JP2001202497A - Method and system for detecting preceding vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect the vehicular gap and the attitude angle to a preceding vehicle by photographing a marker provided in the rear part of the preceding vehicle. SOLUTION: A marker 20 is provided in the rear part of a preceding vehicle 2000 which runs ahead of a user's vehicle 100 and is photographed by a camera of the vehicle 100. The marker 20 consists of four marker elements arranged in horizontal and vertical directions, and the vehicular gap and the attitude angle like a yaw are detected by arrangement in the horizontal direction and that in the perpendicular direction of an obtained image. Since marker elements are not arranged on the same perpendicular line, an influence of smear of a CCD camera is excluded. Since four marker elements are arranged, the vehicular gap and the attitude angle can be detected even if one of them cannot be detected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a system and method for detecting a preceding vehicle, and more particularly to a system and method for detecting a preceding vehicle by identifying a marker provided on the preceding vehicle.

[0002]

2. Description of the Related Art Hitherto, a system has been developed for detecting the preceding vehicle and controlling the traveling of the own vehicle, and is applied to a system in which a group of vehicles travels in a platoon while controlling the distance between the vehicles. Have been.

[0003] For example, Japanese Patent Application Laid-Open No. 10-38560 describes a technique in which two reflectors of a preceding vehicle are viewed in stereo and the inter-vehicle distance to the preceding vehicle is detected from the parallax.

[0004]

However, with only two reflectors, for example, the relative attitude angle (relative yaw angle) between the preceding vehicle and the host vehicle because the preceding vehicle has entered a curved road.
However, this cannot be detected even if is changed, and there is a problem that control becomes complicated. In addition, if one reflector cannot be detected with only two reflectors, the inter-vehicle distance cannot be calculated. Therefore, further improvement in environmental resistance has been required.

[0005] The present invention has been made in view of the above-mentioned problems of the prior art, and has as its object an excellent environmental resistance,
Another object of the present invention is to provide a system and a method capable of detecting a relative attitude angle with respect to a preceding vehicle.

[0006]

According to another aspect of the present invention, there is provided a system for detecting a preceding vehicle by detecting a marker provided on the preceding vehicle, wherein the markers are arranged at predetermined intervals in the horizontal and vertical directions. Imaging means for acquiring a front image, and extracting the images of the at least three marker elements from the front image, and the positional relationship between the images of the marker elements Calculating means for calculating the positional relationship between the preceding vehicle and the own vehicle from the vehicle. The calculated positional relationship is, for example, an inter-vehicle distance and a relative posture angle, and the relative posture angle is specifically a yaw angle, a pitch angle, or the like. When the yaw angle exists, the vertical distance of the marker element image does not change but the horizontal distance changes as compared to the case where the yaw angle does not exist. Also, when the pitch angle exists, the horizontal distance of the marker element image does not change but the vertical distance changes as compared to the case where the pitch angle does not exist. Therefore,
By arranging the marker elements in the horizontal and vertical directions, not only the inter-vehicle distance but also the relative attitude angle can be detected.
In addition, in order to arrange in the horizontal and vertical directions, it is sufficient that at least three or more marker elements are provided. Cannot be detected. Therefore, when emphasis is placed on environmental resistance, the marker element is 4
It is preferable that the number be equal to or more than the number.

Preferably, the marker element is a luminous body. Thereby, the marker element image can be reliably obtained by the imaging means regardless of the traveling environment.

Further, it is preferable that the marker element is a luminous body that emits light at a predetermined cycle. This is because by setting the predetermined period, the light source can be distinguished from other light sources. The predetermined period includes not only a constant period but also a non-constant period, for example, a random period.

It is preferable that the at least three marker elements do not exist on the same vertical line, and the two marker elements exist on the same horizontal line. When a CCD camera is used as an imaging unit, smear may occur when charges accumulated in the CCD element are transferred in the vertical direction. If the two marker elements are positioned on a vertical line, that is, in the direction in which electric charges are transferred in the CCD camera, smears will not be able to detect the marker element image. Therefore, erroneous detection due to smear can be prevented by arranging the marker element images so as not to overlap each other in the charge transfer direction.

[0010] When four marker elements are arranged, it is preferable that the four marker elements be located at the vertices of a trapezoid. This makes it possible to position the two marker elements in the horizontal direction but not in the same vertical direction. Also, in this case, there are two sets of two marker elements arranged in the horizontal direction, both of which can be a reference when calculating the inter-vehicle distance and the relative attitude angle.

