position 21 and in the correct posture.

[Explanation of the third embodiment]
Next, a third embodiment of the present invention will be described. Since the device configuration is the same as the block diagram shown in FIG. 1, the description thereof will be omitted.

In the third embodiment, among the feature points in the environment map 12, in addition to the feature points separated by a predetermined distance or more, the planar and linear feature points included in the environment map 12 are selected, and the selected feature points and features are selected. The position and posture of the vehicle V1 are estimated with high accuracy by collating the feature points and features included in the image captured by the camera 11 with the object. Hereinafter, the processing procedure of the posture estimation device according to the third embodiment will be described with reference to the flowchart shown in FIG.

Since the processing of steps S71 to 76 shown in FIG. 8 is the same as the processing of steps S11 to S16 shown in FIG. 2, the description thereof will be omitted.

In step S77, the control unit 15 selects, among the feature points existing in the environment map 12, the feature points that are separated from the parking target position 21 by a predetermined distance or more. Specifically, as shown in FIG. 9, the feature point p1 outside the region R2 is selected. Further, in step S78, the control unit 15 selects a planar or linear feature existing around the vehicle V1. For example, as shown in FIG. 9, the white line 22 is selected. At this time, the position of the feature (white line 22) may be outside or inside the region R2.

In step S79, the control unit 15 stores the selected feature points and features in the storage unit 14 as feature points used for parking control. After that, in step S80, the control unit 15 collates the selected feature points p1 and the feature (white line 22) with the feature points and features included in the image captured by the camera 11 to determine the position and posture of the vehicle. presume.

In step S81, the control unit 15 controls the vehicle V1 to travel along the target path, and moves the vehicle V1 to the parking target position 21.

In step S82, the control unit 15 determines whether or not the vehicle V1 has reached the parking target position 21 and the parking operation is completed, and if it is completed, the control unit 15 ends this process. When parking is completed, it is possible to detect that the occupant has put the shift position into parking or operated the parking brake and recognize that the vehicle V1 has been parked at the parking target position 21.

In this way, in the posture estimation device according to the third embodiment, among the feature points of the environment map 12, the feature points existing at a position far from the parking target position 21 and the planar or linear feature are selected. Then, the position and posture of the vehicle V1 are estimated. By using the feature points far from the parking target position 21, it is possible to improve the estimation accuracy of the posture of the vehicle V1 on the environment map 12. Further, by using the planar and linear features, it is possible to improve the estimation accuracy of the position and posture of the vehicle on the environment map 12. Therefore, the position and posture of the vehicle V1 with respect to the parking target position 21 can be estimated with high accuracy, and the vehicle V1 can be parked at the correct position and the correct posture of the parking target position 21.

Although the posture estimation method and the posture estimation device of the parking control device of the present invention have been described above based on the illustrated embodiment, the present invention is not limited to this, and the configuration of each part has the same function. It can be replaced with any configuration.

11 Camera 12 Environment map 13 Odometry sensor 14 Storage unit 15 Control unit 16 Actuator 21 Parking target position 22 White line 31 Roadside structure 100 Posture estimation device 151 Feature point / feature selection unit 152 Feature point / feature detection unit (feature point detection) circuit)
153 Posture estimation unit (posture estimation circuit)
p1, p2 Feature point V1 Vehicle R1 neighborhood area

Claims (5)

When parking at the parking target position, an environment map showing the surrounding situation of the parking target position is stored, feature points are stored in the environment map, and the own vehicle in the environment map is stored based on the feature points. It is a posture estimation method that estimates the posture.
Detecting the feature points around the own vehicle,
It is determined whether or not the distance from the own vehicle to the parking target position has reached a nearby region, which is a region less than a preset threshold distance.
When the distance from the vehicle to the parking target position is determined to have reached the vicinity region of the feature points included in the environment map, select the target parking position or al a predetermined distance or more away feature points death,
Parking wherein the predetermined distance or more away feature points, and compares the feature points, the at the periphery of the vehicle corresponding to the feature point, characterized that you estimate the posture of the vehicle in the environment map Control device attitude estimation method. From the feature points around the own vehicle, select the feature point farthest from the own vehicle.
The first aspect of claim 1, wherein the posture of the own vehicle in the environment map is estimated by collating the feature points at the farthest distance with the feature points of the environment map corresponding to the feature points. Posture estimation method for parking control device. From the feature points around the own vehicle, a feature point less than a predetermined distance from the own vehicle is further selected.
Claim 1 or 2 is characterized in that the position of the own vehicle on the environment map is estimated by collating the feature points less than the predetermined distance with the feature points of the environment map corresponding to the feature points. The method of estimating the attitude of the parking control device according to the above. From the feature points around the own vehicle, select a feature that is flat or linear, and select it.
Any of claims 1 to 3, wherein the posture of the own vehicle is estimated by collating the flat or linear feature with the feature of the environment map corresponding to the feature. The posture estimation method for the parking control device according to claim 1. When parking at the parking target position, an environment map showing the surrounding situation of the parking target position is stored, feature points are stored in the environment map, and the own vehicle in the environment map is stored based on the feature points. It is a posture estimation device that estimates the posture.
The feature point detection circuit that detects the feature points around the own vehicle, and the feature point detection circuit.
It is determined whether or not the distance from the own vehicle to the parking target position has reached a neighborhood region which is a region less than a preset threshold distance, and the distance from the own vehicle to the parking target position is the neighborhood region. A posture estimation circuit that selects a feature point that is separated from the target parking position by a predetermined distance or more from the feature points included in the environment map when it is determined that the parking position has been reached is provided.
The posture estimation circuit comprises a predetermined distance or more away feature point, and feature points in the surrounding of the vehicle corresponding to the feature point, by matching, we estimate the posture of the vehicle in the environment map A posture estimation device characterized by this.

Images