JP2002327635A - Vehicular running control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vehicular running control device which detects a distance to an object as much as possible if the object is within a detection zone of an image processing system even if the object is out of a radar detection zone, by totally recognizing the object ahead of a vehicle with greater use of detection result through the radar and the image processing system, and determines whether the detection result is considered as distance information for performing running control for the vehicle, based on certainty degree for the detection result. SOLUTION: In Step 270, an inter-vehicle distance to a preceding vehicle is calculated by three inter-vehicle distance detecting means, using input information from the laser radar 10 and the image processing device; in Step 280, based on the calculated result, each of a detected certainty degree and a total certainty degree are calculated; and in Step 290, when the calculated result of the total certainty degree is not less than a prescribed value, any one of the detection results by the three inter-vehicle distance detecting means is considered as information of the inter-vehicle distance to the preceding vehicle to be used in follow-up control.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle travel control for recognizing a distance to an object ahead of a host vehicle (such as a preceding vehicle) and performing travel control such as follow-up control and vehicle speed control based on the distance information. Belongs to the technical field of equipment.

[0002]

2. Description of the Related Art Conventionally, as a traveling control device for a vehicle, for example, one disclosed in Japanese Patent Application Laid-Open No. 9-264954 is known.

[0003] This publication discloses that image processing is reduced and
In order to improve the accuracy of the recognition distance, in an image processing system having a camera that images the front of the vehicle and an image processing device that processes an image input from the camera,
A radar system that detects the distance to an object in front of the vehicle is provided, and an image processing device that recognizes a road white line in front and a preceding vehicle based on image input is used to generate an image based on the distance to the object detected by the radar system. There is described a technique including an area limiting unit that limits a processing screen by dividing the processing screen into a white line recognition area and a preceding vehicle recognition area.

[0004]

However, in the conventional vehicle traveling control device, although the image processing system for the front image and the radar system are combined, the radar is good at detecting distance and image processing. Since the recognition of the white line and the recognition of the presence of the preceding vehicle are merely combined, there is a problem that if the preceding vehicle exists in an area where the radar cannot detect the distance, the distance to the preceding vehicle cannot be detected.

That is, the detection area of the radar is set according to the relationship between the detection accuracy and the detection cycle, and becomes narrower than the detection area in the image processing system. For example, the horizontal detection area in radar is 12deg (± 6de
g), the horizontal direction detection area in the image processing system is set to about 30 deg (± 15 deg).

Accordingly, although it is possible to detect a white line from an image, detect an area of the own lane and an adjacent lane, and recognize the presence of a preceding vehicle there, for example, in turning or the like, If the vehicle existing in front of the own vehicle is an adjacent lane traveling vehicle and the preceding vehicle traveling in the own lane is not within the detection area of the radar, it is impossible to detect the distance to the preceding vehicle. .

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems. It is an object of the present invention to comprehensively recognize an object ahead of a host vehicle by utilizing detection results of a radar and an image processing system. As a result, even if the object is out of the detection area of the radar, if it is within the detection area of the image processing system, the distance to the object is detected as much as possible, and based on the confidence in the detection result, the vehicle It is an object of the present invention to provide a vehicle travel control device that determines whether or not to use distance information for performing travel control.

[0008]

In order to achieve the above object, according to the first aspect of the present invention, a vehicle for recognizing a distance to an object in front of a host vehicle and performing travel control based on the distance information. In the travel control device, first distance detecting means for detecting a distance between the own vehicle and an object ahead of the own vehicle by a radar, imaging means for imaging the front of the own vehicle, and an image of the front of the own vehicle taken by the imaging means Second distance detecting means for detecting the distance between the own vehicle and the object in front of the own vehicle based on the position of the shadow of the object in front of the own vehicle and the position of the vanishing point in the vehicle, and the own vehicle imaged by the imaging means Third distance detecting means for detecting a distance between the host vehicle and the object ahead of the host vehicle based on the object width ahead of the host vehicle in the image ahead and the actual value of the object width, and the first distance Detection means and second distance detection means Based on the detection result of the third distance detection means, the detection certainty calculating means for calculating each detection certainty for the detected distance between the own vehicle and the object in front of the own vehicle, the detection certainty calculating means Comprehensive confidence calculating means for calculating overall confidence in distance detection between the host vehicle and an object ahead of the host vehicle based on the calculated certainty of each detection, and a calculation result of the total certainty calculating means Is greater than or equal to a predetermined value, one of the detection results of the first distance detecting means, the second distance detecting means, and the third distance detecting means is output as the distance between the own vehicle and the object ahead of the own vehicle. And a distance output unit.

According to a second aspect of the present invention, in the vehicle traveling control apparatus according to the first aspect, the second distance detecting means includes the own vehicle and the own vehicle detected by the first distance detecting means. It is characterized by having road vanishing point correcting means for correcting the height of the road vanishing point in the image based on the distance to the object ahead.

According to a third aspect of the present invention, in the vehicle traveling control device according to the first aspect, the third distance detecting means includes at least a horizontal width of an object ahead of the own vehicle in an image; The distance between the host vehicle and the object in front of the host vehicle is detected based on either the vertical width of the object in front of the host vehicle in the image.

According to a fourth aspect of the present invention, in the vehicle traveling control device according to the first aspect, the third distance detecting means includes the own vehicle and the own vehicle detected by the second distance detecting means. An object width estimating means for estimating an actual value of an object width in front of the host vehicle based on a distance to an object ahead is provided.

According to a fifth aspect of the present invention, in the vehicle traveling control device according to the first aspect, the third distance detecting means includes: the own vehicle and the own vehicle detected by the first distance detecting means. An object width estimation value correcting means for correcting the object width ahead of the host vehicle based on the distance to the object ahead is estimated by the object width estimating means.

According to a sixth aspect of the present invention, in the vehicle traveling control apparatus according to the first aspect, the detection certainty factor calculating means includes a first distance detecting means for determining whether the first distance detecting means is the subject vehicle and the object in front of the subject vehicle. When the distance is detected, the certainty regarding the distance is detected by the third distance detecting means, and the certainty regarding the distance between the own vehicle detected by the second distance detecting means and the object ahead of the own vehicle is detected. It is characterized in that it is set higher than the certainty factor for the distance between the own vehicle and the object in front of the own vehicle.

