JP2004220281A - Obstacle detection device - Google Patents

Abstract

A highly reliable obstacle detection device capable of suppressing erroneous detection of an obstacle in any external environment.
An obstacle detection device according to the present invention can detect a white line on a road and detect a white line from a vehicle based on an imaging unit installed in a vehicle and an image captured by the imaging unit. White line detecting means for obtaining a first distance to a point at which the obstacle cannot be obtained, and obstacle detecting means for detecting the presence or absence of an obstacle based on an image taken by the imaging means and obtaining a second distance from the vehicle to the obstacle. , An obstacle determining unit that compares the first distance with the second distance, and determines whether or not there is an obstacle based on the comparison result and the detection result of the obstacle.
[Selection diagram] Fig. 1

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an obstacle detection device using a camera, and particularly to an obstacle detection device for a vehicle mounted on a vehicle.
[0002]
[Prior art]
Conventionally, in an obstacle detection device using image recognition technology, when the external environment is deteriorated due to rain, fog, or the like, an image obtained by a camera is deteriorated, and the reliability of detecting a preceding vehicle or the like is reduced. was there.
[0003]
In order to solve this problem, Japanese Patent Application Laid-Open No. H11-163873 discloses that information supplied from a wiper switch recognizes a state such as rain or fog, and whether an external environment is suitable for a camera to recognize a preceding vehicle. A technique is described in which the direction of a millimeter wave radar is determined based on image data of a preceding vehicle by a camera when it is determined that the external environment is suitable.
[0004]
[Patent Document 1]
JP-A-6-206507 (pages 3-5, FIG. 1)
[0005]
[Problems to be solved by the invention]
However, the range in which an image can be recognized varies from moment to moment depending on the external environment such as sunshine conditions, and the above-described technique maintains high reliability of the result of image recognition with respect to such a change in the external environment. I couldn't do that. Further, even when it is determined that the external environment is suitable, there is a possibility that erroneous detection may occur because the reliability of the detection result cannot be evaluated.
[0006]
The present invention has been made in order to solve the above problems, and an object of the present invention is to provide a highly reliable obstacle detection device capable of suppressing erroneous detection of an obstacle in any external environment. And
[0007]
[Means for Solving the Problems]
The obstacle detection device according to the present invention detects a white line on a road based on an image pickup unit installed in a vehicle and an image picked up by the image pickup unit up to a point where the vehicle cannot detect the white line. An obstacle detecting means for detecting the presence or absence of an obstacle based on an image captured by the image capturing means and obtaining a second distance from the vehicle to the obstacle; And a second distance, and an obstacle judging means for judging the presence or absence of an obstacle based on the comparison result and the detection result of the obstacle.
[0008]
The reliability of the white line detection result by the white line detection means is high. This is because the contrast between the asphalt and the white line is large, and the white line has a substantially constant shape and is relatively simple. Therefore, it is possible to detect an object with high reliability up to the first distance in an image captured by the imaging unit. The obstacle determining means compares the second distance detected by the obstacle detecting means with the first distance to determine the presence or absence of an obstacle while considering the reliability of the detection result. Therefore, erroneous detection of an obstacle can be suppressed.
[0009]
An obstacle detection device according to the present invention includes: a first and second imaging units installed in a vehicle at predetermined distances with respective optical axes parallel to each other; and a road based on an image captured by the first imaging unit. A white line detecting unit that detects an upper white line and obtains a first distance from the vehicle to a point where the white line cannot be detected; an image captured by the first imaging unit and an image captured by the second imaging unit; Based on the difference between the first and second distances, the obstacle detection means for detecting the presence or absence of an obstacle and calculating a second distance from the vehicle to the obstacle, and comparing the first distance with the second distance. And an obstacle determining means for determining the presence or absence of an obstacle based on the detection result.
[0010]
According to the obstacle detection device of the present invention, since the obstacle detection unit calculates the second distance based on the difference between the image captured by the first imaging unit and the image captured by the second imaging unit. , It is possible to calculate a more accurate distance. In addition, the obstacle determination unit determines the presence or absence of an obstacle by comparing the second distance with the first distance while considering the reliability of the detection result. Therefore, erroneous detection of an obstacle can be suppressed.