When five marker elements are arranged, four marker elements are located at the vertices of the trapezoid, and the other marker is within the trapezoid and on the same horizontal line and vertical line as the four markers. It is preferable to dispose them at a position that does not exist.
It is preferable that the arrangement of the marker elements does not include a similar shape. This is because if a similar shape is included, it becomes impossible to specify which of the actual marker elements the detected marker element image corresponds to. With the above arrangement, the marker elements can be arranged in the horizontal and vertical directions without including the similar shape. Specifically, in such an arrangement, even if one or two marker elements cannot be detected, the arrangement in the horizontal and vertical directions is maintained. Becomes possible.

It is preferable that the apparatus further includes control means for controlling an exposure amount of the image pickup means based on the distance between the vehicles when the distance to the preceding vehicle is calculated. As the inter-vehicle distance increases, the light amount of the marker element incident on the imaging means decreases. Therefore, by controlling the exposure amount in accordance with the inter-vehicle distance, specifically, by increasing the exposure amount as the inter-vehicle distance increases, a marker element image can be reliably obtained regardless of the inter-vehicle distance.

It is preferable that the apparatus further comprises a control means for controlling the amount of exposure of the image pickup means based on the calculated attitude angle when the relative attitude angle with respect to the preceding vehicle is calculated. If the marker element as a light emitter has directivity, the amount of light incident on the imaging means changes according to the attitude angle. Therefore, by controlling the amount of exposure in accordance with the attitude angle, a marker element image can be reliably obtained regardless of the attitude change between the preceding vehicle and the own vehicle.

[0014] It is preferable that the apparatus further comprises a detecting means for detecting an order of the own vehicle with respect to the preceding vehicle based on a detection pattern of the marker element. At least three marker elements of the present invention are arranged in the horizontal and vertical directions. Therefore, when a plurality of vehicles are platooning, the light emission pattern of the marker element according to the order is determined for each vehicle,
By detecting the pattern of the preceding vehicle,
That is, the order of the vehicle in the platoon can be easily confirmed. For example, when the number of marker elements is three and a pattern is generated by emitting any two marker elements,
There are a total of three detection patterns, and it is possible to detect the order in which four vehicles run in platoon.

[0015] Further, a current position detecting means for detecting a current position of the own vehicle, a storage means for storing road shape data, and based on the current position of the own vehicle, the distance to the preceding vehicle and the road shape, Estimating means for estimating the relative attitude angle between the preceding vehicle and the own vehicle; and comparing means for comparing the relative attitude angle estimated by the estimating means with the relative attitude angle calculated by the calculating means. Is preferred. Estimate the attitude angle that would be obtained based on the road shape data,
By comparing the calculated attitude angle with the actually calculated attitude angle, the accuracy of the calculated attitude angle can be evaluated.

The present invention also provides a method for detecting a preceding vehicle by detecting a marker provided on the preceding vehicle. The method includes obtaining a front image, extracting at least three marker element images from the front image, and calculating a positional relationship between the preceding vehicle and the own vehicle from a positional relationship between the marker element images. And steps.

Here, it is preferable to calculate at least a distance to the preceding vehicle and a relative attitude angle with respect to the preceding vehicle from the positional relationship of the image of the marker element. Estimating the relative attitude angle between the preceding vehicle and the own vehicle based on the distance to the preceding vehicle and the road shape, and comparing the estimated relative attitude angle with the calculated relative attitude angle. It is preferable that the method further includes steps.

[0018]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a system conceptual diagram of the present embodiment. The vehicle 100 and the preceding vehicle 200 are present, and a marker 20 is provided at the rear of the preceding vehicle 200. The vehicle 100 has a preceding vehicle detection sensor 10 mounted thereon, and the preceding vehicle detection sensor 10 detects the marker 20 of the preceding vehicle 200 to recognize the preceding vehicle. Although only one preceding vehicle 200 exists in the figure, a plurality of preceding vehicles 200 may exist. The processing in this case will also be described later. The marker 20 is composed of a plurality of marker elements (4 in the figure).
Are arranged in a predetermined arrangement, and the arrangement or positional relationship of the marker elements is determined by the vehicle 100.
Is known.

FIG. 2 is a block diagram showing the configuration of the preceding vehicle detection sensor 10. The preceding vehicle detection sensor 10 includes a wide-angle camera 10a, a telephoto camera 10b, and a steering mechanism 10.
c, a recognition processing ECU 10d. The wide-angle camera 10a and the telephoto camera 10b are for photographing the marker 20, and the obtained wide-angle image and telephoto image are supplied to the recognition processing ECU 10d. Note that the two types of cameras, the wide-angle camera 10a and the telephoto camera 10b, are not necessarily indispensable, and it is also possible to mount only a single camera to shoot a marker image. CCDs can be used as the cameras 10a and 10b.

