JP2008008679A - Object detecting apparatus, collision predicting apparatus and vehicle controlling apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an object detecting apparatus which calculates a reliability value about a lateral displacement (orientation) of an obstruction, and also to provide a collision predicting apparatus and a vehicle controlling apparatus using the same. <P>SOLUTION: The object detecting apparatus for detecting an object existing in its surrounding area comprises: a first detecting means 10 for detecting an orientation and a distance from the surrounding area object; a second detecting means 20 for detecting the orientation and the distance from the surrounding area object; and a reliability calculating means for calculating the reliability about the orientation of the surrounding area object from detection results of the first detecting means 10 and the second detecting means 20. The object detecting apparatus is characterized in that not only the orientation and the distance from the surrounding area object, but also the reliability about the orientation of the surrounding area object are outputted. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an object detection device, a collision prediction device using the same, and a vehicle control device.

2. Description of the Related Art Conventionally, a collision prediction apparatus that detects an obstacle using an image sensor and a millimeter wave sensor and predicts a collision between the host vehicle and the obstacle is known (for example, see Patent Document 1). This device compares the sensor output changes of the image sensor and the millimeter wave sensor to determine the identity of the object detected by both sensors. In addition, the distance to the obstacle is detected by the millimeter wave sensor, and the displacement (azimuth) in the lateral direction of the obstacle is detected by the image sensor to calculate the position of the obstacle. In addition, what improves detection accuracy using such different types of sensors is referred to as sensor fusion.
JP 2004-301567 A

  In general, a radar device such as a millimeter wave sensor can detect the distance to the obstacle relatively accurately, while always detecting the displacement (azimuth) in the lateral direction of the obstacle accurately. It is relatively difficult to do. Furthermore, the reliability of detection may be reduced depending on the external environment, such as when there are many reflective objects around the obstacle.

  Further, even in an image sensor such as a camera, it may be difficult to accurately identify an obstacle depending on the external environment such as the weather or time zone.

  Therefore, in this type of collision prediction apparatus, it is necessary to calculate the reliability that changes depending on the external environment or the like with respect to the position of the obstacle, particularly the lateral displacement (orientation). This is because, in devices such as in-vehicle pre-crash safety systems that are realized using such a collision prediction device, it is important to suppress unnecessary operations due to false detection. This is because more accurate control can be performed by detecting how probable the position is.

  However, in the above-described conventional apparatus, no consideration is given to the reliability related to the lateral displacement (orientation) of the obstacle.

  The present invention is for solving such a problem, and an object detection device capable of calculating the reliability regarding the displacement (azimuth) in the lateral direction of an obstacle, a collision prediction device using the same, and a vehicle The main object is to provide a control device.

  In order to achieve the above object, a first aspect of the present invention is an object detection apparatus for detecting a peripheral object, the first detection means for detecting the distance and direction to the peripheral object, and the distance to the peripheral object. And a second detection means for detecting the orientation, and a reliability calculation means for calculating the reliability related to the orientation of the surrounding object based on the detection results of the first detection means and the second detection means, The reliability regarding the azimuth | direction of a peripheral object is output with the distance and azimuth | direction to a peripheral object.

  According to the first aspect of the present invention, since the reliability calculation means for calculating the reliability related to the orientation of the peripheral object based on the detection results of the first detection means and the second detection means is provided, The reliability related to the direction can be calculated and output.

In the first aspect of the present invention, the reliability calculation means preferably sets the reliability for each detection means related to the orientation of the peripheral object detected by the first detection means and the second detection means, respectively. The reliability calculation means for each detection means to be calculated, the reliability for each detection means calculated by the reliability calculation means for each detection means, and the orientations of the surrounding objects detected by the first detection means and the second detection means And a final reliability calculation means for calculating the reliability related to the orientation of the surrounding object. In this case, the reliability calculation means for each detection means includes a storage means for storing the distance and orientation to the peripheral object detected by each detection means in the past, and the time change of the orientation of the peripheral object detected by each detection means. It may be a means for calculating the reliability for each detection means based on the above, or a means for calculating the reliability for each detection means based on the external environment. Here, the external environment includes, for example, the intensity of reflected waves or the weather when the first detection means and the second detection means are radar devices or camera devices, respectively.
In the first aspect of the present invention, the final reliability calculation means is preferably based on the degree of coincidence of the orientations of the surrounding objects detected by the first detection means and the second detection means, respectively. It is a means for calculating the reliability related to the orientation of the surrounding object.

