JP2002303671A - Object-classification discrimination apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an object-classification discrimination apparatus by which the classification of an object existing at the front of a vehicle can be discriminated accurately. SOLUTION: An image at the front of an own vehicle imaged by a camera 3 is image-processed in an image processing part 4, and a traffic lane at the front of the own vehicle is recognized. In a discrimination part 5, on the basis of a positional relationship with reference to traffic lanes in detection points indicating the object at the front of the own vehicle recognized by a scanning laser radar 2, a detection point indicating a delineator is extracted from the detection points. The remaining detection points excluding the detection point indicating the delineator are grouped, and the classification of the object indicated by the detection points is discriminated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object type discriminating apparatus for discriminating the type of an object mounted on a vehicle and existing in front of the vehicle.

[0002]

2. Description of the Related Art Conventionally, an inter-vehicle distance between a host vehicle and a preceding vehicle traveling ahead of the host vehicle is measured, and the running of the host vehicle is automatically controlled so as to follow the preceding vehicle while keeping the inter-vehicle distance constant. 2. Description of the Related Art A vehicle traveling automatic control device has been proposed. In order to appropriately control the running of the own vehicle using such a vehicle running automatic control device, it is necessary to accurately determine the situation in front of the own vehicle and accurately determine the position of the preceding vehicle in the lane in which the own vehicle runs. It is important to understand.

[0003] In general, this type of vehicle traveling automatic control device measures the position of a preceding vehicle using a laser radar. That is, this type of vehicle traveling automatic control device normally emits a pulsed laser beam toward the front of the host vehicle and detects the reflected laser beam reflected by a reflector attached to the rear of the preceding vehicle. The position of the preceding vehicle with respect to the host vehicle is measured based on the delay time from when the laser light is emitted to when the reflected laser light is detected.

[0004] By the way, on a road on which a vehicle travels and around the road, there are usually various objects such as a delineator which is a reflector beside the road besides the preceding vehicle. Then, for example, when the lane in which the vehicle travels is curved, reflected laser beams from these various objects are detected. At this time, if the reflected laser light reflected by the reflector of the preceding vehicle in the lane in which the own vehicle travels cannot be identified from the reflected laser light reflected by various objects, the presence of the preceding vehicle in front of the own vehicle is determined. In addition, it is impossible to accurately grasp the position of the vehicle, and it is impossible to perform appropriate traveling control.

To cope with such a problem, for example, in Japanese Patent Application Laid-Open No. 6-195600, a scanning laser radar is used to scan a relatively wide area on and around a road, and the detection result is obtained. A technique has been proposed based on which the existence and position of a preceding vehicle are grasped.

In this method, an object including a delineator existing on and around a road is detected by a scanning laser radar, and the delineator is determined based on the movement of these objects and the distribution of reflection intensity. Then, based on the determined delineator, the lane in which the host vehicle travels is estimated, and a moving object located ahead of the estimated lane is recognized as a preceding vehicle, and the position is grasped.

[0007]

However, in the method disclosed in Japanese Patent Application Laid-Open No. HEI 6-195600, a delineator is determined only from the distribution of the motion and reflection intensity of an object detected using a scanning laser radar. Based on this, the lane in which the host vehicle travels and the preceding vehicle traveling in that lane are determined, so that objects other than the delineator and the preceding vehicle, such as adjacent lanes, travel in the detection range of the scanning laser radar. When there are a large number of vehicles, signboards, and the like, it is difficult to accurately distinguish a delineator from these objects. In such a case, there is a problem that the existence and position of the preceding vehicle cannot be properly grasped, and the traveling control of the vehicle cannot be properly performed.

The present invention has been made in order to solve the above-described problems, and it is possible to accurately determine the type of an object existing in front of a vehicle and to more appropriately perform vehicle traveling control. It is an object of the present invention to provide an object type discriminating device that performs

[0009]

According to a first aspect of the present invention, there is provided an object type discriminating apparatus for discriminating a type of an object mounted on a vehicle and existing in front of the vehicle. Detecting an electromagnetic wave reflected from an object present in front of the vehicle, an object recognizing means for recognizing the presence and the position of the object as a detection point, an imaging means for imaging an image in front of the vehicle, Image processing is performed on an image captured by the image capturing unit, and a lane detecting unit that detects a lane ahead of the vehicle, and a positional relationship of a detection point recognized by the object recognizing unit with respect to the lane detected by the lane detecting unit. It is characterized by comprising a determining means for determining the type of the object indicated by the detection point.

In the object type discriminating apparatus according to the first aspect, the presence and the position of the object ahead of the vehicle are recognized as the detection points by the object recognizing means, and the information is supplied to the discriminating means. Further, an image in front of the vehicle is captured by the imaging unit, and the image is subjected to image processing by the lane detection unit to detect a lane in front of the vehicle, and the information is supplied to the determination unit. Then, the type of the object indicated by these detection points is determined by the determination means based on the positional relationship of each detection point with respect to the detected lane.

According to a second aspect of the present invention, in the object type determining apparatus according to the first aspect, the determining means is provided at a side of a road from all the detection points recognized by the object recognizing means. A detection point indicating a roadside reflector is extracted.