The steering mechanism 10c rotates the wide-angle camera 10a and the telephoto camera 10b in a horizontal plane to drive the preceding vehicle 20.
This is for surely photographing the 0 marker. Specifically, when the relative yaw angle between the own vehicle 100 and the preceding vehicle 200 is detected, the marker image is placed at the center of the image based on the yaw angle. , Or rotationally driven so as to fall within a predetermined angle of view. The method for calculating the relative yaw angle will be described later.

Recognition processing ECU (electronic control unit) 10d
Is the image from the wide-angle camera 10a or the telephoto camera 1.
0b, the image of the marker 20 is detected in the image from the position of the preceding vehicle 20 based on the positional relationship of the marker elements in the marker 20.
The distance to 0 and the relative attitude angle are calculated. Regarding the inter-vehicle distance, the positional relationship of the marker elements provided in the preceding vehicle 200 is known, and the constant (focal length) of the optical system is also known. The distance between the vehicle and the preceding vehicle 200 can be easily calculated. The relative attitude angle can be calculated from the ratio between the horizontal distance and the vertical distance in the marker image. That is, the own vehicle 100 and the preceding vehicle 20
If there is a relative yaw angle between 0 and 0, the vertical distance of the marker image is the same as when the yaw angle is 0, but the horizontal distance changes (shortens) depending on the yaw angle. Therefore, the yaw angle can be calculated based on the ratio of the distance between the horizontal direction and the vertical direction. (The yaw angle cannot be calculated in the horizontal direction alone because of the influence of the inter-vehicle distance.
Since the vertical distance is affected by the inter-vehicle distance as in the horizontal direction, the yaw angle can be calculated from the ratio between the two.) The calculated inter-vehicle distance and attitude angle are supplied to an ECU (Electronic Control Unit) 12 mounted on the own vehicle 100, execute vehicle speed control and steering control, and follow the preceding vehicle 200.
The recognition processing ECU 10d is supplied with a signal IG from an ignition switch, and can operate when the ignition is turned on.

FIG. 3 shows the arrangement and positional relationship of marker elements in the marker 20. The marker 20 is composed of four marker elements 20a, 20b, 20c and 20d, and the marker elements 20a and 20b are arranged on the same straight line a in the horizontal direction.
The marker elements 20c and 20d are located on another straight line b in the horizontal direction. The straight line a and the straight line b are parallel to each other. It can be said that the marker elements 20a and 20b and the marker elements 20c and 20d are located in the left-right direction. In addition, it can be said that the marker elements 20a and 20c and the marker elements 20b and 20d are located vertically or vertically. The distance between the marker elements 20a and 20b is shorter than the distance between the marker elements 20c and 20d. Marker elements 20a and 20c and marker element 20b
And 20d are not located on the same vertical line, respectively.
The four marker elements are located at the four vertices of the trapezoid.

Each of the marker elements 20a to 20d can be composed of, for example, an infrared LED that emits infrared light. In this case, as a camera, a camera that receives only infrared light and a camera that receives other visible light components are combined, and by subtracting the latter from the former, other light sources, for example, only the marker image is not affected by sunlight. Can be obtained reliably.

The markers 20 are for allowing a following vehicle viewed from the own vehicle to recognize the own vehicle, and need to be arranged in such a manner that false detection can be suppressed as much as possible. The conditions that the marker 20 should satisfy from such a viewpoint are as follows.

(1) The marker elements are not arranged in the vertical direction in order to avoid the influence of smear that occurs when charges are transferred in the vertical direction in the CCD camera. In FIG. 3, the marker elements 20a and 20c and the marker elements 20b and 20d are not arranged on the same vertical line.

(2) No similar shape is included in the marker arrangement.
This is because including a similar shape provides an erroneous reference when any of the marker elements becomes undetectable. If no similar shape is included, it is possible to correctly determine which marker element in a known positional relationship is the detected marker element.

(3) The arrangement is such that the inter-vehicle distance and the relative attitude angle can be detected even if any of the marker elements cannot be detected. In order to detect the inter-vehicle distance and the relative attitude angle, the horizontal distance and the vertical distance in the marker image are required. For this purpose, at least three marker elements that are not on the same straight line are required. However, it is preferable to use four or more marker elements as a reserve in case any of the three marker elements cannot be detected. In FIG. 3, four marker elements 20a-20
This is why d is provided.