  In the first aspect of the present invention, the azimuth of the peripheral object may be replaced with a displacement in the lateral direction of the peripheral object as viewed from the object detection device.

  According to a second aspect of the present invention, there is provided a collision prediction apparatus that performs a collision prediction with a peripheral object. The collision prediction apparatus includes the object detection apparatus according to the first aspect of the present invention, and the distance to the peripheral object output by the object detection apparatus. , And the reliability of the azimuth and the azimuth of the surrounding object, the collision with the surrounding object is predicted.

  The second aspect of the present invention is preferably characterized in that the possibility of collision with a peripheral object is predicted to be low in accordance with a decrease in reliability related to the orientation of the peripheral object output by the object detection device. Is.

  According to a third aspect of the present invention, there is provided a vehicle control device that is mounted on a vehicle and performs inter-vehicle distance control, including the object detection device according to the first aspect of the present invention, up to a peripheral object output by the object detection device. The inter-vehicle distance control with the preceding vehicle recognized based on the distance, the direction, and the reliability related to the direction of surrounding objects is performed.

  In the third aspect of the present invention, it is preferable that the auxiliary control output for performing the inter-vehicle distance control with the preceding vehicle in accordance with the decrease in the reliability related to the direction of the peripheral object output by the object detection device. The output gain of power and / or auxiliary driving force is made gentle.

  ADVANTAGE OF THE INVENTION According to this invention, the object detection apparatus which can calculate the reliability regarding the displacement (azimuth | direction) of the horizontal direction of an obstruction, the collision prediction apparatus using this, and a vehicle control apparatus can be provided.

  Hereinafter, the best mode for carrying out the present invention will be described with reference to the accompanying drawings.

  Hereinafter, an embodiment of an in-vehicle collision prediction apparatus using the object detection apparatus according to the present invention will be described with reference to FIGS. FIG. 1 is a diagram illustrating an example of the overall configuration of a collision prediction apparatus 1 according to a first embodiment of the present invention. As shown in the figure, the collision prediction apparatus 1 includes a radar device 10, a camera sensor 20, and a collision prediction electronic control unit (hereinafter referred to as a collision prediction ECU) 30 as main components. Note that communication between devices in the vehicle in the following description is performed using an appropriate serial communication protocol such as CAN (Controller Area Network).

  The radar device 10 is, for example, a millimeter wave radar device disposed behind the front grille. The radar apparatus 10 detects the distance, azimuth, and relative velocity of an object using the time until the reflected wave of the millimeter wave returns, the angle of the reflected wave, and the frequency change. Further, as will be described later, the radar apparatus 10 calculates the reliability Rfrd regarding the azimuth of the object in the radar apparatus 10, and transmits the reliability Rfrd to the collision prediction ECU 30 by communication together with the distance, azimuth, and relative speed of the object. Note that the radar device 10 incorporates a storage means such as a calculation CPU, ROM, or RAM.

  FIG. 2 shows the positional relationship between distance, azimuth, and relative speed. As shown in the figure, the “azimuth” can be defined as, for example, an angle measured with reference to a clockwise direction, where the traveling direction of the host vehicle is an angle 0. Further, the “lateral displacement” is obtained as a function of the distance and orientation of the object. Therefore, if the distance of the object is a sufficiently probable value, “azimuth” and “lateral displacement” have a one-to-one relationship. In the present embodiment, explanation is made on the assumption that the reliability related to “azimuth” is calculated. However, even if the reliability related to “lateral displacement” obtained by adding the distance to this is calculated, it is equivalent as an apparatus. is there.

  The camera sensor 20 includes, for example, cameras 22A and 22B and an image processing device 24.

  The cameras 22A and 22B are, for example, CMOS cameras disposed on the left and right ends of the upper part of the windshield. Each optical axis is directed obliquely downward in front of the vehicle, and approximately several tens [m] in front of the vehicle. It is set to cross at the point. Thus, by imaging an object using a plurality of cameras, the distance to the object can be detected more accurately.

  The image processing device 24 is, for example, a computer centered on a CPU, and includes storage means such as a ROM and a RAM. The image processing device 24 receives images captured by the cameras 22A and 22B through communication, and processes the images to detect the distance and direction of the object. Further, as will be described later, the image processing device 24 calculates the reliability Rscm related to the direction of the object in the camera sensor 20, and transmits the reliability and the direction of the object to the collision prediction ECU 30 through communication.

