Object detection device

Abstract

The present invention relates to an object detection device. The object detection device (10) is mounted on a vehicle (40) and detects an object by a reflected wave of a detection wave, and comprises: a speed calculation unit (22) for calculating the relative speed of the object with respect to the vehicle; an orientation estimation unit (23) for estimating the orientation of the object with respect to the vehicle; and an orientation correction unit (30) for correcting the orientation of the object with respect to the vehicle estimated by the orientation estimation unit, the orientation correction unit comprising: a speed orientation curve creation unit (32) for creating a speed orientation curve indicating the relationship between the relative speed of the object and the orientation of the object calculated based on a theoretical mounting angle of the object detection device; a weighting unit (33) for assigning a weight to an orientation error, that is, a difference between the speed orientation curve and the estimated orientation, according to the orientation estimated by the orientation estimation unit; and a correction value calculation unit (34) for calculating a weighted average value of the orientation errors assigned a weight by the weighting unit as an actual mounting angle offset of the object detection device, thereby calculating a correction value of the estimated orientation.

Description

Technical Field

[0001] The present disclosure relates to an object detection device that is mounted on a vehicle and detects an object by a reflected wave of a detection wave. Background Art

[0002] As described in Patent Document 1, when an object detection device such as a vehicle-mounted radar device is installed on the bumper of a vehicle to detect an object by a reflected wave of a detection wave, the detection performance of the orientation of the object is reduced due to the deviation of the mounting angle of the object detection device (deviation from the theoretical mounting angle). In order to improve the detection accuracy of the orientation of the object, it is preferred to calculate the detection error in the orientation of the object caused by the deviation of the mounting angle, that is, the orientation error. In Patent Document 1, the orientation error is calculated by comparing the velocity orientation curve calculated based on the theoretical mounting angle of the object detection device and the observation point with respect to the relationship between the relative speed of the object relative to the vehicle and the orientation of the object. Then, the actual mounting angle is calculated based on the orientation error, and the orientation of the detected object is corrected.

[0003] Patent Document 1: Japanese Patent No. 6358076

[0004] When there is an error in the speed of the vehicle, i.e., a speed error, the speed-direction curve of Patent Document 1 becomes a speed-direction curve including the speed error, and thus a deviation occurs. Therefore, the calculation accuracy of the direction error and the mounting angle may be reduced, and the calculation accuracy of the direction of the detected object may be reduced. Summary of the invention

[0005] In view of the above, an object of the present disclosure is to provide a technology that can detect the orientation of an object with high accuracy even when there is a vehicle speed error in an object detection device.

[0006] The present disclosure provides an object detection device that is mounted on a vehicle and detects an object by a reflected wave of a detection wave. The object detection device includes: a speed calculation unit that calculates the relative speed of the object with respect to the vehicle; an orientation estimation unit that estimates the orientation of the object with respect to the vehicle; and an orientation correction unit that corrects the orientation of the object with respect to the vehicle estimated by the orientation estimation unit. The orientation correction unit includes: a speed orientation curve creation unit that creates a speed orientation curve indicating the relationship between the relative speed of the object and the orientation of the object calculated based on a theoretical mounting angle of the object detection device; a weighting unit that gives a weight to the difference between the speed orientation curve and the estimated orientation, that is, the orientation error, according to the orientation estimated by the orientation estimation unit; and a correction value calculation unit that calculates a weighted average value of the orientation errors to which the weights are given by the weighting unit as an actual mounting angle offset of the object detection device, thereby calculating a correction value of the estimated orientation.

[0007] As a result of intensive research conducted by the present author, the author has obtained the insight that the offset in the speed-azimuth curve caused by the vehicle speed error varies according to the orientation of the object relative to the vehicle. Based on this insight, the object detection device involved in the present disclosure first estimates the orientation of the object relative to the vehicle through an orientation estimation unit. Then, the orientation correction unit assigns weights according to the estimated orientation, calculates the weighted average of the orientation error as the mounting angle offset, and thus calculates the correction value of the orientation. Therefore, the offset in the speed-azimuth curve caused by the vehicle speed error can be taken into account, and the mounting angle offset of the object detection device can be mitigated. As a result, even in the case of a vehicle speed error, the orientation of the object can be detected with good accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above-mentioned purpose and other purposes, features and advantages of the present disclosure will become more apparent through the following detailed description with reference to the accompanying drawings. In the accompanying drawings:

[0009] Figure 1 is a block diagram showing the structure of an object detection device,

[0010] Figure 2 is a diagram illustrating the location of the object detection device, the relative position and orientation of the object,

[0011] Figure 3 is a graph showing the relationship between the observation point and the velocity azimuth curve.