(4) In order to ensure distance accuracy, the distance between marker elements is made as large as possible. However, the marker element is arranged within a range that can be detected at the minimum detection distance (for example, 0.8 m) at the CCD angle of view determined by the minimum turning radius, the total length of the vehicle, the vehicle front overhang, and the like. When detecting the inter-vehicle distance based on the horizontal distance of the marker element, it is preferable in the configuration of FIG. 3 to make the distance between the marker elements 20a and 20b or the marker elements 20c and 20d as large as possible.
The marker elements 20c and 20d are larger than the distance between a and 20b.
Can be detected with higher accuracy.

With the above configuration, the specific processing of the recognition processing ECU 10d will be described below with reference to the processing flowchart of FIG.

First, the recognition processing ECU 10d detects a pair on a horizontal line that is a candidate for a marker element from the input image (S101). Specifically, the following processing is performed. That is, the input image is binarized into a high-luminance part and a low-luminance part, and the position of a high-luminance pixel that is continuous in the horizontal direction and the number of consecutive pixels are obtained. Then, a labeling process is performed on a portion having a high luminance, and the center of gravity, the upper and lower ends, the positions of the left and right ends, and the number of pixels included in the label are obtained.
If the number of labels is too large (5 or more), an error flag is set to set the detection distance to 0, and if the number of pixels forming the label is too large, an error flag is set to set the detection distance. Set to 0. When many labels exist in the vertical direction of the screen, it is determined that smear has occurred, and the label is excluded from detection targets. Then, for the detected label, a combination in which the inclination between the straight line connecting the centers of gravity of the two labels and the horizontal axis is within a predetermined angle (for example, ± 11 degrees) is extracted as a horizontal line pair of the marker candidate. Here, when there is no pair to be extracted, an error flag is set as no detection target and the detection distance is set to 0.

Next, it is determined whether or not there is one horizontal line pair (S102). In the case of the arrangement of the marker elements shown in FIG. 3, two pairs should be originally detected, but if only one pair is detected, the other pair that should be present has one marker element outside the recognition area range. Alternatively, it is assumed that the image is hidden by smear (in the arrangement of FIG. 3, the influence of smear is effectively suppressed as described above, but the possibility of hiding is not zero depending on the width of smear). Execute marker detection logic (S1 described later)
05).

On the other hand, if there are two horizontal line pairs, a combination of four marker element candidates is detected from the combination of the horizontal line pairs (S103). Specifically, which image in the image corresponds to each of the marker elements 20a to 20d is determined from the ratio of the position and the length of the two horizontal line pairs. For example,

Α × actual measurement ratio <lower horizontal line length / upper horizontal line length <β × actual measurement ratio where α and β are predetermined numbers, for example, α = 0.68, β = 1.
Fifteen

Γ × actual measurement ratio <line middle interval / upper horizontal line length <δ × actual measurement ratio where γ and δ are predetermined numbers, for example, γ = 0.68, δ = 1.
If the inclination from the vertical line at the center of the two horizontal lines 35 satisfies all the conditions of a predetermined angle (for example, ± 7 degrees), it is determined that the images are the marker elements 20a to 20d. here,
The actual measurement ratio is a ratio of the actual distance between the marker elements, and is a known value.

Next, a line in a good state is selected from the two combinations (S104). The line in a good state is a line in which noise such as sunlight does not exist. If the two sets of detection states are comparable, the lower line, that is, the line with the longer distance is selected. The reason is that, as described above, the use of the lower line having a longer distance improves the distance accuracy. In addition to the sunlight, the stop lamp may be lit as the noise. In this case, the horizontal distance is significantly larger than the original horizontal distance between the marker elements, so that the noise can be determined.

On the other hand, if YES in S102, that is, if only one horizontal line pair can be detected, three horizontal line pairs and three remaining marker element candidates are detected from the detected horizontal line pair.
A combination of the marker elements is detected (S105).
For example, when a single marker element exists at the upper left of a horizontal line pair, the detected marker elements are 20a and 20a.
c and 20d, and when a single marker element exists at the lower right of the horizontal line pair, the detected marker elements can be specified as 20a, 20b, and 20d. Then, one set of detected horizontal lines is selected.

