Physical object detection device

Abstract

To provide a physical object detection device with which it is possible to improve the accuracy of detecting the shape of a physical object.SOLUTION: A physical object detection device comprises: a transmission unit 40A for transmitting an ultrasonic wave toward a physical object; a reception unit 40B for receiving a reflected wave from the physical object and outputting a reception signal; an amplitude signal generation unit 61 for generating an amplitude signal; and a physical object determination unit 7 for determining the shape of the physical object. The ultrasonic wave transmitted by the transmission unit has a change region whose frequency changes with the passage of time. The reception unit receives a superimposed reflected wave and outputs a reception signal corresponding to the superimposed reflection wave. The amplitude signal generation unit generates an amplitude signal on the basis of the reception signal outputted by the reception unit after receiving the superimposed reflected wave. The physical object determination unit determines the shape of the physical object on the basis of the amplitude signal based on the superimposed reflected wave generated by the amplitude signal generation unit.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present disclosure relates to an object detection device. [Background technology]

[0002] Conventionally, a cruise control controller has been disclosed that determines whether a detected obstacle is a tall obstacle by utilizing the fact that the intensity characteristics of reflected waves differ depending on the height of the obstacle (see, for example, Patent Document 1). The cruise control controller described in Patent Document 1 determines whether an obstacle is a tall obstacle based on an output signal from a sonar that receives waves reflected by the obstacle.

[0003] According to Patent Document 1, for example, if the obstacle is a curb, which is a low obstacle, of the transmission waves sent from the sonar, those sent in the direction of the road surface (i.e., diagonally downward) are reflected by the curb, and these reflected waves are then reflected by the road surface. The sonar receives the reflected waves reflected by the curb and road surface. Furthermore, transmission waves sent from the sonar in the horizontal direction pass above the curb. As a result, the sonar does not receive the reflected waves of these transmission waves sent in the horizontal direction. Therefore, the intensity characteristics of the reflected waves received by the sonar at this time have one peak that exceeds the obstacle detection threshold.

[0004] In contrast, if the obstacle is a tall wall, for example, the transmission wave transmitted horizontally from the sonar is reflected by the wall. As a result, the sonar receives the reflected wave reflected by the wall as is. Therefore, when the obstacle is a wall, the sonar receives two reflected waves within a short period of time: a reflected wave of the transmission wave transmitted diagonally downward and a reflected wave of the transmission wave transmitted horizontally. As a result, the intensity characteristics of the reflected waves received by the sonar at this time have two peaks that exceed the obstacle detection threshold.

[0005] For these reasons, the travel control controller described in Patent Document 1 determines whether an obstacle is a tall obstacle based on whether the reflected wave reflected by the obstacle has one or two peaks. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent No. 5846316 Summary of the Invention [Problem to be solved by the invention]

[0007] However, air turbulence can reduce the signal-to-noise ratio of the reflected waves. Also, two reflected waves, one from a transmission wave emitted diagonally downward and the other from a transmission wave emitted horizontally, can overlap, resulting in a single peak. In these cases, determining the shape of an object, such as the height of an obstacle, based on the number of peaks in the reflected waves can result in an incorrect determination of the object's shape.

[0008] In this way, the method of determining the presence of an object based on the number of peaks of the reflected wave may result in an erroneous determination of the object.

[0009] In view of the above, an object of the present invention is to provide an object detection device that can improve the accuracy of determining the shape of an object. [Means for solving the problem]

[0010] According to one aspect of the present disclosure, An object detection device that detects objects by transmitting and receiving ultrasonic waves a transmitting unit (40A) that transmits ultrasonic waves toward an object; a receiving section (40B) that receives a wave reflected from an object and outputs a reception signal corresponding to the reflected wave; an amplitude signal generator (61) that generates an amplitude signal of a received signal; an object determination unit (7) that determines the shape of an object based on the amplitude signal; The ultrasonic waves transmitted by the transmitting unit have a change region in which the frequency changes over time, the receiving unit receives a superimposed reflected wave including an area where a plurality of reflected waves reflected from a plurality of portions at different heights on the object overlap, and outputs a received signal corresponding to the superimposed reflected wave; the amplitude signal generating unit generates an amplitude signal based on a received signal output by the receiving unit after receiving the superimposed reflected wave; The object determining unit determines the shape of the object based on the amplitude signal generated by the amplitude signal generating unit based on the superimposed reflected waves.

[0011] When a transmitted wave having a change region in which the frequency changes over time is reflected by an object, the reflected wave also has a change region. When the transmitted wave is reflected by different parts of a tall object, the amplitude signal corresponding to the superimposed reflected wave, which includes an area where multiple reflected waves reflected by different parts overlap, will have a mixture of states in which the phases of these multiple reflected waves are in phase and states in which they are out of phase. According to the inventors' extensive research, the waveform of the amplitude signal based on such superimposed reflected waves will change depending on the shape of the object.

[0012] Therefore, the object detection device can accurately determine the shape of an object by receiving a superimposed reflected wave formed when a transmitted wave having a change region is reflected by an object, and determining the shape of the object based on an amplitude signal derived from the superimposed reflected wave.

[0013] The reference symbols in parentheses attached to each component indicate an example of the correspondence between the component and the specific components described in the embodiments described below. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a plan view showing a schematic configuration of a vehicle equipped with an object detection device according to a first embodiment. [Figure 2] 1 is a front view showing a schematic configuration of a vehicle equipped with an object detection device according to a first embodiment. [Figure 3] FIG. 2 is a diagram for explaining the installation position of an ultrasonic sensor according to the first embodiment. [Figure 4]1 is a block diagram of an object detection device according to a first embodiment. [Figure 5] 3A and 3B are diagrams for explaining a top-end reflected wave and a root-reflected wave received by an ultrasonic sensor; [Figure 6] 10A and 10B are diagrams for explaining the phase shift between the top reflected wave and the root reflected wave and the time shift in the reception timing. [Figure 7] FIG. 10 is a diagram illustrating an example of an amplitude signal based on a superimposed reflected wave. [Figure 8] FIG. 10 is a diagram illustrating an example of an amplitude image based on an amplitude signal. [Figure 9] FIG. 10 is a diagram illustrating an example of a combined amplitude image. [Figure 10] FIG. 3 is a diagram showing a control flow of an object detection process executed by the object detection device according to the first embodiment. [Figure 11] FIG. 10 is an explanatory diagram for explaining a convolution process for detecting amplitude features. [Figure 12] FIG. 10 is a diagram showing an example of a simulation result of an amplitude signal based on superimposed reflected waves that change depending on the height of an object. [Figure 13] 10A and 10B are diagrams for explaining the difference between two reflected waves when the mounting height of the ultrasonic sensor is changed. [Figure 14] 10 is a diagram for explaining the difference in time lag between the reception timings of two reflected waves when the mounting height of the ultrasonic sensor is changed. FIG. [Figure 15] FIG. 10 is a block diagram of an object detection device according to a second embodiment. [Figure 16] FIG. 10 is a diagram showing a control flow of an object detection process executed by an object detection device according to a second embodiment. [Figure 17] FIG. 10 is a block diagram of an object detection device according to a third embodiment. [Figure 18] FIG. 11 is a diagram showing a control flow of an object detection process executed by an object detection device according to a third embodiment. [Figure 19] 10A and 10B are diagrams illustrating changes in the positional relationship between an ultrasonic sensor and an object when a vehicle is traveling so as to approach the object. [Figure 20] FIG. 10 is a diagram showing an example of a peak line based on superimposed reflected waves. [Figure 21] FIG. 10 is a diagram illustrating an example of an amplitude image based on peak lines. [Figure 22] FIG. 10 is a diagram illustrating an example of a combined amplitude image. [Figure 23] FIG. 10 is an explanatory diagram for explaining a convolution process for detecting amplitude features. [Figure 24] FIG. 10 is a diagram showing an example of a simulation result of a peak line based on superimposed reflected waves that change depending on the height of an object. [Figure 25] FIG. 10 is a diagram illustrating an example of a feature map. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the following embodiments, parts that are the same as or equivalent to those described in the preceding embodiments will be given the same reference numerals, and their description may be omitted. Furthermore, in the embodiments, when only some of the components are described, the components described in the preceding embodiments can be applied to the remaining components. The following embodiments can be partially combined with each other, even if not specifically stated, as long as there is no particular problem with the combination.

[0016] (First embodiment) This embodiment will be described with reference to Figs. 1 to 14. As shown in Figs. 1 to 3, an object detection device 1 of this embodiment is mounted on a vehicle C as a moving body, and is configured to detect an object B around the vehicle. The vehicle C is a so-called four-wheeled automobile, and is equipped with a box-like body C1 that is formed into a substantially rectangular shape in a plan view. The shape of each part of the vehicle C in a "plan view" refers to the shape of that part when viewed from the line of sight in the same direction as the direction of gravity, with the vehicle C stably placed on a horizontal surface so that it can travel. The vehicle C equipped with the object detection device 1 according to this embodiment will be referred to as the "host vehicle" hereinafter.

[0017] Hereinafter, the imaginary line that passes through the center of the host vehicle in the vehicle width direction in a plan view and is parallel to the vehicle length direction of the host vehicle will be referred to as the vehicle center line LC. The vehicle length direction is a direction that is perpendicular to the vehicle width direction and perpendicular to the vehicle height direction. The vehicle height direction is a direction that defines the vehicle height of the host vehicle and is a direction parallel to the direction of gravity when the host vehicle is stably placed on a horizontal surface so that it can travel. Furthermore, "front," "rear," "left," "right," "up," and "down" are defined as indicated by arrows in Figures 1 to 3. In other words, the vehicle length direction is synonymous with the front-rear direction. Furthermore, the vehicle width direction is synonymous with the left-right direction. Furthermore, the vehicle height direction is synonymous with the up-down direction.

[0018] The object detection device 1 detects an object B by transmitting and receiving ultrasonic waves, and includes an ultrasonic sensor 2 and a control unit 3 that controls the operation of the ultrasonic sensor 2. The ultrasonic sensor 2 is configured to detect an object B by transmitting a search wave, which is an ultrasonic wave, and receiving the search wave reflected by the object B.

[0019] In this embodiment, the host vehicle is equipped with a plurality of ultrasonic sensors 2. Specifically, a first front sensor 2A, a second front sensor 2B, a third front sensor 2C, and a fourth front sensor 2D are mounted on the front bumper of the vehicle body C1, i.e., the front bumper C2 of the host vehicle, as ultrasonic sensors 2. Similarly, a first rear sensor 2E, a second rear sensor 2F, a third rear sensor 2G, and a fourth rear sensor 2H are mounted on the rear bumper of the vehicle body C1, i.e., the rear bumper C3 of the host vehicle, as ultrasonic sensors 2.

[0020] The first front sensor 2A is provided at the right end of the front bumper C2 so as to emit a transmission wave to the right front of the vehicle. The second front sensor 2B is disposed between the first front sensor 2A and the vehicle center line LC in the vehicle width direction so as to emit a transmission wave substantially ahead of the vehicle. The third front sensor 2C is disposed at a position substantially symmetrical to the second front sensor 2B across the vehicle center line LC. The third front sensor 2C is disposed between the vehicle center line LC and the fourth front sensor 2D in the vehicle width direction so as to emit a transmission wave substantially ahead of the vehicle. The fourth front sensor 2D is disposed at a position substantially symmetrical to the first front sensor 2A across the vehicle center line LC. The fourth front sensor 2D is provided at the left end of the front bumper C2 so as to emit a transmission wave to the left front of the vehicle.

[0021] The first front sensor 2A, the second front sensor 2B, the third front sensor 2C, and the fourth front sensor 2D are provided on the front bumper C2 so that their installation positions are substantially the same as the height of the object B, which is the detection target, from the road surface R. For example, as shown in FIG. 3, it is assumed that the height of the object B from the road surface R is around the front bumper C2. In this case, the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, and the fourth front sensor 2D are all provided at substantially the same position, below the center of the front bumper C2 in the vehicle height direction. Specifically, the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, and the fourth front sensor 2D are provided at the lower end of the front bumper C2.

[0022] The first rear sensor 2E is provided at the right end of the rear bumper C3 so as to emit a transmission wave to the right rear of the vehicle. The second rear sensor 2F is disposed between the first rear sensor 2E and the vehicle center line LC in the vehicle width direction so as to emit a transmission wave to approximately rear of the vehicle. The third rear sensor 2G is disposed at a position approximately symmetrical to the second rear sensor 2F across the vehicle center line LC. The third rear sensor 2G is disposed between the vehicle center line LC and the fourth rear sensor 2H in the vehicle width direction so as to emit a transmission wave to approximately rear of the vehicle. The fourth rear sensor 2H is disposed at a position approximately symmetrical to the first rear sensor 2E across the vehicle center line LC. The fourth rear sensor 2H is provided at the left end of the rear bumper C3 so as to emit a transmission wave to the left rear of the vehicle.

[0023] Furthermore, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H are provided at approximately the same installation positions in the vehicle height direction as the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, and the fourth front sensor 2D. That is, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H are provided on the rear bumper C3 so that their installation positions are approximately the same as the height of the object B from the road surface R. Furthermore, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H are provided at approximately the same installation positions in the vehicle height direction, at the lower end of the rear bumper C3.

[0024] Next, the schematic configuration of the ultrasonic sensor 2 will be described with reference to Fig. 4. In Fig. 4, for the sake of simplicity, only one of the multiple ultrasonic sensors 2 mounted on the vehicle is shown, and the others are omitted.

[0025] The ultrasonic sensor 2 is configured to transmit ultrasonic waves as transmission waves toward the outside of the vehicle, and is configured to detect an object B present in the vicinity based on a reception signal corresponding to a reception wave including a wave of the transmission wave reflected by the object B, and to obtain information about the shape of the object B.