  The collision prediction ECU 30 is, for example, a computer centered on a CPU, and includes storage means such as a ROM and a RAM. The collision prediction ECU 30 includes a reliability calculation unit 32 and a collision prediction unit 34 as main functional blocks.

  The reliability calculation unit 32 is based on the distance, azimuth, relative speed, and reliability Rfrd of the object transmitted from the radar device 10 and the distance, azimuth, and reliability Rscm of the object transmitted from the camera sensor 20. The distance, azimuth, and relative speed of the object are specified, and the final reliability Rfsn regarding the azimuth of the object is calculated.

  The collision prediction unit 34 determines whether or not a collision is inevitable based on the information input from the reliability calculation unit 32. When it is determined that a collision is inevitable, an instruction signal is sent to the seat belt ECU, and a motor that removes the seat belt so as to wind up the driver's / passenger's seat belt is driven before the collision. The initial restraint performance of the vehicle is increased, or an instruction signal is transmitted to the brake ECU, and the brake assist is quickly activated simultaneously with the depression of the brake pedal to reduce the collision speed. Further, the airbag may be operated before the collision, or the engine or transmission prepared for the collision (engine braking by fuel cut, etc.) may be performed.

  Here, an outline of a calculation method of the reliability Rfrd regarding the orientation of the object in the radar device 10, the reliability Rscm regarding the orientation of the object in the camera sensor 20, and the final reliability Rfsn regarding the orientation of the object in the present embodiment will be described. .

  The reliability Rfrd regarding the azimuth of the object in the radar apparatus 10 is (1) whether the difference from the previously detected azimuth is within a predetermined range (the detection of the object is repeated), or (2) the external environment is It is calculated centering on whether a predetermined condition that reduces reliability is satisfied. Details will be described later.

  Similarly, the reliability Rfrd relating to the orientation of the object in the camera sensor 20 is calculated around the above (1) and (2).

  The final reliability Rfsn relating to the orientation of the object is calculated based on the degree of coincidence between the orientation of the object detected by the radar device 10 and the orientation of the object detected by the camera sensor 20, It is calculated by taking the minimum value of the reliability Rfrd, reliability Rfrd, and temporary reliability Rtmp calculated as follows. Here, “based on the degree of match” means, for example, the range of the orientation of the object detected by the camera sensor 20 (there is not a single point, and there is a range of orientation that is considered to represent the object on the image) If the orientation of the object detected by the radar device 10 is within the range, the provisional reliability Rtmp is calculated to be the highest, and the radar device 10 detects the orientation range of the object detected by the camera sensor 20. This means that the temporary reliability Rtmp is calculated to be lower as the direction of the object deviates.

  Hereinafter, an operation in which the collision prediction apparatus 1 calculates the final reliability Rfsn regarding the distance, the azimuth, the relative speed, and the azimuth will be described. FIG. 3 is a block diagram illustrating a flow of characteristic processing executed by the collision prediction apparatus 1. This process is executed by the cooperation of the radar apparatus 10, the camera sensor 20, and the reliability calculation unit 32.

  First, processing on the radar device 10 side will be described. In the radar apparatus 10, for example, the distance, direction, and relative speed of the object are detected every predetermined time. Then, the reliability Rfrd regarding the orientation of the object in the radar apparatus 10 is calculated. This calculation is performed, for example, according to the flowchart illustrated in FIG. This flowchart is also used in processing on the camera sensor 20 side.

  First, the difference with the azimuth | direction detected last time is calculated (S100), and it is determined whether this difference exists in the 1st predetermined range (S110). When it is within the first predetermined range, it is set to a first value (for example, 80 [%], etc.). For example, any one of a finite number of levels such as levels 1 to 4 may be selected) as the reliability Rfrd related to the direction of the object in the radar apparatus 10 (S120).

  If the difference from the previously detected azimuth is not within the first predetermined range, it is determined whether or not the external environment satisfies a predetermined condition that may cause a decrease in reliability (S130). This predetermined condition varies depending on the radar device 10 and the camera sensor 20. In the case of the radar device 10, for example, (1) the received wave is lower than a reference value determined based on the distance of the object, that is, the received wave is weak, (2) it is larger in a predetermined region around the object. When any of the following conditions exists: (3) a given area around the object has more than a certain number of reflectors of a certain size or larger, the external environment reduces the reliability. It is determined that a predetermined condition that can be invited is satisfied.

  If it is determined that the external environment satisfies a predetermined condition that may cause a decrease in reliability, it is further determined whether or not the difference from the previously detected orientation is within a second predetermined range (S140). This second predetermined range is a range wider than the first predetermined range used in S110.

