JPH03225513A - Direction controller for self-traveling vehicle - Google Patents

Abstract

PURPOSE:To efficiently control the steering of a self-traveling vehicle by predicting a lane boundary line from a remote area and making corrections repeatedly during the travel of the vehicle. CONSTITUTION:A television camera 10 mounted on the self-traveling vehicle picks up an image of a lanes in front of the vehicle in its traveling direction and outputs image to an image recognition part 11, which processes the image to extract and output road information on a monitor 12. Pieces of sensor information of a vehicle speed sensor, an inertial force sensor, etc., provided on the vehicle are inputted from a sensor processing part 20 to a travel control part 13. A road structure calculation part 15 impants limitation conditions regarding the lane structure of a road structure model 14 upon the image extracted by the image extraction part 11 according to the steering angle of a steering wheel and the speed sensor information and outputs information regarding steering control to a travel command part 16, and the information on the lane structure is outputted to a storage part 18. A steering control means 19 calculates the distance from the road belt boundary line to the vehicle according to the pieces of information from the processing part 20 and calculation part 15 and predicts the road belt boundary line in a remote area to perform steering control.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an autonomous vehicle direction control device that controls the steering of a vehicle autonomously traveling on a road. The road strip structure from a distant area to a nearby area is calculated from the image of the road strip, and the autonomous vehicle direction control device automatically controls the steering based on the calculated road strip structure.

(Conventional technology) Conventionally, as a method for autonomously driving a car, there is a method of continuously installing guidelines such as guiding gables or gold film tares on the road surface along the target driving trajectory, but installing the guidelines requires a large amount of cost. This is not realistic for driving on actual roads.

JP-A-60-3711 [Automatic Driving Steering System for Automobiles] has been proposed as a method of mounting a television camera on a traveling vehicle and driving it on an actual road.

In this method, two television cameras detect road white lines on both the left and right sides, and by comparing the position of the detected boundary line on the image captured by the television cameras, the vehicle is steered. can be controlled automatically,
Although it is necessary to equip autonomous vehicles with equipment for autonomous driving, there is no need for special development of roads.

(Problem to be solved by the invention) However, in the latter method, the television cameras mounted on the autonomous vehicle are located between the left and right sides.
In a certain speed range, the steering can be effectively controlled using the road zone boundary line in the image captured by the television camera, but outside the applicable speed range, for example, in higher-speed driving conditions, the television camera The shape of the road strip boundary line in the area farther than the target area of the road strip detected by the camera is an important factor in steering control.
It is not possible to calculate the distant road strip structure and perform steering control.

An object of the present invention is to provide an autonomous vehicle direction control device that can efficiently control the steering of a traveling vehicle depending on the speed of the traveling vehicle.

(Means for Solving the Problems) The present invention includes: a television camera attached to an autonomously running vehicle facing the direction of the vehicle; an image processing unit that extracts a road zone from an image taken by the television camera; An autonomous vehicle direction control device comprising: a travel control section that controls travel of the vehicle based on information obtained from the road strip image, the travel control section controlling the vehicle based on the road strip image from the image processing section. Calculate the distance from the road strip boundary line to the vehicle, assume the shape of the road strip boundary line in the far region in the road strip area under preset road strip structural conditions, and further calculate the distance from the road strip boundary line to the vehicle. The shape of the road strip boundary line in the area adjacent to the distant area is predicted based on the shape of the road strip boundary line in the far area, and the shape of the road strip boundary line in the area adjacent to the distant area is repeatedly corrected while the vehicle is traveling. The present invention is characterized in that it includes a road construction calculation unit that calculates the shape of a line, and a steering control unit that performs steering control during driving based on the road structure calculated by the road structure calculation unit.

(Operation) As shown in FIG. 1, the road structure calculation section is provided with a condition 1 of a road strip structure model that three-dimensionally defines a road and a speed condition 2 of a traveling vehicle. Condition 1 of the road strip structure model is that the road surface is locally flat. Road zone boundaries are subject to local structural conditions such as being locally parallel, and structural conditions that indicate the relationship between curvature and slope of roads such as expressways. On the other hand, speed condition 2 of the traveling vehicle is such that the road belt area is divided into a far area, an intermediate area, and a middle area depending on the driving speed.
Contains vehicle speed data for dividing into neighboring regions.