[0026] Specifically, the ultrasonic sensor 2 includes a transmitter / receiver 4, a transmission signal generator 5, a reception signal processor 6, and an object determination unit 7. In this embodiment, the transmitter / receiver 4, transmission signal generator 5, reception signal processor 6, and object determination unit 7 are supported by a single sensor housing made of synthetic resin or the like.

[0027] The transceiver 4 has a transmitter 40A and a receiver 40B. The transmitter 40A is configured to be able to transmit a transmission wave, which is a probe wave, toward the object B. The receiver 40B is configured to be able to receive ultrasonic waves including a wave reflected by the object B of the probe wave transmitted from the transmitter 40A.

[0028] The transmitting / receiving unit 4 includes a transducer 41, a transmitting circuit 42, and a receiving circuit 43. The transmitting unit 40A is made up of the transducer 41 and the transmitting circuit 42. The receiving unit 40B is made up of the transducer 41 and the receiving circuit 43.

[0029] The transducer 41 functions as a transmitter that transmits transmitted waves to the outside and as a receiver that receives reflected waves, and is electrically connected to a transmitting circuit 42 and a receiving circuit 43. In other words, the ultrasonic sensor 2 has a so-called integrated transmitting and receiving configuration.

[0030] Specifically, the transducer 41 is configured as an ultrasonic microphone incorporating an electromechanical energy conversion element such as a piezoelectric element. When an electrical signal is input from the transmission circuit 42, the electromechanical energy conversion element vibrates and transmits ultrasonic waves as transmission waves, thereby functioning as a vibration unit that converts the input electrical signal into a sound pressure signal. Furthermore, when the electromechanical energy conversion element receives ultrasonic waves including a reflected wave from the object B and vibrates due to the ultrasonic sound pressure signal, it converts the intensity of the sound pressure signal into an electrical signal and outputs it. For example, a piezoelectric ceramic vibrator, a piezoelectric single crystal vibrator, or a polymer piezoelectric film vibrator can be used as the electromechanical energy conversion element. Hereinafter, the electromechanical energy conversion element will be referred to as a vibration conversion element. The transducer 41 is positioned facing the outer surface of the vehicle so that it can transmit transmission waves to the outside of the vehicle and receive reflected waves from the outside of the vehicle.

[0031] The transmission circuit 42 is configured to drive the transducer 41 based on the input transmission signal, thereby causing the transducer 41 to emit ultrasonic waves. Specifically, the transmission circuit 42 has a digital / analog conversion circuit and the like. That is, the transmission circuit 42 is configured to generate an element input signal by performing signal processing such as digital / analog conversion on the transmission signal output from the transmission signal generation unit 5. The element input signal is an AC voltage signal for driving the transducer 41. The transmission circuit 42 is configured to apply the generated element input signal to the transducer 41 to excite the vibration conversion element in the transducer 41, thereby generating ultrasonic waves as transmission waves. The frequency of the ultrasonic waves generated by the vibration of the vibration conversion element is controlled according to the frequency of the AC voltage signal output from the transmission circuit 42.

[0032] The receiving circuit 43 is configured to generate a receiving signal corresponding to the result of ultrasonic wave reception by the transducer 41 and output the signal to the receiving signal processing unit 6. Specifically, the receiving circuit 43 has an amplifier circuit, an analog-to-digital conversion circuit, etc. That is, the receiving circuit 43 is configured to perform signal processing such as amplification and analog-to-digital conversion on the element output signal output by the transducer 41 to generate a receiving signal including information on the amplitude and frequency of the reflected wave. The element output signal is an AC voltage signal generated by a vibration conversion element provided in the transducer 41 in response to reception of ultrasonic waves. When the transmitting circuit 42 repeatedly transmits ultrasonic waves, the receiving circuit 43 generates a receiving signal corresponding to the received ultrasonic waves and outputs the signal to the receiving signal processing unit 6 each time the transducer 41 receives a reflected wave of the ultrasonic waves.

[0033] The transmission signal generation unit 5 is configured to generate a transmission signal and output it to the transmission circuit 42. The transmission signal is a signal for driving the transducer 41 to emit ultrasonic waves from the transducer 41. The transmission signal generation unit 5 is electrically connected to the control unit 3, and drives the transducer 41 based on a control signal, which is a transmission instruction, sent from the control unit 3. Specifically, the transmission signal generation unit 5 controls the frequency of the AC voltage signal output by the transmission circuit 42 based on the control signal sent from the control unit 3, thereby controlling the frequency of the ultrasonic waves generated by vibrating the vibration conversion element of the transducer 41.

[0034] The transmission signal generation unit 5 transmits one transmission signal to the transmission circuit 42, thereby vibrating the vibration conversion element of the transducer 41 to generate one ultrasonic wave. The transmission signal generation unit 5 also vibrates the vibration conversion element of the transducer 41 so that the frequency of one ultrasonic wave generated by one transmission signal includes a change region in which the frequency changes over time. Specifically, the transmission signal generation unit 5 of this embodiment vibrates the vibration conversion element of the transducer 41 so that the frequency of the ultrasonic wave decreases over time. In other words, the transmission signal generation unit 5 vibrates the vibration conversion element of the transducer 41 so that the ultrasonic wave becomes a down-chirp signal.

[0035] When the ultrasonic waves transmitted by the ultrasonic sensor 2 are down-chirp signals, the frequency of one ultrasonic wave generated by one transmission signal decreases over time from the start of transmission to the completion of transmission. Therefore, the entire region of the ultrasonic wave from the start of transmission to the completion of transmission is a variable region in which the frequency changes over time. In this way, the transmission signal generation unit 5 of this embodiment generates a transmission wave that becomes a down-chirp signal by transmitting a transmission signal to the transmitter 40A. Then, by transmitting the transmission wave that is a down-chirp signal, the transmitter 40A converts the reflected wave reflected by the object B and received by the receiver 40B into a down-chirp signal.

[0036] The received signal processing unit 6 is configured to perform various signal processing such as filtering and quadrature detection on the received signal output from the receiving circuit 43. The received signal processing unit 6 is also configured to output a processed signal resulting from the various signal processing to the object determination unit 7. The received signal processing unit 6 processes the received signal and generates a signal used for detecting an object B in the object determination unit 7. As shown in FIG. 2, the received signal processing unit 6 includes an amplitude signal generation unit 61 and a feature acquisition unit 62.

[0037] The amplitude signal generator 61 generates an amplitude signal of the received signal output by the receiving circuit 43. For example, the amplitude signal generator 61 generates the amplitude signal by using the strength of the received signal as the amplitude. Alternatively, the amplitude signal generator 61 performs quadrature detection on the received signal and generates the amplitude signal by using the amplitude of the complex signal generated thereby as the amplitude of the received signal. Alternatively, the amplitude signal generator 61 generates the amplitude signal by calculating the correlation between the received signal and a reference signal set based on the transmitted signal. When the amplitude signal is generated by such correlation calculation, the signal width of the amplitude signal is reduced due to the effect of pulse compression, improving the S / N ratio and improving the detection accuracy of the reflected signal. The amplitude signal generator 61 outputs the generated amplitude signal to the feature acquirer 62.

[0038] The feature acquisition unit 62 of this embodiment detects amplitude features that indicate the characteristics of the amplitude signal from the amplitude signal output by the amplitude signal generation unit 61, and outputs amplitude feature information that is information related to the detected amplitude features.

[0039] Here, the waveform of the amplitude signal output by the amplitude signal generating unit 61 when the ultrasonic sensor 2 receives one reflected wave will be described with reference to FIGS. 5 to 7. The transmission wave transmitted from the ultrasonic sensor 2 has directionality. If the object B is an obstacle with a certain height, the transmission wave reflected by the object B tends to be strongly reflected by a portion of the object B that is relatively close to the ultrasonic sensor 2 and the base of the object B, as shown in FIG. 5. For example, if the object B is an obstacle that is shorter in height than the installation position of the ultrasonic sensor 2 as shown in FIG. 5, the transmission wave tends to be strongly reflected by the upper end portion of the object B, which is a position relatively close to the ultrasonic sensor 2, and the base of the object B. Furthermore, although not shown, if the obstacle is taller than the installation position of the ultrasonic sensor 2, the transmission wave tends to be strongly reflected by the portion of the object B in front of the ultrasonic sensor 2 and the base of the object B.

[0040] Therefore, if the object B is lower than the installation position of the ultrasonic sensor 2, the ultrasonic sensor 2 receives a single reflected wave that includes an overlapping region between the reflected wave reflected from the top end of the object B and the reflected wave reflected from the base of the object B. The amplitude of the reflected wave, which includes an overlapping region between the reflected wave reflected from the top end of the object B and the reflected wave reflected from the base of the object B, is amplified or attenuated compared to the amplitude of the reflected wave before the overlapping region, depending on the phase of each of the two reflected waves. The transducer 41 receives this reflected wave, which includes an overlapping region between the reflected wave reflected from the top end of the object B and the reflected wave reflected from the base of the object B. Hereinafter, a reflected wave that includes an overlapping region between reflected waves reflected from multiple parts of the object B at different heights is also referred to as a superimposed reflected wave. The reflected wave reflected from the top end of the object B is also referred to as a top reflected wave, and the reflected wave reflected from the base of the object B is also referred to as a base reflected wave.

[0041] Incidentally, the distance from the ultrasonic sensor 2 to the top end of the object B is different from the distance from the ultrasonic sensor 2 to the base of the object B. Specifically, the distance from the ultrasonic sensor 2 to the top end of the object B is shorter than the distance from the ultrasonic sensor 2 to the base of the object B. For this reason, the ultrasonic sensor 2 receives the top reflected wave earlier than the base reflected wave.

[0042] Here, assuming that the ultrasonic sensor 2 transmits one transmission wave and then receives one top-end reflected wave and one root reflected wave independently, an example of the simulation results of the amplitude signals corresponding to these top-end reflected wave and root reflected wave is shown in Fig. 6. In Fig. 6, the solid line indicates the reception time and amplitude intensity when the ultrasonic sensor 2 receives the top-end reflected wave first, and the dashed line indicates the reception time and amplitude intensity when the ultrasonic sensor 2 receives the root reflected wave later.

[0043] As shown in Fig. 6, the top-end reflected wave and the root reflected wave are waveforms whose amplitude intensity periodically increases and decreases, with maximum and minimum amplitude intensity values ​​repeating alternately. The absolute value of the amplitude intensity of both the top-end reflected wave and the root reflected wave gradually increases with the passage of reception time, and after reaching a maximum, the absolute value gradually decreases with the passage of reception time. This is because, when a transmission signal is transmitted so that the vibration conversion element of the transducer 41 vibrates, the vibration amount of the vibration conversion element gradually increases, and the vibration amount attenuates and gradually decreases in the reverberation region after transmission of the transmission signal is stopped.

[0044] As shown in Figure 6, there is a time lag between the timing at which the ultrasonic sensor 2 receives the top-end reflected wave and the timing at which it receives the bottom-end reflected wave. This time lag becomes larger as the distance from the top end to the bottom end of object B increases, and becomes smaller as the distance decreases. In other words, the time lag becomes larger as the height of object B increases, thereby decreasing the difference between the top end of object B and the mounting height of ultrasonic sensor 2, and becomes smaller as the height of object B decreases, thereby increasing the difference between the top end of object B and the mounting height of ultrasonic sensor 2.

[0045] For this reason, the greater the distance from the top end to the base of object B, the smaller the area where the top end reflected wave and the root reflected wave overlap when the ultrasonic sensor 2 receives them. Conversely, the shorter the distance from the top end to the base of object B, the larger the area where the top end reflected wave and the root reflected wave overlap when the ultrasonic sensor 2 receives them. In other words, the area where the amplitudes of the top end reflected wave and the root reflected wave interfere with each other when the ultrasonic sensor 2 receives them changes depending on the height of object B. For this reason, the shape of the superimposed reflected wave, in which the top end reflected wave and the root reflected wave overlap and are received by the ultrasonic sensor 2, changes depending on the height of object B.

[0046] As described above, the transmission signal generating unit 5 of the ultrasonic sensor 2 of this embodiment causes the transmitting unit 40A to transmit a down-chirp signal whose frequency decreases over time as a transmission wave. Therefore, the reflected wave received by the receiving unit 40B also becomes a down-chirp signal whose frequency decreases over time.

[0047] Therefore, as shown in Figure 6, the amplitude signals corresponding to the top end reflected wave and the root reflected wave each have a larger wavelength the later the reception time. Furthermore, a time lag occurs between the reception timing of the top end reflected wave and the reception timing of the root reflected wave. For this reason, when compared at the same reception time, the amplitude signals corresponding to the top end reflected wave and the root reflected wave will sometimes be in phase with each other and sometimes be out of phase with each other. Specifically, the amplitude signals corresponding to the top end reflected wave and the root reflected wave will alternate between being in phase with each other and being out of phase with each other.

[0048] In addition, in order to make it easier to understand the phase states of the amplitude signals corresponding to the top-end reflected wave and the bottom-end reflected wave shown in Figure 6, the frequency is shown decreasing over time even in the reverberation region after the transmission of the transmitted signal has stopped.

[0049] For example, suppose that the ultrasonic sensor 2 first receives only the top-end reflected wave, then simultaneously receives two reflected waves, the top-end reflected wave and the root reflected wave, and then receives only the root reflected wave, as shown in Fig. 6. In this case, in the area where the reflected waves are simultaneously received, the top-end reflected wave and the root reflected wave are initially in phase with each other, then shift in phase with each other to become out of phase, and then again become in phase with each other.

[0050] In such a case, when the ultrasonic sensor 2 receives one reflected wave including a superimposed reflected wave, the waveform of the amplitude signal output by the amplitude signal generating unit 61 is a waveform in which the amplitude intensity periodically increases and decreases, and in which maximum and minimum values ​​of the amplitude intensity alternately repeat, as shown in Fig. 7. Specifically, in the region where only the upper-end reflected wave is received, the waveform of the amplitude signal is a waveform corresponding to the amplitude of the upper-end reflected wave, and in which maximum and minimum values ​​of the amplitude intensity of the upper-end reflected wave alternately repeat.