[0012] Figure 4 is a flowchart of the object detection process,

[0013] Figure 5 is a flowchart of the azimuth correction value calculation process.

[0014] Figure 6 This is a diagram for explaining weighting according to directions. DETAILED DESCRIPTION

[0015] Figure 1 The object detection device 10 shown is a device for detecting an object by using the reflected wave of the detection wave. Figure 2 As shown, the object detection device 10 is mounted on the vehicle 40 by being installed on the rear bumper 41 of the vehicle 40. The rear bumper 41 is made of a material that transmits electromagnetic waves such as radar waves, and the object detection device 10 is installed in the rear bumper 41 near the right end.

[0016] The object detection device 10 includes an antenna unit 11, a transceiver unit 12, and a signal processing unit 20. The object detection device 10 is communicably connected to other devices mounted on the vehicle via an in-vehicle local area network (LAN) (not shown).

[0017] The antenna unit 11 includes a plurality of antennas arranged in a row in the horizontal direction. The antenna unit 11 transmits and receives radar waves composed of multi-frequency CW (continuous wave) as detection waves.

[0018] The transceiver 12 periodically transmits and receives radar waves as detection waves at a constant time interval via the antenna 11. In addition, the transceiver 12 generates a beat signal consisting of a frequency component of the difference between the received signal and the transmitted signal according to each received signal received by each antenna constituting the antenna 11, and supplies the received data after A / D conversion to the signal processing unit 20. In addition, the multi-frequency CW is composed of a plurality of continuous waves of the GHz level with frequencies differing by about 1 MHz.

[0019] The signal processing unit 20 is a well-known microcomputer composed mainly of a CPU, a ROM, and a RAM. The signal processing unit 20 performs at least a main process of detecting an object that reflects radar waves and generating information related to the object according to a program stored in the ROM. In addition, a part of the RAM is composed of a non-volatile memory that retains the contents of the memory even when the power of the object detection device 10 is turned off. The non-volatile memory stores a azimuth correction table that indicates the correspondence between the relative speed with the object (here, the frequency window obtained by frequency analysis) and the azimuth error at the relative speed.

[0020] The signal processing unit 20 includes an object detection unit 21 , a velocity calculation unit 22 , a direction estimation unit 23 , and a direction correction unit 30 .

[0021] The object detection unit 21 can detect an object by, for example, acquiring a reflected wave of a radar wave received by the transmission / reception unit 12 and analyzing the reflected wave.

[0022] When an object is detected, the speed calculation unit 22 calculates the relative speed of the detected object with respect to the host vehicle 40. For example, the relative speed can be calculated based on the reflected wave of the radar wave received by the transceiver 12. More specifically, the relative speed of the object with respect to the host vehicle 40 can be calculated based on the frequency of the reflected wave of the radar wave reflected by the object that changes due to the Doppler effect.

[0023] When an object is detected, the direction estimation unit 23 estimates the direction of the detected object relative to the host vehicle 40. For example, the direction of the object can be calculated based on the phase difference of the reflected waves of the radar waves received by the multiple antennas of the transceiver unit 12. In addition, the distance between the host vehicle 40 and the object can be calculated based on the transmission time of the radar wave and the reception time of the reflected wave. In addition, if the position and direction of the object can be calculated, the relative position of the object relative to the host vehicle 40 can be determined.

[0024] The orientation correction unit 30 corrects the orientation of the object relative to the host vehicle 40 estimated by the orientation estimation unit 23 (hereinafter referred to as the estimated orientation). The orientation correction unit 30 includes an observation point distribution creation unit 31 , a speed orientation curve creation unit 32 , a weighting unit 33 , and a correction value calculation unit 34 .

[0025] When an object is detected by the object detection unit 21, the observation point distribution creation unit 31 creates object information including at least the relative speed calculated by the speed calculation unit 22 and the direction estimated by the direction estimation unit 23 for the detected object, and creates Figure 3 The observation point distribution P shown in FIG. The observation point distribution P includes N observation points about the object, and for the kth (k=1, ..., N) observation point, the relative speed Vok of the object and the estimated orientation θok are associated.