According to a seventh aspect of the present invention, in the vehicle traveling control apparatus according to the second aspect, the detection certainty calculating means includes: the own vehicle detected by the second distance detecting means; When the height of the road vanishing point in the image is corrected by the road vanishing point correction means, the certainty factor for the distance to the object is set higher than when there is no correction.

According to an eighth aspect of the present invention, in the vehicle traveling control device according to the fifth aspect, the detection certainty calculating means includes: the own vehicle detected by the third distance detecting means; When the estimated value of the object width ahead of the own vehicle is corrected by the object width estimated value correcting means, the certainty factor for the distance to the object is set higher than the case without the correction.

According to a ninth aspect of the present invention, in the vehicle travel control device according to any one of the first to fifth aspects, a road yaw angle for detecting a relative yaw angle of the road with respect to the own vehicle at an object position ahead of the own vehicle. Detecting means, wherein the detection certainty calculating means calculates the certainty regarding the distance between the host vehicle and the object in front of the host vehicle based on the vertical edge-to-edge distance detected by the third distance detecting means. When the angle is equal to or larger than a predetermined value, the relative yaw angle is set lower than when the angle is smaller than the predetermined value.

According to a tenth aspect of the present invention, in the vehicle travel control device according to any one of the first to fifth aspects, a road yaw angle for detecting a relative yaw angle of the road with respect to the own vehicle at an object position ahead of the own vehicle. Detecting means; and the detection certainty calculating means calculates the certainty regarding the distance between the host vehicle and the object ahead of the host vehicle based on the horizontal edge-to-edge distance detected by the third distance detecting means. It is characterized in that it is set independently of the relative yaw angle detected by the angle detecting means.

According to the eleventh aspect of the present invention, in the vehicle travel control device according to the first to tenth aspects,
When the calculation result of the overall certainty calculation means is equal to or more than a predetermined value, based on the output result of the distance output means, performs a follow-up control to follow a preceding vehicle traveling ahead of the host vehicle,
When the calculation result of the total certainty calculating means is less than a predetermined value, the vehicle further comprises a traveling control means for controlling a speed of the vehicle traveling at a preset speed.

[0019]

According to the first aspect of the present invention, at the time of traveling control, the first distance detecting means detects the distance between the host vehicle and the object ahead of the host vehicle by the radar, In the distance detecting means, the distance between the own vehicle and the object ahead of the own vehicle is detected based on the position of the shadow of the object ahead of the own vehicle and the position of the road vanishing point in the image ahead of the own vehicle captured by the imaging means. The distance between the host vehicle and the object ahead of the host vehicle based on the object width ahead of the host vehicle in the image ahead of the host vehicle imaged by the imaging unit and the actual value of the object width. Is detected. Then, in the detection certainty calculating means, based on the detection results of the first distance detecting means, the second distance detecting means, and the third distance detecting means, the detected distance between the own vehicle and the object in front of the own vehicle is detected. Is calculated, and the total certainty calculating means calculates the total certainty in the distance detection between the own vehicle and the object in front of the own vehicle based on the certainty of each detection calculated by the detecting certainty calculating means. Is calculated, and in the distance output means, when the calculation result of the total certainty calculation means is equal to or more than a predetermined value, detection of the first distance detection means, the second distance detection means, and the third distance detection means One of the results is output as the distance between the host vehicle and the object ahead of the host vehicle. Then, the distance output from the distance output means is recognized as a distance to an object in front of the host vehicle, and travel control is performed based on the distance information.

That is, the distance to the object ahead is detected by three distance detecting means using the radar and the image, and based on the detection result, the calculation of the detection certainty and the calculation of the total certainty are performed. When the calculation result of the certainty factor is equal to or more than the predetermined value, one of the detection results by the three distance detection means is used as the distance information between the host vehicle and the object ahead of the host vehicle used in the traveling control. For example, when an object ahead of the own vehicle is present in the detection area of the radar, the detection result of the first distance detecting unit is set as the distance to the object ahead of the own vehicle, and the object ahead of the own vehicle is detected in the detection area of the radar. However, if the object in front of the own vehicle is present in the detection area of the image processing system, the detection result of the second distance detection unit or the detection result of the third distance detection unit Distance to the object.

Therefore, by using the detection results of the radar and the image processing system to comprehensively recognize the object ahead of the host vehicle, even if the object is out of the radar detection area, the detection area of the image processing system can be reduced. If so, the distance to the object is detected as much as possible, and it can be determined based on the degree of certainty about the detection result whether or not to use the distance information for controlling the traveling of the vehicle.

According to the second aspect of the present invention, in the road vanishing point correcting means of the second distance detecting means, the distance between the own vehicle detected by the first distance detecting means and the object ahead of the own vehicle is detected. , The height of the road vanishing point in the image is corrected.

Therefore, the accuracy of distance detection with the object ahead of the host vehicle by the second distance detection means, which depends on the position accuracy of the road vanishing point, can be improved.

According to a third aspect of the present invention, in the third distance detecting means, at least a horizontal width of the object ahead of the own vehicle in the image and a vertical width of the object ahead of the own vehicle in the image. The distance between the host vehicle and the object in front of the host vehicle is detected based on either of the width.

Therefore, even if one of the horizontal width and the vertical width of the object in front of the own vehicle cannot be obtained from the image, as long as one of the widths can be obtained from the image, the own width can be obtained. The distance to the object ahead of the vehicle can be detected.

According to the invention described in claim 4, the object width estimating means of the third distance detecting means determines the distance between the own vehicle and the object ahead of the own vehicle detected by the second distance detecting means. Based on this, the actual value of the object width ahead of the host vehicle is estimated.

Therefore, if the estimation of the object width is performed in advance when the distance is detected by the second distance detecting means, then when the means using the detection result is shifted to the third distance detecting means, The distance between the host vehicle and the object ahead of the host vehicle can be immediately detected by the third distance detecting means using the actual object width value previously estimated for the same preceding vehicle.

According to a fifth aspect of the present invention, in the object width estimation value correcting means of the third distance detecting means, the self-vehicle detected by the first distance detecting means and the object in front of the own vehicle are compared. Based on the distance, the object width in front of the host vehicle estimated by the object width estimating means is corrected.