In the object type discriminating apparatus according to the second aspect, the discriminating means grasps the positional relationship of all the detection points with respect to the lane, and indicates the roadside reflector from all the detection points based on this. Detection points are extracted.

According to a third aspect of the present invention, in the object type determining apparatus according to the second aspect, the determining means extracts a plurality of detection points located near the lane detected by the lane detecting means. When a line connecting the plurality of detected points is along the lane, the object indicated by each of the detected points is recognized as a roadside reflector.

In the object type discriminating apparatus according to the third aspect, a plurality of detection points located near the lane are extracted by the discriminating means. Then, when the line connecting these detection points is along the lane detected by the lane detecting means, the object indicated by the plurality of detection points located near these lanes is recognized as the roadside reflector.

According to a fourth aspect of the present invention, in the object type discriminating apparatus according to the second aspect, the object recognizing means continuously recognizes the detection points at predetermined time intervals, and the discriminating means sets the time. When the motion vector connecting the detection points indicating the same object whose position has changed according to the change is along the lane detected by the lane detection means, the object indicated by these detection points is recognized as a roadside reflector. It is characterized by the following.

In the object type discriminating apparatus according to the fourth aspect, the detection points indicating the object in front of the vehicle are continuously recognized at predetermined time intervals by the object recognizing means, and the information is supplied to the discriminating means. Then, the motion vector connecting the detection points indicating the same object whose position has changed in accordance with the time change, that is, the detection point indicating the previous position of the same object and the detection point indicating the current position are determined by the determination unit. A vector is required. Then, when this motion vector is along the lane recognized by the lane detecting means, the object indicated by these detection points is recognized as a roadside reflector.

According to a fifth aspect of the present invention, in the object type discriminating apparatus according to the second aspect, the object recognizing means continuously recognizes the detection point at predetermined time intervals, A plurality of vectors connecting the respective detection points that may be the same object whose position has changed according to the change are respectively detected for each of a plurality of detection ranges, and the directions are substantially parallel to each other between the plurality of detection ranges. A combination of vectors along the lane detected by the lane detecting means and having substantially the same length is determined, and the detection points connected by these vectors are determined to indicate the same object, and these objects are determined. Each is characterized by being recognized as a roadside reflector.

In the object type discriminating apparatus according to the fifth aspect, the detection points indicating the object ahead of the vehicle are continuously recognized by the object recognizing means at predetermined time intervals, and the information is supplied to the discriminating means. Then, the plurality of vectors connecting the respective detection points that may be the same object whose position has changed in accordance with the time change are detected by the determination means for each of the plurality of detection ranges. That is, a plurality of detection ranges are set by the discriminating means, and a plurality of vectors connecting the respective detection points which may be the same object whose position has changed in accordance with time change in the plurality of detection ranges are respectively detected. Is done.
Then, among a plurality of vectors detected for each of the plurality of detection ranges, a combination of vectors having directions substantially parallel to each other, along a lane, and substantially equal in length is obtained between the detection ranges. It is determined that the connected detection points indicate the same object, and the objects indicated by these detection points are respectively recognized as roadside reflectors.

The invention described in claim 6 is the invention according to claims 2 to 5
In the object type determination device according to any one of the above, the determination unit sets an area where the roadside reflector will be present based on the lane detected by the lane detection unit,
A detection point indicating the roadside reflector is extracted from the detection points located in this area.

In the object type discriminating apparatus according to the sixth aspect, the lane detected by the lane detecting means is used as a reference.
The area where the roadside reflector will be located is set by the determining means. Then, a detection point indicating the roadside reflector is extracted from the detection points located in the region where the roadside reflector will be located.

The invention according to claim 7 is the invention according to claims 2 to 6
In the object type determination device according to any one of the above, the determination unit groups detection points excluding the detection point indicating the roadside reflector from all the detection points recognized by the object recognition unit. Based on the result, the type of the object indicated by a detection point other than the detection point indicating the roadside reflector is determined.

In the object type discriminating device according to the seventh aspect, the discriminating means grasps the positional relationship of all the detection points with respect to the lane, and indicates the roadside reflector from all the detection points based on this. Detection points are extracted. Then, a plurality of remaining detection points excluding the detection point indicating the roadside reflector are grouped, and based on the result, the type of the object indicated by the detection point other than the detection point indicating the roadside reflector is determined. Are respectively determined.

According to an eighth aspect of the present invention, in the object type discriminating apparatus according to the seventh aspect, the discriminating means considers an intensity distribution of reflected electromagnetic waves from the object detected by the object recognizing means, The type of an object indicated by a detection point other than the detection point indicating the roadside reflector is determined.

In the object type discriminating apparatus according to the eighth aspect, the intensity distribution of the reflected electromagnetic wave from the object in front of the vehicle is detected by the object recognizing means, and the information is supplied to the discriminating means. The type of the object indicated by the detection point other than the detection point indicating the roadside reflector is determined by the determination unit after considering the intensity distribution of the electromagnetic wave reflected from the object.

The ninth aspect of the present invention is the invention according to the seventh or eighth aspect.
In the object type determination device according to the above, the determination means,
When the object recognizing means cannot detect reflected electromagnetic waves from a specific object other than the roadside reflector, the height position of the object is detected from the distance to the object, and this height position is taken into consideration. And determining the type of the object.