[0051] In a region where the top-end reflected wave and the root reflected wave are received simultaneously, the waveform of the amplitude signal is a waveform in which the absolute values ​​of the top-end reflected wave and the root reflected wave alternate between maximum and minimum values, where the absolute values ​​are amplified, or a waveform in which the absolute values ​​of the top-end reflected wave and the root reflected wave alternate between maximum and minimum values, where the absolute values ​​are attenuated. Also, in a region where the top-end reflected wave and the root reflected wave are received simultaneously, the waveform is a waveform in which there are alternate regions where the absolute values ​​of the maximum and minimum values ​​are gradually amplified and regions where the absolute values ​​of the maximum and minimum values ​​are gradually attenuated. In a region where only the root reflected wave is received, the waveform of the amplitude signal is a waveform corresponding to the amplitude of the root reflected wave, and is a waveform in which the amplitude intensity of the root reflected wave alternates between maximum and minimum values.

[0052] Hereinafter, as shown in Fig. 7, among the regions where the top-end reflected wave and the root reflected wave are simultaneously received, the region where the absolute values ​​of the maximum and minimum values ​​are amplified due to the overlap of their amplitudes is referred to as the amplification region RA. Also, among the regions where the top-end reflected wave and the root reflected wave are simultaneously received, the region where the absolute values ​​of the maximum and minimum values ​​are attenuated due to the cancellation of their amplitudes is referred to as the attenuation region RB.

[0053] When the ultrasonic sensor 2 receives amplitude signals corresponding to the top-end reflected wave and the root reflected wave shown in FIG. 6 , the amplitude signal generator 61 outputs an amplitude signal with alternating maximum and minimum values, as shown in FIG. 7 . Specifically, the waveform of the amplitude signal output by the amplitude signal generator 61 has two amplification regions RA and an attenuation region RB between them. In the amplification region RA, where the amplitude intensity is amplified by the overlapping of amplitudes, the absolute values ​​of the maximum and minimum values ​​of the amplitude intensity tend to be large. In contrast, in the attenuation region RB, where the amplitude intensity is attenuated by the cancellation of amplitudes, the absolute values ​​of the maximum and minimum values ​​of the amplitude intensity tend to be smaller than those in the amplification region RA. The amplitude signal generator 61 generates an amplitude signal for each reflected wave, including the overlapping reflected wave, and outputs the generated amplitude signal to the feature acquisition unit 62.

[0054] The feature acquisition unit 62 detects amplitude features that indicate the characteristics of the amplitude signal from such an amplitude signal output by the amplitude signal generation unit 61, and outputs amplitude feature information that is information on the detected amplitude features to the object determination unit 7. The feature acquisition unit 62 converts one amplitude signal into an image as shown in Fig. 8 that has information on the magnitude of the amplitude intensity of the amplitude signal, and outputs information including the converted image as amplitude features to the object determination unit 7. Specifically, the feature acquisition unit 62 converts the magnitude of the amplitude intensity according to the reception timing of the received reflected wave into a brightness value of a predetermined gradation, and outputs image information including information on the brightness value according to the amplitude intensity for each reception timing to the object determination unit 7 as amplitude features.

[0055] 8, the feature acquisition unit 62 converts the magnitude of the amplitude intensity for each reception time of a received amplitude signal into a straight line image of a predetermined gradation, and generates an image in which the straight line images converted for each reception time are arranged in order of reception time. That is, the feature acquisition unit 62 digitizes the magnitude of the amplitude intensity for each reception timing of the amplitude signal using a number of predetermined steps, and generates an image in which straight line images of densities corresponding to the calculated number are arranged in order of reception time.

[0056] In this embodiment, the feature acquisition unit 62 generates a black-and-white image in which the greater the amplitude signal, the lighter the shading of the line, and the smaller the amplitude signal, the darker the shading of the line. As a result, the color of the line corresponding to the magnitude of the amplitude intensity is closer to white for amplitude signals with greater amplitude intensity, and closer to black for amplitude signals with smaller amplitude intensity.

[0057] In the region receiving only the top-end reflected wave, the region receiving only the root reflected wave, and the attenuation region RB where the amplitude intensity is attenuated, the absolute values ​​of the maximum and minimum amplitude intensity values ​​tend to be smaller than those in the amplification region RA. Therefore, as shown in FIG. 8, the image regions based on the region receiving only the top-end reflected wave, the region receiving only the root reflected wave, and the attenuation region RB produce images with alternating straight lines with relatively small differences in shading. In contrast, in the amplification region RA where the amplitude intensity is amplified by overlapping amplitudes, the absolute values ​​of the maximum and minimum amplitude intensity values ​​tend to be large. Therefore, the image region based on the amplification region RA produces images with alternating straight lines with relatively large differences in shading. Hereinafter, the image information, which is the amplitude feature generated by the feature acquisition unit 62 as shown in FIG. 8, is also referred to as an amplitude image.

[0058] The gradation of the intensity of the amplitude may be, for example, 4 bits (that is, 16 gradations) or 8 bits (that is, 256 gradations).

[0059] The amplitude image is created by converting each amplitude signal received by the feature acquisition unit 62 for each reception timing into a single linear image and arranging the images in order of reception time. The amplitude signal is generated by the amplitude signal generation unit 61 when the ultrasonic sensor 2 receives a reflected wave from object B. The timing at which the amplitude intensity increases, i.e., the reception time of the amplitude signal, changes depending on the distance between the ultrasonic sensor 2 and object B. Therefore, the amplitude image based on the amplitude signal contains information about the distance to object B that reflected the reflected wave received by the ultrasonic sensor 2. Therefore, the information about the reception time of the amplitude signal contained in the amplitude image also corresponds to the distance between the ultrasonic sensor 2 and object B. Hereinafter, the distance between the ultrasonic sensor 2 and object B, which is the detection target, will also be referred to as the ranging distance.

[0060] Then, as shown in Fig. 9, the feature acquisition unit 62 generates one image by arranging multiple amplitude images generated from one reflected wave. Specifically, when the direction in which lines indicating the amplitude intensity of the amplitude signal for each reception timing in the amplitude image are aligned is defined as the signal direction, the feature acquisition unit 62 generates one image by arranging a predetermined number of amplitude images in a direction perpendicular to the signal direction. The signal direction is the direction in which information indicating the amplitude intensity for each reception timing is aligned, and is the direction in which information indicating the amplitude intensity according to the ranging distance is aligned. When the amplitude image is an image displayed in two-dimensional coordinates of X and Y coordinates, if the X direction is the signal direction, the direction perpendicular to the signal direction is the Y direction.

[0061] The feature acquisition unit 62 outputs information about an image generated by arranging a plurality of amplitude images in a direction perpendicular to the signal direction as amplitude feature information to the object determination unit 7. Hereinafter, an image generated by arranging a plurality of amplitude images in a direction perpendicular to the signal direction will be referred to as a concatenated amplitude image. The feature acquisition unit 62 generates a concatenated amplitude image by arranging, for example, ten amplitude images in a direction perpendicular to the signal direction. Note that the number of amplitude images constituting the concatenated amplitude image is not limited to ten, and may be more or less than ten.

[0062] When the transmitter 40A repeatedly transmits a transmission wave, the feature acquirer 62 repeatedly generates, as a concatenated amplitude image, the amplitude features of the respective amplitude signals output by the amplitude signal generator 61 each time the transducer 41 receives a wave reflected by the object B. Then, the feature acquirer 62 outputs the repeatedly generated concatenated amplitude image to the object determiner 7 as amplitude feature information.

[0063] The object determination unit 7 determines the shape of the object B to be detected based on the amplitude feature information output from the feature acquisition unit 62. Specifically, the object determination unit 7 determines the height of the shape of the object B based on a combined amplitude image generated from amplitude signals based on superimposed reflected waves.

[0064] 4, the object determination unit 7 has a storage unit 71 that stores amplitude feature information output from the feature acquisition unit 62, and a determination unit 72 that determines the height of the object B. The object determination unit 7 transmits the determined object detection determination result to the control unit 3. The method of determining the height of the object B by the object determination unit 7 will be described in detail later.

[0065] The transmission signal generator 5, the reception signal processor 6, and the object determiner 7 are configured with a DSP programmed with functions such as transmission signal generation, amplitude signal generation, amplitude feature detection, and object detection / determination. DSP is an abbreviation for Digital Signal Processor.

[0066] The control unit 3 is connected to the ultrasonic sensor 2 via an in-vehicle communication line so as to be able to communicate information with the ultrasonic sensor 2, and is configured to control the transmission and reception operations of the ultrasonic sensor 2. The control unit 3 is provided as a so-called sonar ECU, and is equipped with an in-vehicle microcomputer having a CPU, ROM, RAM, non-volatile rewritable memory, etc. (not shown). ECU is an abbreviation for Electronic Control Unit. Examples of non-volatile rewritable memory include EEPROM and flash ROM. EEPROM is an abbreviation for Electronically Erasable and Programmable Read Only Memory.

[0067] As described above, the control unit 3 controls the frequency of the ultrasonic waves by controlling the frequency at which the vibration conversion element of the transducer 41 vibrates by sending a control signal to the transmission signal generation unit 5. The control unit 3 also vibrates the vibration conversion element of the transducer 41 so that one ultrasonic wave generated by one transmission signal includes a change region in which the frequency changes over time. More specifically, the control unit 3 vibrates the vibration conversion element of the transducer 41 so that the transmission wave transmitted by the ultrasonic sensor 2 and the reflected wave received by it become down-chirp signals.

[0068] When the ultrasonic sensor 2 receives the ultrasonic waves, the object determination unit 7 transmits the detection determination result of the object B to the control unit 3. This detection determination result is used for processing such as obstacle notification and automatic parking. For example, when the control unit 3 receives the detection determination result from the object determination unit 7 while the host vehicle is moving due to automatic parking, the control unit 3 reduces the speed of the host vehicle or stops the host vehicle.

[0069] Next, the operation of the object detection device 1 will be described with reference to the object detection process shown in FIG. 10. In this embodiment, for simplicity of explanation, only the object detection process executed by one of the multiple ultrasonic sensors 2 mounted on the host vehicle will be described, and explanations of the processes executed by the others will be omitted. Furthermore, in this embodiment, an example of operation when the object detection device 1 executes the object detection process to detect object B when the host vehicle is traveling so as to approach object B will be described. When the object detection device 1 executes the object detection process, first, the control unit 3 outputs a control signal and outputs a transmission instruction to the transmission signal generation unit 5. Specifically, the control unit 3 outputs a control signal to the transmission signal generation unit 5, which causes the ultrasonic waves generated from the transducer 41 to become a chirp signal.

[0070] Then, the transmission signal generating unit 5 generates a transmission signal based on the control signal and outputs the generated transmission signal to the transmitting unit 40A. The transmitting unit 40A is driven by the transmission signal. The transmitting circuit 42 vibrates the vibration converting element of the transducer 41 based on the input transmission signal so that the frequency at which the vibration converting element vibrates decreases over time. As a result, ultrasonic waves of a chirp signal are transmitted from the transducer 41, which functions as a transmitter, to the outside of the vehicle.

[0071] Then, when the ultrasonic waves are reflected by the object B and a reflected wave is generated, the object detection device 1 executes the object detection process shown in Fig. 10. Every time the transducer 41 receives a reflected wave including a superimposed reflected wave, the object detection device 1 repeatedly executes the object detection process shown in Fig. 10.

[0072] Here, it is assumed that the transducer 41 receives a superimposed reflected wave including an upper end reflected wave reflected from the upper end of the object B and a base reflected wave reflected from the base. In the object detection process, when the transducer 41 receives the superimposed reflected wave, the vibration conversion element of the transducer 41 is excited. When the vibration conversion element of the transducer 41 is excited, a voltage signal according to the excited state is output from the transducer 41.

[0073] The receiving circuit 43 generates a received signal by performing signal processing such as amplification and analog-to-digital conversion on the voltage signal output from the transducer 41, and outputs this received signal to the received signal processing unit 6. Here, if the transmitted wave is a chirp signal, the vibration conversion element of the transducer 41, which is excited when the transducer 41 receives the superimposed reflected wave, has a vibration frequency that increases over time. Therefore, the received signal received by the received signal processing unit 6 becomes a signal whose signal strength increases over time as the time of reception. When the received signal is received, the received signal processing unit 6 executes the following processing.

[0074] First, in step S10, the received signal processing unit 6 performs threshold processing on the received signal. Specifically, the amplitude signal generation unit 61 generates an amplitude signal of the received signal by performing various signal processing such as filtering and quadrature detection on the received signal. The amplitude signal generation unit 61 then determines whether the generated amplitude signal contains a signal with a signal strength equal to or greater than a predetermined determination threshold. The determination threshold is a threshold that is set in advance to determine whether or not there is a possibility that an object B, which is the detection target, is present, and is a reference signal value that is set to avoid erroneous detection due to noise, etc.

[0075] If it is not determined that the amplitude signal includes a signal with a signal strength equal to or greater than the determination threshold, the processes from step S20 onward are skipped and the object detection device 1 ends the object detection process. On the other hand, if it is determined that the amplitude signal includes a signal with a signal strength equal to or greater than the determination threshold, the received signal processing unit 6 performs a reflected wave detection process in step S20. In the reflected wave detection process, the amplitude signal generation unit 61 determines that the received signal received from the receiving circuit 43 includes a reflected wave from object B, and outputs the generated amplitude signal to the feature acquisition unit 62.

[0076] In step S30, the feature acquisition unit 62 detects amplitude features from the amplitude signal output by the amplitude signal generation unit 61, and outputs amplitude feature information to the object determination unit 7. Specifically, the feature acquisition unit 62 converts the magnitude of the amplitude intensity for each reception timing of the acquired amplitude signal into a luminance value corresponding to the amplitude intensity for each reception timing, and generates an amplitude image including information on the luminance value corresponding to the amplitude intensity for each reception timing. Then, the feature acquisition unit 62 outputs a concatenated amplitude image in which multiple amplitude images are arranged to the object determination unit 7 as amplitude feature information.