Therefore, in the third distance detecting means for detecting a distance based on the change in the object width on the image, the object width estimated value is calculated based on the detection result by the first distance detecting means which is highly accurate distance data. By performing the correction, it is possible to increase the accuracy of detecting the distance to the object ahead of the host vehicle by the third distance detecting means.

In the invention according to claim 6, when the first distance detecting means detects the distance between the own vehicle and the object ahead of the own vehicle, the detection certainty degree calculating means detects the certainty degree for the distance. Is certainty with respect to the distance between the own vehicle and the object in front of the own vehicle detected by the second distance detecting means, and the confidence between the own vehicle and the object in front of the own vehicle detected by the third distance detecting means. It is set higher than the certainty factor.

Therefore, when the object ahead of the host vehicle is present in the detection area of the radar and the distance to the object ahead of the host vehicle is detected by the first distance detecting means, the first distance using the radar information is used. The distance detection accuracy from the detection means is higher than the distance detection accuracy from the second and third distance detection means, and a high confidence evaluation is given to the detection result from the first distance detection means. Can be.

[0032] In the invention according to claim 7, in the detection certainty calculating means, the certainty regarding the distance between the own vehicle and the object in front of the own vehicle detected by the second distance detecting means is determined by the road disappearance. When the height of the road vanishing point in the image is corrected by the point correction unit, the height is set higher than when there is no correction.

Therefore, the accuracy of detecting the distance to the object ahead of the host vehicle by the second distance detecting means depends on the position accuracy of the vanishing point of the road, which is higher than when the height of the vanishing point of the road is not corrected. When the height of the road vanishing point is corrected, a higher confidence evaluation can be given.

According to the present invention, in the detection certainty calculating means, the certainty regarding the distance between the own vehicle and the object ahead of the own vehicle detected by the third distance detecting means is determined by the object width. When the estimated value of the object width in front of the host vehicle is corrected by the estimated value correcting means, the value is set higher than when there is no correction.

Therefore, the distance detection accuracy of the third distance detection means with respect to the object in front of the host vehicle depends on the estimation accuracy of the object width estimation value, and the distance detection accuracy is higher than when the object width estimation value is not corrected. A higher confidence rating can be given to the case where the width estimation value is corrected.

According to a ninth aspect of the present invention, the road yaw angle detecting means detects the relative yaw angle of the road relative to the own vehicle at the object position ahead of the own vehicle, and the detection certainty calculating means calculates the third yaw angle. If the certainty factor for the distance between the host vehicle and the object ahead of the host vehicle detected based on the vertical edge-to-edge distance by the distance detecting means is equal to or greater than a predetermined value, the relative yaw angle is smaller than a predetermined value. It is set lower than the case.

Therefore, the distance detection accuracy of the third distance detection means with respect to the object ahead of the host vehicle corresponds to the fact that the detection accuracy decreases as the vertical edge of the object ahead of the host vehicle becomes slanted. When the front of the host vehicle is a curve, a lower confidence evaluation can be given as compared with the case where the front of the host vehicle is not a curve.

According to the tenth aspect of the present invention, the road yaw angle detecting means detects the relative yaw angle of the road relative to the own vehicle at the object position in front of the own vehicle, and the detection certainty calculating means calculates the third yaw angle. Is set irrespective of the relative yaw angle detected by the relative yaw angle detecting means, based on the distance between the own vehicle and the object ahead of the own vehicle detected based on the horizontal edge-to-edge distance by the distance detecting means. .

Therefore, the distance detection accuracy of the third distance detecting means with respect to the object in front of the host vehicle is almost constant even if the horizontal edge of the object in front of the host vehicle is slanted. Correspondingly, the same reliability evaluation can be given both when the front of the host vehicle is a curve and when the front of the host vehicle is not a curve.

According to the eleventh aspect of the present invention, the traveling control means travels ahead of the host vehicle on the basis of the output result of the distance output means when the result of calculation by the total certainty calculating means is equal to or greater than a predetermined value. Following control is performed to follow the preceding vehicle, and when the result of the calculation by the total certainty calculating means is less than a predetermined value, vehicle speed control for running at a preset speed is performed.

Therefore, only for the preceding vehicle for which the detection results from the radar and the image processing system are likely to be reliable, it is possible to perform tracking control using the distance information between the correctly detected preceding vehicle and the preceding vehicle.

[0042]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of a vehicle traveling control apparatus according to the present invention;
1 will be described based on the first embodiment.

(First Embodiment) First, the configuration will be described.

The vehicle travel control device of the first embodiment combines a radar and an image processing system to recognize the presence or absence of a preceding vehicle on the own lane, and to maintain the set vehicle speed when there is no preceding vehicle. When there is a preceding vehicle, the so-called inter-vehicle control type constant speed traveling device (Adaptive Cruise Control: ACC) controls the inter-vehicle distance to the preceding vehicle to an appropriate value based on the detected inter-vehicle distance information. ).

FIG. 1 is a system diagram showing the configuration of the vehicle travel control device of the first embodiment. FIG. 2 is a perspective view showing a vehicle equipped with the vehicle travel control device of the first embodiment. 2, reference numeral 1 denotes a vehicle travel control device, 10 denotes a laser radar (radar), 20 denotes a front camera (imaging means), 3
0 is an image processing device, 40 is a vehicle speed sensor, 50 is a follow-up controller, 80 is a throttle actuator, 90
Is a brake actuator.

The laser radar 10 is mounted on a grill portion or a bumper or the like in front of the vehicle, propagates an infrared light pulse while scanning in a horizontal direction, and transmits a plurality of reflectors in front (usually at the rear end of a preceding vehicle). ) Is measured, the inter-vehicle distance to a plurality of preceding vehicles and the direction of its existence are detected based on the arrival time of the reflected wave, and the detected inter-vehicle distance and direction are output to the following controller 50. The front area scanned by the laser radar 10 is about ± 6 degrees with respect to the front of the vehicle, and a front object existing in this area is detected.

The front camera 20 is a small CCD camera, a CMOS camera, or the like attached to the upper part of the front window, detects the state of the road ahead as an image, and outputs it to the image processing device 30. The detection area by the front camera 20 is ± 30 de in the horizontal direction.
It is about g, and the front road scenery included in this area is captured.

The image processing device 30 includes a front camera 20
Performs various image processing such as filter processing and recognition processing on the image input from the camera, and interlocks with the follow-up controller 50 from the image of the front camera 20 to obtain a plurality of feature amounts (distance between the lower shadow position of the object and the plurality of edges). ) Is calculated.