In the object type discriminating apparatus according to the ninth aspect, when the object recognizing means cannot detect reflected electromagnetic waves from a specific object other than the roadside reflector, the discriminating means also determines the position of the object at that time. Then, the distance to the object is calculated, and the height position of the object is detected based on the calculated distance. Then, the type of the object is determined based on the height position of the object.

According to a tenth aspect of the present invention, in the object type determining apparatus according to any one of the second to ninth aspects, the determining means determines the position of the detected roadside reflector based on the position of the detected roadside reflector. The lane position detected by the lane detecting means is corrected.

In the object type discriminating apparatus according to the tenth aspect, the discriminating means grasps the positional relationship of all the detected points with respect to the lane, and indicates the roadside reflector from all the detected points based on this. Detection points are extracted. Then, the position of the lane detected by the lane detecting means is corrected based on the position of the detection point indicating the roadside reflector.

[0029]

According to the object type discriminating apparatus according to the first aspect, the lane ahead of the vehicle is detected based on the image ahead of the vehicle, and the lane ahead of the vehicle is detected based on the positional relationship of each detection point with respect to the detected lane. Since the type of the object indicated by these detection points is determined, it is necessary to accurately determine the type of the object existing in front of the vehicle even when the situation in front of the vehicle greatly changes due to a change in the lane, for example. Can be.

According to the object type discriminating apparatus of the second aspect, the detection point indicating the roadside reflector is extracted from the positional relationship with the lane detected based on the image in front of the vehicle. A detection point indicating a roadside reflector can be extracted very accurately from all the detection points indicating an existing object.

According to the object type discriminating apparatus according to the third aspect, when a line connecting a plurality of detection points located near the lane is along the lane, the object indicated by these detection points is used as a roadside reflector. As a result, even when the vehicle is stopped, that is, in a state where the roadside reflector does not move, the roadside reflector can be extracted very accurately from all the objects existing in front of the vehicle.

According to the object type discriminating apparatus according to the fourth aspect, the motion vector connecting the detection points indicating the same object whose position has changed with time changes along the lane.
Since this object is recognized as a roadside reflector, when the vehicle is traveling, especially when the distance between adjacent roadside reflectors is relatively wide, the roadside reflector is selected from all objects existing in front of the vehicle. The reflector can be extracted very accurately.

According to the object type discriminating apparatus according to the fifth aspect, a plurality of vectors connecting each detection point which may be the same object whose position has changed with time change are respectively provided for each of a plurality of detection ranges. Among the plurality of vectors detected for each of the plurality of detection ranges, between the respective detection ranges,
A combination of vectors whose directions are substantially parallel to each other along the lane and whose lengths are substantially equal is determined, and the detection points connected by these vectors are determined as detection points indicating the same object, and the objects indicated by these detection points are determined. Are recognized as roadside reflectors, respectively, so that the distance between adjacent roadside reflectors is relatively small during traveling of the vehicle, and it is difficult to specify the roadside reflectors for each detection range. Even in this case, the roadside reflector can be extracted with extremely high accuracy from all the objects existing in front of the vehicle.

According to the object type discriminating apparatus of the sixth aspect, an area where a roadside reflector is likely to be set based on a lane detected based on an image in front of the vehicle, is set in this area. Since the detection point indicating the roadside reflector is extracted from the detection points located, the detection point indicating the roadside reflector can be extracted very efficiently.

According to the object type discriminating apparatus of the present invention, the detection points indicating the roadside reflector are excluded from all the detection points, and the remaining detection points are grouped, and based on the result, Since the type of the object indicated by these detection points is determined, the inconvenience of not being able to perform appropriate grouping due to the presence of a detection point indicating a roadside reflector when performing grouping is effectively avoided. The type of the object other than the side reflector can be accurately determined.

According to the object type discriminating apparatus according to the eighth aspect, the types of objects other than the roadside reflectors are determined in consideration of the intensity distribution of the electromagnetic waves reflected from the objects. It is possible to determine with high accuracy.

According to the object type discriminating apparatus of the ninth aspect, the types of the objects other than the roadside reflectors are determined in consideration of the height position of the objects, so that the types of these objects can be more accurately determined. It is possible to do.

According to the object type discriminating apparatus according to the tenth aspect, the position of the lane detected based on the image in front of the vehicle is determined by:
Since the correction is performed based on the position of the extracted roadside reflector, it is possible to accurately detect the position of a distant lane that is difficult to accurately detect only from the image in front of the vehicle, and as a result, It is possible to more accurately determine the type of the object existing in front of the vehicle.

[0039]

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

FIG. 1 shows a schematic configuration of an object type discriminating apparatus to which the present invention is applied. The object type discriminating apparatus 1 shown in FIG.
Is mounted on a vehicle (hereinafter referred to as a host vehicle) and determines a type of an object existing in front of the host vehicle. The scanning laser radar 2, the camera 3, the image processing unit 4, and the determination unit 5 are provided. It is composed.

The scanning laser radar 2 emits laser light, two-dimensionally scans the front of the host vehicle in a direction parallel to the road surface, and receives the reflected laser light. The position is recognized as a detection point. The scanning laser radar 2 continuously recognizes the presence or absence and the position of an object in front of the host vehicle at a predetermined time interval ΔT. The time interval ΔT at this time is appropriately set according to the data processing capability of the determination unit 5, the accuracy required for the object type determination device 1, and the like.