[0077] As described above, when the ultrasonic sensor 2 receives a superimposed reflected wave in which the top-end reflected wave and the base reflected wave are overlapped, the waveform of the amplitude signal changes depending on the height of the object B, i.e., the shape of the object B. For this reason, in an amplitude image in which each amplitude signal at each reception timing is converted into a single line image and then arranged, the shading of each line image forming the amplitude image changes depending on the height of the object B. As a result, in a concatenated amplitude image generated by arranging multiple amplitude images, the concatenated amplitude image changes depending on the height of the object B. The feature acquisition unit 62 outputs the concatenated amplitude image, which changes depending on the height of the object B, to the object determination unit 7 as amplitude feature information.

[0078] In step S40, the object determination unit 7 stores the amplitude feature information acquired from the feature acquisition unit 62 in the storage unit 71.

[0079] Then, in step S50, the determination unit 72 of the object determination unit 7 performs a shape determination process to determine the shape of object B based on the amplitude features stored in the storage unit 71 and a trained learning model acquired in advance by performing machine learning. The determination unit 72 inputs amplitude feature information into the learning model, thereby obtaining information on the height of object B, which is information on the shape of object B, from the learning model. The learning model is configured as a neural network, and is stored in the determination unit 72.

[0080] The learning model is generated by detecting amplitude features for each of various objects B, which are detection targets, using the ultrasonic sensor 2, and inputting amplitude features corresponding to the shape of object B detected for each shape of object B, i.e., a concatenated amplitude image corresponding to the height of object B. For example, the learning model may be generated by detecting features corresponding to the shape of object B through machine learning convolution processing that extracts feature portions in the concatenated amplitude image. When generating a learning model by performing convolution processing, for example, as shown in the boxed area in FIG. 11, portions in the concatenated amplitude image that tend to have large differences in shading, including the amplification region RA and the attenuation region RB, can be used as features.

[0081] The learning model is a trained model that has been machine-learned using multiple pieces of training data that combine the shape and amplitude features of object B. The learning model stores, as information used to determine the shape of object B, parameter information obtained from multiple linked amplitude images that are input during machine learning, linked to the shape of object B.

[0082] 12 shows an example of a simulation result of an amplitude signal based on superimposed reflected waves received by the ultrasonic sensor 2 when an object B of a different height is placed at a predetermined position from the ultrasonic sensor 2. The height of the object B used to obtain this simulation result is set to be equal to or lower than the position at which the ultrasonic sensor 2 is installed. The distance measured to obtain this simulation result is set to a distance from the object B that allows the top end reflected wave and the base reflected wave to be sufficiently detected, and is set to a distance that allows the reception of superimposed reflected waves having an area where the top end reflected wave and the base reflected wave overlap.

[0083] 12, even if the distance measured from the ultrasonic sensor 2 to the object B is constant, the waveform of the amplitude signal differs depending on the height of the object B. The reason for the different waveforms of the amplitude signals is that, although the amplitude signal is generated based on superimposed reflected waves, the region in which the amplitudes of the top-end reflected wave and the base reflected wave interfere with each other when the ultrasonic sensor 2 receives them changes depending on the height of the object B.

[0084] For example, the smaller the height of object B, i.e., the shorter the distance from the top end to the base of object B, the larger the region where the amplitudes of the top end reflected wave and the base reflected wave interfere when the ultrasonic sensor 2 receives the top end reflected wave and the base reflected wave. On the other hand, the larger the height of object B, i.e., the greater the distance from the top end to the base of object B, the larger the region where the amplitudes of the top end reflected wave and the base reflected wave interfere when the ultrasonic sensor 2 receives the top end reflected wave and the base reflected wave.

[0085] Therefore, even if the distance measured from the ultrasonic sensor 2 to the object B is constant, the waveform of the amplitude signal based on the superimposed reflected waves changes depending on the height of the object B. The change in the waveform of the amplitude signal changes the concatenated amplitude image, which is an amplitude feature. The learning model stores information about the amplitude features obtained from the amplitude signal that changes depending on the height of the object B.

[0086] In step S50, the determination unit 72 determines the shape of the object B based on the feature changes and the learning model stored in the storage unit 71. Specifically, the determination unit 72 detects the height of the object B using information on parameters calculated from the amplitude features stored in the storage unit 71 and the amplitude features stored in the learning model.

[0087] The ultrasonic sensors 2 of this embodiment, that is, the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, and the fourth front sensor 2D, are provided below the center in the vehicle height direction of the front bumper C2. The ultrasonic sensors 2 of this embodiment, that is, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H, are provided below the center in the vehicle height direction of the rear bumper C3.

[0088] That is, the transmitter 40A and receiver 40B of the transmitter / receiver 4 supported by the sensor housing of the ultrasonic sensor 2 are provided on the front bumper C2 or the rear bumper C3. The transmitter 40A and receiver 40B are attached below the center of the bumpers C2 and C3 in the vehicle height direction. By attaching the transmitter 40A and receiver 40B below the center of the bumpers C2 and C3 in the vehicle height direction in this way, it becomes easy to install the transmitter 40A and receiver 40B at a height within a range of 20 to 35 cm from the road surface R.

[0089] This makes it possible to easily position transmitter 40A and receiver 40B closer to the height of object B, even if object B is a relatively low object such as a curb or a wheel stopper. The reason for positioning transmitter 40A and receiver 40B closer to the height of object B, which is a low object, will be described with reference to FIGS. 13 and 14.

[0090] 13 shows the difference in the positional relationship of the ultrasonic sensor 2 between when the ultrasonic sensor 2 is placed at a position approximately the same as the height of the object B, as in this embodiment, and when the ultrasonic sensor 2 is installed at a higher position compared to the mounting position of this embodiment. Here, the mounting height of the ultrasonic sensor 2 in this embodiment is defined as a first height H1, and the mounting height of the ultrasonic sensor 2 when the ultrasonic sensor 2 is installed at a higher position than the ultrasonic sensor 2 in this embodiment is defined as a second height H2. The first height H1 is set to be approximately the same as the height of the object B. In contrast, the second height H2 is set to be higher than the height of the object B.

[0091] Here, when the ultrasonic sensor 2 receives a reflected wave from an object B having a height, as described above, the ultrasonic sensor 2 receives a reflected wave with a relatively strong signal strength from a portion of the object B that is relatively close to the ultrasonic sensor 2 and from a portion at the base of the object B.

[0092] When the ultrasonic sensor 2 is placed at a position equal to or lower than the height of the object B, the part relatively close to the ultrasonic sensor 2 is the front part of the object B in front of the ultrasonic sensor 2. Therefore, when the ultrasonic sensor 2 is placed at a position equal to or lower than the height of the object B, the ultrasonic sensor 2 receives a reflected wave with a strong signal strength from the front part of the object B in front of the ultrasonic sensor 2 as a reflected wave from a part relatively close to the ultrasonic sensor 2.

[0093] In contrast, when the ultrasonic sensor 2 is placed at a position higher than the height of the object B, the ultrasonic sensor 2 receives a reflected wave with a relatively strong signal strength from the upper end portion of the object B as a reflected wave from a location relatively close to the ultrasonic sensor 2. However, the signal strength of the reflected wave reflected from the upper end portion of the object B tends to be weaker than the signal strength of the reflected wave reflected from the front portion of the object B.

[0094] The further the angle of the direction from the ultrasonic sensor 2 toward the top of object B deviates from the horizontal direction, which is the direction perpendicular to the vehicle height direction, the weaker the signal strength of the reflected wave reflected at the top of object B. In other words, the greater the difference between the mounting height of ultrasonic sensor 2 and the height of object B, and the greater the angle of the transmission wave transmitted from ultrasonic sensor 2 to object B relative to the horizontal direction, the weaker the signal strength of the reflected wave from the top of object B.

[0095] Therefore, the closer the mounting height of the ultrasonic sensor 2 is to the height of the object B, the greater the S / N ratio of the amplitude signal obtained from the reflected wave. Also, the greater the difference between the mounting height of the ultrasonic sensor 2 and the height of the object B, that is, the farther the installation position of the ultrasonic sensor 2 is from the height of the object B, the smaller the S / N ratio of the amplitude signal. Therefore, the closer the mounting height of the ultrasonic sensor 2 is to the height of the object B, the more easily amplitude features appear in the amplitude signal obtained from the reflected wave.

[0096] Furthermore, when ultrasonic sensor 2 receives a reflected wave from a portion of object B that is relatively close to ultrasonic sensor 2 and a reflected wave from the base of object B, as described above, there is a time lag between the timing at which these two reflected waves are received. This time lag is caused by the difference between the distance from ultrasonic sensor 2 to the portion of object B that is relatively close to ultrasonic sensor 2 and the distance from ultrasonic sensor 2 to the base of object B. For this reason, the closer the mounting height of ultrasonic sensor 2 is to the height of object B, the smaller the time lag between the timing at which the two reflected waves are received, and the greater the mounting height of ultrasonic sensor 2 is relative to the height of object B, the greater the time lag between the timing at which the two reflected waves are received.

[0097] 14 shows the difference between the lag in reception timing of the two reflected waves when the mounting height of the ultrasonic sensor 2 is set to the first height H1 and the lag in reception timing of the two reflected waves when the mounting height of the ultrasonic sensor 2 is set to the second height H2. In Fig. 14, the time lag when the mounting height of the ultrasonic sensor 2 is set to the first height H1 is shown by a solid line, and the time lag when the mounting height of the ultrasonic sensor 2 is set to the second height H2 is shown by a dashed line.

[0098] As shown in Figure 14, regardless of the measured distance, when the mounting height of the ultrasonic sensor 2 is set to the first height H1, the time lag is smaller than when the mounting height of the ultrasonic sensor 2 is set to the second height H2.

[0099] The smaller the difference in the reception timing of the two reflected waves, the larger the overlapping area between the reflected wave from a portion of object B relatively close to ultrasonic sensor 2 and the reflected wave from the root of object B. Conversely, the larger the difference in the reception timing of the two reflected waves, the smaller the overlapping area between the reflected wave from a portion of object B relatively close to ultrasonic sensor 2 and the reflected wave from the root of object B. Therefore, the smaller the difference in the reception timing of the two reflected waves, the more likely amplitude features will appear in the amplitude signal obtained from the superimposed reflected waves having an overlapping area between the two reflected waves.

[0100] For these reasons, the closer the mounting height of the ultrasonic sensor 2 is to the height of the object B, the more likely the amplitude characteristics appear in the amplitude signal. Therefore, even if the height of the object B is relatively low, the height of the object B can be determined with high accuracy by providing the ultrasonic sensor 2 below the center of the bumpers C2 and C3 in the vehicle height direction.

[0101] As described above, the object detection device 1 of this embodiment includes a transmitter 40A that transmits ultrasonic waves toward object B, a receiver 40B that receives reflected waves from object B and outputs a received signal corresponding to the reflected waves, an amplitude signal generator 61 that generates an amplitude signal of the received signal, a feature acquirer 62 that outputs amplitude feature information, and an object determination unit 7 that determines the shape of object B based on the amplitude feature information. The ultrasonic waves transmitted by the transmitter 40A have a change region in which the frequency changes over time. The receiver 40B receives superimposed reflected waves including a region where top-end reflected waves and bottom-end reflected waves reflected from multiple locations at different heights on object B overlap, and outputs a received signal corresponding to the superimposed reflected waves. The amplitude signal generator 61 generates an amplitude signal based on the received signal output by the receiver 40B after receiving the superimposed reflected waves. The feature acquirer 62 outputs the amplitude feature information acquired based on the amplitude signal to the object determination unit 7. The object determination unit 7 determines the shape of object B based on the amplitude feature information.

[0102] As described above, when a transmitted wave having a change region in which the frequency changes over time is reflected by object B, the reflected wave also has a change region. When the transmitted wave is reflected by different parts of object B having a height, the amplitude signal corresponding to the superimposed reflected wave, which includes an area where the reflected waves reflected by different parts overlap, will have a mixture of states where the reflected waves reflected by the different parts are in phase and states where they are out of phase. The waveform of the amplitude signal based on such superimposed reflected waves will change depending on the shape of object B.

[0103] Therefore, the object detection device 1 receives a superimposed reflected wave formed when a transmitted wave having a change region is reflected by the object B, and determines the shape of the object B based on the amplitude signal based on the superimposed reflected wave, thereby enabling the object detection device 1 to accurately determine the shape of the object B.

[0104] One method for determining the height of object B is to determine that the height of object B is lower than the mounting height of ultrasonic sensor 2 if there is one peak in the amplitude signal based on the two reflected waves received by ultrasonic sensor 2, and to determine that the height of object B is higher than the mounting height of ultrasonic sensor 2 if there are two peaks in the amplitude signal. However, when using such a method to determine the height of object B, it can only be determined whether or not the height of object B is higher than the mounting height of ultrasonic sensor 2. Furthermore, the criterion for determining the height of object B is the mounting height of ultrasonic sensor 2, and the determination is dependent on the mounting height of ultrasonic sensor 2.

[0105] Furthermore, even if the height of object B is higher than the mounting height, if the distance between the ultrasonic sensor 2 and object B is relatively short, two reflected waves received by the ultrasonic sensor 2 may overlap, resulting in a single peak in the amplitude signal. In such a case, if the height of object B is determined based on the number of peaks in the amplitude signal, there is a risk that the height of object B will be determined erroneously.