[0026] The speed-azimuth curve creating unit 32 obtains the relative speed Vo of the object calculated by the speed calculating unit 22 and the vehicle speed Vc, and creates a speed-azimuth curve C. As the vehicle speed Vc, for example, a detection value of a vehicle speed sensor or the like can be obtained. When there is a stationary object such as a wall on the right side of the host vehicle 40, a reflected wave can be obtained from various positions of the wall. In addition, the direction at which the observation point of the wall exists and the relative speed detected for the observation point have a relationship represented by the speed-azimuth curve C.

[0027] like Figure 2 As shown, the example object detection device 10 is mounted on the vehicle in a direction of a right-turned tilted mounting angle θr [deg] relative to the front-rear direction of the vehicle 40 indicated by the dotted line 50 when viewed from above the vehicle 40 indicated by the dotted line 51. When the object reflecting the radar wave is a stationary object, there is a relationship between the relative speed of the side stationary object and the direction in which the stationary object exists. Figure 2 The corresponding relationship shown in the following formula (2) is also recorded. Figure 2 The following formula (1) is also described to calculate the mounting angle deviation θd of the object detection device 10 .

[0028] [Number 1]

[0029]

[0030]

[0031] In addition, θr is the actual mounting angle of the object detection device 10. θb is the theoretical mounting angle (the mounting angle without angle offset). θik is the direction of the kth observation point of the object calculated based on θb. θok is the observed direction of the kth observation point of the object (i.e., the estimated direction). Vci is the actual vehicle speed of the vehicle 40. Vok is the relative speed of the kth observation point of the object. Ve is the vehicle speed error (the error of Vc). N is the total number of observation points of the object. The units of θd, θr, θik, θok, and θb are all degrees (rad). In addition, the detected vehicle speed Vc of the vehicle 40 can be expressed by Vc=Vci×Ve.

[0032] The velocity direction curve creating unit 32 is based on Figure 2 The above-described equation (2) is also used to create a velocity-azimuth curve C. That is, the velocity-azimuth curve C is a curve indicating the relationship between the relative velocity Vo of the object and the azimuth θi of the object calculated based on the theoretical mounting angle θb of the object detection device 10 .

[0033] When the laser detection range of the object detection device 10 is 0 to 180 degrees, the detection range is the range behind the dotted line 51 relative to the host vehicle 40 (in Figure 2 The dotted line 52 represents the orientation of the detected object, i.e., the estimated orientation θo, which is equivalent to the angle between the dotted line 51 and the dotted line 52. Figure 2 The detection range at the lower left of the dotted line 51 is set to 0 [deg], and the detection range at the upper right is set to 180 [deg]. Then, the positive side direction of the vehicle 40 (the right direction of the vehicle 40 perpendicular to the vehicle speed Vc) is 90+θr [deg].

[0034] like Figure 2 As shown, the relative speed Vo of the detected object can be expressed by Vo = -Vc × cos (θo - θr). When the side directly opposite the position where the object detection device 10 is mounted is the observation point, since θo = 90 + θr [deg], the relative speed Vo of the observation point is 0. That is, Figure 3 The azimuth of the intersection of the velocity azimuth curve C and the azimuth axis (θ axis) is 90+θr[deg]. The relative velocity Vo of the observation point located closer to the traveling direction of the host vehicle 40 than the observation point on the positive side becomes a positive value indicating that the vehicle is approaching the host vehicle 40. In addition, the relative velocity Vo of the observation point located closer to the traveling direction opposite to the host vehicle 40 than the observation point on the positive side becomes a negative value indicating that the vehicle is moving away from the host vehicle 40.

[0035] That is, when θ>90+θr[deg], the relative speed Vo is Vo>0, and when θ<90+θr[deg], the relative speed Vo is Vo<0. In addition, when the direction directly behind the host vehicle 40 is the observation point, θ=θr, so the speed direction curve C becomes a minimum value, and the relative speed Vo=-Vc. That is, the speed direction curve C becomes a shape that is symmetrical about the direction θr directly behind the host vehicle 40.