  When a positive determination is obtained in both S130 and S140, the second value (for example, 40 [%] or the like) is calculated as the reliability Rfrd related to the direction of the object in the radar apparatus 10 (S150). In this way, when the external environment is relatively poor, it is judged that there is a possibility that the difference from the previously detected orientation is relatively large, and this is allowed, while the reliability is The lower value is calculated.

  When a negative determination is obtained in any one of S130 and S140, the third value (for example, 0 [%] or the like) is calculated as the reliability Rfrd regarding the azimuth of the object in the radar apparatus 10 (S160). In this way, even if the external environment is relatively good, if the difference from the previously detected orientation is relatively large, it is determined that the probability of false detection is high, and the reliability is further increased. The low value is calculated. In this embodiment, when the reliability of 0 [%] is output, it is considered that the object has not been detected.

  When the reliability Rfrd regarding the azimuth of the object in the radar device 10 is calculated in this way, this is transmitted to the reliability calculation unit 32 of the collision prediction ECU 30 together with the distance, azimuth, and relative speed of the object.

  Next, processing on the camera sensor 20 side will be described. For example, the camera sensor 20 detects the distance and direction of the object at predetermined time intervals. Then, the reliability Rscm related to the orientation of the object in the camera sensor 20 is calculated.

  This calculation is also performed according to the flowchart illustrated in FIG. Since the difference from the processing in the radar apparatus 10 is only the predetermined condition used in S130, description of other processing is omitted.

  In the case of the camera sensor 20, for example, when either (1) rainy or snowing or (2) nighttime is satisfied, it is determined that the external environment satisfies a predetermined condition that may cause a decrease in reliability. Here, the information about the weather may be input from, for example, a navigation device, or may be obtained from an operating state of the wiper. The time zone may be determined using an in-vehicle timer.

  When the reliability Rscm related to the azimuth of the object in the camera sensor 20 is calculated in this way, it is transmitted to the reliability calculation unit 32 of the collision prediction ECU 30 together with the distance, azimuth, and relative speed of the object.

  The reason why the camera sensor 20 does not detect the relative velocity of the object and does not transmit it to the collision prediction ECU 30 is based on the fact that the radar device 10 is superior in performance in simply detecting the relative velocity.

  Further, it is desirable that the information transmission from the radar device 10 and the camera sensor 20 is performed in synchronization at the same timing. Therefore, for example, a request signal may be transmitted from the collision prediction ECU 30, and information from the radar device 10 or the camera sensor 20 may be transmitted as a response signal to the request signal. As another method, information is transmitted as needed from the radar device 10 or the camera sensor 20 at a specific cycle, and the latest input information for each predetermined cycle may be adopted in the collision prediction ECU 30.

  Hereinafter, the description will be given with reference back to the functional block of FIG. When information is received from the radar device 10 and the camera sensor 20, the reliability calculation unit 32 calculates the temporary reliability Rtmp. For example, as described above, the temporary reliability Rtmp has a value such as 80 [%] if the orientation of the object detected by the radar device 10 is within the range of the orientation of the object detected by the camera sensor 20. Each time the orientation of the object detected by the radar apparatus 10 deviates from the range of the orientation of the object detected by the camera sensor 20 by 1 degree, it is calculated by decreasing by 0 comma number [%] to several [%]. The

  Then, the final reliability Rfsn is calculated by taking the minimum values of Rfrd, Rscm, and Rtmp. By doing so, it is possible to output more highly reliable reliability as compared with the case where the reliability related to the orientation of the object in the radar device 10 or the camera sensor 20 is simply used. In addition, when the distance and azimuth | direction which the radar apparatus 10 and the camera sensor 20 output differ, what is necessary is just to discard information by the method of employ | adopting the information which transmitted the higher reliability.

  When the final reliability Rfsn is calculated, it is output to the collision prediction unit 34 together with the distance, direction, and relative speed of the object. The collision prediction unit 34 determines whether or not a collision is unavoidable using a collision prediction map stored in advance in a ROM or the like. In this determination, the final reliability Rfsn relating to the orientation of the object is used.