[0106] In contrast, the ultrasonic sensor 2 of this embodiment determines the shape of the object B based on an amplitude signal based on the superimposed reflected waves, and therefore can determine not only whether the height of the object B is higher or lower than the mounting height of the ultrasonic sensor 2, but also the specific height of the object B, without depending on the mounting height of the ultrasonic sensor 2. Furthermore, even if the distance between the ultrasonic sensor 2 and the object B is relatively short and the two reflected waves received by the ultrasonic sensor 2 overlap, the height of the object B can be determined with high accuracy.

[0107] Furthermore, according to the above embodiment, the following effects can be obtained.

[0108] (1) In the above embodiment, a feature acquisition unit 62 is provided which detects amplitude features of the amplitude signal and outputs amplitude feature information, which is information relating to the amplitude features. The amplitude signal generation unit 61 outputs an amplitude signal based on the superimposed reflected wave to the feature acquisition unit 62. Based on the amplitude signal, the feature acquisition unit 62 generates an amplitude image, which is an image in which information indicating the magnitude of the amplitude intensity for each reception timing of the amplitude signal is converted into brightness values ​​of predetermined gradations and arranged, and outputs information relating to the generated amplitude image to the object determination unit 7 as amplitude feature information. The object determination unit 7 determines the shape of the object B based on the amplitude feature information.

[0109] This makes it easier to detect the features of the amplitude signal by generating the amplitude features of the amplitude signal as an amplitude image that indicates the magnitude of the amplitude intensity, which makes it easier for the object determination unit 7 to determine the shape of the object B based on the amplitude feature information.

[0110] (2) In the above embodiment, the feature acquisition unit 62 arranges multiple generated amplitude images in a direction perpendicular to the signal direction to generate one concatenated amplitude image, and outputs the generated concatenated amplitude image to the object determination unit 7 as amplitude feature information.

[0111] Incidentally, a single amplitude image in which information indicating the magnitude of the amplitude intensity for each reception timing of the amplitude signal is arranged can detect a change in the magnitude of the amplitude intensity for each reception timing. In other words, a single amplitude image can detect a change in the magnitude of the amplitude intensity due to a change in the reception time.

[0112] In contrast, a single concatenated amplitude image generated by arranging a plurality of amplitude images in a direction perpendicular to the signal direction can detect not only changes in amplitude intensity for each reception timing but also changes in amplitude intensity in the signal direction and in a direction intersecting the direction perpendicular to the signal direction. Therefore, by determining the shape of object B based on the concatenated amplitude image, the object determination unit 7 can more easily determine the shape of object B compared to when determining the shape of object B based on amplitude images.

[0113] (3) In the above embodiment, the object determination unit 7 stores a learning model generated by machine learning amplitude features corresponding to the shape of object B, and determines the shape of object B based on the amplitude feature information and the learning model.

[0114] This allows the height of object B to be determined with high accuracy based on a previously acquired learning model.

[0115] (4) In the above embodiment, the object detection device 1 is applied to a vehicle C. The transmitter 40A and the receiver 40B are provided on the bumpers C2 and C3. The transmitter 40A and the receiver 40B are attached to the bumpers C2 and C3 below the center of the bumpers C2 and C3 in the vehicle height direction.

[0116] As described above, the closer the mounting height of the ultrasonic sensor 2 is to the height of the object B, the higher the S / N ratio of the amplitude signal obtained from the reflected wave. Furthermore, the closer the mounting height of the ultrasonic sensor 2 is to the height of the object B, the smaller the difference in reception timing of the two reflected waves from a portion of the object B relatively close to the ultrasonic sensor 2 and from the base of the object B, and the larger the area where these two reflected waves overlap when received.

[0117] Therefore, the closer the mounting height of the ultrasonic sensor 2 is to the height of the object B, the more likely it is that amplitude characteristics will appear in the amplitude signal obtained from the reflected wave. Therefore, even if the object B is a relatively low-profile object, the height of the object B can be determined with high accuracy by providing the transmitter 40A and the receiver 40B below the center of the bumpers C2 and C3 in the vehicle height direction.

[0118] (Modification of the first embodiment) In the above-described first embodiment, the object determination unit 7 stores a learning model generated by machine learning amplitude features corresponding to the shape of object B, and an example has been described in which the shape of object B is determined based on amplitude feature information and the learning model, but this is not limiting.

[0119] For example, the object determining unit 7 may be configured not to store a learning model, but to determine the shape of the object B based only on the amplitude feature information.

[0120] (Second embodiment) Next, the second embodiment will be described with reference to FIGS. 15 and 16. This embodiment differs from the first embodiment in that the received signal processing unit 6 includes a distance calculation unit 63. Also, this embodiment differs from the first embodiment in the content of the object detection processing of the object detection device 1. Other than this, this embodiment is similar to the first embodiment. Therefore, in this embodiment, differences from the first embodiment will be mainly described, and descriptions of similar parts to the first embodiment may be omitted.

[0121] 15, the received signal processing unit 6 of this embodiment includes a distance calculation unit 63. The amplitude signal generation unit 61 of this embodiment outputs an amplitude signal generated based on the received signal output by the receiving circuit 43 to the feature acquisition unit 62, and also outputs the amplitude signal to the distance calculation unit 63.

[0122] The distance calculation unit 63 calculates the distance from the ultrasonic sensor 2 to the object B from the amplitude signal output by the amplitude signal generation unit 61, and outputs information related to this calculated distance. Specifically, the distance calculation unit 63 calculates a measured distance, which is the distance between the transducer 41 included in the transmission unit 40A and the reception unit 40B of the ultrasonic sensor 2 and the object B to be detected, and outputs measured distance information, which is information related to the calculated measured distance, to the object determination unit 7. The distance calculation unit 63 calculates the measured distance using the TOF method, for example, based on the time elapsed from when the transducer 41 transmits a transmitted wave until when it receives a reflected wave reflected by the object B, and the speed of sound. TOF is an abbreviation for Time of Flight.

[0123] When the transmitter 40A repeatedly transmits a transmission wave, the distance calculation unit 63 repeatedly calculates the measured distance at the timing when the transducer 41 receives the reflected wave based on the amplitude signal each time the transducer 41 receives the reflected wave reflected by the object B. The distance calculation unit 63 then outputs measured distance information including the repeatedly calculated measured distance to the object determination unit 7.

[0124] Then, the object determination unit 7 determines the shape of the object B to be detected based on the amplitude feature information output from the feature acquisition unit 62 and the distance measurement information output from the distance calculation unit 63.

[0125] The object detection process of this embodiment executed by the object determination unit 7 will be described with reference to Fig. 16. Note that the processes of steps S10, S20, and S30 shown in Fig. 16 are similar to the processes of the respective steps described with reference to Fig. 10 of the first embodiment, and therefore description thereof will be omitted.

[0126] After the amplitude signal generated by the received signal processing unit 6 is output to the distance calculation unit 63 in step S20, the distance calculation unit 63 executes the process of step S25. Specifically, in step S25, the distance calculation unit 63 calculates the measured distance from the amplitude signal output by the amplitude signal generation unit 61, and outputs the measured distance information to the object determination unit 7. The distance calculation unit 63 calculates the measured distance based on the speed of sound and the time elapsed from when the transducer 41 transmits an ultrasonic wave until when it receives a reflected wave of the transmitted wave.

[0127] Then, in step S30, the feature acquisition unit 62 generates a connected amplitude image from the amplitude signals output by the amplitude signal generation unit 61, and outputs the generated connected amplitude image to the object determination unit 7 as amplitude feature information.

[0128] In step S42, the object determination unit 7 associates the amplitude feature information acquired from the feature acquisition unit 62 and the ranging distance information acquired from the distance calculation unit 63, and stores the associated information in the storage unit 71. That is, the storage unit 71 stores a combined amplitude image, which is an amplitude feature corresponding to the ranging distance.

[0129] Then, in step S52, the determination unit 72 of the object determination unit 7 performs a shape determination process to determine the shape of object B based on the amplitude features stored in the storage unit 71, the measured distances linked to these amplitude features, and a trained learning model acquired in advance by performing machine learning. The determination unit 72 inputs the amplitude features and the measured distances into the learning model, thereby obtaining information on the height of object B, which is information on the shape of object B, from the learning model. The learning model is configured as a neural network and is stored in the determination unit 72.

[0130] The learning model of this embodiment is generated by detecting, for each object B, a concatenated amplitude image that changes according to the ranging distance using the ultrasonic sensor 2, and inputting, for each ranging distance, the detected ranging distance and the concatenated amplitude image in association with each other. The learning model is also generated by inputting, for each object B, the ranging distance detected while varying the ranging distance to various magnitudes, and the concatenated amplitude image in association with each other. In other words, the learning model of this embodiment is a trained model that has been machine-learned using multiple pieces of training data that combine the shape of object B and concatenated amplitude images corresponding to the ranging distance. The learning model stores, as information used to determine the shape of object B, information on parameters obtained from the multiple ranging distances and concatenated amplitude images input during machine learning, linked to the shape of object B.

[0131] In step S52, the determination unit 72 determines the shape of the object B based on the measured distances stored in the storage unit 71, the concatenated amplitude images linked to the measured distances, and the learning model. Specifically, the determination unit 72 detects the height of the object B using the measured distances stored in the storage unit 71, the concatenated amplitude images, and information on parameters calculated from the measured distances and amplitude features stored in the learning model.

[0132] As described above, the object detection device 1 of this embodiment includes a distance calculation unit 63 that calculates the measured distance, which is the distance between the receiving unit 40B and the object B, based on the amplitude signal and outputs measured distance information, which is information related to the measured distance. The object determination unit 7 determines the shape of the object B based on the amplitude feature information and the measured distance information.

[0133] As described above, the waveform of the amplitude signal based on the superimposed reflected waves changes depending on the shape and ranging distance of object B. The change in the waveform of the amplitude signal changes the amplitude characteristics. Therefore, object detection device 1 can accurately determine the shape of object B by determining the shape of object B based on the ranging distance and amplitude characteristics of the amplitude signal based on the superimposed reflected waves.

[0134] Furthermore, in this embodiment, the object determination unit 7 stores a trained model that has been machine-learned using a plurality of pieces of training data that combine the shape of object B and a concatenated amplitude image according to the distance measurement. The object determination unit 7 then determines the shape of object B based on the distance measurement information, the concatenated amplitude image, and the trained model. This allows the height of object B to be determined with high accuracy based on the trained model acquired in advance.

[0135] (Third embodiment) Next, a third embodiment will be described with reference to FIGS. 17 to 19. In this embodiment, the object determination unit 7 includes an identity determination unit 73, and the processing content of the object detection process of the object detection device 1 differs from that of the second embodiment. Other than this, the third embodiment is the same as the second embodiment. Therefore, in this embodiment, the differences from the second embodiment will be mainly described, and a description of the same parts as the second embodiment may be omitted.

[0136] 17, the object determination unit 7 of this embodiment includes an identity determination unit 73. The identity determination unit 73 determines whether or not the multiple superimposed reflected waves received by the transducer 41 are superimposed reflected waves from the same object B. The identity determination unit 73 determines whether or not the multiple superimposed reflected waves received by the transducer 41 are superimposed reflected waves from the same object B.

[0137] In this embodiment, when the transducer 41 receives superimposed reflected waves multiple times, the object determination unit 7 determines whether the received superimposed reflected waves are superimposed reflected waves from the same object B, and then determines the height of the shape of object B based on multiple amplitude feature information and multiple ranging distance information.

[0138] The object detection process of this embodiment executed by the object determination unit 7 will be described with reference to Fig. 18. Note that the processes of steps S10, S20, S25, S30, and S42 shown in Fig. 18 are similar to the processes of the respective steps described with reference to Fig. 16 of the second embodiment, and therefore description thereof will be omitted.

[0139] In step S25, the distance calculation unit 63 calculates the measured distance from the amplitude signal output by the amplitude signal generation unit 61, and outputs the measured distance information to the object determination unit .

[0140] When the transducer 41 receives a superimposed reflected wave including a top-end reflected wave and a root reflected wave, the distance calculation unit 63 calculates the distance from the ultrasonic sensor 2 to the top of the object B based on the top-end reflected wave. Furthermore, when the transducer 41 receives a superimposed reflected wave including a top-end reflected wave and a root reflected wave, the distance calculation unit 63 calculates the distance from the ultrasonic sensor 2 to the root of the object B based on the root reflected wave.

[0141] As described above, when the ultrasonic sensor 2 receives two reflected waves, a top-end reflected wave and a root reflected wave, it receives the top-end reflected wave earlier than the root reflected wave. Therefore, when two reflected waves are received, the distance calculation unit 63 calculates the distance calculated based on the reflected wave received first as the distance from the ultrasonic sensor 2 to the top of object B. Then, the distance calculation unit 63 calculates the distance calculated based on the reflected wave received later as the distance from the ultrasonic sensor 2 to the root of object B. Hereinafter, the distance calculated based on the reflected wave received earlier is referred to as the first distance, and the distance calculated based on the reflected wave received later is referred to as the second distance. When the transducer 41 receives superimposed reflected waves including two reflected waves received at different times, the distance calculation unit 63 calculates the first distance and the second distance in step S25.

[0142] In step S30, the feature acquisition unit 62 generates a connected amplitude image from the amplitude signals output by the amplitude signal generation unit 61, and outputs the generated connected amplitude image to the object determination unit 7 as amplitude feature information.

[0143] In step S42, the object determination unit 7 associates the amplitude feature information acquired from the feature acquisition unit 62 and the ranging distance information acquired from the distance calculation unit 63 and stores them in the storage unit 71.

[0144] In step S44, the object determination unit 7 determines whether the quantity of amplitude feature information and ranging distance information stored in the storage unit 71 is equal to or greater than a predetermined determination count. The determination count is set to determine whether the quantity of amplitude feature information, i.e., the quantity of connected amplitude images, stored in the storage unit 71 is equal to or greater than the quantity necessary to execute the shape determination process in step S52, which will be described later, and is set to, for example, four times. The determination count can be determined, for example, by an experiment that investigates the quantity of amplitude feature information necessary to accurately determine the shape of object B.