The following control controller 50 calculates the following distance and the likelihood of the preceding vehicle on the own lane from the input of the following distance from the laser radar 10 and the result of recognition from the image processing device 30, and controls the running control. I do. The contents of the control are based on the following distance and the vehicle speed to the preceding vehicle, and output appropriate command values to the throttle actuator 80 and the brake actuator 90 so as to follow the preceding vehicle at a vehicle speed that is set separately or less, and output the command value of the own vehicle. Controls acceleration and deceleration.

Next, the operation will be described.

[Image Processing] FIG. 3 is a flowchart showing the flow of image processing executed by the image processing apparatus 30, and each step will be described below. Note that this process
For example, it is continuously performed once every 100 msec.

In step S100, a front image from the front camera 20 is read.

In step S110, a white line recognition process is performed to specify an area of the front road where the own vehicle travel lane and the adjacent travel lane are not located on the screen. Here, the positions of the white lines (lane markers) on both sides of the own vehicle traveling lane are detected.

In step S120, step S110
From the white lines recognized in the above, the area within the screen of the own vehicle traveling lane (inside the white lines on both sides) and the adjacent lane (outside the white lines on both sides) are calculated. This is used in the processing by the following controller 50 to determine whether the recognized preceding vehicle is a preceding vehicle on the own lane or an adjacent lane traveling vehicle on an adjacent lane.

In step S130, step S110
The position of the road vanishing point on the screen and the height in the screen are calculated from the white line recognized in step (1). Figure 5 shows the position of the vanishing point of the road.
As shown in (1), it is calculated by approximating a region near the recognized white line on both sides with a straight line and finding the intersection of the two straight lines.

In step S140, the shadow position of the preceding vehicle is detected in the lane area of the own vehicle and the adjacent lane area on the screen. The processing here is, for example, assuming that the density value changes darker upward from the lower end of the screen and that can extract the edge component in the horizontal direction as a candidate for a shadow of the preceding vehicle, and the position of the horizontal edge May be detected as the lower shadow position.

FIG. 6 is a diagram schematically showing the relative relationship between the host vehicle and the preceding vehicle, and the layout of the front camera 20 and the laser radar 10 from the side. FIG. 7 shows a camera image in the same situation. The image has an X axis and a Y axis as shown in FIG. Here, the meaning of each symbol in the figure is as follows. Hc: Height at which the camera 20 is mounted (fixed value; m) DL: Distance between vehicles (m) DO: Distance between the front end of the vehicle and the mounting position of the camera 20 in the front-rear direction (fixed value; m) DC ': After the vehicle ahead of the camera 20 Distance to end (m) DC '= DL + DO (1) YO: Y coordinate value indicating road vanishing point position in screen YE: Y coordinate value of horizontal edge A due to lower shadow in screen dY: In screen The relative Y coordinate value of the horizontal edge A due to the shadow below the road vanishing point position dY = YO-YE (2) Here, it is assumed that the angular resolution per Y coordinate value of the camera 20 is △ θ (rad). According to the geometric relationships in FIGS.
The following equation (3) holds.

DY = Hc / DC ′ / △ θ (3) From equations (1) to (3), DL = Hc / (YO−YE) / △ θ−DO (4) As described above, if the Y coordinate value of the lower shadow and the Y coordinate value of the road vanishing point are obtained, the inter-vehicle distance can be estimated.

In step S150, an area slightly larger than the width of the lower shadow is set above the detected lower shadow position.
Vertical edge components generated at both ends of the preceding vehicle are detected, and their width (vertical edge width △ X) is calculated. This vertical edge width ΔX is inversely proportional to the inter-vehicle distance as shown in FIG. However, if the actual width of the detected vertical edge (both ends of the vehicle) is not known, a correct inter-vehicle distance cannot be calculated.

In step S160, based on the position of the detected vertical edge in the left-right direction (X direction) and the road area calculated in step S120, the detected traveling lane indicates which traveling lane is running (own lane, right adjacent lane). The lane and the left adjacent lane are determined and stored.

In step S170, step S140
Another horizontal edge is detected from the information on the position of the lower shadow detected in the above step, and is combined with the horizontal edge due to the lower shadow.
Y is calculated. This horizontal edge width ΔY is also inversely proportional to the inter-vehicle distance, as shown in FIG. However, if the actual width of the detected horizontal edge (upper and lower shadows of the vehicle) is not known, a correct inter-vehicle distance cannot be calculated.

In step S180, the relative yaw angle of the road at those positions (under shadow position) with respect to the detected preceding vehicles is calculated (relative yaw angle detecting means). This is because when the front of the road is curved, the preceding vehicle traveling there is oblique relative to the own vehicle, and when the angle is too large, the vertical edge width detected in step S150 is used. There is a high possibility that ΔX does not correctly capture both ends of the preceding vehicle, and the accuracy of calculating the inter-vehicle distance from the vertical edge width ΔX may deteriorate.

In step S190, the detected results such as the detected lower shadow height YE, the vertical edge width △ X, and the horizontal edge width △ Y are output to the follow-up controller 50, and the current processing ends.

[Traveling Control Processing] FIG. 4 is a flowchart showing the flow of the traveling control processing executed by the follow-up controller 50. Each step will be described below. This processing is continuously performed, for example, once every 100 msec.

In step S200, the image processing device 30
Is read. Subsequently, in step S210, a radar detection result from the laser radar 10 is read.

In step S220, the laser radar 10
It is determined whether or not there is a detection object due to. As shown in FIG. 8, the detection area in the horizontal direction by the laser radar 10 is set to be smaller than the detection area by the image. In this step S220, the laser radar 10
If there is an object detection (distance detection), the process proceeds to step S230 (determination of the presence of the corresponding shadow), and if there is no object detection (distance detection) by the laser radar 10, the process proceeds to step S270 (inter-vehicle distance calculation). move on.

