JP4864450B2 - Vehicle driving support device - Google Patents

Description

  The present invention relates to a vehicle driving support apparatus that determines the possibility of contact with a front three-dimensional object detected by a stereo camera, a monocular camera, a millimeter wave radar, or the like, and performs driving support control.

  In recent years, various driving assistance control devices have been proposed and put into practical use in vehicles, in which a driving environment is photographed by a camera mounted on the vehicle, a three-dimensional object is recognized, and the possibility of a collision with the host vehicle is estimated. ing. In such an apparatus for estimating the possibility of a collision between the front three-dimensional object and the host vehicle, it is required to accurately grasp the position of the front three-dimensional object, but detection by a camera, a radar device, etc. includes an error. It is necessary to specify the position of the front three-dimensional object in consideration of errors.

For example, in Japanese Patent Application Laid-Open No. 2004-330950, a radar device detects a preceding vehicle in the traveling direction of the host vehicle, calculates an overlap amount in which the preceding vehicle overlaps the predicted course of the host vehicle, and the overlap amount is predetermined. A technique is disclosed in which when a time exceeding a value exceeds a predetermined value, it is determined that the host vehicle may come into contact with a preceding vehicle, and a safety device including an alarm and automatic braking is activated.
JP 2004-330950 A

  That is, Patent Document 1 described above is intended to prevent the driver from feeling uncomfortable by operating the safety device when the overlap amount temporarily increases for some reason such as a sensor error. There is a possibility of causing a delay. Obstacle detection errors may occur depending on the driving environment and driving conditions in addition to the errors of the sensors themselves. Natural control is not performed until driving assistance control is performed while absorbing such errors.

  The present invention has been made in view of the above circumstances, and provides a vehicle driving support control apparatus capable of performing driving support control with high response and high accuracy while properly absorbing and detecting obstacle detection errors. It is an object.

The present invention provides a three-dimensional object detection means for detecting a three-dimensional object existing on the traveling area by determining a traveling area based on the traveling path of the own vehicle, and at least an area based on statistical processing according to a preset condition in front of the own vehicle. A vehicle driving support device comprising: an area setting means for setting a position; and a control execution means for executing a preset control in accordance with the position of the three-dimensional object with respect to the area determined by the statistical processing set by the area setting means. The three-dimensional object detection means extracts the three-dimensional object as a determination target for determining the possibility of contact with the host vehicle, and the area setting means determines the left end, the right end, and the front end of the area. Variable according to at least one of the own vehicle speed, the width of the three-dimensional object, the type of the three-dimensional object detection means, and the turning motion of the own vehicle, and set by at least statistical processing. Up It is characterized by computing the probability of contact from the area.

  The vehicle driving support control apparatus according to the present invention can perform driving support control with high response and high accuracy while properly absorbing and detecting obstacle detection errors.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
1 to 7 show an embodiment of the present invention, FIG. 1 is a schematic configuration diagram of a driving support control device mounted on a vehicle, FIG. 2 is a flowchart of a driving support control program, and FIG. FIG. 4 is an explanatory diagram of a first gain characteristic map for setting the lateral median, FIG. 5 is an explanatory diagram of a third gain characteristic map for setting the forward region length, and FIG. Is an explanatory diagram of a fourth gain characteristic map for setting a control gain, and FIG. 7 is a schematic explanatory diagram of a region using a contact probability calculation formula set in front of the host vehicle.

  In FIG. 1, reference numeral 1 denotes a vehicle (host vehicle) such as an automobile, and a vehicle driving support control device 2 is mounted on the vehicle 1. The vehicle driving support control device 2 mainly includes a stereo camera 3, a stereo image recognition device 4, and a control unit 5. The vehicle driving support control device 2 basically has the following description. Control is performed according to the driving support control program, and either alarm control or automatic brake control is executed according to the probability of contact with the front three-dimensional object (details of the calculation will be described later).

  That is, when the contact probability with the front three-dimensional object is low, it is left as it is, and when the contact probability is increased from this state, a sound alarm is generated from the monitor 6 or the speaker 7 (alarm control), and further from this alarm control state. When the contact probability increases, a deceleration signal is output to the automatic brake control device 8 to operate a certain automatic brake (deceleration control).

  The stereo camera 3 is composed of a pair of (left and right) CCD cameras using a solid-state image sensor such as a charge coupled device (CCD) as a stereo optical system. Are attached at regular intervals, and an object outside the vehicle is stereo-photographed from different viewpoints, and an imaging signal is output to the stereo image recognition device 4.

  Further, the host vehicle 1 is provided with a vehicle speed sensor 9 for detecting the host vehicle speed Vown, and the host vehicle speed Vown is output to the stereo image recognition device 4 and the control unit 5.