[0145] If it is not determined in step S44 that the quantity of amplitude feature information and ranging distance information stored in storage unit 71 is equal to or greater than the determination count, the processes of steps S10 to S42 are executed again. That is, the processes of steps S10 to S42 are repeatedly executed until the quantity of acquired amplitude feature information and ranging distance information is equal to or greater than the determination count.

[0146] Each time the feature acquisition unit 62 receives a repetitive amplitude signal from the amplitude signal generation unit 61, it repeatedly generates a concatenated amplitude image based on each amplitude signal output by the amplitude signal generation unit 61, and repeatedly outputs the concatenated amplitude image to the object determination unit 7 as amplitude feature information. Furthermore, each time the distance calculation unit 63 receives a repetitive amplitude signal from the amplitude signal generation unit 61, it repeatedly calculates the measured distance at the timing when the transducer 41 receives the superimposed reflected wave, and repeatedly outputs the measured distance information to the object determination unit 7. This measured distance information includes information on the first distance and information on the second distance.

[0147] When object detection processing is performed while the vehicle is traveling so that the vehicle approaches object B, the measured distance decreases over time. Therefore, the object determination unit 7 stores the amplitude feature that changes in accordance with the change in the measured distance in the storage unit 71 until the number of acquired amplitude feature information and measured distance information becomes equal to or greater than the number of determinations.

[0148] If it is determined in step S44 that the number of amplitude feature information and ranging distance information stored in the storage unit 71 is equal to or greater than the predetermined number of determinations, the process proceeds to step S46. In step S46, the identity determination unit 73 determines whether the superimposed reflected waves received by the transducer 41 the number of determinations or more are superimposed reflected waves from the same object B.

[0149] Here, the reason and method for determining whether or not the object B that reflected all of the multiple superimposed reflected waves is the same when the transducer 41 receives multiple superimposed reflected waves will be described with reference to Fig. 19. Fig. 19 shows the change in the positional relationship between the ultrasonic sensor 2 and the object B as the host vehicle travels toward the object B. The upper side of Fig. 19 shows a side view of the ultrasonic sensor 2 and the object B as seen in the width direction of the host vehicle, i.e., from one side in the left-right direction, and the lower side of Fig. 19 shows a top view of the ultrasonic sensor 2 and the object B as seen from the higher side in the vehicle height direction of the host vehicle, i.e., from the upper side in the up-down direction.

[0150] 19, when the height of object B and the installation position of ultrasonic sensor 2 in the height direction are relatively close, the transmitted wave reflected by object B tends to be strongly reflected from the position in front of ultrasonic sensor 2 on object B and the position at the base of object B. For this reason, transducer 41 receives two reflected waves: one reflected from the front of ultrasonic sensor 2 on object B and the other reflected from the base of object B.

[0151] Therefore, when the transducer 41 receives two reflected waves, the object detection device 1 can calculate the measured distance to object B based on these two reflected waves. For example, suppose that the transducer 41 receives a reflected wave reflected from the front of the ultrasonic sensor 2 and then a reflected wave reflected from the base of object B. In this case, the distance from the ultrasonic sensor 2 to the base of object B can be calculated based on the elapsed time until the later received reflected wave reflected from the base of object B is received and the speed of sound. Then, the measured distance to object B can be geometrically calculated based on the calculated distance from the ultrasonic sensor 2 to the base of object B and the mounting height of the ultrasonic sensor 2.

[0152] However, due to external factors, etc., the transducer 41 may not be able to receive one of the two reflected waves, the reflected wave reflected from the front of the ultrasonic sensor 2 on the object B and the reflected wave reflected from the base of the object B. In such a case, as shown in Fig. 19, assume that a non-target object B2, which is an object B that is not to be detected, exists around a target object B1, which is an object B that is to be detected. In this case, the transducer 41 may receive the reflected wave from the non-target object B2, which may cause the object detection device 1 to erroneously determine the distance to the target object B1.

[0153] For example, suppose that transducer 41 receives a reflected wave reflected by the front surface of ultrasonic sensor 2 on non-target object B2 and then receives a reflected wave reflected by the front surface of ultrasonic sensor 2 on target object B1. In this case, the distance from ultrasonic sensor 2 to the front surface of target object B1, which is calculated based on the elapsed time until the later received reflected wave reflected by the front surface of ultrasonic sensor 2 on target object B1, is erroneously calculated as the distance from ultrasonic sensor 2 to the base of target object B1. Therefore, if the measured distance is calculated based on the erroneously calculated distance from ultrasonic sensor 2 to the base of target object B1 and the mounting height of ultrasonic sensor 2, the calculated measured distance will be incorrect.

[0154] If the object detection device 1 determines the shape of object B through object detection processing based on such an erroneously determined distance measurement, there is a risk that the object detection device 1 will erroneously determine the shape of object B. For this reason, it is desirable for the object detection device 1 to accurately determine the distance measurement when determining the shape of object B based on the distance measurement. To achieve this, when the ultrasonic sensor 2 receives multiple superimposed reflected waves, it is necessary to determine whether the object B that reflected all of these multiple superimposed reflected waves is the same.

[0155] Therefore, the object determination unit 7 of the object detection device 1 of this embodiment determines, based on the acquired multiple pieces of ranging distance information, whether or not the superimposed reflected waves received by the transducer 41 the number of times equal to or greater than the determination number are superimposed reflected waves from the same target object B1. Specifically, the identity determination unit 73 of the object determination unit 7 calculates the amount of change between the first distance information and the second distance information included in the multiple pieces of amplitude feature information stored in the storage unit 71. Then, the identity determination unit 73 determines, based on the amount of change between the two distances, whether or not the superimposed reflected waves received by the transducer 41 the number of times equal to or greater than the determination number are superimposed reflected waves from the same target object B1.

[0156] As described above, the first distance is the distance calculated based on the reflected wave received first when the transducer 41 receives two reflected waves. In contrast, the second distance is the distance calculated based on the reflected wave received last when the transducer 41 receives two reflected waves. As shown in Fig. 19, when the host vehicle travels closer to the target object B1 and the measured distance between the ultrasonic sensor 2 and the target object B1 decreases, the values ​​of the multiple pieces of first distance information and the multiple pieces of second distance information acquired by the object determination unit 7 decrease over time.

[0157] Furthermore, when the measured distance decreases over time as the host vehicle approaches the target object B1, there is a correlation between the amount of change in the distance from the ultrasonic sensor 2 to the top of the target object B1, the amount of change in the distance from the ultrasonic sensor 2 to the base of the target object B1, and the travel distance of the host vehicle. Therefore, when the host vehicle approaches the target object B1, there is a correlation between the amount of decrease in the multiple first distances stored in the memory unit 71, the amount of decrease in the multiple second distances, and the travel distance of the host vehicle. In other words, when the transducer 41 receives multiple superimposed reflected waves from the same target object B1, there is a correlation between the amount of decrease in the multiple first distances calculated when the multiple superimposed reflected waves are received, the amount of decrease in the multiple second distances, and the travel distance of the host vehicle.

[0158] In contrast, when transducer 41 receives superimposed reflected waves from target object B1 and non-target object B2, there is no such correlation between the calculated decrease amounts of the first distances, the calculated decrease amounts of the second distances, and the traveling distance of the host vehicle. For example, the first distance information is assumed to be information included in the ranging distance information calculated when transducer 41 receives superimposed reflected waves reflected from target object B1. Furthermore, the second distance information is assumed to be information included in the ranging distance information calculated when transducer 41 receives superimposed reflected waves reflected from non-target object B2.

[0159] In this case, the correlation between the decrease in the first distance, the decrease in the multiple second distances, and the traveling distance of the host vehicle is different from the correlation between the decrease in the multiple first distances, the decrease in the multiple second distances, and the traveling distance of the host vehicle when the transducer 41 receives superimposed reflected waves multiple times from the same target object B1. In other words, the correlation between the decrease in the multiple first distances, the decrease in the multiple second distances, and the traveling distance of the host vehicle is different when the transducer 41 receives superimposed reflected waves multiple times from the same target object B1 and when the transducer 41 receives superimposed reflected waves multiple times from different objects B1.

[0160] Therefore, the identity determination unit 73 can determine, based on the correlation between the multiple first distance change amounts, the multiple second distance change amounts, and the traveling distance of the host vehicle, whether or not the superimposed reflected waves received by the transducer 41 the number of times equal to or greater than the determination number are superimposed reflected waves from the same object B. The traveling distance of the host vehicle can be calculated, for example, based on the vehicle speed detected by a vehicle speed sensor provided on the host vehicle.

[0161] In step S46, if the identity determination unit 73 does not determine that the superimposed reflected waves received by the transducer 41 the number of times equal to or greater than the determination number are superimposed reflected waves from the same object B, the processes of steps S10 to S44 are executed again. On the other hand, in step S46, if the identity determination unit 73 determines that the superimposed reflected waves received by the transducer 41 the number of times equal to or greater than the determination number are superimposed reflected waves from the same object B, the process of step S52 is executed.

[0162] Then, in step S52, the determination unit 72 of the object determination unit 7 performs a shape determination process to determine the shape of object B based on the amplitude features stored in the storage unit 71, the distance measurements associated with these amplitude features, and a trained learning model acquired in advance by performing machine learning. The determination unit 72 inputs the amplitude features of the number of determinations and the distance measurements for the number of determinations into the learning model, thereby obtaining information on the height of object B, which is information on the shape of object B, from the learning model.

[0163] Specifically, the determination unit 72 extracts one of the multiple connected amplitude images associated with multiple measurement distances stored in the storage unit 71. Then, the determination unit 72 detects the height of the object B using the extracted connected amplitude image, the measurement distance associated with the extracted connected amplitude image, and information on parameters calculated from the measurement distance and amplitude features stored in the learning model.

[0164] For example, suppose the number of determinations is set to four, and information on four concatenated amplitude images linked to the measured distances is stored in the storage unit 71. In this case, the determination unit 72 extracts one concatenated amplitude image from the four concatenated amplitude images linked to the four measured distances stored in the storage unit 71. The determination unit 72 then detects the height of object B using the extracted concatenated amplitude image, the measured distances linked to the extracted concatenated amplitude image, and information on parameters calculated from the measured distances and amplitude features stored in the learning model. The determination unit 72 detects the height of object B for each measured distance based on the four measured distances and concatenated amplitude images stored in the storage unit 71.

[0165] Here, when detecting the height of object B based on each of the four concatenated amplitude images stored in the storage unit 71 and the learning model, if the heights of the four object B detected based on the ranging distances do not match, the determination unit 72 sets a weighting according to the ranging distance. Specifically, if the heights of object B detected based on the four concatenated amplitude images do not match, the determination unit 72 assigns a lower priority to the value of the height of object B detected from the concatenated amplitude image, the greater the ranging distance associated with the concatenated amplitude image used to detect the height of object B.

[0166] In contrast, the determination unit 72 assigns a higher priority to the value of the height of object B detected from the concatenated amplitude image, the smaller the distance measurement associated with the concatenated amplitude image used to detect the height of object B. In other words, the determination unit 72 assigns a higher weight to the value of the height of object B detected from the concatenated amplitude image, the smaller the distance measurement associated with the concatenated amplitude image used to detect the height of object B, and determines the shape of object B.

[0167] The following explains why it is determined that the longer the ranging distance, the lower the priority of the height value of object B detected from the concatenated amplitude image, and the shorter the ranging distance, the higher the priority of the height value of object B detected from the concatenated amplitude image.

[0168] The amplitude strength of the amplitude signal generated by the ultrasonic sensor 2 receiving the superimposed reflected waves from the object B varies depending on the distance measured. Specifically, the longer the distance measured, the smaller the amplitude strength of the amplitude signal, and the shorter the distance measured, the larger the amplitude strength of the amplitude signal. Therefore, the longer the distance measured, the smaller the SN ratio of the amplitude signal, and the shorter the distance measured, the larger the SN ratio of the amplitude signal.

[0169] Furthermore, the amplitude signal generated when the ultrasonic sensor 2 receives the superimposed reflected waves is generated based on the reflected waves that include the area where the top-end reflected wave and the root reflected wave overlap, and the area where the top-end reflected wave and the root reflected wave overlap varies depending on the ranging distance. Specifically, the longer the ranging distance, the smaller the time difference between the top-end reflected wave and the root reflected wave, so similar amplitude signals overlap. On the other hand, the shorter the ranging distance, the larger the time difference between the top-end reflected wave and the root reflected wave, so different amplitude signals overlap. This is because the difference between the distance from the ultrasonic sensor 2 to the top of object B and the distance from the ultrasonic sensor 2 to the root of object B becomes smaller as the ranging distance increases.

[0170] For this reason, the greater the ranging distance, the less likely differences in shading will appear in the concatenated amplitude image based on the amplitude signal, and the shorter the ranging distance, the more likely differences in shading will appear in the concatenated amplitude image based on the amplitude signal. Therefore, the determination unit 72 assigns a lower priority to the detected height of object B the greater the ranging distance associated with the concatenated amplitude image used to detect the height of object B, and assigns a higher priority to the detected height value of object B the shorter the ranging distance. As a result, even if there is a discrepancy in the heights of object B detected based on multiple concatenated amplitude images associated with the ranging distances, the height of object B determined with a higher priority can be used to accurately determine the height of object B.

[0171] As described above, in this embodiment, the object determination unit 7 determines whether the multiple superimposed reflected waves received by the receiving unit 40B are superimposed reflected waves from the same object B, based on the multiple pieces of ranging distance information and the multiple pieces of amplitude feature information that are based on the multiple amplitude signals generated by the receiving unit 40B receiving the superimposed reflected waves multiple times. If the object determination unit 7 determines that the multiple superimposed reflected waves are from the same object B, it determines the shape of the object B based on the multiple pieces of amplitude feature information and the multiple pieces of ranging distance information.