The scanning laser radar 2 has the configuration shown in FIG.
As shown in FIG. 2A and FIG. 2B, the host vehicle 100 is controlled so that the scanning center axis is along the traveling direction of the host vehicle 100 and the scanning origin is located at the center of the host vehicle 100 in the width direction. It is attached. Hereinafter, the scanning origin of the scanning laser radar 2 is set as the origin, the Z axis (indicating the distance to the object in front of the host vehicle 100) is set in the scanning center axis direction of the scanning laser radar 2, and the X axis (the object in front of the host vehicle 100). Is shown in the horizontal direction of the scanning laser radar 2, a direction perpendicular to the scanning center axis of the scanning laser radar 2 and parallel to the scanning plane, and a Y axis (showing the height position of an object in front of the host vehicle 100) is shown. A coordinate system that is a direction perpendicular to the scanning plane 2 will be described as a reference coordinate system.

The camera 3 captures an image in front of the host vehicle. This camera 3 is shown in FIGS.
As shown in (B), the optical axis is along the traveling direction of the host vehicle 100 similarly to the scanning center axis of the scanning laser radar 2, and its origin is the same as the scanning origin of the scanning laser radar 2. The scanning laser radar 2 is attached to the vehicle 100 so as to be located at the center in the width direction of the scanning laser 100 and different from the scanning laser radar 2 in a direction perpendicular to the road surface.

The above is the scanning laser radar 2
And an ideal example of how to attach the camera 3 to the own vehicle 100 has been described.
The scanning center axis of the camera and the optical axis of the camera 3 are
Or the scanning origin of the scanning laser radar 2 or the origin of the camera 3
Even if the scanning laser radar 2 and the camera 3 are not at the center in the width direction of the
If the mounting angle and the mounting position with respect to 0 are taken into account during the geometric calculation, the scanning laser radar 2
And the camera 3 can be handled in the same manner as when the camera 3 is ideally mounted.

The image processing section 4 performs image processing on an image ahead of the host vehicle captured by the camera 3 to detect a lane ahead of the host vehicle.

The image in front of the host vehicle captured by the camera 3 is, for example, a two-dimensional image in which the X axis and the Y axis perpendicular to the running direction of the host vehicle are coordinate axes, as shown in FIG. The image processing unit 4 first extracts a plurality of feature points of a white line image that separates lanes from the two-dimensional image captured by the camera 3. Then, the position coordinates of these feature points are shown in FIG.
As shown in (2), the coordinates are converted into two-dimensional coordinates using the Z axis and the X axis as coordinate axes, that is, position coordinates on two-dimensional coordinates parallel to the scanning surface of the scanning laser radar 2.

Next, the image processing unit 4 converts the plurality of feature points converted into points on two-dimensional coordinates parallel to the scanning plane of the scanning laser radar 2 using a method such as a least square error. The shape of the white line is determined by fitting with the function that most closely matches the distribution of the plurality of feature points, thereby detecting the lane ahead of the host vehicle.

Specifically, for example, assuming that the shape of the white line is a circular arc approximation, as shown in FIG. 5, the distribution best fits the distribution of a plurality of feature points converted into points on two-dimensional coordinates. A function of an arc having a radius of curvature, and the position and orientation of the vehicle 100 with respect to the lane, that is, R (radius of curvature of the arc), Xc (horizontal distance from the vehicle to the white line), θ (lane in FIG. 5) The value of the YAW angle of the own vehicle with respect to the vehicle 10 is updated each time a feature point is extracted.
The lane ahead of 0 is detected.

The method of detecting a lane from an image captured by a camera mounted on a vehicle is described in detail in
No. 0-11580.

The information indicating the position of the object ahead of the host vehicle recognized by the scanning laser radar 2 and the information of the lane detected by the image processing unit 4 are supplied to the discriminating unit 5. Has become. The position of the object ahead of the host vehicle recognized by the scanning laser radar 2 is determined by the determination unit 5 as a plurality of detection points on two-dimensional coordinates parallel to the scanning surface of the scanning laser radar 2 as shown in FIG. Be recognized.
The lane detected by the image processing unit 4 is also recognized as a pair of white line images on the same coordinates.

The discriminating unit 5 obtains the positional relationship of the object existing ahead of the host vehicle with respect to the lane based on the plurality of detection points indicated on the two-dimensional coordinates and the image of the pair of white lines. Is used to determine the type of the object existing in front of the host vehicle. Note that this discriminating unit 5
Is continuously performed with a time interval (ΔT) equal to the processing time interval by the scanning laser radar 2.

Hereinafter, the processing in the determination section 5 will be specifically described.

When the detection unit 5 is supplied with the detection points indicating the object in front of the host vehicle and the lane information, first, based on the positional relationship of each detection point with respect to the lane, the discriminator 5 selects a delineator from these detection points. Is performed to identify a detection point indicating “1”.
The delineator is a rod-shaped reflector with a reflective surface installed on the side of the road so that the state of the lane can be properly judged even at night, and is usually installed continuously at predetermined intervals along the lane. Is what is being done.

As a method of identifying a detection point indicating a delineator from a plurality of detection points indicating an object ahead of the host vehicle, a method of identifying a detection point indicating a delineator based on a connection between adjacent detection points can be considered.