  The stereo image recognition device 4 receives the image from the stereo camera 3 and the own vehicle speed Vown from the vehicle speed sensor 9 and detects the front information of the three-dimensional object data and the white line data ahead of the own vehicle 1 based on the image from the stereo camera 3. Then, the traveling path of the host vehicle 1 (the host vehicle traveling path) is estimated (for example, estimated according to the driving state of the host vehicle 1, the white line, etc.). Then, based on the own vehicle traveling path, a traveling area is defined (for example, a 1 m wide area centered on the own traveling path), and the three-dimensional object closest to the own vehicle 1 on the traveling area is Extracted as a determination target for determining the possibility of contact with the vehicle 1. Furthermore, as will be described later, at least a region by statistical processing is set in accordance with a preset condition, this region is expressed by the contact probability calculation formula PPxz, and the contact probability PP calculated based on this contact probability calculation formula PPxz is calculated. To the control unit 5.

  Here, the processing of the image from the stereo camera 3 in the stereo image recognition device 4 is performed as follows, for example. First, distance data is generated based on the principle of triangulation from a corresponding positional shift amount for a stereo image in front of the host vehicle 1 captured by the stereo camera 3. And based on this distance data, compared with well-known grouping processing, pre-stored three-dimensional road shape data, solid object data, etc., white line data, guardrails, curbs, etc. existing along the road Side wall data and three-dimensional object data such as vehicles and pedestrians are extracted. In the three-dimensional object data, the distance to the three-dimensional object and the temporal change (relative speed with respect to the host vehicle 1) of this distance are obtained, and in particular, the three-dimensional object that is the determination target is the closest three-dimensional object on the travel region described above. The contact probability PP is calculated by the contact probability calculation formula PPxz.

  Thus, the stereo image recognition device 4 is provided with the functions of the three-dimensional object detection means and the area setting means.

  The control unit 5 receives the contact probability PP from the stereo image recognition device 4, and when the contact probability PP is equal to or higher than a preset threshold value C <b> 1, generates a sound alarm from the speaker 7 and performs alarm control. Further, when the threshold value is equal to or greater than a preset threshold value C2 (> C1), a deceleration signal is output to the automatic brake control device 8 to operate a certain automatic brake and perform deceleration control. That is, the control unit 5 is provided as control execution means. The function of calculating the contact probability PP by the contact probability calculation formula PPxz included in the stereo image recognition device 4 constitutes a control execution unit.

Next, the driving support control program according to the present embodiment executed by the vehicle driving support control device 2 will be described with reference to the flowchart of FIG.
First, in step (hereinafter abbreviated as “S”) 101, necessary parameters (image information, own vehicle speed Vown) are read, and the process proceeds to S102, where the stereo image recognition device 4 performs the three-dimensional object recognition process as described above. Execute.

  Next, the process proceeds to S103, and the stereo image recognition device 4 determines whether or not the three-dimensional object closest to the host vehicle 1 on the travel region set based on the host vehicle traveling path may come into contact with the host vehicle 1. Extract as target.

  Next, when proceeding to S104, it is determined whether or not there is a three-dimensional object to be determined. If there is no three-dimensional object to be determined, the program is left as it is, and if there is a three-dimensional object, the process proceeds to S105.

  In S105, the contact probability calculation formula PPxz is set as a region according to the flowchart shown in FIG. The coordinate system used in the present embodiment is the own vehicle 1 in which the left and right (width) direction of the own vehicle 1 is the X coordinate, the up and down direction of the own vehicle 1 is the Y coordinate, and the front and rear direction of the own vehicle 1 is the Z coordinate. Each processing is performed using a three-dimensional coordinate system in real space with reference to. In this case, with the road surface directly below the center of the two CCD cameras constituting the stereo camera 3 as the origin, the right side of the own vehicle 1 is the + side of the X axis, the upper side of the own vehicle 1 is the + side of the Y axis, and the own vehicle. The front of 1 is set as the + side of the Z axis. Further, the own vehicle traveling path is expressed as a function of z as f (z).

In the flowchart shown in FIG. 3, first, in S201, the lateral median dx0 (see FIG. 7) is set by the following equation (1). FIG. 7 illustrates a case where the own vehicle traveling path is a straight line (that is, x = 0) in order to facilitate understanding.
dx0 = k1, k2, dx0b (1)
Here, dx0b is a basic value of the lateral median value dx0 set in advance by experiments and calculations, k1 is a first gain for setting the lateral median value, and k2 is a lateral median value. This is the second gain.

  The first gain k1 for setting the horizontal median value is variably set on a map as shown in FIG. 4, for example, and is set to a larger value as the width of the determination target becomes wider. That is, when a narrow object such as a motorcycle is a determination target, the lateral median value dx0 is narrow, and when a wide object such as a truck is a determination object, the lateral median value dx0 is set wide. Thus, a large value is obtained as the contact probability PP.