[0011]
Further, in the obstacle detection device according to the present invention, the obstacle determination means affirms the detection result of the obstacle when the first distance is longer than the second distance, and detects the obstacle when the first distance is shorter than the second distance. It is preferable to suspend the judgment on the presence or absence.
[0012]
In this case, when the first distance is longer than the second distance, the reliability of the obstacle detection result is high, so that the detection result is determined to be correct. When the first distance is shorter than the second distance, the reliability of the obstacle detection result is high. Is low, the judgment on the presence or absence of the obstacle is suspended. Therefore, erroneous detection of an obstacle can be suppressed.
[0013]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements will be denoted by the same reference symbols, without redundant description.
[0014]
First, the configuration of the present embodiment will be described with reference to FIG.
[0015]
The obstacle detection device 1 captures a scene in front of the vehicle and obtains image data by a camera (first imaging unit) 10 and a camera (second imaging unit) 11, and performs image processing on the image data. And an image processing ECU 2 that outputs a determination result regarding the presence or absence of an obstacle.
[0016]
The image processing ECU 2 includes a white line detection unit (white line detection unit) 20 that detects a white line based on image data, an obstacle detection unit (obstacle detection unit) 30 that detects an obstacle based on the image data, An obstacle determining unit (obstacle determining means) 40 for determining whether or not there is an obstacle is provided.
[0017]
Each of the camera 10 and the camera 11 is, for example, a CCD camera, and is installed in front of the own vehicle so as to face forward. The optical axes of both cameras are parallel to each other (substantially parallel). In addition, they are installed so that the horizontal axis of the imaging surface is aligned on the same line. Then, each of the camera 10 and the camera 11 captures an image of a scene in front of the vehicle and acquires image data. The image data acquired by the camera 10 is output to each of the white line detection unit 20 and the obstacle detection unit 30, and the image data acquired by the camera 11 is output only to the obstacle detection unit 30.
[0018]
The white line detection unit 20 detects a white line on a road based on the image data obtained by the camera 10 and further detects a distance from the own vehicle to a point where the white line cannot be detected (hereinafter referred to as a white line detectable distance). ) Is calculated. Note that, in this specification, all lines (including dotted lines and the like) drawn on a road to indicate the delimitation of each lane (lane) are referred to as “white lines”.
[0019]
The obstacle detection unit 30 detects an obstacle based on a difference between the image data obtained by the camera 10 and the image data obtained by the camera 20, and further calculates a distance from the own vehicle to the obstacle.
[0020]
The obstacle determination unit 40 compares the white line detectable distance calculated by the white line detection unit 20 with the distance information to the obstacle output from the obstacle detection unit 30, and compares the comparison result with the obstacle detection result. , The presence or absence of various obstacles (for example, other vehicles, falling objects, pedestrians, etc.) in the traveling direction of the own vehicle is determined. Specifically, when the white line detectable distance is longer than the distance to the obstacle, the detection result of the obstacle is affirmed, and when the white line detectable distance is shorter than the distance to the obstacle, the determination about the presence or absence of the obstacle is suspended. .
[0021]
A vehicle control ECU 50 is connected to the obstacle detection device 1. The vehicle control ECU 50 to which information on the presence or absence of an obstacle (obstacle information) is input, controls the vehicle speed and the steering detected by the vehicle speed sensor 51. Based on the steering angle detected by the angle sensor 52, the yaw rate detected by the yaw rate sensor 53, and obstacle information, it is determined whether to change lanes (or avoidance) or to stop, and the steering actuator 55, the throttle actuator 56 and the brake actuator 57 are controlled.
[0022]
Next, the operation of the obstacle detection device 1 will be described with reference to the flowchart shown in FIG.
[0023]
In step S100, image data captured by the camera 10 is captured by the white line detection unit 20, and a white line on a road is detected based on the image data.
[0024]
In the detection of a white line, first, image data obtained from the camera 10 is binarized based on luminance, and a pixel considered to correspond to a white line portion for each horizontal line in the binarized image (hereinafter, referred to as a white line candidate point). ).