In step S230, it is determined based on the result of the image processing whether or not an under shadow corresponding to the laser detected object has been detected. That is, the headway distance DL is estimated from the image-detected shadow detection position by using the equation (4), the value is compared with the detection distance by the laser radar 10, and the direction of the object detected by the laser radar 10 is lowered. Compare with the X direction of the shadow detection position. Regarding the detection of the lower shadow at which these are almost the same position, the lower shadow is regarded as the lower shadow corresponding to the radar detected object, the process proceeds to step S240 (road vanishing point correction), and it is determined that the lower shadow corresponding to the laser detected object is not detected. If so, the process proceeds to step S250 (determination of presence / absence of a corresponding edge width). .

In step S240, the radar detection distance D
Based on L 'and the lower shadow height YE, the Y coordinate value of the road vanishing point is corrected to the following YO' (road vanishing point correcting means).

YO ′ = Hc / (DL ′ + DO) / △ θ + YE (5) By using the corrected Y coordinate value YO ′ of the road vanishing point instead of YO in equation (4), (4) The accuracy of the estimated value of the inter-vehicle distance DL calculated by the equation can be improved.

In step S250, it is determined whether there is a corresponding edge width. Based on the image processing result, it is determined whether the vertical edge width △ X or the horizontal edge width △ Y has been detected in the corresponding object, and if the edge width has been detected, the process proceeds to step S260 (actual width estimation value correction). ), And if the edge width has not been detected, the process proceeds to step S270 (inter-vehicle distance calculation).

In step S260, an estimated value of the actual width serving as a reference for calculating the inter-vehicle distance is calculated (object width estimating means, object width estimated value correcting means). The processing here is performed separately for each of the vertical edge and the horizontal edge. The distance between the camera 20 and the preceding vehicle calculated by the equation (1) is defined as DC ′ (= DL + DO). Next, assuming that the vertical edge width calculated in step S150 in FIG. 3 is △ X and the horizontal edge width calculated in step S170 is △ Y, these vertical edge widths △ X
And the horizontal edge width ΔY are inversely proportional to the distance DC ′, and the following two equations hold.

X_width / (DL + DO) = △ X / fH (A1) Y_width / (DL + DO) = △ Y / fV (A2) where X_width and Y_width correspond to the vertical and horizontal edges of the preceding vehicle, respectively. Are the horizontal and vertical distances (m), and fH and fV are the horizontal and vertical focal lengths (pixels; constant values) determined by the specifications of the image processing device 30 of the front camera 20, respectively. The actual width estimation values corresponding to the vertical and horizontal edges correspond to X_width and Y_width described above. Therefore, the inter-vehicle distance DL ′ to the preceding vehicle detected by the laser radar 10 and the vertical and horizontal edge widths △ X, △ Y calculated from the image are calculated by using the following 2 which is obtained by modifying the equations (A1) and (A2).
The correction value of the estimated actual width can be obtained by substituting into the equation.

X_width = △ X / fH * DL '(A3) Y_width = △ Y / fV * DL' (A4) When the distance is not detected by the laser radar 10, the equations (1) and (2) are used. The distance DC ′ is obtained by the equation (4), and (A1), (A
The actual width estimation values X_width and Y_width before correction are obtained by the equation (2). In this step S260, the actual width estimated value estimated by the above equation is stored, and the next and subsequent steps are executed in step S2.
At 70, it is used to calculate the following distance for the same preceding vehicle.

In step S270, from all road vehicles detected from the image processing result road radar detection result,
The inter-vehicle distance is calculated (first distance detecting means, second distance detecting means, third distance detecting means). Here, the following steps are taken. For a preceding vehicle whose distance has been detected by the radar, the radar detection distance DL ′ is set as the inter-vehicle distance. For vehicles ahead that are not detected by radar,
(i) If the horizontal edge width △ Y is detected and the actual width estimation value Y_width of the horizontal edge has been previously calculated for the same preceding vehicle, the horizontal edge width △ Y and the actual width estimation value Y_wid
From the th, an inter-vehicle distance DC 'is calculated by the following equation. DC ′ = Y_width * fV / △ Y (6) (ii) If the vertical edge width ΔX is detected and the actual width estimation value X_width of the vertical edge has previously been calculated for the same preceding vehicle, Vertical edge width △ X and actual width estimated value X_wid
From the th, an inter-vehicle distance DC 'is calculated by the following equation. DC '= X_width * fH / △ X (7) If none of (iii) (i) and (ii) has been detected and the lower shadow position has been detected, the Y coordinate value YE of the lower shadow and From the road vanishing point Y coordinate value YO, an inter-vehicle distance DL is calculated by equation (4).

At step S280, the detection certainty is calculated for all the detected preceding vehicles (detection certainty calculating means, total certainty calculating means). Here, a calculation as shown in FIG. 11 is performed. That is, the degree of certainty regarding the lower shadow height is calculated. If undercast is detected and road vanishing point correction is performed, the value is set to 1.0. If the shadow is detected but the road vanishing point correction process is not performed, 0.
Assume 5. Otherwise 0.0. Calculate the certainty factor regarding the edge width in the screen. If the horizontal edge width in the screen is calculated and the correction value of the actual width estimated value of the horizontal edge has been calculated before, it is set to 1.0, and the actual width estimated value of the horizontal edge (before correction) is calculated before. If it is, set it to 0.5. Even when the horizontal edge width in the screen is not detected, the vertical yaw width is detected and the correction value of the actual width estimated value of the vertical edge has been previously calculated, and the change in the road yaw angle at the position of the preceding vehicle is predetermined. If it is less than the value (if the front is not a curve), it is set to 1.0. When the vertical edge width is detected and the correction value of the actual width estimated value of the vertical edge has been calculated before, and the change in the road yaw angle at the position of the preceding vehicle exceeds a predetermined value (when the front is a curve), Is 0.5. Otherwise, it is set to 0.0. From these certainty factors, a comprehensive detection certainty factor is calculated. Here, a weight of 0.4 for whether or not the radar is detected, a weight of 0.3 for the certainty regarding the lower shadow height, and a 0.3 for the certainty regarding the in-screen edge width.
Are obtained by adding all the degrees of certainty.

In step S290, of all the detected preceding vehicles, the one present on the own lane and having the shortest inter-vehicle distance is regarded as the preceding vehicle, and the overall detection certainty is determined to be a predetermined value (for example, 0.5) is determined. If YES in step S290, the process proceeds to step S300. In step S300, the preceding vehicle is set as a control target of the following control, and in step S310, the following vehicle is controlled to follow. If NO in step S290, that is, if the total detection certainty is less than 0.5,
Since the detection of the preceding vehicle is not certain, tracking control is not performed.
Proceeding to step S320, vehicle speed control for traveling at the set vehicle speed is performed.