  The second gain k2 for setting the lateral median value is a gain that is changed according to the type of sensor. For example, if the camera is used as a sensor, the second gain k2 is set to a large value. If it is used as a sensor, a value smaller than the value for the camera is set, and if a millimeter wave radar device is used as the sensor, a value smaller than the value for the laser radar device is set. . This is determined by reflecting the accuracy (reliability) of the dimension in the width direction when each device is used as a sensor.

  Next, in S202, setting of the lateral contact probability calculation formula PPx is executed.

The lateral contact probability calculation formula PPx is set in the following three regions (Equations (2), (3), and (4)), respectively. Note that σx ² is a dispersion obtained in advance by experiment, calculation, or the like.

When x ≦ (f (z) − (dx0 / 2)) PPx = (1 / ((2 · π) ^1/2 · σx))
Exp (− (x− (f (z) − (dx0 / 2))) ² / (2 · σx ² )) (2)
When (f (z) − (dx0 / 2)) <x <(f (z) + (dx0 / 2)) PPx = 1.0 (3)
When x ≧ (f (z) + (dx0 / 2)) PPx = (1 / ((2 · π) ^1/2 · σx))
Exp (− (x− (f (z) + (dx0 / 2))) ² / (2 · σx ² )) (4)
That is, Equations (2) and (4) are probability density functions of Gaussian distribution, and the left and right ends of the region ahead of the host vehicle are regions by statistical processing.

Next, in S203, the value of the forward area length dz0 is set according to the following equation (5).
dz0 = k3 · dz0b (5)
Here, dz0b is a basic value of the forward area length dz0 set in advance by experiments and calculations, and k3 is a third gain for setting the forward area length.

  For example, the third gain k3 for setting the forward region length is variably set on a map as shown in FIG. 5, and as the vehicle speed Vown becomes higher, naturally, a longer region is set forward. ing.

  Next, in S204, the setting of the forward contact probability calculation formula PPz is executed.

The forward contact probability calculation formula PPz is set in the following two regions (Equations (6) and (7)), respectively. Note that σz ² is a dispersion obtained in advance by experiment, calculation, or the like.

When z ≧ dz0, PPz = (1 / ((2 · π) ^1/2 · σz))
Exp (-(z-dz0) ² / (2 · σz ² )) (6)
When z <dz0, PPz = 1.0 (7)
That is, Equation (6) is a probability density function of Gaussian distribution, and the front end of the area in front of the host vehicle is an area by statistical processing.

Next, in S205, the control gain rm is set by the following equation (8).
rm = k4 · k5 · k6 · rmb (8)
Here, k4 is a fourth gain for setting the control gain, k5 is a fifth gain for setting the control gain, and k6 is a sixth gain for setting the control gain.

  The fourth gain k4 for setting the control gain is variably set on a map as shown in FIG. 6, for example, and is set to a larger value as the width of the determination target becomes wider. That is, when a narrow object such as a motorcycle is a determination target, the control gain rm is narrowed. When a wide object such as a truck is a determination target, the control gain rm is set wide and the contact gain is reduced. A large value can be obtained as the probability PP.

  The fifth gain k5 for setting the control gain is variably set in advance according to the type of sensor, and is set according to the accuracy of each sensor. For example, when the detection target is detected by the camera, the fifth gain k5 is 0. .8, 0.9 is set for the millimeter wave radar device, 1.0 is set for the mixed camera and millimeter wave radar device.

  The sixth gain k6 for setting the control gain is variably set according to the operating condition of the turn signal switch. When the turn signal switch is ON, it is set in advance larger than when the turn signal switch is OFF. Value is set (for example, 70% when ON). That is, when making a turn, it is preferable to reduce the contact probability PP by reducing the area to be determined.

In S206, the contact probability calculation formula PPxz is set according to the following equation (9).
PPxz = PPx · PPz · rm (9)

  That is, as shown in FIG. 7, the contact probability calculation formula PPxz according to the present embodiment is formed by a Gaussian distribution at the left and right ends and the front end of the region, and the type of sensor that detects the determination target, the vehicle speed Vown, and the determination target The width and turning operation (turn signal switch ON / OFF) are variably set.

  Returning again to the flowchart of FIG. 2, after setting the contact probability calculation formula PPxz as described above in S105, the process proceeds to S106, where the contact probability PP is calculated.

  This is obtained by substituting the coordinate of the center of the width of the determination target in the above equation (9). That is, as described with reference to FIG. 7, it is obtained by substituting the coordinates (dxf, dzf) of the determination target rear part.

  Note that the processes of S101 to S106 described above are processes performed on the stereo image recognition apparatus 4 side, and the subsequent processes of S107 to S110 are processes performed on the control unit 5 side.