[0036] like Figure 3 As shown in FIG. 1 , the speed direction curve C changes according to the value of the vehicle speed error Ve. Figure 3 In the figure, the speed direction curve C shown by the solid line represents the speed direction curve when the vehicle speed error Ve is 1 (Ve=1). The speed direction curve C shown by the dashed line represents the speed direction curve when the vehicle speed error Ve is greater than 1 (Ve>1). The speed direction curve C shown by the dotted line represents the speed direction curve when the vehicle speed error Ve is less than 1 (Ve<1). Compared with the case of Ve=1, in the case of Ve>1, for the speed direction curve C, the rate of change of the relative speed Vo with respect to the direction becomes larger, and in the case of Ve<1, for the speed direction curve C, the rate of change of the relative speed Vo with respect to the direction becomes smaller. In addition, in any case of Ve>1 and Ve<1, the further the relative speed Vo is from zero, the greater the offset of the speed direction curve C.

[0037] like Figure 3 As shown, the azimuth axis direction of each point included in the velocity azimuth curve C and the observation point distribution P ( Figure 3 The difference (θik-θok) in the left and right directions (in the left and right directions) represents the azimuth error of each observation point. The mounting angle deviation θd can be calculated by calculating the sum average of the azimuth errors of each observation point using the above formula (1). It can be understood that if the speed azimuth curve C is offset due to the vehicle speed error Ve, the azimuth error is also offset, and the mounting angle deviation θd calculated by the above formula (1) is also offset.

[0038] If compared with the velocity-azimuth curve in the case of Ve=1, in the case of Ve>1, the velocity-azimuth curve shifts toward the direction close to the observation point distribution P as the relative velocity increases toward the approaching side, and the velocity-azimuth curve shifts toward the direction away from the observation point distribution P as the relative velocity increases toward the departing side. Therefore, in the case of Ve>1, the azimuth error is calculated to be smaller as the relative velocity increases toward the approaching side, and the azimuth error is calculated to be larger as the relative velocity increases toward the departing side. In contrast, in the case of Ve<1, the velocity-azimuth curve shifts toward the direction away from the observation point distribution P as the relative velocity increases toward the departing side, and the velocity-azimuth curve shifts toward the direction close to the observation point distribution P as the relative velocity increases toward the departing side. Therefore, in the case of Ve<1, the azimuth error is calculated to be larger as the relative velocity increases toward the approaching side, and the azimuth error is calculated to be smaller as the relative velocity increases toward the departing side. In any case of Ve>1 and Ve<1, the farther the relative speed is from zero, the larger the offset amount of the azimuth error is, and the larger the error in the calculated value of the mounting angle offset θd is.

[0039] On the other hand, in CW radar, it is theoretically impossible to measure an object whose relative velocity Vo is zero, and the measurement accuracy near the relative velocity Vo is also low. Therefore, it is difficult to accurately measure an object at a relative velocity Vo where the offset of the velocity azimuth curve C is almost zero, and it is necessary to correct the offset of the azimuth error based on the offset of the velocity azimuth curve C.

[0040] Therefore, in the weighting unit 33, a weight is assigned according to the estimated azimuth θok estimated by the azimuth estimation unit 23 based on the following insight: in either case of Ve>1 or Ve<1, the farther the azimuth θ is from 90+θr[deg], the greater the offset of the azimuth error (θik-θok).

[0041] For example, the weighting unit 33 may also make the weight smaller as the offset of the direction error caused by the vehicle speed error Ve of the host vehicle 40 is larger. Figure 3 As shown, since the offset of the direction error caused by the vehicle speed error Ve increases as the relative speed is further away from zero, the weight wk may be made smaller as the estimated direction θok of the object's relative speed Vok is further away from zero.

[0042] Alternatively, the weighting unit 33 may increase the weight as the estimated orientation θok of the object is closer to the normal direction orthogonal to the traveling direction of the host vehicle 40. Figure 3As shown, the direction of the normal line perpendicular to the traveling direction of the host vehicle 40 (the direction of ±90+θr[deg]) is the direction where the offset of the speed direction curve C caused by the vehicle speed error Ve is the smallest. Since the offset of the speed direction curve C is smaller as the value of θok is closer to the direction of the normal line perpendicular to the traveling direction of the host vehicle 40, it is preferable to increase the weight wk assigned.