In this example (hereinafter, referred to as a first example),
First, from among a plurality of detection points indicating an object in front of the vehicle,
As shown in FIG. 7, each detection point located near the lane is extracted. At this time, the size of the detection point is also determined, and only the detection point that is considered to have a size corresponding to the delineator is extracted. Then, it is determined whether the line connecting these detection points is along the lane.If the line connecting these detection points is along the lane, each of these detection points is recognized as a delineator. I do.

In the first example, since the detection point indicating the delineator is identified without considering the movement of the detection point over time, the detection point indicating the delineator is appropriately identified by a simple process. can do. This first example is very effective when, for example, the own vehicle is stopped and the detection point does not move.

As a method for identifying a detection point indicating a delineator from a plurality of detection points indicating an object in front of the host vehicle, a movement of the detection point over time is obtained, and the movement of the detection point is determined based on the movement of the detection point. A method of identifying a detection point indicating a delineator is also conceivable.

In this example (hereinafter referred to as a second example),
First, the processing data of this time (time T) is superimposed on the processing data of the previous time (time T-ΔT), and as shown in FIG. It is shown as a different detection point. Then, a detection point located near the lane is extracted from the plurality of detection points. At this time, a range (detection range) in which the position of the object can be changed during the processing time interval ΔT is set based on the traveling speed of the host vehicle 100 and the processing time interval ΔT. And the previous point (time T
-ΔT) is recognized.

Then, for each detection range, this time (time T)
A vector (motion vector) connecting the detection point of (1) and the detection point of the previous time (time T-ΔT) is obtained, and based on this motion vector, it is determined whether or not the extracted detection point is a detection point indicating a delineator. . That is, since the delineator is a stationary object installed along the lane, the relative motion of the delineator viewed from the own vehicle is indicated by a motion vector of the same length in the opposite direction to the motion vector of the own vehicle, and , The direction of the motion vector is along the lane. Therefore, the current (time T) detection point and the previous (time T−Δ
By confirming that the motion vector connecting the detection points of T) is of the same length in the opposite direction to the motion vector of the own vehicle, it is determined that the objects indicated by these detection points are stationary objects. By confirming that the motion vector is along the lane, it is possible to determine that the object indicated by these detection points is a delineator. At this time, the first
As in the above example, if the determination of the size of the detection point is also performed at the same time, the identification of the delineator can be performed more accurately and efficiently.

In the second example, since the detection point indicating the delineator is identified based on the movement of the detection point with the passage of time, an extremely accurate method for identifying the delineator when the host vehicle is running is very important. It is effective for It should be noted that the second example is combined with the first example described above so that the detection point indicating the delineator is identified based on both the movement of the detection point over time and the connection between adjacent detection points. Then, more accurate identification can be performed.

By the way, in the second example described above, when the distance between adjacent delineators is narrow, FIG.
As shown in (A), the previous (time T-ΔT) detection point indicating the adjacent delineator enters the detection range set for each detection point indicating the delineator, and the motion vector of the delineator is reduced. It may be difficult to determine uniquely.

In such a case, first, it is determined whether or not the interval between adjacent delineators is equal to or larger than a predetermined value for uniquely obtaining a motion vector of each delineator. If the interval is smaller than a predetermined value, the motion vector of the delineator is obtained by the following method.

That is, first, as shown in FIG. 9B, the current (time T) detection point located in each detection range and the previous (time T-ΔT) detection point are connected, respectively.
A plurality of vectors that may be motion vectors of the delineator are obtained for each detection range. Next, a plurality of vectors obtained for each detection range are compared between the detection ranges, and a combination of vectors having directions substantially parallel to each other, along a lane, and substantially equal in length is obtained. More specifically, a plurality of vectors are evaluated for each detection range, and vectors that are considered to be along the lane are extracted from the plurality of vectors. If the vectors extracted for each detection range are substantially parallel to each other and have substantially the same length, these vectors are determined as motion vectors of the delineator.

If the motion vector of the delineator is determined by such a method, even if the interval between adjacent delineators is narrower than a predetermined value for uniquely determining the motion vector of each delineator. , The motion vector of the delineator can be determined appropriately.
Then, if the same processing as in the above-described second example is performed based on the motion vector of the delineator, it is possible to appropriately determine whether the object indicated by the detection point within the detection range is the delineator. .

Since the delineator is installed on the side of the road as described above, if the position of the lane is known, it is possible to recognize the area where the delineator will exist based on the lane. Therefore, when a detection point indicating a delineator is identified from a plurality of detection points indicating an object ahead of the host vehicle recognized by the scanning laser radar 2, image processing is performed in advance as shown in FIG. It is desirable to set an area where the delineator will be located based on the lane detected by the unit 4, and to extract a detection point indicating the delineator from the detection points located in this area. If a region where a delineator will be present is set in advance and a detection point indicating a delineator is extracted in this region, efficient processing can be performed.

However, since the lane used as a reference when setting the area where the delineator will be located is the one detected by the function fitting in the image processing section 4, it is not necessarily a faithful reproduction of the actual lane. It is not. Therefore, when setting the area where the delineator will be located, it is necessary to allow this area to have a certain width in accordance with the lane detection accuracy, that is, the fitting accuracy in the image processing unit 4. .