  In S107, the contact probability PP calculated in S106 is compared with a preset threshold C1, and if the contact probability PP is smaller than the threshold C1 (in the case of PP <C1), the program is directly exited.

  Conversely, if the contact probability PP is equal to or greater than the threshold value C1 (when PP ≧ C1), the process proceeds to S108 and, for example, a prerecorded voice indicating that the vehicle is approaching the determination target (obstacle) ahead. To warn.

  In step S109, the contact probability PP is compared with a preset threshold C2 (> C1). If the contact probability PP is smaller than the threshold C2 (in the case of PP <C2), the program is exited as it is.

  Conversely, if the contact probability PP is greater than or equal to the threshold C2 (when PP ≧ C2), the process proceeds to S110, the automatic brake control is executed so as to add a preset amount of brake, and the program is exited.

  As described above, according to the present embodiment, alarm control and automatic brake control are not simply performed based on the overlap amount with the determination target, but alarm control or automatic control is performed based on the positional relationship between the statistical area and the determination target. Since the brake control is performed, it is possible to perform driving support control with high response and high accuracy while appropriately absorbing and absorbing obstacle detection errors.

  In contrast to statistically setting the determination region as in the present embodiment, it is also conceivable that the determination target itself is statistically set and calculated. However, when the calculation is performed with the determination target itself set statistically, many detection and calculation parameters, such as the speed and distance of the determination target, are required. On the other hand, the error may increase and the calculation amount increases. The response may be worse. On the other hand, in this embodiment, since the statistical element is introduced not on the determination target but on the side of the region, the above calculation for each determination target is not necessary, and the calculation result can be obtained easily and quickly. Therefore, accurate control with good response is possible.

  In this embodiment, an area based on a Gaussian distribution is set and controlled at both the left and right ends and the front end of the front area. It may be what you do.

  In this embodiment, an example in which both alarm control and deceleration control can be performed based on the contact probability has been described, but only one of them may be performed.

  Furthermore, although the region in consideration of the statistical elements proposed in the present embodiment has been described as an example in the case of using it as a region for determining warning and deceleration control, other regions for determining the preceding vehicle ( It can also be used as an area for determining whether the preceding vehicle is a preceding vehicle used for follow-up control or the like.

  In this embodiment, the preceding vehicle is recognized on the basis of the image from the stereo camera. However, other technologies, for example, recognition based on information from the millimeter wave radar and the monocular camera are performed. It may be.

Schematic configuration diagram of a driving support control device mounted on a vehicle Flow chart of driving support control program Flowchart of contact probability calculation formula setting routine Explanatory drawing of the characteristic map of the 1st gain which sets a horizontal direction median value Explanatory drawing of the characteristic map of the 3rd gain which sets up front direction field length Explanatory drawing of the characteristic map of the 4th gain which sets control gain Schematic explanatory diagram of the area using the contact probability calculation formula set in front of the host vehicle

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Own vehicle 2 Driving assistance control apparatus for vehicles 3 Stereo camera 4 Stereo image recognition apparatus (3D object detection means, area | region setting means, control execution means)
5 Control unit (control execution means)
6 Monitor 7 Speaker 8 Automatic brake control device 9 Vehicle speed sensor

Claims (4)

A three-dimensional object detecting means for determining a traveling area based on the own vehicle traveling path and detecting a three-dimensional object existing on the traveling area;
An area setting means for setting at least an area by statistical processing in accordance with a condition set in front of the host vehicle;
Control execution means for executing control preset according to the position of the three-dimensional object with respect to the area by the statistical processing set by the area setting means;
A vehicle driving support device comprising:
The three-dimensional object detection means extracts the three-dimensional object as a determination target for determining the possibility of contacting the host vehicle,
The region setting means can vary the left end, the right end, and the front end of the region according to at least one of the own vehicle speed, the width of the three-dimensional object, the type of the three-dimensional object detection means, and the turning operation of the own vehicle. , set at least statistical processing,
The vehicle driving support apparatus, wherein the control execution means calculates a contact probability from the solid object and the region.   When the area set by the area setting means is compared to the preset forward distance from the host vehicle and the position of the three-dimensional object, and the three-dimensional object exists beyond the preset front distance, The front end is formed with a statistical distribution from the preset forward distance to the front, and the horizontal position set in advance based on the own vehicle traveling path is compared with the presence position of the three-dimensional object. 2. The method according to claim 1, wherein when the object is outside the preset lateral position, the side edges are formed in a statistical distribution from the preset lateral position to the side. The vehicle driving support device according to claim 1.   The vehicle driving support device according to claim 1 or 2, wherein the region by the statistical processing set by the region setting means is a region using a Gaussian distribution. 4. The vehicle driving support device according to claim 1 , wherein the control execution unit executes at least one of alarm control and deceleration control according to the value of the contact probability. 5. .

Images