[0043] Specifically, for example, the weighting unit 33 may also make the weight wk smaller when θok is larger when 0≤θok<θr, and make the weight wk larger when θr<θok<90+θr, and make the weight wk smaller when θok is larger when 90+θ≤θok≤180 (the unit is [deg]). In addition, the weight wk may change in a step-like manner relative to the estimated direction θok, or may change continuously. The weight wk may also be obtained in advance based on simulations, experiments, etc. that change the vehicle speed error Ve, and set it to a mapping or formula that corresponds to the estimated direction θok, and stored in the RAM of the signal processing unit 20.

[0044] The correction value calculation unit 34 calculates the mounting angle deviation θdw by calculating the weighted average value of the azimuth errors (θik-θok) given the weight wk by the weighting unit 33. Then, based on the mounting angle deviation θdw, the azimuth correction value θa is calculated, which is a correction value of the azimuth θo, which is the azimuth of the object estimated by the azimuth estimation unit 23.

[0045] Next, use Figure 4 The object detection process, which is the main process executed by the CPU of the signal processing unit 20, is described in the flowchart of FIG. 1 . The signal processing unit 20 periodically executes the object detection process in each measurement cycle of transmitting and receiving radar waves. Figure 4 The object detection process is shown.

[0046] First, in step S101, it is determined whether an object is detected. Specifically, the sampling data of the beat signal of one measurement cycle obtained by transmitting and receiving radar waves by the transceiver 12 is obtained to determine whether there is an object that reflects the radar wave. In addition, one measurement cycle includes sampling data related to all transmission frequencies of the multi-frequency CW. In step S101, if it is determined that an object is detected, the process proceeds to step S102. If it is not determined that an object is detected, the process ends.

[0047] In step S102, the relative speed Vok of the object detected at each observation point relative to the vehicle 40 is calculated. Specifically, by performing frequency analysis on the acquired sampled data, the spectrum is calculated according to each transmission frequency of the multi-frequency CW and according to each antenna constituting the antenna unit 11. The frequency window of the spectrum thus obtained represents the relative speed with the object that reflected the radar wave. In addition, a fast Fourier transform (FFT) can be used as a frequency analysis. And the sum average value Vo of the relative speed Vok is calculated for N observation points. After step S102, step S103 is entered.

[0048] In step S103, the direction θok of the object detected at each observation point relative to the vehicle 40 is estimated. Specifically, based on the spectrum obtained in step S102, the average spectrum is calculated for each antenna. Then, the frequency window in which the peak whose reception intensity is above a preset threshold is detected is extracted from the average spectrum, and the direction estimation process is performed for each frequency window. In addition, although it is preferred that the direction estimation process uses a high-resolution estimation process such as MUSIC (Multiple Signal Classification), beamforming and the like may also be used. And, for N observation points, the sum average value θo of the estimated direction θok is calculated. After step S103, proceed to step S104.

[0049] In step S104, the azimuth θo estimated in step S103 is corrected to calculate the azimuth correction value θa. Figure 5 The azimuth correction value calculation process shown is performed to calculate the azimuth correction value θa.

[0050] exist Figure 5 In the direction correction value calculation process shown in FIG. 1 , first, in step S201, the vehicle speed Vc of the host vehicle 40 is acquired via the in-vehicle LAN, and it is determined whether Vc exceeds a predetermined vehicle speed threshold value V1. The vehicle speed threshold value V1 is set to a speed direction curve C (see FIG. 1 ) showing the relationship between the relative speed Vo of the object and the observed direction θo. Figure 3 ) has a sufficiently large slope. In the case of Vc>V1, proceed to step S202. In the case of Vc≤V1, the process ends and proceeds to Figure 4 Step S105 shown.

[0051] In step S202 , a distribution of two-dimensional data consisting of the relative speed Vok and the estimated orientation θok calculated in steps S102 and S103 , that is, an observation point distribution P is created. Thereafter, the process proceeds to step S203 .

[0052] In step S203, based on Figure 2The speed-azimuth curve C is calculated using the above-described equation (2). The speed-azimuth curve C is a curve indicating the relationship between the relative speed of the object and the azimuth of the object, calculated based on the theoretical mounting angle θb of the object detection device 10. Then, the process proceeds to step S204.