  Specifically, when the reliability Rfsn is a relatively high value, the determination criterion (threshold value) that the collision is unavoidable is made relatively gentle, and it is determined that the collision is more unavoidable. This determination criterion may be a function value that continuously changes against the reliability Rfsn, or may be a value that changes stepwise. By doing so, the control in preparation for the above-described collision can be more actively performed, which can contribute to safe driving. On the other hand, when the reliability Rfsn is a low value, the criterion (threshold) for determining that a collision is unavoidable is made relatively strict so that it is difficult to determine that the collision is unavoidable. This is because erroneous detection is likely to occur when the reliability Rfsn is a low value, and if the control for the collision is performed each time the erroneous detection is made, the driver feels bothered and burdens the device. Because.

  Further, the collision prediction by the collision prediction unit 34 is not limited to the one that alternatively determines whether or not the collision is unavoidable. For example, the collision possibility may be calculated as a percentage display. In this case, for example, a threshold for different operation is provided for each seat belt, brake, and airbag of the vehicle, and the operation can be performed with a time difference.

  It should be noted that the numerical values such as 80 [%], 40 [%], and 0 [%] exemplified above are not personalities that have special meanings in themselves, but can be changed in any way. Nor.

  Thus, since the collision prediction apparatus 1 according to the embodiment of the present invention calculates the reliability Rfrd and Rscm related to the orientation of the object detected by the radar apparatus 10 and the camera sensor 20, respectively, in these sensor apparatuses, It can reflect that the reliability of detection changes according to the external environment. Further, since the final reliability Rfsn is calculated based on the reliability Rfrd and Rscm, it is possible to output a reliability with higher accuracy than when the reliability Rfrd and Rscm are simply used. And since this final reliability Rfsn is used for collision prediction, more accurate collision prediction can be performed to contribute to safe driving and malfunction can be suppressed.

  In the collision prediction apparatus 1, the radar apparatus 10, the camera sensor 20, and the reliability calculation unit 32 constitute an object detection apparatus in the claims.

  Hereinafter, an embodiment of a vehicle control apparatus using the object detection apparatus according to the present invention will be described with reference to FIG. FIG. 5 is a diagram illustrating an example of the overall configuration of the vehicle control device 2 according to the second embodiment of the present invention. As illustrated, the vehicle control device 2 includes a radar device 10, a camera sensor 20, and a cruise control electronic control unit (hereinafter referred to as a cruise control ECU) 40 as main components.

  The radar apparatus 10 and the camera sensor 20 have the same configuration and function as the collision prediction apparatus 1 according to the first embodiment, and thus description thereof is omitted.

  The cruise control ECU 40 is, for example, a computer centered on a CPU, and includes storage means such as a ROM and a RAM. The cruise control ECU 40 includes a reliability calculation unit 42 and a cruise control unit 44 as main functional blocks.

  The reliability calculation unit 42 has the same function as the reliability calculation unit 32 according to the first embodiment. That is, the final reliability Rfsn is calculated in cooperation with the radar device 10 and the camera sensor 20, and is output to the cruise control unit 44 together with the distance, direction, and relative speed of the object (preceding vehicle).

  The cruise control unit 44 controls the brake independent of the driver's brake operation and accelerator operation so that the set inter-vehicle distance is maintained with respect to the preceding vehicle when there is a preceding vehicle by an instruction to the brake ECU or the engine ECU. Outputs power and driving force (hereinafter referred to as auxiliary braking / driving force) to execute inter-vehicle distance control for controlling the vehicle speed of the host vehicle and to perform constant speed running control for maintaining the vehicle speed of the host vehicle at a set vehicle speed. To do. Since the basic control method of this inter-vehicle distance control and constant speed traveling control is known to those skilled in the art, a detailed description thereof will be omitted.

  In this embodiment, the cruise control unit 44 performs the inter-vehicle distance control in consideration of the reliability Rfsn regarding the azimuth in addition to the distance, the azimuth, and the relative speed of the preceding vehicle. Specifically, when the reliability Rfsn is relatively low, the auxiliary braking / driving force is recalculated so that the output gain is lower than the auxiliary braking / driving force calculated in the basic inter-vehicle distance control, and the brake ECU and engine An instruction signal is transmitted to the ECU. That is, when the possibility of erroneous detection is high, the output of the auxiliary braking / driving force is made more gradual. As a result, the auxiliary braking / driving force is output every time erroneous detection is performed, and it can be suppressed that the driver feels bothersome and places a burden on the device.

  As mentioned above, although the best mode for carrying out the present invention has been described using one embodiment, the present invention is not limited to such one embodiment, and within the scope not departing from the gist of the present invention, Various modifications and substitutions can be made to the above-described embodiment.