When there are a plurality of lanes on the road, the delineator is located near and along the left white line of the leftmost lane. Therefore, when the host vehicle 100 is traveling in the leftmost lane, FIG.
As shown in, an area where the delineator will be located near the left white line detected by the image processing unit 4 may be set, but when the own vehicle 100 is traveling in the right lane, As shown in FIG. 10 (B), a position translated by a corresponding lane width based on the number of lanes from the vicinity of the white line to the leftmost lane based on the left white line detected by the image processing unit 4. It is necessary to set the area where the delineator will be located in the.

After performing the processing described above and identifying the detection point indicating the delineator from the plurality of detection points indicating the object ahead of the host vehicle recognized by the scanning laser radar 2, the determination unit 5 determines Then, a grouping process is performed on the remaining detection points excluding the detection point indicating the delineator to determine the type of the object in front of the host vehicle.

Normally, in a vehicle ahead of the host vehicle, reflected laser beams from left and right reflectors attached to the rear of the vehicle are detected as detection point groups, respectively. Instead, the combination is recognized as a combination of a pair of detection point groups that make the same movement. In addition, the signboard installed in front of the vehicle is recognized as a set of detection points in a relatively large range because reflected laser light from a surface having a relatively large area is detected as a group of detection points.

In view of this, the discriminating unit 5 takes into account various information serving as indices for object identification, such as the shapes of the above-described groups of detected points and their temporal movements, and detects the detected points that are considered to be the same object. The groups are grouped, and the position of the object existing in front of the own vehicle is determined based on the positional relationship with respect to the lane from the grouping result, and the size, center position, movement, etc. of the grouped detection points are also determined. The type of the object ahead of the vehicle is determined. At this time, if the intensity of each detection point, that is, the intensity distribution of the reflected laser light detected by the scanning laser radar 2 is also taken into consideration, the type of the object ahead of the host vehicle can be determined with higher accuracy. Become.

In order to confirm whether or not the object ahead of the host vehicle is an object that hinders the running of the host vehicle, the discriminating unit 5 determines whether the object is on the lane in which the host vehicle runs, that is, the image processing. A determination is made as to whether the object is on the lane detected by the unit 4, or is an object on another lane or outside the lane, a signboard overhead, or the like.

Here, as a method of discriminating the type of an object existing in front of the host vehicle, conventionally, a grouping process is performed on all detection points including a detection point indicating a delineator, and a group of detected points is grouped. Based on the distribution of the movement and reflection intensity, etc., the detection point indicating the delineator is determined from these, and the lane is estimated from the detection point determined as indicating the delineator, and In general, the type of an object represented by another detection point group is determined based on the positional relationship.

However, in such a method, FIG.
As shown in FIG. 1 (A) and FIG. 11 (B), when the traffic is heavy and the vehicle in front of the host vehicle 100 hides the delineator, there is a high possibility that the lane cannot be properly estimated. . Then, even if the type of the object indicated by the detection point group is determined based on the positional relationship with the lane that was guessed incorrectly, correct determination cannot be made. If the vehicle in front of the host vehicle 100 overlaps and only one of the left and right reflectors is detected, the detection point cannot be determined to be a reflector, and in particular, as shown in FIG. When the detection point indicating the reflector is located in the vicinity of the detection point indicating the delineator, the detection point indicating the one reflector is erroneously recognized as the detection point indicating the delineator, or the detection point indicating the one reflector and the detection point indicating the delineator are different. Are inconveniently grouped, and these are erroneously recognized as left and right reflectors of the vehicle in front of the host vehicle 100. further,
As described above, when an incorrect grouping process is performed without properly estimating a lane, an object such as a signboard located at a position distant from the lane is likely to be adversely affected.

On the other hand, in the object type discriminating apparatus 1 to which the present invention is applied, the image in front of the own vehicle taken by the camera 3 is image-processed by the image processing unit 4 so that the lane ahead of the own vehicle is recognized. The determination unit 5 determines the type of the object indicated by each detection point based on the positional relationship of each detection point with respect to the lane recognized by this image processing. By effectively avoiding the problem, it is possible to appropriately determine the type of the object ahead of the host vehicle.

That is, in the object type discriminating apparatus 1 to which the present invention is applied, first, as shown in FIG.
Since the detection point indicating the delineator is extracted from all the detection points based on the lane recognized by the image processing, for example, a part of the delineator is hidden behind another object. Even in this case, the detection point indicating the delineator can be appropriately extracted. Then, since the remaining detection points excluding the detection point indicating the delineator are grouped from all the detection points, the type of the object indicated by the detection point other than the detection point indicating the delineator is determined, The inconvenience of erroneously recognizing the detection point indicating the delineator as indicating another object is prevented beforehand, and the type of the object other than the delineator can be determined with extremely high accuracy.

In the example described above, since the scanning laser radar 2 having a scanning surface in a direction parallel to the road surface recognizes the object ahead of the host vehicle as a detection point, the height position of the object ahead of the host vehicle is determined. No information about is available. Therefore, for example, when a signboard is installed above the lane in which the vehicle runs,
The detection point indicating this signboard is recognized as a detection point on the lane where the host vehicle runs, and the object indicated by the detection point is an obstacle that hinders the running of the host vehicle, or is installed above where it does not become an obstacle. I can't tell if it's a sign that was done.