After selecting the line, the distance to the preceding vehicle 200 is calculated from the distance between the centers of gravity of the marker elements of the selected line (S106). Specifically, the focal length of the camera is f, and the actual distance between marker elements (for example, the distance between the marker elements 20c and 20d) attached to the preceding vehicle 200 is D,
Assuming that the line length between the images of the detected marker elements 20c and 20d is d, the distance L from the camera to the preceding vehicle 200 is
Is

L = D × f / d

After calculating the inter-vehicle distance, the attitude angle of the vehicle, specifically, the relative yaw angle is calculated based on the distance between the selected horizontal line and the vertical direction (S107). The yaw angle can be calculated from the horizontal distance and the vertical distance, for example, the center distance between lines / the lower line length, as described above.
As relative attitude angles, there are pitch angles in addition to yaw angles,
The pitch angle is generated by the curvature of the vertical section of the road, but is sufficiently large (60R or more) as compared with the horizontal curvature (minimum 20R), and can be ignored at the inter-vehicle target distance (for example, within 20m).

As described above, in the present embodiment, the distance to the preceding vehicle 200 and the relative posture of the preceding vehicle 200 are obtained by photographing the marker 20 by arranging the four marker elements in a predetermined arrangement as the marker 20. Corners can be detected.

In the present embodiment, if there are a plurality of combinations that match the four marker elements, it is conceivable that there are a plurality of preceding vehicles 200 and those markers 20 exist on the same image. Therefore, in such a case, the marker element having the largest horizontal line length may be selected. Thereby, the inter-vehicle distance and the attitude angle with respect to the nearest preceding vehicle 200 can be detected.

In the present embodiment, if the error flag is set, the inter-vehicle distance cannot be detected, but in such a case, the inter-vehicle distance is measured using another measuring means, for example, a millimeter wave radar device. Should be measured. That is, a detection system using a preceding vehicle detection sensor using a camera and a detection system using a millimeter wave radar device are installed in parallel, and if the preceding vehicle cannot be detected by the preceding vehicle detection sensor, the system switches to the millimeter wave radar device immediately and automatically. It is preferable to control the running of the vehicle.

In this embodiment, the case where the marker 20 is composed of four marker elements has been described. However, it is sufficient that at least three or more marker elements are provided as described above. Can also.

FIG. 5 shows an example in which the marker 20 is formed by arranging five marker elements. The arrangement of the four marker elements 20a to 20d is the same as in the case of FIG. 3, but one more marker element 20e is arranged at substantially the center of the trapezoid formed by the four marker elements. The marker element 20e is not on the same vertical line as any of the marker elements 20a to 20d, and is not on the same horizontal line as the marker elements 20a to 20d. Also, these five marker elements do not constitute any analog. Therefore, it can function as the marker 20 as in the case of FIG. Further, in the configuration of FIG. 5, even if any two marker elements cannot be detected for some reason, the remaining three marker elements are detected to detect the inter-vehicle distance and the attitude angle to the preceding vehicle 200. Therefore, the arrangement is robust with respect to environmental resistance.

FIG. 6 shows an example in which two marker elements cannot be detected in the marker element arrangement of FIG. (A) is a case where the marker elements 20e and 20d cannot be detected, the marker elements 20a, 20b and 20c are detected, and the distance and the posture angle are calculated based on these. The inter-vehicle distance is calculated from the horizontal line length of the marker elements 20a and 20b.
0b and the vertical distance and the horizontal distance of the marker element 20c. (B) is a marker element 20
In the case where a and 20b cannot be detected, the inter-vehicle distance is calculated from the horizontal line length of the marker elements 20c and 20d, and the attitude angle is calculated from the ratio between the vertical distance and the horizontal distance of the marker element 20e. (C) shows marker elements 20b and 20c
Is not detectable, and in this case, as can be seen from the figure, there is no pair on the horizontal line. However, since it is possible to uniquely determine which marker element these marker elements correspond to from the three detection marker elements, for example, the inter-vehicle distance can be calculated from the horizontal distance between the marker element 20a and the marker element 20d, The attitude angle is calculated from the ratio of the vertical distance and the horizontal distance between the marker element 20a and the marker element 20d. (D) is a case where the marker elements 20a and 20c cannot be detected,
In this case as well, there is no pair on the horizontal line. However, as in the case of (c), it is possible to uniquely determine which of the three images corresponds to each of the marker elements 20a to 20e.
The inter-vehicle distance can be calculated from the horizontal distance between 0e and 20d, and the attitude angle can be calculated from the ratio of the vertical distance and the horizontal distance between the marker elements 20b and 20d.

Further, in the present embodiment, the inter-vehicle distance to the preceding vehicle 200 is calculated from the horizontal distance of the marker element image, and the attitude angle is detected from the ratio of the horizontal distance to the vertical distance. The inter-vehicle distance to the preceding vehicle 200 can be calculated from the distance, and the yaw angle can be calculated from the horizontal distance.