[0025]
Next, it is determined whether or not the white line candidate point selected for each line is a point corresponding to the white line portion. Here, this determination is made based on, for example, whether or not the respective white line candidate points are linearly arranged, or whether or not the position of the white line candidate point is close to a position recognized as a white line in the previous processing. Done.
[0026]
When it is determined that the white line candidate point selected for each line corresponds to the white line portion, a white line model (white line template) is applied to the column of the white line candidate points. Here, as the white line model, for example, a straight line, a curve, an arc, a quadratic curve, or the like is used.
[0027]
In step S110, the white line detectable distance is calculated for the white line detected in step S100. With reference to FIG. 3, a method of calculating the white line detectable distance will be briefly described.
[0028]
First, for a white line candidate point determined to correspond to a white line portion in step S100, a white line candidate point located at the uppermost position in the height direction H of the image is extracted. Here, when there are white line candidate points on the left and right of the position of the own vehicle, for example, with respect to the white line candidate points existing on the left side of the own vehicle, the uppermost white line candidate point is extracted as point A, Of the white line candidate points existing on the right side of the vehicle, the uppermost white line candidate point is extracted and set as point B.
[0029]
Next, the distance from the own vehicle to the point A and the distance from the own vehicle to the point B are calculated. Here, when the mounting position of the camera 10 (the height from the ground to the camera 10) and the mounting angle in the vertical direction are predetermined, the position in the height direction H in the image acquired by the camera 10 and the position The relationship with the distance is uniquely determined. Therefore, the distance from the own vehicle to the point A or the point B can be obtained according to the position in the height direction H on the image. For example, in FIG. 3, a line L30 corresponds to a line at a position 30 m ahead of the own vehicle, a line L50 corresponds to a line at a position 50 m ahead of the own vehicle, and a line L70 corresponds to a line at a position 70 m ahead of the own vehicle. I do.
[0030]
Then, the distance from the own vehicle to the point A is compared with the distance from the own vehicle to the point B, and the longer distance is defined as a distance at which a white line can be detected (white line detectable distance). Here, the reason for adopting the longer distance is to consider a white line such as a dotted line.
[0031]
According to the example shown in FIG. 3, the distance from the own vehicle to the point A is 50 m, and the distance from the own vehicle to the point B is also 50 m. Therefore, it is determined that the white line can be detected up to 50 m in front of the own vehicle, but that the white line cannot be detected farther than 50 m. In this case, the white line detectable distance is 50 m. Hereinafter, returning to FIG. 2, the operation of the obstacle detection device 1 will be continuously described.
[0032]
In step S120, an obstacle is detected based on the image data input from the cameras 10 and 11. In the detection of an obstacle, the presence or absence of a height can be determined by, for example, a stereo vision technique to separate the obstacle region from the road region. That is, the camera 10 and the camera 11 are arranged on the left and right, a point that is the same point in the three-dimensional space is associated between the left and right images, and the three-dimensional position of the point is obtained in the manner of triangulation. If the positions and postures of the cameras 10 and 11 with respect to the road plane are determined in advance, the height of an arbitrary point in the image from the road plane can be determined by stereo vision.
[0033]
In step S130, a determination is made as to whether an obstacle has been detected in step S120. Here, when no obstacle is detected, the process proceeds to step S140. On the other hand, when an obstacle is detected, the process proceeds to step S150.
[0034]
In step S140, a determination is made as to the presence or absence of an obstacle according to the white line detectable distance. Here, since the detection result of the obstacle within the white line detectable distance is considered to be highly reliable, it is determined that no obstacle exists within this distance. On the other hand, since the detection result of an obstacle at a distance where the white line cannot be detected is considered to have low reliability, the processing is performed without making a determination as to whether or not an obstacle is present farther than the distance at which the white line can be detected. finish.
[0035]
In step S150, the distance from the vehicle to the obstacle is calculated based on the difference between the image data input from the camera 10 and the image data input from the camera 11. The distance to the obstacle can be calculated, for example, by using stereoscopic image data and the principle of triangulation. Next, a method of calculating the distance will be briefly described with reference to FIG.