The road structure calculation section calculates the shape of a road strip boundary line sufficiently far away from the road strip extracted by the image processing section based on condition 1 of the road strip structure model and speed condition 2 of the traveling vehicle. '', and then repeating ``Verification'' 1, ``Modify'', / without driving, to define the shape of the road strip boundary line in the intermediate region, and similarly define the shape of the road strip boundary line in the intermediate region. By defining the shape of the road strip boundary line up to the nearby area as "temporary construction", a road strip consisting of a far area road strip boundary line shape 4, an intermediate area road strip boundary line shape 5, and a near area road strip boundary line shape 6 is created. Since the region boundary shape 3 is detected, the fin of the road strip boundary line for controlling the direction according to the traveling speed is based on the condition 1 of the road strip structure model, the traveling vehicle speed condition 2, and the road strip region boundary shape 3.
It is possible to determine the appropriate area. The road strip boundary line shape of the optimal area determined by the road structure calculation unit is given to the steering control means, and the steering control means can perform direction control by performing steering control based on the road strip line shape. The speed conditions for autonomous vehicles can be expanded.

(Example) Hereinafter, an example of the present invention will be described in detail based on the drawings.

FIG. 2 is a block diagram showing an autonomous vehicle using an embodiment of the autonomous vehicle direction control device of the present invention. In FIG. 2, a television camera 10 mounted on an autonomous vehicle captures an image of a road strip ahead in the direction of travel, and the road strip image is output to an image recognition unit 11. The image recognition unit 11 performs image processing such as color extraction, edge detection, and labeling to extract the road strip head, road strip boundary line, and road strip white line portion from the road strip image, and outputs the extraction results as an output image. The image is output to the monitor 12 together with the input image.

In addition to the television camera 10, the autonomous vehicle includes an operation angle sensor (not shown) that detects the operation angle of the steering wheel.
, a vehicle speed sensor (
(not shown), an inertial force sensor (not shown), etc. that detect moment force applied to the autonomous vehicle while it is running. Each sensor information is output from the various sensor processing units 20 to the travel control unit 13.The travel control unit 13 includes a road strip structure model 14, a road strip structure calculation unit 15, a travel command unit 16, a steering control unit 17, and a road strip structure model 14, a road strip structure calculation unit 15, It is composed of a structure storage section 18. The road structure calculation unit 15 provides limiting conditions regarding the road band structure in the road band structure model 14 to the image extracted by the image recognition unit 11 based on the steering angle and speed sensor information, and gives the driving command unit 16 steering control conditions. Outputs information related to control,
Additionally, information on the road strip structure is output to the road strip structure storage section 18.

The driving command unit 16 checks the driving state of the autonomous vehicle based on the information from the various sensor processing units 20 and the information from the road strip structure calculation unit 15, and issues control commands according to the road conditions to the steering control unit 17. Output to. The steering control section 17 outputs a steering signal to the actuator control section 21 based on the control command from the travel command section 16 . The travel command section 16 and the steering control section 17 constitute a steering control means 19. The actuator control unit 21 controls a plurality of actuators such as a built-in steering actuator, an accelerator actuator, and a brake actuator.
Based on the input from 7, operations such as steering operation, accelerator operation, and brake operation are executed in response to this steering signal.

The road strip structure storage unit 18 is a road strip! The information on the road strip structure from the f4 structure calculation unit 15 is stored.

Next, the process of calculating the road structure in the road strip structure calculation unit 15 will be explained.

FIG. 3 is a schematic diagram from a road image in the traveling direction taken by a camera mounted on the autonomous vehicle shown in FIG. 2 to window settings. The area where the horizon 30 is the farthest area dividing line from the farthest area dividing line 34 that can be captured by the television camera 10 becomes the driving target area. In the travel target area, area dividing lines 31 and 3 are drawn according to the travel speed of the autonomous vehicle.
2.33 varies in the vertical direction, and the driving target area is divided into a far driving area 35, an intermediate driving area 36, 37, and a near driving area 38 by area dividing lines 31, 32, and 33.