Therefore, in such a case, first, based on the intensity of the detection point, that is, the intensity distribution of the reflected laser light detected by the scanning laser radar 2, the object indicated by the detection point on the lane is detected. Judge whether it is a signboard. That is, since the signboard has a strong reflection intensity from the entire surface and has substantially the same reflection intensity over the entire surface, it is highly likely that the signboard having such a reflection intensity is a signboard. Since it is unlikely that the signboard is installed on the road, it can be recognized as a signboard installed above.

However, when the preceding vehicle is traveling in the vicinity of the own vehicle, the laser beam reflected by the vehicle body of the preceding vehicle may be detected by the scanning laser radar 2 and its intensity distribution is indicated by a signboard. May be close to the intensity distribution of the reflected laser light from Further, even when there is a falling object having a relatively large reflecting surface on the lane, the intensity distribution of the reflected laser light from this falling object may be close to the intensity distribution of the reflected laser light from the signboard.

Therefore, in such a case, information on the height of the object ahead of the host vehicle is indirectly obtained by the method described below, and the object is used as an obstacle to hinder the running of the host vehicle. It is desirable to determine whether there is a sign or a signboard installed above that does not become an obstacle.

That is, the scanning laser radar 2 having a scanning surface in a direction parallel to the road surface has a detection angle of usually about 2 to 3 degrees even in a direction perpendicular to the road surface. Therefore, the signboard installed above the lane can be detected by the scanning laser radar 2 when it is far away from the vehicle 100 as shown in FIG. When the signboard installed above the lane relatively approaches the host vehicle 100 as the host vehicle 100 moves forward, the signboard is out of the detection range of the scanning laser radar 2. From this, when the distance between the target object and the host vehicle 100 becomes a predetermined value, it is determined whether the object is out of the detection range of the scanning laser radar 2 and the height of the object is determined. The position can be indirectly recognized, and based on this, it can be determined whether the target object is an obstacle that hinders traveling of the vehicle 100 or a signboard installed above that does not become an obstacle. Can be determined.

As described above, the object type discriminating apparatus 1 to which the present invention is applied recognizes the lane ahead of the own vehicle by processing the image ahead of the own vehicle captured by the camera 3 by the image processing unit 4. Then, the determination unit 5 determines the type of the object indicated by each detection point based on the positional relationship of each detection point with respect to the lane recognized by the image processing. The delineator, the preceding vehicle, a signboard above the lane, an object outside the lane, and the like can be appropriately determined.

However, since the lane used as a reference for discriminating the type of the object is obtained by performing image processing (function fitting) on an image in front of the host vehicle, the lane far away from the host vehicle is used. May not always be accurate. That is, as the image captured by the camera 3 becomes farther, the reproducibility deteriorates, and the lane detection accuracy also decreases. It is usually up to about 50 m that a high-precision image can be obtained by a general vehicle-mounted camera. On the other hand, a scanning laser radar that is generally mounted on a vehicle can accurately detect an object located at a distance of about 120 to 130 m.

Therefore, in order to determine the type of the object existing in front of the own vehicle with higher accuracy, after the delineator is detected by the above-described processing, the lane obtained by the image processing based on the detected delineator is determined. It is desirable to correct the position of the vehicle, in particular, the position of the lane far away from the host vehicle.

As described above, the scanning laser radar 2
If the lane position obtained by the image processing is corrected based on the position of the delineator recognized as the detection point of, the position of the lane far away from the vehicle can be accurately recognized, and Since the type of the object is determined based on the lane, the type of the object can be determined with higher accuracy.

As described above, the object type discriminating apparatus 1 to which the present invention is applied can determine the type of the object ahead of the host vehicle with extremely high accuracy. If the preceding vehicle is detected and the speed control and the braking control of the own vehicle are performed so as to keep the inter-vehicle distance to the preceding vehicle constant, the traveling control of the vehicle can be performed extremely appropriately. Further, if the object type discriminating apparatus 1 detects an obstacle or the like existing on the traveling lane and notifies the driver of the obstacle by a warning sound or the like,
The driving operation of the vehicle by the driver can be extremely appropriately assisted.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an object type discrimination device to which the present invention is applied.

2A and 2B are diagrams illustrating a scanning laser radar and a camera that constitute the object type discriminating apparatus and a mounting state of the camera with respect to a host vehicle; FIG. 2A schematically illustrates a state of the host vehicle viewed from a side; (B) is a figure which shows typically the mode seen from the upper part of the own vehicle.

FIG. 3 is a diagram showing an image in front of a host vehicle taken by the camera.

FIG. 4 is a diagram illustrating a method of recognizing a lane ahead of the host vehicle from an image captured by the camera.

FIG. 5 is a diagram specifically illustrating a method of recognizing a lane in an image processing unit when a lane ahead of the host vehicle is assumed to be a circular arc approximation.

FIG. 6 is a diagram illustrating a state in which an object ahead of the host vehicle is recognized as a detection point on two-dimensional coordinates by a determination unit, and a lane is recognized as an image of a pair of white lines on the same coordinates.

FIG. 7 is a diagram illustrating a first method for identifying a detection point indicating a delineator from a plurality of detection points indicating an object ahead of the host vehicle.