As shown in FIG. 7, a plurality of cameras are arranged in a horizontal plane such that the respective photographing ranges partially overlap each other, and the marker 20 is photographed. It is also possible to align as shown to substantially increase the angle of view.

When the marker element is formed of a light-emitting body such as an infrared LED, it is preferable to make the marker element emit light continuously and blink it at a predetermined cycle. This makes it possible to recognize, as a marker image, an image whose luminance increases at a predetermined cycle when photographed by a camera. Further, in the case of blinking at a predetermined cycle, it is also preferable to blink at a unique cycle assigned to each vehicle. For example, by setting the marker element provided in the vehicle A to blink at a cycle Ta and the marker element provided to the vehicle B to blink at a cycle Tb, the marker images of a plurality of preceding vehicles 200 are included in the captured image. Even if they exist, the two can be easily identified. The period may be constant, but it is also preferable to use a random code from the viewpoint of discrimination from noise.

Further, the inter-vehicle distance and attitude angle, for example, the yaw angle detected in the present embodiment can be used for vehicle running control, but can also be used for other controls, specifically, for controlling the exposure of a camera. it can. That is, when the marker element is an illuminant such as an LED, as the distance increases, the amount of light emitted from the LED as viewed from the camera decreases. The amount of light emitted from the LED decreases. Therefore, by detecting the inter-vehicle distance and the yaw angle, and by increasing the exposure amount of the camera according to the inter-vehicle distance and the yaw angle detected by the recognition processing ECU 10d, the marker 20 is always reliably detected regardless of the inter-vehicle distance and the yaw angle. It is possible to do.

In the case where a plurality of vehicles travel in a row, it is preferable that each vehicle travels while recognizing the position of the current vehicle (if it does not recognize it). In this case, even if the marker 20 of the first vehicle is detected instead of the second vehicle despite the fact that the marker is originally the third vehicle, the travel control is performed so as to follow the first vehicle. Thus, the marker 20 can be made to function as a means for recognizing the number of the vehicle in this way.

FIG. 9 shows a marker 2 as such means.
A processing flowchart when 0 is made to function is shown. First, prior to the processing, a unique blinking pattern is assigned to each vehicle that forms a platoon. An example is shown in Table 1.

[0050]

[Table 1] In the table, it is assumed that 10 vehicles run in a platoon. The first vehicle turns off the marker element 1 (marker element 20a) and turns on the other marker elements. 2
The second vehicle turns off the marker element 2 (marker element 20b) and turns on the other marker elements. The same applies to the following, for example, the tenth vehicle is marker element 2 (marker element 20b)
And the marker element 3 (marker element 20c) is turned off, and the other marker elements are turned on. In this way, marker elements to be turned off (or turned on) are determined in accordance with the order in the platoon, and by detecting this turn-off pattern (or turn-on pattern), the position of the own vehicle is determined. Is to judge. Assuming that there are five marker elements, one light-out turns off 5
As a result, it is possible to assign 10 patterns with two lights off.

In FIG. 9, first, the first to ninth cars turn on all marker elements of their own marker 20 (S201).
The subsequent recognition processing ECU confirms the lighting of the marker (S202). If you cannot confirm the lighting of the marker,
It is determined that the marker of the preceding vehicle has failed (S20).
3). When the lighting of the marker can be confirmed, the first car turns off the marker element according to the rules shown in Table 1 (S204). If this turn-off can be confirmed for the second car (S205), it means that the second car is surely confirmed to be its own car. Then, if it can be confirmed, the first car turns on all the marker elements again, and then the second car
The car turns off according to the rules shown in Table 1.
This turning off is confirmed by the third car, and the same processing is repeated until the ninth car (S208). As described above, by detecting the turn-off pattern of the preceding vehicle, it is possible to easily confirm the number of the own vehicle (S209). Here, when the first car is turned off, the third car detects only the lighting marker of the second car (or the turning-off marker of the first car and the lighting marker of the second car), so that it is erroneously determined that the own car is the second car. I do not recognize. In addition, for example, if the second car cannot confirm this even if the first car is turned off, it is determined that an error has occurred in the first car and the second car (S207). Of course, Table 1
, A blinking cycle may be assigned to each vehicle instead of the light-off pattern.

FIG. 10 shows a vehicle 1 according to another embodiment.
00 is shown in FIG. Although the preceding vehicle detection sensor 10 including the camera and the vehicle ECU 12 are provided in the same manner as the above-described embodiment, in this embodiment, the road-to-vehicle communication device 22, the vehicle-to-vehicle communication device 24, the magnetic sensor 2
6 and a road database 28. The road-to-vehicle communication device 22 is a communication device that transmits and receives data to and from a beacon provided on a road surface, and can detect the position of the vehicle with the accuracy of the installation interval of the beacon based on information from the beacon.