[0172] As described above, if, due to an external factor or the like, the object detection device 1 is unable to receive some of the reflected waves from parts of object B at different heights and receives reflected waves from object B that is not the detection target, there is a risk that the object detection device 1 will erroneously determine the measured distance to object B that is the detection target. If the shape of object B is determined based on the erroneous measured distance, there is a risk that the object detection device 1 will erroneously determine the shape of object B.

[0173] In contrast, by determining whether the multiple superimposed reflected waves received by transducer 41 of receiving unit 40B are superimposed reflected waves from the same object B, it is possible to avoid erroneous calculation of the distance measurement based on reflected waves from object B, which is not the detection target. Therefore, object detection device 1 can accurately determine the shape of object B.

[0174] Furthermore, by determining the shape of object B based on multiple amplitude feature information and multiple ranging distance information, the object detection device 1 can determine the shape of object B with greater accuracy than when determining the shape of object B based on a single amplitude feature information and ranging distance information.

[0175] (First modified example of the third embodiment) In the above-described second embodiment, an example was described in which, in step S52, the judgment unit 72 detects the height of object B for each ranging distance based on the amplitude feature of the number of judgments linked to the ranging distance of the number of judgments stored in the memory unit 71 and the learning model, but this is not limited to this.

[0176] For example, the determination unit 72 may be configured to detect the height of a single object B based on the amplitude features of the determination counts associated with the respective ranging distances of the determination counts stored in the storage unit 71 and a learning model. In this case, the determination unit 72 may detect the height of object B without setting a weight according to the ranging distance for the height of object B detected from the amplitude features of the determination counts associated with the respective ranging distances of the determination counts.

[0177] (Second modified example of the third embodiment) In the above-described third embodiment, an example has been described in which, in step S52, if the heights of object B detected based on the distance measurement do not match for the number of determinations, the determination unit 72 sets a weighting according to the distance measurement. Specifically, an example has been described in which the determination unit 72 assigns a lower priority to the value of the height of object B detected from the amplitude feature as the distance measurement distance increases, and assigns a higher priority to the value of the height of object B detected from the amplitude feature as the distance measurement distance decreases. However, the method of assigning a weighting according to the distance measurement is not limited to this.

[0178] For example, the determination unit 72 may be configured to set different weights for the heights of object B detected at the respective detection determination times, regardless of whether the heights of object B for the detection determination times match, and detect the height of object B. In this case, the determination unit 72 may be configured to detect the height of object B by a weighted average (i.e., a weighted average) in which the greater the measured distance, the greater the weight assigned to the height of object B detected from the amplitude feature associated with this measured distance, and the smaller the measured distance, the less weight assigned to the height of object B detected from the amplitude feature associated with this measured distance.

[0179] (Third Modification of the Third Embodiment) In the above-described third embodiment, in step S52, if the heights of object B detected based on the distance measurement for the number of determinations do not match, the determination unit 72 sets a weight according to the distance measurement and detects the height of object B, but this is not limited to this.

[0180] For example, when the heights of object B detected based on the distance measurement and the number of times of determination do not match, the determination unit 72 may be configured to detect the height of object B by performing a majority decision without setting a weight according to the distance measurement. Specifically, when any of the heights of object B detected based on the distance measurement and the number of times of determination do not match, the height of object B with the greatest number of matches among the heights of object B detected based on the distance measurement and the number of times of determination may be used as the height of object B.

[0181] (Fourth embodiment) Next, a fourth embodiment will be described with reference to Figs. 20 to 24. In this embodiment, the connected amplitude image that the feature acquisition unit 62 outputs to the object determination unit 7 as amplitude feature information is different from that in the first embodiment. Other than this, the fourth embodiment is the same as the first embodiment. Therefore, in this embodiment, differences from the first embodiment will be mainly described, and descriptions of similar parts to the first embodiment may be omitted.

[0182] In the first embodiment, an example was described in which the feature acquisition unit 62 generates an amplitude image based on an amplitude signal in which maximum and minimum amplitude values ​​alternate, and then arranges a plurality of the generated amplitude images in a direction perpendicular to the signal direction to generate a concatenated amplitude image. In contrast, the feature acquisition unit 62 of this embodiment finds a peak line PL shown in FIG. 20 by tracing the maximum values ​​in the amplitude signal in which maximum and minimum amplitude values ​​alternate. Then, the feature acquisition unit 62 generates an amplitude image as shown in FIG. 21 based on the found peak line PL. The peak line PL is a so-called envelope that indicates the characteristics of the amplitude change of the amplitude signal.

[0183] Specifically, the feature acquisition unit 62 converts the magnitude of the amplitude intensity at each reception time on the peak line PL of the amplitude signal into a straight line image of predetermined gradation, and generates an amplitude image by arranging the straight line images converted for each reception time in the order of the received amplitude signals. That is, the feature acquisition unit 62 digitizes the magnitude of the amplitude intensity in the envelope of the waveform of the amplitude signal using numbers of predetermined stages, and generates an amplitude image by arranging straight line images of densities corresponding to the calculated numbers in the order of the received amplitude signals.

[0184] In this embodiment, the feature acquisition unit 62 generates an image in which the shading of the lines is lighter in areas where the amplitude intensity in the envelope of the waveform of the amplitude signal is greater, and darker in areas where the amplitude intensity is smaller. The amplitude image generated in this way includes ranging distance information.

[0185] When the subject vehicle is traveling and the measured distance changes over time, the feature acquisition unit 62 repeatedly obtains an amplitude image based on each amplitude signal output by the amplitude signal generation unit 61 each time the transducer 41 repeatedly receives a reflected wave reflected by the object B. In other words, the amplitude signal generation unit 61 repeatedly outputs an amplitude signal to the feature acquisition unit 62 based on the received signal output by the receiving unit 40B each time the receiving unit 40B receives a superimposed reflected wave.

[0186] The feature acquisition unit 62 arranges the generated amplitude images in a direction perpendicular to the signal direction in accordance with the order in which the superimposed reflected waves were received by the receiving unit 40B to generate one concatenated amplitude image as shown in Fig. 22. Then, the feature acquisition unit 62 outputs the generated concatenated amplitude image to the object determination unit 7 as an amplitude feature.

[0187] The concatenated amplitude image thus generated includes information on the measured distance and the magnitude of the amplitude intensity of the amplitude signal, as well as information on the change in the magnitude of the amplitude intensity according to the measured distance that changes over time. That is, the concatenated amplitude image includes information on the change in the magnitude of the amplitude intensity according to the measured distance when the measured distance changes over time.

[0188] The concatenated amplitude image shown in Figure 22 is an example in which 10 amplitude images are arranged side by side and generated based on the amplitude signal output by the amplitude signal generating unit 61 when the subject vehicle travels toward object B and the ranging distance gradually decreases over time.

[0189] Furthermore, the storage unit 71 of the object determination unit 7 of this embodiment stores a learning model in which changing measured distances are linked to changes in the magnitude of amplitude intensity corresponding to these changing measured distances. The learning model stored in the storage unit 71 is generated by inputting, for each of various objects B, information on concatenated amplitude images that change over time, linked to the measured distances.

[0190] The learning model may be generated by performing a machine learning convolution process to extract a feature portion from the concatenated amplitude image, as in the first embodiment. When generating the learning model by performing a convolution process, the feature portion may be, for example, a portion of the concatenated amplitude image that is likely to have a large difference in shading, including the amplification region RA and the attenuation region RB, as shown in the boxed portion in Fig. 23 .

[0191] As described above, the waveform of the amplitude signal differs depending on the height of object B. Therefore, as shown in Fig. 24, the waveform of the peak line PL changes depending on the measured distance from the ultrasonic sensor 2 to object B. The learning model stores, in association with the height of object B, information used to determine the shape of object B, such as parameter information obtained from amplitude features obtained from the amplitude signal that changes depending on the height of object B.

[0192] The object detection process performed by the object detection device 1 of this embodiment is similar to the object detection process performed by the object detection device 1 of the first embodiment described with reference to Fig. 10. Therefore, a description of the object detection process performed by the object detection device 1 of this embodiment will be omitted.

[0193] As described above, the amplitude signal generation unit 61 of this embodiment repeatedly outputs amplitude signals to the feature acquisition unit 62 based on the received signals output by the receiving unit 40B each time the receiving unit 40B receives a superimposed reflected wave. The feature acquisition unit 62 generates an amplitude image for each amplitude signal repeatedly acquired from the amplitude signal generation unit 61, and generates one concatenated amplitude image by arranging multiple amplitude images in a direction perpendicular to the signal direction in correspondence with the order in which the receiving unit 40B received the superimposed reflected waves. The feature acquisition unit 62 then outputs the generated concatenated amplitude image as amplitude feature information.

[0194] According to this, the combined amplitude image includes information on the measured distance and the magnitude of the amplitude intensity of the amplitude signal, as well as information on the change in the magnitude of the amplitude intensity according to the measured distance that changes over time. Therefore, if the combined amplitude image generated changes each time the receiving unit 40B repeatedly receives the superimposed reflected waves, the object determination unit 7 can determine the shape of the object B more easily by determining the shape of the object B based on the combined amplitude image.

[0195] (Modification of the fourth embodiment) In the above-described fourth embodiment, an example has been described in which the received signal processing unit 6 does not include the distance calculation unit 63, as in the first embodiment, but the present invention is not limited to this.

[0196] For example, the received signal processing unit 6 may be configured to include a distance calculation unit 63, as in the second and third embodiments. In this case, the object detection process performed by the object detection device 1 of this embodiment can determine the height of the object B by performing the same object detection process as the object detection process performed by the object detection device 1 of the second embodiment described with reference to Fig. 16. Alternatively, the object detection process performed by the object detection device 1 of this embodiment can determine the height of the object B by performing the same object detection process as the object detection process performed by the object detection device 1 of the third embodiment described with reference to Fig. 18.

[0197] (Fifth embodiment) Next, the fourth embodiment will be described. In this embodiment, the amplitude feature information output by the feature acquisition unit 62 to the object determination unit 7 is different from that of the second to fourth embodiments. Other than this, the fourth embodiment is the same as the second, third, or fourth embodiment. Therefore, in this embodiment, differences from the second to fourth embodiments will be mainly described, and descriptions of similar parts to the second to fourth embodiments may be omitted.

[0198] In the second and third embodiments, the feature acquisition unit 62 generates an amplitude image based on an amplitude signal in which maximum and minimum amplitude values ​​alternate, and then arranges the generated amplitude images to generate a concatenated amplitude image as an amplitude feature. In contrast, the feature acquisition unit 62 of this embodiment, like the fourth embodiment, determines a peak line PL shown in FIG. 20 from an amplitude signal in which maximum and minimum amplitude values ​​alternate. The feature acquisition unit 62 then detects waveform features that indicate characteristic portions of the amplitude waveform of the amplitude signal as amplitude features, and outputs the detected waveform features to the object determination unit 7 as amplitude feature information. For example, the feature acquisition unit 62 acquires, as the waveform feature, at least one of the amplitude waveform, waveform area, waveform slope, number of maximum and minimum waveform points, and waveform peak ratio of the amplitude signal.

[0199] The amplitude waveform is the waveform of an amplitude signal that corresponds to the signal strength of the superimposed reflected wave. The waveform area is the area of ​​the portion surrounded by the peak line PL and is calculated by integrating the peak line PL. The waveform slope is the slope of the curve indicated by the peak line PL and is calculated by differentiating the peak line PL. The number of maximum and minimum points of the waveform is the number of vertices in the curve indicated by the peak line PL that are greater than a predetermined threshold or smaller than a predetermined threshold. The waveform peak ratio is the ratio of the number of maximum and minimum points on the peak line PL.

[0200] As described above, the amplitude signal when the superimposed reflected wave is received changes depending on the height of object B, i.e., the shape of object B. For this reason, at least one of the shape of the amplitude waveform, the area of ​​the amplitude waveform, the slope of the amplitude waveform, the number of maximum and minimum points of the amplitude waveform, the peak ratio of the amplitude waveform, etc. of the amplitude signal changes depending on the height of object B. For this reason, the feature acquisition unit 62 detects the shape of the amplitude waveform, the area of ​​the amplitude waveform, the slope of the amplitude waveform, the number of maximum and minimum points of the amplitude waveform, the peak ratio of the amplitude waveform, etc. of the amplitude signal that change depending on the height of object B, and outputs the detected information to the object determination unit 7 as amplitude feature information.

[0201] Like the object detection device 1 of the second and third embodiments, the object determination unit 7 of this embodiment includes a distance calculation unit 63. The amplitude signal generation unit 61 of this embodiment outputs an amplitude signal generated based on the reception signal output by the reception circuit 43 to the feature acquisition unit 62, and also outputs the amplitude signal to the distance calculation unit 63.

[0202] Each time the transducer 41 repeatedly receives the reflected wave reflected by the object B, the distance calculation unit 63 repeatedly calculates the measured distance at the timing when the transducer 41 receives the reflected wave. Then, the distance calculation unit 63 outputs measured distance information including the repeatedly calculated measured distance to the object determination unit 7.

[0203] Furthermore, the storage unit 71 of the object determination unit 7 of this embodiment stores a learning model in which, for each object B, the measured distance and the amplitude waveform corresponding to this measured distance are linked and input. The learning model stored in the storage unit 71 is generated by inputting, for each of various objects B, information on the amplitude waveform that changes over time, linked to the measured distance. That is, the learning model is a trained model that has been machine-learned using multiple pieces of training data that combine the shape of the object B and feature changes that indicate waveform features according to the measured distance. The learning model stores at least one waveform feature used to determine the shape of the object B, such as the shape of the amplitude waveform in the amplitude signal, the area of ​​the amplitude waveform, the slope of the amplitude waveform, the number of maximum and minimum points in the amplitude waveform, and the peak ratio of the amplitude waveform, linked to the shape of the object B and the measured distance.