[Driving Control Action] At the time of driving control started by the driver's setting operation, when there is no preceding vehicle in the own lane, the flow chart of FIG. 4 shows steps S200 → S210 → S220.
→ Step S270 → Step S280 → Step S2
The flow proceeds from 90 to step S320. In step S320, vehicle speed control for traveling at the set vehicle speed is performed.

On the other hand, if there is a preceding vehicle in the own lane, the process proceeds to step S200 in the flowchart of FIG.
→ Step S210 → Step S220 → Step S2
30 → (Step S240 →) Step S250 → (Step S260 →) Step S270 → Step S28
0 → the flow proceeds to step S290.
In 290, for example, when it is determined that the own vehicle and the preceding vehicle are far apart and the certainty factor of the preceding vehicle (the preceding vehicle to be controlled) is less than 0.5, the process proceeds from step S290 to step S320, and the set vehicle speed is set. Is performed to control the vehicle speed.

However, if the inter-vehicle distance between the host vehicle and the preceding vehicle is sufficiently short, and it is determined in step S290 that the certainty of the preceding vehicle is 0.5 or more, step S290 is performed.
From step S300 to step S310, and, based on the inter-vehicle distance calculated in step S270 (if there is more than one, select the one with high detection certainty), for example, the own vehicle and the preceding vehicle obtained from the inter-vehicle distance and the relative speed Following control is performed so that the own vehicle follows the preceding vehicle so as to keep the inter-vehicle time with the vehicle constant.

Here, a method of calculating an inter-vehicle distance based on a feature amount from a radar and an image will be described with reference to FIG. FIG. 12 shows the principle of calculating the distance from the radar, the height of the shadow, and the width in the screen, and the drawbacks of each method. That is, the inter-vehicle distance calculation method using a radar calculates the inter-vehicle distance with the preceding vehicle based on the time required for the reflected wave from the preceding vehicle to return, and the detection accuracy of the inter-vehicle distance is high.
The detection area is small and the driving lane cannot be determined. The method of calculating the inter-vehicle distance based on the height of the lower shadow calculates the inter-vehicle distance based on the difference between the Y coordinate value of the vanishing point in the screen and the Y coordinate value of the lower shadow. The distance cannot be calculated, and the accuracy of detecting the inter-vehicle distance depends on the accuracy of the vanishing point coordinates. The inter-vehicle distance calculation method based on the width in the screen is based on estimating the vehicle width based on highly accurate distance data and calculating the change in the inter-vehicle distance from the change in vehicle width in the screen. It is necessary to calibrate the absolute distance, and the detection accuracy of the inter-vehicle distance is poor at the vertical edge when the rear surface of the vehicle is inclined. An inter-vehicle distance calculation process in step S270,
In addition, the processing of calculating the detection certainty factor in step S280 is appropriately set based on the principle of each of these methods so as to complement the defect.

With the above processing, the distance to the preceding vehicle is detected by the laser radar 10 and the front camera 20.
By using the detection results of the laser radar 10 and the image processing device 30 to perform comprehensive recognition, the laser radar 1
For vehicles that are out of the zero detection area, the inter-vehicle distance is calculated as accurately as possible, the detection confidence is calculated for the detection result, and the preceding vehicle detected based on the detection confidence is determined. It is determined whether or not to be a control target, and the tracking control can be performed only when the detection is correctly performed.

Next, the effects will be described.

(1) In step S270, an inter-vehicle distance to a preceding vehicle is calculated by three inter-vehicle distance detecting means using input information from the laser radar 10 and the image processing apparatus, and in step S280, based on the calculation result, If the calculation result of the total certainty is equal to or more than a predetermined value in step S290, one of the detection results by the three inter-vehicle distance detecting means is determined by the follow-up control. The information is used as inter-vehicle distance information with respect to the preceding vehicle to be used, that is, since the system uses a detection result obtained by using the laser radar 10 and the image processing device 30 to comprehensively recognize the preceding vehicle, it deviates from the detection area of the laser radar 10. As far as possible, if the vehicle is in the detection area of the image processing device 30, the inter-vehicle distance to the vehicle in front is detected as much as possible. It can be determined whether or not to use inter-vehicle distance information for performing follow-up control based on the degree of certainty with respect to the output result.

(2) In step S240, the Y coordinate value of the vanishing point of the road is corrected to YO ′ by the above equation (5) based on the radar detection distance DL ′ and the lower shadow height YE. By using the corrected Y coordinate value YO ′ of the road vanishing point instead of YO in equation (4), the inter-vehicle distance DL (second The distance) can be improved.

(3) In step S270, when the horizontal edge width ΔY is detected, and if the actual width estimation value Y_width of the horizontal edge has been previously calculated for the same preceding vehicle, the horizontal edge width ΔY is The inter-vehicle distance DC ′ is calculated from the actual width estimation value Y_width, and the vertical edge width ΔX is detected, and the actual width estimation value X_w of the vertical edge has previously been obtained for the same preceding vehicle.
When idth is calculated, the inter-vehicle distance DC 'is calculated from the vertical edge width △ X and the actual width estimated value X_width. Therefore, one of the horizontal edge width △ Y and the vertical edge width △ X is calculated. Even if the width of the image cannot be obtained from the image, as long as one width is obtained from the image, the inter-vehicle distance DC ′ (third distance) by the edge can be detected.

(4) In step S260, the actual width estimation values X_width and Y_width are obtained from the above equations (A1) and (A2) based on the inter-vehicle distance DL (second distance) due to the shadow. If the actual width estimation values X_width and Y_width are obtained when the inter-vehicle distance DL is detected by the shadow below, when the process shifts to the detection of the inter-vehicle distance DC ′ by the edge, the same preceding vehicle is previously estimated. By using the estimated actual widths X_width and Y_width, the inter-vehicle distance DC ′ can be immediately detected by the edge.

(5) In step S260, the correction values X_width and Y_width of the actual width estimation value are calculated based on the inter-vehicle distance DL ′ (first distance) to the preceding vehicle detected by the laser radar 10 in the above (A3). , (A4), the detection accuracy of the inter-vehicle distance DC ′ (third distance) by the edge can be improved by correction based on highly accurate distance data.