The inter-vehicle communication device 24 is a communication device for transmitting and receiving data to and from a preceding vehicle and a following vehicle, and transmits traveling data of the own vehicle and traveling data of another vehicle (acceleration, deceleration,
Vehicle speed).

The magnetic sensor 26 detects the magnetism of the magnetic nail embedded in the road surface, and can detect the position of the vehicle with the accuracy of the installation interval of the magnetic nail. By combining the road-vehicle communication device 22 and the magnetic sensor 26, the position of the own vehicle can be detected with high accuracy.

The road database 28 is a database for storing road shapes.
The curve radius and the like are stored.

The vehicle ECU 12 reads the road shape corresponding to the vehicle position detected by the road-to-vehicle communication device 22 and the magnetic sensor 26 from the road shape database, and detects the preceding vehicle detection sensor 10 based on the read road shape data. It is evaluated whether the relative attitude angle with respect to the preceding vehicle is correct. Specifically, the relative posture angle between the preceding vehicle and the own vehicle is estimated based on the road shape data at a predetermined distance ahead of the own vehicle position (for example, up to 100 m ahead) and the detected inter-vehicle distance. For example, if the road shape data is a curved road having a curve radius R and the inter-vehicle distance is L, the relative yaw angle is

Yaw angle = COS ⁻¹ {(lower line length / center distance between lines) · (upper / lower marker interval / lower marker interval)} The lower line length / center distance between lines is a ratio on the image, and the upper / lower marker interval / lower marker interval is the ratio of the actual length. Then, the estimated yaw angle is compared with the yaw angle detected by the preceding vehicle detection sensor 10 to evaluate the detected yaw angle.

FIG. 11 shows a vehicle EC according to this embodiment.
A detailed processing flowchart of U12 is shown. First, the number, inter-vehicle distance, and attitude angle of the preceding vehicle are acquired based on the detection result from the preceding vehicle detection sensor 10. The number of preceding vehicles is the number of the markers 20 in the image detected by the preceding vehicle detection sensor 10, and the inter-vehicle distance and the attitude angle are obtained for each marker 20 (S301). Next, the own vehicle position is acquired from the road-vehicle communication device 22 and the magnetic sensor 26 (S302).
Of course, the vehicle position may be detected using DGPS instead of these. Then, the road shape data (road linear information) located a predetermined distance ahead of the vehicle position is read from the road database 28 (S303). If the posture angle to be evaluated is a yaw angle, the curve radius may be read as road shape data, and if the posture angle is a pitch angle, the longitudinal gradient may be read. The distance from which the road shape data should be read is variable according to the vehicle speed of the own vehicle (increase as the vehicle speed increases)
It is also preferable to determine according to the detected inter-vehicle distance.

After reading the road shape data, the attitude angle of the preceding vehicle is estimated from the detected inter-vehicle distance and the road shape data. The attitude angle is preferably estimated as a range in consideration of an error in the position of the vehicle, an error in road shape data, and the like (S304). Then, the posture angle acquired in S301 is S
It is determined whether it is within the range estimated in 304 (S3).
05). If it is within the range, it is determined that the accuracy of the attitude angle and the inter-vehicle distance detected by the preceding vehicle detection sensor 10 are sufficient, and the data of the inter-vehicle distance and the attitude angle to the nearest preceding vehicle are output (S306). A control actuator (not shown) controls the speed and steering of the own vehicle based on the inter-vehicle distance and the attitude angle. On the other hand, if the detected attitude angle does not exist within the estimation range, an error is output as an erroneous detection (S307).

Through the above processing, the reliability of the attitude angle and the inter-vehicle distance detected by the preceding vehicle detection sensor 10 can be evaluated using the road shape data.

[0060]

According to the present invention, the distance between the vehicle and the attitude angle such as the yaw angle can be easily detected by detecting the marker provided on the preceding vehicle.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram of a system according to an embodiment of the present invention.

FIG. 2 is a configuration block diagram of a host vehicle in FIG.

FIG. 3 is a configuration diagram of a marker provided at a rear part of a preceding vehicle in FIG. 1;

FIG. 4 is a processing flowchart of a recognition processing ECU in FIG. 2;

5 is another configuration diagram of a marker provided at a rear portion of a preceding vehicle in FIG. 1. FIG.

FIG. 6 is an explanatory diagram of a detection marker element in the marker of FIG.