It is known that there is a vanishing point in the road zone image in the direction of travel captured by the television camera 10 while driving, but because of this, the windows 40 and 44 for detecting road boundaries in distant areas are Even if it is fixed at the boundary, it is considered possible to detect the road strip boundary line or the road white line. In the window 40, using the window position determination method in the next area shown in FIG.
Determine the position of window 41, and then move window 42 .
The same applies to windows 44, 45, 46, and 47 for determining the position of window 43 and detecting road boundary lines.

FIG. 4 shows a method for determining the position of each window. The road strip boundary line is detected in the window 51, and the image coordinates (X
(PI), Y(PI)).

Using <X (P2>, Y (P2>), the gradient tan θ of the tangent at point P2 is determined. Using this gradient, the X coordinate of the intersection of the tangent at P2 and the next neighboring area division line 55, P3) X (P3)= (D1+D2)*X (P2>-D2
From NeX<Pl)l/Dl and the fixed value W, the vertical width D2 and the horizontal width (X (P2) -X(
A window of Pi))*D2/D1+W is determined, where DI, D2 are the vertical position differences of the area division lines and are determined by the speed sensor information. W is a constant. In this embodiment, automatic steering control is performed based on recognition of both center lines and side lines. Further, by recognizing only the center line, side line, or one road strip boundary line, the position may be kept at a predetermined distance from the reference position. In addition, when there is an area where the driving road zone and the road zone 39 that intersect with each other exist, the shape of the road zone boundary line is predicted from a distant area, so the road zone boundary detected in the distant area adjacent to the intersection area is Based on the line recognition results, it is possible to predict the shape of road strip boundaries and the shape of road strip white lines that are suitable for directional control for autonomous driving.

(Effects of the Invention) As explained above, according to the present invention, by predicting the shape of the road strip boundary line from a distant area,
Even if there is an area where the driving road zone intersects with a road zone that intersects the driving road zone, the road zone is determined to be suitable for direction control for autonomous driving based on the recognition results of the road zone boundary line detected in an adjacent distant area. The shape of the boundary line and the shape of the road strip white line can be predicted, and in the direction control of the autonomous vehicle, the road strip boundary line calculated in the optimal area for the traveling speed of the autonomous vehicle can be used to predict the shape of the road strip white line. By performing steering control using the shape of the vehicle, it is possible to expand the speed conditions for autonomous vehicles that can control direction, so it is possible to efficiently control the steering of the vehicle according to the speed of the vehicle. can.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of the processing contents in the road crossing calculation unit, FIG. 2 is a block diagram showing an autonomous vehicle in which an embodiment of the autonomous vehicle direction control device of the present invention is used, and FIG. Figure 2 is an overview of the process from the road image in the direction of travel captured by the camera mounted on the autonomous vehicle to the window setting, and Figure 4 is the process from the window that detects the road boundary line to determining the size and position of the adjacent window. FIG. 10... Television camera, 11... Image recognition unit, 13
... Travel control section, 15... Road strip structure calculation section, 16
... Travel command section, 17... Steering control section, 19...
Steering control means.

Claims (1)

[Claims] a television camera attached to an autonomously running vehicle facing the direction of travel; an image processing unit that extracts road strips from images taken by the television camera; An autonomous vehicle direction control device comprising a travel control section that controls travel, the travel control section calculating a distance from a road strip boundary line to the vehicle based on a road strip image from the image processing section. , assuming the shape of the road strip boundary line in the far region in the road strip region under preset structural conditions of the road strip, and further based on the shape of the road strip boundary line in the far region. a road structure calculation unit that predicts the shape of a road zone boundary line in an area adjacent to the distant area, and repeatedly makes corrections while the vehicle is running to calculate the shape of the road zone boundary line up to the area near the vehicle; What is claimed is: 1. An autonomous vehicle direction control device comprising: steering control means for performing steering control during travel based on the road structure calculated by the road structure calculation section.