  For example, although an example of detecting an object using a millimeter wave radar device and a camera sensor has been illustrated, the present invention is not limited to this, and any detection device may be combined. For example, a laser radar device may be used. In this case, the predetermined condition related to the external environment that may cause a decrease in reliability is, for example, that either (1) rainy or snowing, (2) a predetermined number or more of reflectors exist on the road surface, or the like is satisfied. That's fine.

  Further, the present invention can be applied to a device that detects an object by combining three or more types of sensor devices.

  The present invention can be used for an apparatus that detects at least an object. When mounted on a vehicle, the appearance, weight, size, running performance, etc. of the mounted vehicle are not limited.

It is a figure which shows an example of the whole structure of the collision prediction apparatus 1 which concerns on 1st Example of this invention. It is a figure which shows the positional relationship of distance, an azimuth | direction, and relative velocity. It is a block diagram which shows the flow of the characteristic process which the collision prediction apparatus 1 which concerns on 1st Example of this invention performs. It is a flowchart figure which shows the flow of the characteristic process which the radar apparatus 10 and the camera sensor 20 perform. It is a figure which shows an example of the whole structure of the vehicle control apparatus 2 which concerns on 2nd Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Collision prediction apparatus 2 Vehicle control apparatus 10 Radar apparatus 20 Camera sensor 22A, 22B Camera 24 Image processing apparatus 30 Collision prediction electronic control unit 32, 42 Reliability calculation part 34 Collision prediction part 40 Cruise control electronic control unit 44 Cruise control Part

Claims (11)

An object detection device for detecting surrounding objects,
First detection means for detecting a distance and direction to the surrounding object;
Second detection means for detecting a distance and direction to the surrounding object;
Reliability calculation means for calculating reliability on the orientation of the surrounding object based on the detection results of the first detection means and the second detection means,
An object detection apparatus that outputs a reliability related to an orientation of the peripheral object together with a distance and an orientation to the peripheral object. The object detection device according to claim 1,
The reliability calculation means is:
A reliability calculation means for each detection means for calculating the reliability for each detection means regarding the orientation of the surrounding object detected by each of the first detection means and the second detection means;
Based on the reliability of each detection means calculated by the reliability calculation means for each detection means and the orientation of the peripheral object detected by the first detection means and the second detection means, A final reliability calculation means for calculating the reliability related to the bearing,
Object detection device. The object detection device according to claim 2,
The detection means-specific reliability calculation means includes storage means for storing the distance and orientation to the peripheral object detected by each detection means in the past, and measures the temporal change in the orientation of the peripheral object detected by each detection means. Based on the above, it is a means for calculating the reliability for each detection means,
Object detection device. The object detection device according to claim 2 or 3,
The detection means-specific reliability calculation means is a means for calculating the reliability for each detection means based on the external environment.
Object detection device. The object detection device according to claim 4,
Each of the first detection means and the second detection means is a radar device or a camera device,
The external environment includes the intensity of reflected waves, or weather,
Object detection device. The object detection device according to any one of claims 2 to 5,
The final reliability calculation means is means for calculating a reliability related to the orientation of the peripheral object based on the degree of coincidence of the orientation of the peripheral object detected by the first detection means and the second detection means, respectively. is there,
Object detection device.   The object detection device according to claim 1, wherein the orientation of the peripheral object is replaced with a lateral displacement of the peripheral object viewed from the object detection device. A collision prediction device for predicting collisions with surrounding objects,
An object detection device according to any one of claims 1 to 7,
Predicting a collision with the surrounding object based on the distance to the surrounding object, the orientation, and the reliability related to the orientation of the surrounding object output by the object detection device,
Collision prediction device. The collision prediction device according to claim 8,
The collision prediction apparatus according to claim 1, wherein the possibility of collision with the peripheral object is predicted to be low in accordance with a decrease in reliability related to the orientation of the peripheral object output from the object detection apparatus. A vehicle control device that is mounted on a vehicle and performs inter-vehicle distance control,
An object detection device according to any one of claims 1 to 7,
Vehicle control characterized by performing inter-vehicle distance control with a preceding vehicle recognized based on the distance to the surrounding object, the direction, and the reliability related to the direction of the surrounding object output by the object detection device. apparatus. The vehicle control device according to claim 10,
The output gain of the auxiliary braking force and / or auxiliary driving force that is output to perform the inter-vehicle distance control with the preceding vehicle in response to a decrease in the reliability related to the direction of the surrounding object that is output by the object detection device. A vehicle control device characterized by being loosened.

Images