(6) In step 280, a weight is set to 0.4 with respect to the certainty regarding the inter-vehicle distance DL ′ (first distance) detected by the radar, and the certainty about the inter-vehicle distance DL (second distance) due to the shadow is calculated. Since the weight is set to 0.3 for the degree and the weight is set to 0.3 for the certainty regarding the inter-vehicle distance DC ′ (third distance) based on the edge width, the preceding vehicle exists in the detection area of the laser radar 10 and the When the distance DL 'is detected, a high confidence evaluation can be given, corresponding to the fact that the distance detection accuracy of the inter-vehicle distance DL' is higher than the distance detection accuracy of the inter-vehicle distance DL or the inter-vehicle distance DC '.

(7) In step S280, if a shadow is detected and the road vanishing point correction is performed, the confidence is set to 1.0 and the shadow is detected.
If the road vanishing point correction process is not performed, the confidence is set to 0.5, so that the detection accuracy of the inter-vehicle distance DL (second distance) due to the shadow depends on the positional accuracy of the road vanishing point. Correspondingly, it is possible to give a higher confidence evaluation when the height of the road vanishing point is corrected than when the height of the road vanishing point is not corrected.

(8) In step S280, the horizontal edge width ΔY in the screen is calculated, and if the correction value Y_width of the estimated actual value of the horizontal edge has been calculated before, the certainty factor is set to 1.0, If the horizontal edge width △ Y in the screen is calculated and the actual width estimated value Y_width of the horizontal edge has been calculated before, the certainty factor is set to 0.5. Distance)
Corresponding to the estimation accuracy of the actual width estimation value Y_width,
It is possible to give a higher certainty evaluation when the actual width estimation value Y_width is corrected than when the actual width estimation value Y_width is not corrected.

(9) In step S280, when the vertical edge width △ X is detected and the correction value X_width of the estimated actual value of the vertical edge has been calculated before, the change in the road yaw angle at the position of the preceding vehicle is determined. If it is less than the predetermined value, the confidence is 1.
In the case where the vertical edge width 検 出 X is detected and the correction value X_width of the estimated actual value of the vertical edge has been calculated before, and the change in the road yaw angle at the position of the preceding vehicle exceeds a predetermined value, Since the certainty factor is set to 0.5, the detection accuracy of the inter-vehicle distance DC ′ (third distance) based on the edge width corresponds to the detection accuracy decreasing as the vertical edge of the preceding vehicle becomes slanted. When the front of the vehicle is a curve, a lower confidence evaluation can be given as compared with the case where the front of the host vehicle is not a curve.

(10) In step S280, when the horizontal edge width ΔY is detected and the correction value Y_width of the estimated actual value of the horizontal edge has been previously calculated, the change in the road yaw angle at the position of the preceding vehicle is determined. If it is less than the predetermined value,
1.0, the horizontal edge width △ Y is detected, and the correction value Y_width of the actual width estimated value of the horizontal edge has been calculated before, and if the change in the road yaw angle at the preceding vehicle position exceeds a predetermined value, Since the certainty factor is set to 1.0, the detection accuracy of the inter-vehicle distance DC ′ (third distance) based on the edge width hardly changes even if the horizontal edge of the object in front of the host vehicle is inclined. Accordingly, the same certainty factor evaluation can be given both when the front of the host vehicle is a curve and when the front of the host vehicle is not a curve.

(11) If the calculation result of the total certainty factor is equal to or more than the predetermined value in step S290, step S300
Then, the process proceeds from step S310 to step S310, in which the following control for causing the own vehicle to follow the preceding vehicle is performed.
20 to control the speed of the vehicle traveling at a preset speed, the laser radar 10 and the image processing device 3
Following only the preceding vehicle for which the detection result at 0 is likely to be correct, the following control using the inter-vehicle distance information with the correctly detected preceding vehicle can be performed.

(Other Embodiments) The traveling control device for a vehicle according to the present invention has been described based on the first embodiment. However, the specific configuration is not limited to the first embodiment. Changes and additions of the design are permitted without departing from the gist of the invention according to each claim of the claims.

For example, in the first embodiment, a laser radar is shown as an example of a radar, but it is needless to say that the same effect can be obtained even with another type of radar such as a millimeter wave radar.

[Brief description of the drawings]

FIG. 1 is a system diagram showing a configuration of a vehicle travel control device according to a first embodiment.

FIG. 2 is a perspective view showing a vehicle equipped with the vehicle travel control device of the first embodiment.

FIG. 3 is a flowchart illustrating a flow of image processing executed by an image processing device in the vehicle travel control device of the first embodiment.

FIG. 4 is a flowchart showing a flow of a follow-up control process executed by a follow-up control controller in the vehicle travel control device of the first embodiment.

FIG. 5 is an operation explanatory view showing a road vanishing point.

FIG. 6 is a diagram showing a relationship between a camera detection area and each week distance.

FIG. 7 is a diagram showing a preceding vehicle image on a straight road where one preceding vehicle exists.

FIG. 8 is a diagram illustrating a front vehicle image on a curved road where two front vehicles exist.

FIG. 9 is a diagram illustrating a forward vehicle image representing a vertical edge width.

FIG. 10 is a diagram showing a preceding vehicle image representing a horizontal edge width.

FIG. 11 is a diagram showing each certainty factor depending on the presence or absence of detection.

FIG. 12 is a table showing a principle and a defect of each inter-vehicle distance detection based on a laser, a shadow height, and a width in a screen.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Vehicle traveling control apparatus 10 Laser radar (radar) 20 Front camera (imaging means) 30 Image processing apparatus 40 Vehicle speed sensor 50 Tracking controller 80 Throttle actuator 90 Brake actuator S240 Vanishing point correction step (road vanishing point correcting means) S260 Actual Width estimation value correction step (object width estimation means,
S270 Inter-vehicle distance calculating step (first distance detecting means, second distance detecting means, third distance detecting means) S280 Detection certainty calculating step (detecting certainty calculating means, total certainty) Calculation means) S290 preceding vehicle certainty determination step S300 control object setting step (distance output means) S310 preceding vehicle follow-up control step (travel control means) S320 vehicle speed control step (travel control means)