[0053] In step S204, a weight wk is assigned according to the estimated direction θok. The weight wk is assigned in such a way that the smaller the offset amount of the direction error caused by the vehicle speed error Ve of the host vehicle 40 is, the larger the weight is. Figure 6 The following formula (3) is also described to calculate the weighted average value θdw.

[0054] [Number 2]

[0055]

[0056] Figure 6 The dashed line represents the speed-direction curve without speed error, and the solid line represents the speed-direction curve with speed error. Figure 6 In the above, the difference between the speed axis (V axis) direction of the two speed direction curves represents the vehicle speed error. Also, the difference between the direction of the azimuth axis (θ axis) of the two speed direction curves represents the azimuth error.

[0057] The offset of the heading error caused by the vehicle speed error varies according to θ, for example, the offset of the heading error θe1 is greater than the offset of the heading error θe2 (θe1>θe2). In this case, the smaller the offset of the heading error, the larger the value of the weight assigned, and the larger the offset of the heading error, the smaller the value of the weight assigned. By assigning weights in this way, the contribution of the observation point with a larger offset of the heading error can be reduced, while the contribution of the observation point with a smaller offset of the heading error can be increased. In the RAM of the signal processing unit 20, the weight wk set based on the offset of the heading error as described above is mapped and stored corresponding to the estimated heading θok. The weight wk is read from the mapping based on the estimated heading θok estimated in step S103. Thereafter, step S205 is entered.

[0058] In step S205, the orientation correction value θa is calculated based on the weighted average value θdw calculated in step S204. θdw is the mounting angle deviation θdw of the object detection device 10 calculated by assigning a weight wk to the orientation error (θik-θok) at each observation point. Specifically, the orientation correction value θa can be calculated according to θa=acos(Vo / -Vc)+θdw. After that, the process ends and the process proceeds to Figure 4 Step S105 shown.

[0059] In step S105, object information including at least the relative speed Vo calculated in step S102 and the orientation correction value θa calculated in step S205 is created and output to each device on the vehicle via the vehicle LAN. Figure 4 Processing shown.

[0060] As described above, according to the present embodiment, a weighted average value θdw is calculated by assigning a weight wk according to the estimated direction θok. The larger the offset of the direction error (θik-θok) caused by the vehicle speed error Ve, the smaller the weight wk. Then, the weighted average value θdw is used as the mounting angle offset, and the estimated direction is corrected by the direction correction value. Therefore, even in the case of a vehicle speed error Ve, the offset of the mounting angle can be mitigated, and the direction of the object can be calculated with good accuracy. The offset of the mounting angle caused by the vehicle speed error Ve can be mitigated without actually measuring the vehicle speed error Ve, so there is no need to add new equipment to the vehicle 40 and the object detection device 10. Therefore, it can also be relatively easily applied to existing object detection devices.

[0061] According to each of the above-mentioned embodiments, the following effects can be obtained.

[0062] The object detection device 10 is mounted on the host vehicle 40 and detects the object by the reflected wave of the detection wave. The object detection device 10 includes: a speed calculation unit 22 that calculates the relative speed Vo of the object with respect to the host vehicle 40, an orientation estimation unit 23 that estimates the orientation of the object with respect to the host vehicle 40, and an orientation correction unit 30 that corrects the orientation θo estimated by the orientation estimation unit 23. The orientation correction unit 30 includes: a speed orientation curve creation unit 32, a weighting unit 33, and a correction value calculation unit 34.

[0063] The speed-azimuth curve creating unit 32 generates a velocity-azimuth curve by, for example, Figure 2 The above-mentioned formula (2) also recorded creates a velocity-azimuth curve C representing the relationship between the relative velocity Vo of the object and the azimuth θi of the object calculated based on the theoretical mounting angle θb of the object detection device 10. The weighting unit 33 assigns a weight wk to the difference between the velocity-azimuth curve C and the estimated azimuth θok, that is, the azimuth error (θik-θok), for each observation point, based on the estimated azimuth θok estimated by the azimuth estimation unit 23. The correction value calculation unit 34 Figure 6 The above-mentioned formula (3) also calculates the weighted average value θdw of the azimuth errors given the weight wk by the weighting unit 33. Then, the weighted average value θdw is used as the mounting angle deviation of the object detection device 10 to calculate the correction value of the estimated azimuth θik.

[0064] According to the object detection device 10, based on the knowledge that the offset amount generated on the speed and direction curve C due to the vehicle speed error Ve changes according to the estimated direction θok of the object relative to the host vehicle 40, a weighted average value θdw calculated by multiplying the direction error (θik-θok) by the weight wk is calculated as the mounting angle of the object detection device 10. Therefore, the calculation deviation of the mounting angle of the object detection device 10 caused by the vehicle speed error Ve can be alleviated. Then, the orientation correction value θa of the object is calculated using the mounting angle offset θdw, so that even in the case of the vehicle speed error Ve, the orientation of the object can be detected with good accuracy.

[0065] The weighting unit 33 may also increase the weight wk as the error in the vehicle speed Vc of the host vehicle 40, that is, the smaller the offset of the heading error (θik-θok) caused by the vehicle speed error Ve, is. Since the contribution of the observation point with a smaller offset of the heading error caused by the vehicle speed error Ve can be increased and the loading angle offset θdw can be calculated, the offset of the loading angle caused by the vehicle speed error Ve can be reduced.

[0066] The weighting unit 33 may also increase the weight wk as the estimated direction θok estimated by the direction estimation unit 23 is closer to the normal direction orthogonal to the traveling direction of the host vehicle 40. The contribution of the observation point corresponding to the estimated direction θok having a smaller deviation of the speed direction curve C due to the vehicle speed error Ve can be increased, and the loading angle deviation θdw can be calculated, so that the deviation of the loading angle due to the vehicle speed error Ve can be reduced.

[0067] In addition, in the above-mentioned embodiment, the case where the antenna unit 11, the transceiver unit 12, and the signal processing unit 20 are integrated in the object detection device 10 provided on the bumper of the vehicle is described, but the present invention is not limited thereto. For example, the signal processing unit 20 may be configured in the ECU of the vehicle 40, and only the antenna unit 11 and the transceiver unit 12 may be provided on the bumper of the vehicle. In addition, when the vehicle 40 is equipped with a camera, a GNSS receiving device, etc., the signal processing unit 20 may also be configured to be able to use the data obtained from these devices for object detection, vehicle speed detection, etc.

[0068] The control unit and method described in the present disclosure may also be implemented by a special-purpose computer provided by a processor and a memory programmed to execute one or more functions embodied by a computer program. Alternatively, the control unit and method described in the present disclosure may also be implemented by a special-purpose computer provided by a processor composed of one or more special-purpose hardware logic circuits. Alternatively, the control unit and method described in the present disclosure may also be implemented by one or more special-purpose computers composed of a combination of a processor and a memory programmed to execute one or more functions and a processor composed of one or more hardware logic circuits. In addition, the computer program may also be stored as an instruction executed by a computer in a non-migratable tangible recording medium that can be read by a computer.

[0069] The present disclosure is described based on the embodiments, but it should be understood that the present disclosure is not limited to the embodiments and configurations. The present disclosure also includes various modifications and variations within the scope of equality. In addition, various combinations and methods, and other combinations and methods that only include one element, more or less, also fall within the scope and scope of the present disclosure.

Claims

1. An object detection device is an object detection device mounted on a vehicle and detects an object by a reflected wave of a detection wave, comprising: A speed calculation unit, which calculates a relative speed of the object with respect to the vehicle; a direction estimating unit, which estimates the direction of the object relative to the vehicle; as well as a direction correction unit for correcting the direction of the object relative to the vehicle estimated by the direction estimation unit, The above-mentioned orientation correction unit comprises: a speed-azimuth curve creating unit that creates a speed-azimuth curve indicating a relationship between a relative speed of the object and an azimuth of the object calculated based on a theoretical mounting angle of the object detection device; a weighting unit that assigns a weight to a difference between the velocity azimuth curve and the estimated azimuth, that is, an azimuth error, according to the azimuth estimated by the azimuth estimating unit; as well as a correction value calculation unit that calculates a correction value of the estimated orientation by calculating a weighted average value of the orientation errors to which the weights are assigned by the weighting unit as an actual mounting angle offset of the object detection device, The weighting unit increases the weight as the amount of deviation of the direction error caused by a vehicle speed error is smaller, and the vehicle speed error is an error in the vehicle speed of the vehicle.

2. The object detection device according to claim 1, wherein: The weighting unit increases the weight as the estimated direction is closer to a normal direction perpendicular to the traveling direction of the vehicle.

Images