[0036]
When the obstacle is in front of the own vehicle, when the image obtained from the camera 10 and the image obtained from the camera 11 are superimposed, the obstacle is shifted to the left and right and lateral directions. Thus, the position where one of the images overlaps is obtained while shifting one image by one pixel. The number of pixels shifted at this time is defined as n. Further, as shown in FIG. 4, if the focal length of the lens F constituting the camera 10 or 11 is f, the distance between the optical axes of the left and right cameras 10 and 11 is L, and the pixel pitch is d, obstacles from the own vehicle will occur. The distance R to the object can be calculated by the following equation (1).
[0037]
R = (f · L) / (nd · d) (1)
Hereinafter, returning to FIG. 2, the operation of the obstacle detection device 1 will be continuously described.
[0038]
In step S160, it is determined whether the distance R from the own vehicle to the obstacle calculated in step S150 is equal to or less than the white line detectable distance calculated in step S110. Here, when it is determined that the distance R is shorter than the white line detectable distance, the process proceeds to step S180. On the other hand, when it is determined that the distance R is longer than the white line detectable distance, the process proceeds to step S170.
[0039]
In step S170, a determination is made as to whether there is an obstacle. Here, since the detection result of the obstacle farther than the white line detectable distance is considered to have low reliability, the determination as to whether an obstacle exists or not is suspended and the processing is terminated.
[0040]
On the other hand, in step S180, since the detection result of the obstacle within the white line detectable distance is considered to be highly reliable, it is determined that the obstacle exists. Then, the process ends.
[0041]
Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and various modifications are possible. For example, in the present embodiment, white line detection is performed by binarizing the obtained image data, obtaining white line candidate points, and applying a white line model, but the white line detection method is not limited to this. Alternatively, other methods may be used. Further, the number of cameras is not limited to two.
[0042]
【The invention's effect】
As described in detail above, according to the present invention, a configuration is provided in which the presence / absence of an obstacle is determined by comparing the detectable distance of the white line with the distance to the obstacle while taking into account the reliability of the detection result. Therefore, it is possible to provide a highly reliable obstacle detection device capable of suppressing erroneous detection of an obstacle in any external environment.
[Brief description of the drawings]
FIG. 1 is a block diagram showing the overall configuration of the present embodiment.
FIG. 2 is a flowchart illustrating an overall operation of the obstacle detection device according to the embodiment;
FIG. 3 is a diagram illustrating a method for calculating a white line detectable distance.
FIG. 4 is a diagram illustrating a principle of obtaining a distance by a triangulation principle using a stereo image.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Obstacle detection apparatus, 2 ... Image processing ECU, 10 and 11 ... Camera, 20 ... White line detection part, 30 ... Obstacle detection part, 40 ... Obstacle judgment part.

Claims (3)

Imaging means installed on the vehicle;
A white line detecting unit that detects a white line on a road based on an image captured by the image capturing unit and obtains a first distance from the vehicle to a point where the white line cannot be detected;
Obstacle detection means for detecting the presence or absence of an obstacle based on the image taken by the imaging means and obtaining a second distance from the vehicle to the obstacle;
An obstacle determination unit that compares the first distance and the second distance, and determines whether the obstacle is present based on the comparison result and the detection result of the obstacle;
An obstacle detection device comprising: First and second image pickup means which are arranged on a vehicle at a predetermined distance with their optical axes parallel to each other,
A white line detection unit that detects a white line on a road based on an image captured by the first imaging unit, and determines a first distance from the vehicle to a point where the white line cannot be detected;
Obstacles are detected based on a difference between an image captured by the first imaging unit and an image captured by the second imaging unit, and a second distance from the vehicle to the obstacle is determined. Object detection means;
An obstacle determination unit that compares the first distance and the second distance, and determines whether the obstacle is present based on the comparison result and the detection result of the obstacle;
An obstacle detection device comprising: The obstacle determining means affirms the detection result of the obstacle when the first distance is longer than the second distance, and makes a determination as to the presence or absence of the obstacle when the first distance is shorter than the second distance. Hold,
The obstacle detection device according to claim 1 or 2, wherein:

Images