FIG. 8 is a diagram illustrating a second method for identifying a detection point indicating a delineator from a plurality of detection points indicating an object in front of the host vehicle.

9A and 9B are diagrams illustrating a method of appropriately obtaining a motion vector of a delineator in a case where a distance between adjacent delineators is narrow in the second example, and FIG. 9A illustrates detection of each detection point indicating a delineator. FIG. 3B is a diagram illustrating a state in which a range is set, and FIG. 3B is a diagram illustrating a portion A in FIG.

FIG. 10 is a diagram showing a state in which a region where a delineator will be present is set based on a lane recognized by image processing, and FIG. 10 (A) shows a case where the own vehicle is traveling in a left lane; FIG. 7B is a diagram showing an area set when the host vehicle is traveling in the right lane.

11A and 11B are diagrams illustrating a state in front of the host vehicle when the host vehicle is traveling in a lane with a large amount of traffic, where FIG. 11A is a diagram illustrating an actual state in front of the host vehicle, and FIG. FIG. 4 is a diagram illustrating detection points recognized by a laser radar.

FIG. 12 is a diagram showing a situation in which a lane ahead of the host vehicle is erroneously estimated and grouping of each detection point is not properly performed.

FIG. 13 is a diagram illustrating a state in which a lane ahead of the host vehicle is appropriately recognized and detection points are appropriately grouped.

FIG. 14 is a diagram illustrating a method of indirectly recognizing a height position of an object in front of a host vehicle by a scanning laser radar having a scanning surface in a direction horizontal to a road surface.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Object type discrimination device 2 Scanning laser radar 3 Camera 4 Image processing part 5 Discrimination part

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2F065 AA03 AA54 BB00 BB28 CC11 DD04 FF01 FF04 FF11 JJ03 JJ09 JJ19 JJ26 MM02 MM07 MM16 QQ18 QQ25 QQ32 QQ38 SS09 SS15 UU05 5H180 AA01 CC03 CC04 A04 AB04 AB04 AB01 AB04 FA03

Claims (10)

[Claims] 1. An object type discriminating device mounted on a vehicle for discriminating a type of an object existing in front of the vehicle, wherein the electromagnetic wave is emitted in front of the vehicle and reflected from an object existing in front of the vehicle. By detecting the object, an object recognizing unit that recognizes the presence and the position of the object as a detection point, an image capturing unit that captures an image in front of the vehicle, and performing image processing on an image captured by the image capturing unit. A lane detecting unit that detects a lane ahead of the vehicle, and a positional relationship between a detection point recognized by the object recognizing unit and a lane detected by the lane detecting unit. An object type discriminating device comprising: a discriminating unit for discriminating a type. 2. The method according to claim 1, wherein the determination unit extracts a detection point indicating a roadside reflector installed on a roadside from all the detection points recognized by the object recognition unit. 2. The object type determination device according to 1. 3. The method according to claim 1, wherein the determining unit extracts a plurality of detection points located near the lane detected by the lane detecting unit, and determines whether a line connecting the extracted plurality of detection points is along the lane. 3. The object type discriminating apparatus according to claim 2, wherein the object indicated by each of the detection points is recognized as a roadside reflector. 4. The object recognizing means continuously recognizes the detection points at a predetermined time interval, and the discriminating means moves the detection points indicating the same object whose position has changed with time. The object type discriminating apparatus according to claim 2, wherein when the vector is along the lane detected by the lane detecting means, the object indicated by these detection points is recognized as a roadside reflector. 5. The object recognizing means continuously recognizes the detection points at predetermined time intervals, and the discriminating means determines whether each of the detected objects is likely to be the same object whose position has changed with time. A plurality of vectors connecting the detection points are detected for each of a plurality of detection ranges, and between these detection ranges, directions are substantially parallel to each other, along the lane detected by the lane detection means, and have substantially equal lengths. 3. A combination of the following vectors is determined, and detection points connected by these vectors are determined to indicate the same object, and these objects are respectively recognized as roadside reflectors. Object type discrimination device. 6. The discriminating means sets an area where the roadside reflector will be located based on the lane detected by the lane detecting means, and sets the area from among the detection points located in this area. 6. The object type discriminating apparatus according to claim 2, wherein a detection point indicating a roadside reflector is extracted. 7. The determination means groups detection points excluding the detection points indicating the roadside reflector from all the detection points recognized by the object recognition means, and based on the result, The object type determination device according to claim 2, wherein the type of the object indicated by a detection point other than the detection point indicating the roadside reflector is determined. 8. The type of the object indicated by a detection point other than the detection point indicating the roadside reflector, by taking into account the intensity distribution of the reflected electromagnetic wave from the object detected by the object recognition unit. The object type determination device according to claim 7, wherein the object type is determined. 9. When the object recognizing means cannot detect reflected electromagnetic waves from a specific object other than the roadside reflector, the object recognizing means determines a height position of the object from a distance from the object. The object type discriminating apparatus according to claim 7, wherein the object type is detected and the type of the object is discriminated in consideration of the height position. 10. The method according to claim 2, wherein the determining unit corrects the position of the lane detected by the lane detecting unit based on the extracted position of the detection point indicating the roadside reflector. An object type determination device according to any one of the above.