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) B60R 21/00 B60R 21/00 624G 5H180 626 626G 5J084 627 627 5L096 G01B 11/00 G01B 11/00 B G01C 3 / 06 G01C 3/06 Z G01S 17/93 G06T 1/00 280 G06T 1/00 280 330B 330 7/60 180B 7/60 180 G08G 1/16 E G08G 1/16 G01S 17/88 A F term (reference) 2F065 AA06 BB15 CC11 DD00 FF01 FF04 FF61 JJ03 JJ26 MM02 MM07 QQ23 QQ25 QQ31 QQ33 RR07 SS03 UU05 2F112 AD05 BA20 CA05 FA21 FA32 FA35 FA45 GA10 3D044 AA25 AA31 AB01 AC26 AC56 AC01 AD04 AEB AE3A01 AE3 EC01 EC04 FA02 FA04 FA11 FB01 FB02 5B057 AA16 BA0 2 CA08 CA13 CA16 DB02 DB03 DB09 DC02 DC03 DC08 DC16 5H180 AA01 CC03 CC04 CC14 LL01 LL04 LL09 5J084 AA05 AB01 AB20 AC02 BA03 FA03 5L096 AA06 AA09 BA04 CA04 FA06 FA64 FA66 FA67 GA30

Claims (11)

[Claims] 1. Recognizing a distance to an object in front of a host vehicle,
A travel control device for a vehicle that performs travel control based on the distance information, wherein a first distance detecting means for detecting a distance between the own vehicle and an object ahead of the own vehicle by a radar, and an imaging means for imaging the front of the own vehicle. And, based on the position of the shadow of the object ahead of the vehicle and the position of the road vanishing point in the image ahead of the vehicle captured by the imaging unit, detecting a distance between the vehicle and the object ahead of the vehicle. A distance detection unit, and detects a distance between the host vehicle and an object ahead of the host vehicle based on an object width ahead of the host vehicle in an image ahead of the host vehicle captured by the imaging unit and an actual value of the object width. Third distance detection means, Based on the detection results of the first distance detection means, the second distance detection means, and the third distance detection means, the detected distance between the vehicle and the object in front of the vehicle Each detection confidence for Calculating the total certainty in the distance detection between the own vehicle and the object in front of the own vehicle based on the certainty of each detection calculated by the detection certainty calculating means to be output. An overall certainty factor calculating means, and if the calculation result of the overall certainty factor calculating means is equal to or more than a predetermined value, one of the detection results of the first distance detecting means, the second distance detecting means, and the third distance detecting means And a distance output unit that outputs the distance as a distance between the host vehicle and an object in front of the host vehicle. 2. The travel control device for a vehicle according to claim 1, wherein the second distance detecting means determines a distance between the own vehicle and an object ahead of the own vehicle detected by the first distance detecting means. A travel control device for a vehicle, comprising: a road vanishing point correcting unit that corrects a height of a road vanishing point in an image based on the road vanishing point. 3. The traveling control device for a vehicle according to claim 1, wherein the third distance detecting means includes at least a horizontal width of an object in front of the own vehicle in the image and an object in front of the own vehicle in the image. A distance between the host vehicle and an object in front of the host vehicle based on the vertical width of the vehicle. 4. The travel control device for a vehicle according to claim 1, wherein the third distance detecting means determines a distance between the own vehicle and an object ahead of the own vehicle detected by the second distance detecting means. A vehicle travel control device comprising an object width estimating means for estimating an actual value of an object width in front of the host vehicle based on the vehicle width. 5. The travel control device for a vehicle according to claim 1, wherein the third distance detecting means determines a distance between the own vehicle and an object ahead of the own vehicle detected by the first distance detecting means. A vehicle travel control device comprising: an object width estimation value correction unit that corrects an object width ahead of the host vehicle based on the object width estimation unit based on the object width estimation unit. 6. The traveling control device for a vehicle according to claim 1, wherein the detection certainty calculating means detects when the first distance detecting means detects a distance between the host vehicle and an object in front of the host vehicle. The degree of certainty regarding the distance, the degree of certainty regarding the distance between the own vehicle detected by the second distance detecting means and the object in front of the own vehicle, and the own vehicle and the own vehicle detected by the third distance detecting means A travel control device for a vehicle, wherein the confidence level is set higher than a certainty factor regarding a distance to a forward object. 7. The travel control device for a vehicle according to claim 2, wherein the detection certainty calculating means is convinced of the distance between the own vehicle and the object ahead of the own vehicle detected by the second distance detecting means. A travel control device for a vehicle, wherein the degree is set higher when the height of the road vanishing point in the image is corrected by the road vanishing point correcting means than when there is no correction. 8. The travel control device for a vehicle according to claim 5, wherein the detection certainty calculating means is convinced of the distance between the own vehicle and the object ahead of the own vehicle detected by the third distance detecting means. A vehicle travel control device, wherein the degree is set higher when the estimated value of the object width ahead of the own vehicle is corrected by the object width estimated value correction means than when the correction is not made. 9. The vehicle travel control device according to claim 1, further comprising road yaw angle detection means for detecting a relative yaw angle of the road relative to the vehicle at an object position ahead of the vehicle. The certainty factor calculating means calculates the certainty factor for the distance between the own vehicle and the object ahead of the own vehicle detected based on the vertical edge-to-edge distance by the third distance detecting means, when the relative yaw angle is a predetermined value or more. A travel control device for a vehicle, wherein the relative yaw angle is set to be lower than a case where the relative yaw angle is less than a predetermined value. 10. The traveling control device for a vehicle according to claim 1, further comprising a road yaw angle detecting means for detecting a relative yaw angle of a road at an object position in front of the own vehicle with respect to the own vehicle, The certainty factor calculating means detects the certainty factor for the distance between the own vehicle and the object ahead of the own vehicle detected based on the horizontal edge-to-edge distance by the third distance detecting means, and is detected by the relative yaw angle detecting means. A travel control device for a vehicle, wherein the travel control device is set independently of a relative yaw angle. 11. The travel control device for a vehicle according to claim 1, wherein when a calculation result of said total certainty calculating means is equal to or more than a predetermined value, the vehicle is controlled based on an output result of said distance output means. Carry out following control to follow the preceding vehicle traveling ahead,
A travel control device for a vehicle, comprising: travel control means for controlling a vehicle speed at which the vehicle travels at a preset speed when a calculation result of the total certainty calculation means is less than a predetermined value.