FIG. 7 is an explanatory diagram of a camera arrangement.

FIG. 8 is an explanatory diagram of imaging in the arrangement of FIG. 7;

FIG. 9 is a flowchart of a car recognition process during platooning.

FIG. 10 is a configuration block diagram of a host vehicle according to another embodiment.

11 is a processing flowchart of a vehicle ECU in the configuration of FIG.

[Explanation of symbols]

10 preceding vehicle detection sensor, 12 vehicle ECU, 100
Own car, 200 preceding car.

Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat II (reference) B60R 21/00 628 B60R 21/00 628B 5H301 F02D 29/02 301D F02D 29/02 301 G01B 11/00 H G01B 11/00 G08G 1/16 C G08G 1/16 E G05D 1/02 K // G05D 1/02 G06F 15/62 380 F F-term (reference) 2F065 AA04 AA06 AA37 BB28 BB29 CC11 DD11 FF04 FF63 JJ03 JJ05 JJ07 A31AQ4 A4Q1 AA45 AB01 AC26 AC55 AC56 AC59 AE03 AE04 AE15 AE27 3G093 BA25 DB00 DB16 5B057 AA16 BA02 CH11 DA07 DB02 DC02 DC03 DC05 DC36 5H180 AA01 CC02 CC04 CC12 CC14 CC19 LL01 LL04 LL09 5H301 AA01 CC06 EE06 GG09 LL09 GG01 LL06

Claims (14)

[Claims] 1. A system for detecting a preceding vehicle by detecting a marker provided on the preceding vehicle, wherein the marker includes at least three marker elements arranged at predetermined intervals in the horizontal and vertical directions. Imaging means for acquiring a front image; extracting images of the at least three marker elements from the front image; and determining a positional relationship between the preceding vehicle and the own vehicle from a positional relationship of the image of the marker elements. A preceding vehicle detection system, comprising: calculation means for calculating. 2. The preceding vehicle detection system according to claim 1, wherein the marker element is a light emitter. 3. The preceding vehicle detection system according to claim 1, wherein the marker element is a light emitter that emits light at a predetermined cycle. 4. The preceding vehicle detection system according to claim 1, wherein the at least three marker elements are not on the same vertical line, and two are on the same horizontal line. 5. The preceding vehicle detection system according to claim 1, wherein there are four said marker elements, and said four marker elements are located at the vertices of a trapezoid. 6. The system of claim 1, wherein there are five said marker elements, four marker elements are located at the vertices of a trapezoid, and one other marker is within said trapezoid, A preceding vehicle detection system, which is arranged at a position that is not on the same horizontal line and vertical line as the number of markers. 7. The system according to claim 1, wherein the calculating means calculates at least a distance to the preceding vehicle and a relative attitude angle with respect to the preceding vehicle from a positional relationship of the image of the marker element. Preceding vehicle detection system. 8. The preceding vehicle detection system according to claim 7, further comprising: control means for controlling an exposure amount of said imaging means based on a distance to said preceding vehicle. 9. The preceding vehicle detection system according to claim 7, further comprising: control means for controlling an exposure amount of said imaging means based on an attitude angle to said preceding vehicle. 10. The preceding vehicle detecting system according to claim 1, further comprising: detecting means for detecting an order of the own vehicle with respect to the preceding vehicle based on a detection pattern of the marker element. . 11. The system according to claim 7, further comprising: a current position detecting means for detecting a current position of the own vehicle; a storage means for storing road shape data; and a current position of the own vehicle and the preceding vehicle. Estimating means for estimating a relative attitude angle between the preceding vehicle and the own vehicle based on the distance and the road shape; and a relative attitude angle estimated by the estimating means and a relative attitude calculated by the calculating means. A preceding vehicle detection system, comprising: comparison means for comparing an angle with the vehicle. 12. A method for detecting a preceding vehicle by detecting a marker provided on the preceding vehicle, comprising: acquiring a front image; and extracting at least three marker element images from the front image. Calculating the positional relationship between the preceding vehicle and the host vehicle from the positional relationship of the image of the marker element. 13. The preceding vehicle detection method according to claim 12, wherein at least a distance to the preceding vehicle and a relative attitude angle with respect to the preceding vehicle are calculated from a positional relationship of the image of the marker element. Method. 14. The method according to claim 13, further comprising: estimating a relative attitude angle between the preceding vehicle and the own vehicle based on a current position of the own vehicle, a distance to the preceding vehicle, and a road shape. Comparing the estimated relative attitude angle with the calculated relative attitude angle. A preceding vehicle detection method, comprising: