Vehicle control device and program

Abstract

To provide a vehicle control device capable of appropriately detecting an object.SOLUTION: An ECU 30 as a vehicle control device acquires detection positions of objects existing in the periphery of an own vehicle 50 with the use of object detection devices 11-13, and performs parking support control to support the operation of the own vehicle to a parking space based on the acquired detection positions. The ECU 30 includes: an angle calculation section for calculating detection angles of the detection positions with respect to the reference direction of the own vehicle 50; a distribution calculation section for calculating a frequency distribution of the detection positions with respect to the calculated detection angles; and a determination section for determining that the opposing surfaces of the objects exist in the directions of the detection angles corresponding to ridges when the plurality of ridges having a peak value at the different detection angles exist in the calculated frequency distribution.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a vehicle control device and a program. [Background technology]

[0002] One such type of vehicle control device, as described in Patent Document 1, for example, uses an object detection device to obtain the detection position of an object present around the vehicle, and performs parking assistance control to assist the vehicle in driving into a parking space based on the obtained detection position. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2014-34321 A Summary of the Invention [Problem to be solved by the invention]

[0004] In parking assistance control, information corresponding to the actual environment may not be obtained. In detail, when determining whether or not an object has an opposing surface to the vehicle in parking assistance control, it may be determined that an opposing surface exists only in one direction of the vehicle due to the size or shape of the object, even though the opposing surface actually exists in two different directions of the vehicle. Thus, there is still room for improvement in object detection technology in parking assistance control.

[0005] The present invention has been made in consideration of the above circumstances, and a main object of the present invention is to provide a vehicle control device and a program that can appropriately detect an object. [Means for solving the problem]

[0006] The present invention relates to a vehicle control device that uses an object detection device to acquire a detection position of an object present in the vicinity of a vehicle, and performs parking assistance control to assist in driving the vehicle into a parking space based on the acquired detection position.The vehicle control device includes an angle calculation unit that calculates a detection angle of the detection position relative to a reference direction of the vehicle, a distribution calculation unit that calculates a frequency distribution of the detection position relative to the calculated detection angle, and a determination unit that, when the calculated frequency distribution includes multiple peaks having peak values ​​at different detection angles, determines that the opposing surface of the object is present in the direction of the detection angle corresponding to the peaks.

[0007] According to the present invention, it is determined that there is an opposing surface of an object relative to the vehicle, corresponding to each peak of the frequency distribution for the detection angle. This makes it possible to detect each opposing surface that exists in different directions relative to the vehicle, and as a result, it is possible to appropriately detect the object. [Brief description of the drawings]

[0008] [Figure 1] 1 is a diagram showing a schematic configuration of a parking assistance system according to a first embodiment. [Diagram 2] FIG. 4 is a diagram showing a situation in which parking assist control is performed. [Diagram 3] FIG. 13 is a diagram showing an object detection result when clustering processing is performed. [Figure 4] FIG. 4 is a diagram showing a situation in which parking assist control is performed. [Diagram 5] 13A and 13B are diagrams showing object detection results when straight line segment extraction processing is performed. [Figure 6] FIG. 4 is a diagram showing a situation in which parking assist control is performed. [Figure 7] FIG. 11 is a diagram showing an object detection result in a comparative example. [Figure 8] FIG. 4 is a diagram showing a situation in which parking assist control is performed. [Figure 9] FIG. 11 is a diagram showing an object detection result in a comparative example. [Figure 10] 5A and 5B are diagrams showing a method for calculating the detection angle of each detection position. [Figure 11] 5A and 5B are diagrams showing a method for calculating the detection angle of each detection position. [Figure 12] 5A and 5B are diagrams showing a method for calculating the detection angle of each detection position. [Figure 13] FIG. 13 is a diagram showing a frequency distribution of each detection position with respect to the detection angle. [Figure 14] 13A and 13B are diagrams showing object detection results when a detection angle calculation process is performed. [Figure 15] 11 is a flowchart showing the procedure of an object detection process. [Figure 16] FIG. 11 is a diagram showing a method for setting a frequency threshold according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] First Embodiment Hereinafter, an embodiment of the present invention will be described with reference to the drawings. This embodiment is a parking assistance system mounted on a vehicle. The parking assistance system uses a vehicle control device (hereinafter referred to as ECU) as a central unit to detect objects (e.g., other vehicles, road structures, etc.) present around the vehicle and perform parking assistance control.

[0010] First, a schematic configuration of a parking assistance system 10 for a host vehicle 50 according to this embodiment will be described with reference to Fig. 1. The host vehicle 50 has object detection devices 11 to 13 and an ECU 30.

[0011] The object detection devices 11-13 are sensors that acquire the distance and direction to an object present around the vehicle 50. In this embodiment, the object detection devices 11-13 are provided with ultrasonic sensors such as sonar, which transmit a search wave at every predetermined control period and acquire the distance and direction of an object present around the vehicle 50 by receiving a reflected wave of the transmitted search wave. Note that the object detection devices 11-13 are not limited to ultrasonic sensors, and may be sensors such as laser sensors and millimeter wave sensors.

[0012] The object detection devices 11-13 are mounted at multiple locations on the vehicle 50. In this embodiment, as shown in Fig. 1, the multiple object detection devices 11-13 are attached at predetermined intervals to the front bumper, the rear bumper, and the vehicle body side portions of the vehicle 50.

[0013] Specifically, the vehicle 50 is equipped with, as object detection devices 11-13, a front sensor 11 and a rear sensor 12 attached near a central axis 50a that passes through the center position of the vehicle width and is an axis extending in the vehicle length direction, and side sensors 13 attached respectively to the left and right sides of the vehicle 50. The sensors 11-13 are attached at positions symmetrical to the left and right with respect to the central axis 50a.

[0014] The ECU 30 is a computer equipped with a CPU, ROM, RAM, I / O, etc., and the CPU executes programs installed in the ROM to realize various functions for performing parking assistance control of the vehicle 50.

[0015] Specifically, information acquired by each of the sensors 11-13 provided on the host vehicle 50 is input to the ECU 30. The ECU 30 calculates a detection position P of an object present around the host vehicle 50 based on the information acquired by each of the sensors 11-13. The detection position P is indicated as a point in a two-dimensional coordinate system defined by the vehicle length direction and the vehicle width direction of the host vehicle 50, for example, with the center of the host vehicle 50 as the origin.

[0016] The ECU 30 performs an object detection process to detect an object existing around the vehicle 50. In the object detection process, the ECU 30 determines the presence of an opposing surface to the vehicle 50 based on the acquired detection position P. Here, in the object detection process, when the ECU 30 recognizes a plurality of objects separately, the ECU 30 determines that an opposing surface to the vehicle 50 exists for each of the recognized objects. In addition, when the ECU 30 recognizes one recognized object by separating it into a plurality of parts, the ECU 30 determines that an opposing surface to the vehicle 50 exists for each of the recognized parts of the object.

[0017] The ECU 30 performs vehicle control based on the result of the object detection process. In the vehicle control, a parking space and a parking route are calculated so as to leave a distance W in a direction perpendicular to an opposing surface determined to exist by the object detection process, and the vehicle 50 is controlled to park the vehicle 50 in the calculated parking space. In this embodiment, the ECU 30 corresponds to a "control unit."

[0018] Specifically, the ECU 30 controls a steering device 21, an accelerator device 22, and a brake device 23 of the host vehicle 50. The steering device 21 is a device for steering the host vehicle 50, and is controlled by commands from the ECU 30. The accelerator device 22 is an engine or a motor as a vehicle power source, and applies driving force to the host vehicle 50 according to control commands from the ECU 30. The brake device 23 is provided on each wheel, and applies braking force to the host vehicle 50 according to control commands from the ECU 30.

[0019] Next, the parking assist control performed by the ECU 30 will be described in detail for each specific situation of the host vehicle 50.

[0020] FIG. 2 shows a case where parking assist control is performed in a situation where another vehicle 51 exists to the right of the host vehicle 50 and a rear wall 52 exists in the vehicle width direction.

[0021] Each side sensor 13 on the right side of the vehicle 50 transmits a search wave toward a range 13a extending to the right side of the vehicle 50, and receives a reflected wave from the left side of the other vehicle 51. As a result, each side sensor 13 obtains information such as the distance and direction to the other vehicle 51. Also, each rear sensor 12 transmits a search wave toward a range 12a extending behind the vehicle 50, and receives a reflected wave from the surface of the rear wall 52. As a result, each rear sensor 12 obtains information such as the distance and direction to the rear wall 52.

[0022] 2, the ECU 30 performs a clustering process to separately recognize the other vehicle 51 and the rear wall 52. The clustering process is a process for associating, among the acquired multiple detection positions P, detection positions P that form the same object with each other.

[0023] For example, the clustering process may be performed as follows. The ECU 30 calculates the detection position P of an object present in the vicinity of the vehicle 50 based on the information acquired by the sensors 12, 13. In the clustering process, the ECU 30 calculates the distance between the acquired detection positions P. The ECU 30 associates two detection positions P that are determined to be shorter than a predetermined distance threshold Dth among the calculated distances as a group G forming the same object. Furthermore, when the ECU 30 associates one of the two detection positions P as the group G and determines that the distance between the other detection position P is shorter than the distance threshold Dth, the ECU 30 also associates the other detection position P as a group G forming the same object.

[0024] FIG. 3 is a diagram showing an object detection result when clustering processing is performed in the situation shown in FIG.

[0025] In Fig. 3, when the first to m-th detection positions P1 to Pm and the m+1-th to n-th detection positions Pm+1 to Pn are acquired, the distance between adjacent detection positions P is calculated. Since the first distance D1 between the first and second detection positions P1 and P2 is shorter than the distance threshold Dth, the first and second detection positions P1 and P2 are associated as a first group G1. Next, since the second distance D2 between the second and third detection positions P2 and P3 is shorter than the distance threshold Dth, the third detection position P3 is also associated as the first group G1. Similarly, the fourth to m-th detection positions P4 to Pm are also associated as the first group G1.

[0026] Because the third distance D3 between the m-th detection position Pm and the (m+1)-th detection position Pm+1 is equal to or greater than the distance threshold Dth, the (m+1)-th detection position Pm+1 is not associated as the first group G1. Meanwhile, because the fourth distance D4 between the (m+1)-th detection position Pm+1 and the (m+2)-th detection position Pm+2 is shorter than the distance threshold Dth, the (m+1)-th and (m+2)-th detection positions Pm+1 and Pm+2 are associated as a second group G2 different from the first group G1. Similarly, the (m+3)-th to (n)-th detection positions Pm+3-Pn are associated as the second group G2.

[0027] The ECU 30 recognizes that the detection positions P1 to Pm included in the first group G1 are formed by the same object, and recognizes that the detection positions Pm+1 to Pn included in the second group G2 are formed by the same object. In other words, the ECU 30 recognizes the detection positions P1 to Pm included in the first group G1 and the detection positions Pm+1 to Pn included in the second group G2 separately.

[0028] The ECU 30 determines the presence of an opposing surface S facing the host vehicle 50 for each of the separated groups G1 and G2. Specifically, for the first group G1, it determines that a first opposing surface S1 facing the right side of the host vehicle 50 exists, and for the second group G2, it determines that a second opposing surface S2 facing the rear of the host vehicle 50 exists. In this case, the ECU 30 calculates the parking space 50b so as to leave a first distance W1 from the first opposing surface S1 in the vehicle width direction and a second distance W2 from the second opposing surface S2 in the vehicle length direction, and performs vehicle control to park the host vehicle 50 in the calculated parking space 50b.

[0029] FIG. 4 shows a case where parking assist control is performed in a situation where a surrounding wall 53 having a bent shape is present on the right side and rear of the host vehicle 50.

[0030] Each side sensor 13 receives a wave reflected from a surface of the surrounding wall 53 on the right side with respect to the vehicle 50. Additionally, each rear sensor 12 receives a wave reflected from a surface of the surrounding wall 53 on the rear side with respect to the vehicle 50. In this way, each sensor 12, 13 obtains information such as the distance and direction to the surrounding wall 53.

[0031] 4, the ECU 30 performs a straight-line segment extraction process to separate and recognize the surrounding wall 53. The straight-line segment extraction process is a process for extracting a straight-line segment L of a line segment connecting the acquired multiple detection positions P.

[0032] For example, the straight-line section extraction process may be performed as follows. The ECU 30 calculates the detection position P of an object present around the vehicle 50 based on the information acquired by the sensors 12, 13. In the straight-line section extraction process, the ECU 30 connects adjacent detection positions P with line segments and calculates a section in which the difference in slope between the adjacent line segments falls within a predetermined value. The ECU 30 extracts, from among the calculated sections, a section whose length is equal to or greater than a length threshold value Lth as a straight-line section L.

[0033] FIG. 5 is a diagram showing an object detection result when the clustering process and the straight line segment extraction process are performed in the situation shown in FIG.

[0034] The ECU 30 acquires the detection positions P, and in the clustering process, associates the detection positions P with each other as being in the same group G because the distance between each of the adjacent detection positions P is shorter than the distance threshold Dth.

[0035] The ECU 30 performs straight section extraction processing on each detection position P associated as a group G. As shown in Fig. 5, the ECU 30 extracts a first straight section L1 and a second straight section L2 having a length equal to or greater than a length threshold Lth. The first straight section L1 is a straight section extending in the vehicle length direction of the host vehicle 50, and the second straight section L2 is a straight section extending in the vehicle width direction of the host vehicle 50. In this case, the ECU 30 recognizes each detection position P included in the first straight section L1 separately from each detection position P included in the second straight section L2.

[0036] The ECU 30 determines the presence of an opposing surface S facing the host vehicle 50 for each of the separated straight sections L1, L2. Specifically, for the first straight section L1, it determines that a first opposing surface S1 facing the right side of the host vehicle 50 exists, and for the second straight section L2, it determines that a second opposing surface S2 facing the rear of the host vehicle 50 exists. In this case, the ECU 30 calculates the parking space 50b so as to leave a first distance W1 from the first opposing surface S1 in the vehicle width direction and a second distance W2 from the second opposing surface S2 in the vehicle length direction, and performs vehicle control to park the host vehicle 50 in the calculated parking space 50b.

[0037] As described above, by performing the clustering process and the straight line section extraction process, objects are recognized separately, and it is possible to determine the presence of the opposing surface S in two different directions of the vehicle 50. This allows the vehicle to be appropriately controlled so as to provide a distance W from each opposing surface S.

[0038] However, when an object is not recognized separately due to its size, shape, or the like, it may be determined that an opposing surface S exists only in one direction of the vehicle 50, even though an opposing surface S actually exists in two different directions of the vehicle 50. In this case, the vehicle cannot be controlled to keep a distance W from each of the opposing surfaces S that exist in the two different directions of the vehicle 50, which causes a problem that the vehicle 50 cannot be parked appropriately.

[0039] Here, an object detection process of a comparative example, which is different from the present embodiment in that the presence of the opposing surface S cannot be appropriately determined, will be described.

[0040] 6 shows a case where parking assist control is performed in a situation where a straight side wall 54 exists on the right side of the vehicle 50 and a pillar 55 exists behind the vehicle 50. Here, the distance between the straight side wall 54 and the pillar 55 is shorter than the distance threshold Dth. Also, the length of the pillar 55 in the vehicle width direction is shorter than the length threshold Lth.

[0041] Each side sensor 13 on the right side of the vehicle 50 transmits a search wave toward an area 13a extending to the right side of the vehicle 50, and receives a reflected wave from the surface of the straight side wall 54. As a result, each side sensor 13 obtains information such as the distance to and direction of the straight side wall 54. Also, each rear sensor 12 transmits a search wave toward an area 12a extending behind the vehicle 50, and receives a reflected wave from the surface of the pillar 55. As a result, each rear sensor 12 obtains information such as the distance to and direction of the pillar 55.

[0042] 6, even if the ECU 30 performs a clustering process, the distance between the linear side wall 54 and the pillar 55 is shorter than the distance threshold Dth, so the linear side wall 54 and the pillar 55 cannot be separately recognized. In addition, even if the ECU 30 performs a straight section extraction process, the length of the pillar 55 in the vehicle width direction is shorter than the length threshold Lth, so the ECU 30 cannot separate the pillar 55 from the linear side wall 54.

[0043] Fig. 7 is a diagram showing a detection result in an object detection process of a comparative example different from this embodiment in the situation shown in Fig. 6. Note that in the comparative example of Fig. 7, a clustering process is performed.

[0044] The ECU 30 acquires the detection positions P of the linear side wall 54 and the pillar 55. The ECU 30 associates each detection position P as a group G by performing a clustering process. The ECU 30 determines that an opposing surface S exists on the right side of the vehicle 50 for the group G. In this case, the ECU 30 performs vehicle control so as to leave a distance W from the opposing surface S in the vehicle width direction. Here, since the pillar 55 is recognized as a protruding part in the vehicle width direction, it is determined that the opposing surface S exists in the vehicle length direction from the protruding part. The ECU 30 performs vehicle control so as to calculate the parking space 50b so as to leave a distance W in the vehicle width direction from the pillar 55, which is a protruding part, and can perform vehicle control so as to park the vehicle 50 in the calculated parking space 50b. However, because the ECU 30 cannot determine that an opposing surface S exists behind the vehicle 50, the ECU 30 is unable to calculate the parking space 50b so as to leave a gap from the pillar 55 in the vehicle length direction, and is unable to perform vehicle control to park the vehicle 50 in the calculated parking space 50b.

[0045] 8 shows a case where parking assist control is performed in a situation where a curved side wall 56 exists on the right side of the vehicle 50 and a pillar 55 exists behind the vehicle 50. Here, the distance between the curved side wall 56 and the pillar 55 is shorter than the distance threshold Dth. Also, the length of the pillar 55 in the vehicle width direction is shorter than the length threshold Lth.

[0046] In the situation shown in FIG. 8, as in the case described in FIG. 6, even if the ECU 30 performs the clustering process, the ECU 30 associates each detection position P with the same group G. As a result, the ECU 30 cannot recognize the curved side wall 56 and the pillar 55 separately. Furthermore, even if the ECU 30 performs the straight line section extraction process, the ECU 30 cannot extract the straight line section L from the curved side wall 56, and cannot extract the straight line section L from the pillar 55. As a result, the ECU 30 cannot recognize the curved side wall 56 and the pillar 55 separately.

[0047] Fig. 9 is a diagram showing a detection result in an object detection process of a comparative example different from this embodiment in the situation shown in Fig. 8. Note that in the comparative example of Fig. 9, a clustering process is performed.

[0048] The ECU 30 acquires the detection positions P of the curved side wall 56 and the pillar 55. The ECU 30 associates each detection position P as a group G by performing a clustering process. The ECU 30 determines that an opposing surface S exists on the right side of the vehicle 50 for the group G. In this case, the ECU 30 recognizes the pillar 55 as a protruding part in the vehicle width direction, as described in FIG. 7, and determines that the opposing surface S exists in the vehicle length direction from the protruding part. Therefore, the ECU 30 can calculate the parking space 50b so as to leave a distance W in the vehicle width direction from the pillar 55, which is the protruding part, and can perform vehicle control to park the vehicle 50 in the calculated parking space 50b. However, the ECU 30 cannot determine that the opposing surface S exists behind the vehicle 50, and cannot perform vehicle control to calculate the parking space 50b so as to leave a distance W in the vehicle length direction from the pillar 55 and park the vehicle 50 in the calculated parking space 50b.

[0049] Therefore, in this embodiment, in order to prevent the occurrence of a situation in which the vehicle 50 is not parked in an appropriate position due to the object not being recognized separately, the presence of the opposing surface S of the object is determined based on the detection angle of each detection position P. The following describes a case in which the detection angle is calculated in the situation shown in FIG.

[0050] 10 to 12 are diagrams illustrating a method for calculating the detection angle of each detection position P. In FIG.

[0051] 10, the ECU 30 estimates the detection direction of each detection position P indicated by an arrow based on each detection position P acquired from each sensor 12, 13 and the time when each detection position P was acquired. The detection direction of each detection position P indicates the direction in which the search wave is reflected by an object, and is estimated based on the direction in which each sensor 12, 13 receives the reflected wave.

[0052] 11 and 12, the ECU 30 calculates the detection direction of each detection position P as a detection angle. The detection angle is the angle of the direction in which each sensor 12, 13 receives the reflected wave. In this embodiment, the detection angle is calculated with the forward direction of the host vehicle 50 in the vehicle length direction as a reference direction, and a clockwise direction is calculated as a positive detection angle and a counterclockwise direction is calculated as a negative detection angle.

[0053] 11 is a diagram showing an example of a case where detection angles α1 to α3 are calculated for each right side detection position Pr1 to Pr3 acquired by the lateral sensor 13 on the right side of the host vehicle 50. In this case, the ECU 30 calculates the detection angle of the second right side detection position Pr2 as α2=-90°, calculates the detection angle of the first right side detection position Pr1 as α1<α2, and calculates the detection angle of the third right side detection position Pr3 as α3>α2.

[0054] 12 is a diagram showing an example of a case where detection angles β1 to β3 are calculated for each of rear detection positions Pb1 to Pb3 acquired by the rear sensor 12. In this case, the ECU 30 calculates the detection angle at the second rear detection position Pb2 as β2=0°, calculates the detection angle at the first rear detection position Pb1 as β1>β2, and calculates the detection angle at the third rear detection position Pb3 as β3<β2.

[0055] The ECU 30 calculates the frequency distribution of each detection position P with respect to the detection angle. FIG. 13 is a diagram showing the frequency distribution in the situation shown in FIG. 6. In this embodiment, the interval width of the detection angle in the frequency distribution is set to 5°. The ECU 30 detects peaks having peak values ​​at different detection angles in the calculated frequency distribution. In detail, when the ECU 30 determines that there are peak values ​​at different detection angles and that there is a detection angle at which the frequency of the detection position P is not detected between the peak values ​​of the different detection angles, the ECU 30 detects peaks having peak values ​​at each detection angle.

[0056] In the frequency distribution shown in FIG. 13, the ECU 30 detects peaks having peak values ​​at two different detection angles. The first peak is detected as a peak having a peak value Fp1 at a detection angle including −90°. The first peak is a peak detected corresponding to the linear side wall 54. On the side of the first peak with a larger detection angle, a second peak is detected across a detection angle at which the detection position P is not detected. The second peak has a peak value Fp2 at a detection angle including 0°. The second peak is a peak detected corresponding to the pillar 55.

[0057] The ECU 30 judges whether or not an opposing surface S of an object relative to the vehicle 50 exists in correspondence with each detected mountain. In detail, the ECU 30 judges that an opposing surface S corresponding to each mountain exists on the condition that an angle width θr between adjacent mountains where the frequency of the detection position P is not detected is determined to be larger than an angle width threshold θth. The angle width threshold θth may be set based on the mounting positions of the sensors 11 to 13 and the range in which the exploration waves of the sensors 11 to 13 spread. In this embodiment, the angle width threshold θth is set to a value within a range of 50° to 70°, specifically, set to 60°. In addition, the ECU 30 judges that an opposing surface S corresponding to each mountain exists on the condition that the peak values ​​Fp1, Fp2 of each mountain are determined to be larger than a frequency threshold Fth. In this embodiment, the frequency threshold Fth is set to 10.

[0058] The number of detected peaks is not limited to two, and peaks having peak values ​​at three different detection angles may be detected due to noise in the reflected waves of the sensors 11 to 13. Even in this case, the presence of the opposing surface S corresponding to each peak can be determined on the condition that the angular width θr between the peaks is greater than the angular width threshold θth and the peak value of each peak is greater than the frequency threshold Fth.

[0059] FIG. 14 is a diagram showing an object detection result when the clustering process and the above-mentioned detection angle calculation process are performed in the situation shown in FIG.

[0060] The ECU 30 calculates the detection angle for the detection positions P associated as a group G by the clustering process, and calculates the frequency distribution of each detection position P with respect to the detection angle. As shown in FIG. 13, the ECU 30 detects mountains having peak values ​​Fp at different detection angles. As shown in FIG. 14, the ECU 30 determines that a right facing surface Sα and a rear facing surface Sβ exist corresponding to each mountain. The right facing surface Sα corresponds to the first mountain, and the rear facing surface Sβ corresponds to the second mountain. In this case, the ECU 30 performs vehicle control so as to leave a right side interval Wα in the vehicle width direction from the right facing surface Sα, and a rear side interval Wβ in the vehicle length direction from the rear facing surface Sβ. This allows the host vehicle 50 to be parked appropriately.

[0061] FIG. 15 shows the procedure of the object detection process performed by the ECU 30.

[0062] In step S10, information such as the distance and direction to the object is obtained from each of the sensors 12, 13. Based on the obtained information from each of the sensors 12, 13, a detection position P is calculated.

[0063] In step S11, a clustering process is performed. By performing the clustering process, each detection position P whose distance between the detection positions P is shorter than the distance threshold Dth is associated with the same group G. This makes it possible to separate and recognize objects for each different group G.

[0064] In step S12, a straight line segment extraction process is performed. By performing the straight line segment extraction process, a straight line segment L is extracted from each detection position P. This makes it possible to separate and recognize an object for each extracted straight line segment L. In this embodiment, the clustering process in step S11 and the straight line segment extraction process in step S12 correspond to a "separation recognition unit."

[0065] In step S13, the detection angle is calculated. The detection position P of an object that is separately recognized by at least one of the clustering process and the straight line segment extraction process is set as the detection angle calculation target. This makes it possible to calculate the detection angle for the detection position P of an object that cannot be separately recognized by the clustering process and the straight line segment extraction process. In step S14, the frequency distribution of the detection position P for the calculated detection angle is calculated. In this embodiment, the process of step S13 corresponds to the "setting unit" and the "angle calculation unit", and the process of step S14 corresponds to the "distribution calculation unit".

[0066] In step S15, it is determined whether or not there are two or more peaks having a peak value Fp at different detection angles in the calculated frequency distribution. For example, if there are peak values ​​Fp at different detection angles in the calculated frequency distribution, and there is a detection angle at which the detection position P is not detected between the peak values ​​of the different detection angles, it is determined that there is a peak having a peak value Fp at each detection angle. If it is determined that there is one peak, this process is terminated. On the other hand, if there are two or more peaks having a peak value Fp at different detection angles, the process proceeds to step S16.

[0067] In step S16, it is determined whether the angle width θr between adjacent mountains, where the detection position P is not detected, is greater than the angle width threshold θth. If it is determined that the angle width θr is equal to or smaller than the angle width threshold θth, this process ends. On the other hand, if it is determined that the angle width θr is greater than the angle width threshold θth, the process proceeds to step S17.

[0068] In step S17, it is determined whether the peak value Fp of each mountain is greater than the frequency threshold Fth. If it is determined that the peak value Fp of each mountain is equal to or less than the frequency threshold Fth, this process ends. On the other hand, if it is determined that the peak value Fp of each mountain is greater than the frequency threshold Fth, the process proceeds to step S18.

[0069] In step S18, it is determined that there is an opposing surface S of an object for the vehicle 50 corresponding to each mountain. In this case, the parking space 50b is calculated so as to leave a distance W in the perpendicular direction from each opposing surface S, and the vehicle is controlled so as to park the vehicle 50 in the calculated parking space 50b. In this embodiment, the processes in steps S15 to S18 correspond to a "determination unit."

[0070] According to the present embodiment described above in detail, the following effects can be obtained.

[0071] Corresponding to each peak of the frequency distribution for the detection angle, it is determined that there is an opposing surface S of an object with respect to the vehicle 50. This makes it possible to detect each opposing surface S that exists in different directions with respect to the vehicle 50. As a result, it is possible to appropriately detect objects.

[0072] It is determined that there is an opposing surface S corresponding to each mountain, provided that the angle width θr, within which the detection position P is not detected, is greater than the angle width threshold θth between adjacent mountains. This makes it possible to accurately determine whether each mountain corresponds to an opposing surface S that exists in a different direction relative to the host vehicle 50.

[0073] On the condition that the peak value Fp of each mountain is greater than the frequency threshold value Fth, it is determined that there is an opposing surface S corresponding to each mountain. This makes it possible to accurately determine whether or not each mountain corresponds to an opposing surface S that exists in a different direction with respect to the host vehicle 50.

[0074] <Second embodiment> The second embodiment will be described below with reference to the drawings, focusing on the differences from the first embodiment. In this embodiment, the frequency threshold Fth is changed.

[0075] The sensors 11-13 are attached to the front, rear and side of the vehicle 50 at a certain height from the ground. Therefore, the search waves of the sensors 11-13 are not transmitted to the space below the mounting positions of the sensors 11-13. In this case, the reflected waves are difficult to receive near low objects such as wheel chocks installed in the parking space. This reduces the number of calculated detection positions P, which may make it difficult to detect the presence of the opposing surface S corresponding to the object such as the wheel chock.

[0076] Therefore, in this embodiment, the lower the height of the object, the lower the frequency threshold Fth is set. In detail, the ECU 30 calculates the height of the object based on the detection position P acquired by each of the sensors 11 to 13 before the host vehicle 50 approaches the object. As shown in FIG. 16, when the calculated height of the object is higher than the height threshold ha, the ECU 30 sets the frequency threshold Fth to a predetermined frequency value Fpa. For example, the predetermined frequency value Fpa is 10. When the height of the object is equal to or less than the height threshold ha, the ECU 30 sets the frequency threshold Fth to a smaller value as the height of the object becomes lower.

[0077] Note that the frequency threshold Fth is not limited to being set so as to continuously decrease as the height of the object decreases, as shown in Fig. 16. For example, when the height of the object is higher than the height threshold ha, the frequency threshold Fth may be set to a predetermined frequency value Fpa=10, and when the height of the object is equal to or less than the height threshold ha, the frequency threshold Fth may be set to 5, which is smaller than the predetermined frequency value Fpa.

[0078] According to this embodiment, even for an object of low height, the presence of the opposing surface S corresponding to each mountain can be accurately determined.

[0079] <Other embodiments> The above embodiment may be modified as follows.

[0080] Instead of a sonar or other sound wave sensor, an on-board camera may be provided as the object detection device. The on-board camera may be, for example, a CCD camera, a CMOS image sensor, a near-infrared camera, or the like. The on-board camera captures an image of an area that spreads over a predetermined angle range, and acquires the captured image. The on-board camera may be a monocular camera or a stereo camera.

[0081] The ECU 30 acquires an image captured by an on-board camera. The ECU 30 calculates the detection position P and the detection angle of the detection position P by, for example, performing a three-dimensional reconstruction process on the acquired image.

[0082] The object detection devices are not limited to being installed on the front bumper, rear bumper, and vehicle body side of the vehicle 50, but may be installed on each of the right and left corners of the vehicle 50. The object detection device installed on the right corner transmits a search wave diagonally forward or backward to the right of the vehicle 50 and receives the reflected wave. The object detection device installed on the left corner transmits a search wave diagonally forward or backward to the left of the vehicle 50 and receives the reflected wave.

[0083] In detecting a peak in the calculated frequency distribution, instead of determining whether a detection angle at which the frequency of the detection position P is not detected exists between the peak values ​​of different detection angles, it may be determined whether the frequency of the detection position P between the peak values ​​of different detection angles is less than a predetermined value. Here, the predetermined value is a value larger than 0 and smaller than the frequency threshold value Fth, for example, 1 to 3. [Explanation of symbols]

[0084] 11...front sensor, 12...rear sensor, 13...side sensor, 30...ECU.

Claims

1. A vehicle control device (30) that uses an object detection device (11-13) that transmits detection waves to objects (51-56) present around a vehicle (50) and receives reflected waves of the detection waves to obtain distances and directions to the objects, obtains a plurality of detection positions of the objects in two-dimensional coordinates defined by the vehicle length direction and vehicle width direction of the vehicle, and performs parking assistance control to assist the vehicle in driving into a parking space based on each of the obtained detection positions, performing a clustering process for associating adjacent detection positions, among the acquired detection positions, whose distance between adjacent detection positions is shorter than a distance threshold, into the same group; an angle calculation unit that calculates a detection angle of each of the detection positions associated as a same group, the detection angle being an angle of a direction in which the reflected wave is received relative to a reference direction of the vehicle; a distribution calculation unit that calculates a frequency distribution of each of the detection positions associated as a same group with respect to the detection angle set to a predetermined interval based on the calculated detection angle and each of the detection positions associated as a same group; and A determination unit; Equipped with The determination unit is When a plurality of peaks having peak values ​​at different detection angles are present in the calculated frequency distribution, it is determined that the opposing surface of the object is present in the direction of the detection angle corresponding to the peak; The vehicle control device does not perform the determination when the calculated frequency distribution has one mountain having a peak value for the detection angle.

2. The object detection device is installed at a plurality of locations on the host vehicle, The vehicle control device according to claim 1 , wherein the detection positions obtained using a plurality of the object detection devices are used in the clustering process.

3. 3. The vehicle control device according to claim 1, wherein the determination unit determines that the opposing surface exists in the direction of the detection angle corresponding to each of the peaks, on the condition that the width of the detection angle between the peaks is determined to be greater than an angle width threshold for two adjacent peaks.

4. The vehicle control device according to any one of claims 1 to 3, wherein the determination unit determines that the opposing surface exists in the direction of the detection angle corresponding to the mountain, on the condition that it determines that the peak value of the mountain is greater than a frequency threshold value.

5. 5. The vehicle control device according to claim 1, further comprising a control unit that controls the host vehicle so as to keep a predetermined distance from the opposing surface of the object in the parking assistance control.

6. A program applied to a computer (30) that performs parking assistance control to assist in driving the vehicle into a parking space, using an object detection device (11-13) that transmits detection waves to objects (51-56) present around the vehicle (50) and receives reflected waves of the detection waves to obtain distances and directions to the objects, and obtains a plurality of detection positions of the objects in two-dimensional coordinates in the vehicle length direction and vehicle width direction of the vehicle, based on each of the obtained detection positions, the program comprising: The computer includes: a clustering process for associating adjacent detection positions, among the acquired detection positions, whose distance between adjacent detection positions is shorter than a distance threshold, into the same group; an angle calculation process for calculating a detection angle of each of the detection positions associated as a same group, the detection angle being an angle of a direction in which the reflected wave is received relative to a reference direction of the vehicle; a distribution calculation process for calculating a frequency distribution of each of the detection positions associated as a same group with respect to the detection angle set to a predetermined interval based on the calculated detection angle and each of the detection positions associated as a same group; A judgment process; Perform a process including In the determination process, When a plurality of peaks having peak values ​​at different detection angles are present in the calculated frequency distribution, it is determined that the opposing surface of the object is present in the direction of the detection angle corresponding to the peak; The program does not perform the determination if there is one mountain having a peak value for the detection angle in the calculated frequency distribution.

Images