[0204] The object detection process performed by the object detection device 1 of this embodiment is similar to the object detection process performed by the object detection device 1 of the second embodiment described with reference to Fig. 16. Therefore, a description of the object detection process performed by the object detection device 1 of this embodiment will be omitted.

[0205] As described above, the feature acquisition unit 62 of this embodiment detects waveform features that indicate characteristic portions of the amplitude waveform in the amplitude signal as amplitude features, and outputs the detected waveform features as amplitude feature information to the object determination unit 7. For example, the feature acquisition unit 62 acquires, as the waveform feature, at least one of the shape of the amplitude waveform in the amplitude signal, the area of ​​the amplitude waveform, the slope of the amplitude waveform, the number of maximum and minimum points in the amplitude waveform, and the peak ratio of the amplitude waveform. Then, the object determination unit 7 determines the shape of object B based on the waveform features and ranging distance information, which are amplitude feature information.

[0206] As described above, the amplitude waveform of the amplitude signal based on the superimposed reflected waves changes depending on the shape and measuring distance of object B. The change in the amplitude waveform also changes the waveform characteristics. Therefore, the object detection device 1 can accurately determine the shape of object B by determining the shape of object B based on the measuring distance and waveform characteristics of the amplitude signal based on the superimposed reflected waves.

[0207] Furthermore, in this embodiment, the object determination unit 7 stores a trained model that has been machine-learned using multiple pieces of teaching data that combine the shape of object B and amplitude waveforms according to the measured distance. The object determination unit 7 then determines the shape of object B based on the measured distance information, waveform features, and the trained model. This allows the height of object B to be determined with high accuracy based on the trained model acquired in advance.

[0208] (Modification of the fifth embodiment) In the above-described fourth embodiment, an example has been described in which the received signal processing unit 6 includes the distance calculation unit 63, but the present invention is not limited to this.

[0209] For example, like the first embodiment, the received signal processing unit 6 may be configured not to include the distance calculation unit 63. In this case, the object detection process executed by the object detection device 1 of this embodiment can determine the height of the object B by executing the same object detection process as the object detection process executed by the object detection device 1 of the first embodiment described with reference to FIG.

[0210] (Sixth embodiment) Next, a sixth embodiment will be described with reference to Fig. 25. In this embodiment, the shape determination process for object B differs from that in the fifth embodiment. The rest of the process is the same as in the fifth embodiment. Therefore, in this embodiment, the differences from the fifth embodiment will be mainly described, and a description of the same parts as in the fifth embodiment may be omitted.

[0211] The object determination unit 7 of this embodiment stores in the storage unit 71 a feature map indicating the correspondence between the measurement distance and the amplitude feature for each object B. The feature map stores amplitude features for each measurement distance for various objects B that are detection targets. The feature map stores information such as the amplitude waveform of the amplitude signal, the waveform area, the waveform slope, the number of maximum and minimum points of the waveform, and the waveform peak ratio as amplitude features, linked to the height and measurement distance, which are the shape of the object B.

[0212] Here, for example, the feature map stores information such as the shape of the amplitude waveform, the area of ​​the amplitude waveform, the slope of the amplitude waveform, the number of maximum and minimum points of the amplitude waveform, and the peak ratio of the amplitude waveform, which are amplitude features, sorted into four categories, as shown in Fig. 25. In this sorting, the various pieces of information on the amplitude features are sorted into feature 1 and feature 2 according to the priority when determining the height of object B.

[0213] Specifically, information on similar amplitude features among the shape of the amplitude waveform, the area of ​​the amplitude waveform, the slope of the amplitude waveform, the number of maximum and minimum points of the amplitude waveform, and the peak ratio of the amplitude waveform are classified into Feature 1 and Feature 2. For example, information such as the shape of the amplitude waveform, the area of ​​the amplitude waveform, and the slope of the amplitude waveform are classified into Feature 1, and information such as the number of maximum and minimum points of the amplitude waveform and the peak ratio of the amplitude waveform are classified into Feature 2.

[0214] Furthermore, the various pieces of information classified into feature 1 and feature 2 are classified according to the priority for determining the height of object B. For example, when determining the height of object B, if information on the shape of the amplitude waveform is given priority over information on the area of ​​the amplitude waveform, the information on the shape of the amplitude waveform is set to category 1, and the information on the area of ​​the amplitude waveform is set to category 2. Furthermore, if information on the quantity of maximum and minimum points of the amplitude waveform is given priority over information on the peak ratio of the amplitude waveform, the information on the quantity of maximum and minimum points of the amplitude waveform is set to category 4, and the information on the peak ratio of the amplitude waveform is set to category 3.

[0215] The object determination unit 7 determines the height of the object B based on such a feature map stored in the storage unit 71.

[0216] Next, a description will be given of the object detection process executed by the object detection device 1 of this embodiment. The object detection process executed by the object detection device 1 of this embodiment differs from the object detection processes executed by the object detection devices 1 of the second and third embodiments in the content of the shape determination process in step S52.

[0217] In step S52, the determination unit 72 of this embodiment determines the height of object B based on the amplitude feature information stored in the storage unit 71, the ranging distance information linked to this amplitude feature information, and a feature map previously stored in the storage unit 71. Specifically, the determination unit 72 extracts ranging distances and amplitude features corresponding to the ranging distances and amplitude features stored in the storage unit 71 in step S52 from the ranging distances and amplitude features stored in the feature map. Then, the determination unit 72 determines the height of object B based on the information of object B corresponding to the extracted ranging distances and amplitude features.

[0218] When the determination unit 72 extracts ranging distance and amplitude features corresponding to the ranging distance and amplitude features stored in the memory unit 71 from the ranging distance and amplitude features stored in the feature map, the determination unit 72 may determine the height of the object B based on the priority set in the feature map.

[0219] For example, when determining the height of object B based on feature 1 of the feature map, if the height of object B determined from the shape of the amplitude waveform differs from the height of object B determined from the area of ​​the amplitude waveform, the height of object B may be determined based on the shape of the amplitude waveform with high priority classified as category 1. Furthermore, when determining the height of object B based on feature 2 of the feature map, if the height of object B determined from the number of maximum and minimum points in the amplitude waveform differs from the height of object B determined from the peak ratio of the amplitude waveform, the height of object B may be determined based on the number of maximum and minimum points in the amplitude waveform with high priority classified as category 4.

[0220] As described above, the object determination unit 7 of this embodiment stores a feature map indicating the correspondence between the measured distance and the amplitude feature for each object B. Then, the object determination unit 7 determines the height of the object B based on the amplitude feature information, the measured distance information, and the feature map. The other configurations are the same as those of the fifth embodiment. The object detection device 1 of this embodiment can obtain the same effects as those of the fifth embodiment, which are achieved by a configuration similar to or equivalent to that of the fifth embodiment.

[0221] Therefore, the object detection device 1 receives a superimposed reflected wave formed when a transmitted wave having a change region is reflected by the object B, and determines the shape of the object B based on the ranging distance and amplitude characteristics of the amplitude signal based on the superimposed reflected wave, thereby being able to accurately determine the shape of the object B.

[0222] (Other embodiments) Representative embodiments of the present disclosure have been described above, but the present disclosure is not limited to the above-described embodiments and can be modified in various ways, for example, as follows.

[0223] In the above-described embodiments, the received signal processing unit 6 includes at least a feature acquisition unit 62, and in the second, third, fifth, and sixth embodiments, the received signal processing unit 6 includes a distance calculation unit 63, but this is not limiting.

[0224] For example, the received signal processing unit 6 may be configured not to include the feature acquisition unit 62 and the distance calculation unit 63. In this case, the object determination unit 7 may be configured to receive the amplitude signal directly from the amplitude signal generation unit 61 and determine the shape of the object B to be detected based on the amplitude signal received from the amplitude signal generation unit 61.

[0225] In addition, the object determination unit 7 may be configured to store a learning model generated by inputting an amplitude signal corresponding to the ranging distance in the memory unit 71, and to determine the shape of object B based on the amplitude signal and the learning model.

[0226] In the above embodiment, an example has been described in which the transmitter 40A transmits a down-chirp signal in which the frequency of the change region decreases over time, but the present invention is not limited to this.

[0227] For example, the transmitting section 40A may be configured to transmit an up-chirp signal in which the frequency of the change region increases over time.

[0228] In the above-described embodiment, an example was described in which the ultrasonic sensors 2, namely the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, the fourth front sensor 2D, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H, all transmit the same transmission wave, but this is not limited to this.

[0229] For example, any one or more of the ultrasonic sensors 2, i.e., the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, the fourth front sensor 2D, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H, may be configured to transmit transmission waves with different frequencies in the change range. Also, the ultrasonic sensors 2, i.e., the first front sensor 2A, the second front sensor 2B, the third front sensor 2C, the fourth front sensor 2D, the first rear sensor 2E, the second rear sensor 2F, the third rear sensor 2G, and the fourth rear sensor 2H, may all be configured to initiate transmission waves at the same timing, or may be configured to transmit transmission waves at different timings.

[0230] In the above-described embodiment, an example was described in which the transmitter 40A and the receiver 40B are provided on the bumpers C2 and C3, and the mounting positions of the transmitter 40A and the receiver 40B are set below the center of the bumpers C2 and C3 in the vehicle height direction, but this is not limited to this.

[0231] For example, the transmitter 40A and the receiver 40B may be provided at a location other than the bumpers C2 and C3 of the vehicle C. Furthermore, the attachment positions of the transmitter 40A and the receiver 40B may be set above the center of the bumpers C2 and C3 in the vehicle height direction.

[0232] In the above embodiment, an example has been described in which the object detection device 1 of the present disclosure is applied to the vehicle C, but the object detection device 1 can also be applied to devices other than vehicles.

[0233] In the above embodiment, the transducer 41 has a function as both a transmitter and a receiver, and the ultrasonic sensor 2 has an integrated transmission and reception configuration, but this is not limiting. For example, the transducer 41 constituting the transmission unit 40A and the transducer 41 constituting the reception unit 40B may be provided as separate bodies.

[0234] In the above-described first to third embodiments, examples have been described in which the feature acquisition unit 62 acquires amplitude feature information based on an amplitude signal in which maximum and minimum amplitude intensity values ​​alternately repeat. Furthermore, in the above-described fourth to sixth embodiments, examples have been described in which the feature acquisition unit 62 acquires amplitude feature information based on a peak line PL of the amplitude signal. However, the method by which the feature acquisition unit 62 of each embodiment acquires amplitude feature information is not limited to these.

[0235] For example, in the first to third embodiments, the feature acquisition unit 62 may be configured to acquire amplitude feature information based on the peak line PL. Also, in the fourth to sixth embodiments, the feature acquisition unit 62 may be configured to acquire amplitude feature information based on an amplitude signal in which maximum and minimum values ​​of amplitude intensity alternately repeat. Alternatively, in the first to sixth embodiments, the feature acquisition unit 62 may be configured to acquire amplitude feature information based on a pulse-compressed amplitude signal.

[0236] In the above-described embodiments, it goes without saying that the elements constituting the embodiments are not necessarily essential unless they are specifically stated as essential or are clearly considered essential in principle.

[0237] In the above-described embodiments, when numerical values ​​such as the number, values, amounts, ranges, etc. of components of the embodiments are mentioned, they are not limited to the specific numbers unless they are specifically stated as essential or are clearly limited to a specific number in principle.

[0238] In the above-described embodiments, when referring to the shapes, positional relationships, etc. of components, etc., the shapes, positional relationships, etc. are not limited to those unless otherwise specified or when they are fundamentally limited to specific shapes, positional relationships, etc.

[0239] The control unit 3 and the method of the present disclosure may be implemented on a special-purpose computer by configuring a processor and memory programmed to perform one or more functions embodied in a computer program. The control unit 3 and the method of the present disclosure may be implemented on a special-purpose computer by configuring a processor with one or more dedicated hardware logic circuits. The control unit 3 and the method of the present disclosure may be implemented on one or more special-purpose computers configured with a combination of a processor and memory programmed to perform one or more functions and a processor configured with one or more hardware logic circuits. The computer program may also be stored on a computer-readable non-transitory tangible storage medium as instructions executed by a computer. [Explanation of symbols]

[0240] 7 Object determination section 40A Transmitter 40B Receiver 61 Amplitude signal generation section

Claims

1. An object detection device that detects an object by transmitting and receiving ultrasonic waves, a transmitting unit (40A) that transmits the ultrasonic waves toward the object; a receiving section (40B) that receives a reflected wave from the object and outputs a received signal corresponding to the reflected wave; an amplitude signal generator (61) for generating an amplitude signal of the received signal; an object determination unit (7) that determines the shape of the object based on the amplitude signal, the ultrasonic waves transmitted by the transmitting unit have a frequency change region in which the frequency changes over time, the receiving unit receives a superimposed reflected wave including an area where a plurality of the reflected waves reflected from a plurality of portions of the object at different heights overlap, and outputs the received signal corresponding to the superimposed reflected wave; the amplitude signal generating unit generates the amplitude signal based on the received signal output by the receiving unit after receiving the superimposed reflected wave; The object determination unit determines the shape of the object based on the amplitude signal generated by the amplitude signal generation unit based on the superimposed reflected waves.

2. a feature acquisition unit (62) that detects amplitude features of the amplitude signal and outputs amplitude feature information that is information about the amplitude features; the amplitude signal generation unit outputs the amplitude signal based on the superimposed reflected wave to the feature acquisition unit; the feature acquisition unit generates an amplitude image, which is an image in which information indicating the magnitude of amplitude intensity for each reception timing of the amplitude signal is converted into brightness values ​​of predetermined gradations and arranged based on the amplitude signal, and outputs information about the generated amplitude image to the object determination unit as the amplitude feature information; The object detection device according to claim 1 , wherein the object determination unit determines the shape of the object based on the amplitude feature information.

3. 3. The object detection device according to claim 2, wherein, when a direction in which information indicating the magnitude of the amplitude intensity for each reception timing in the amplitude image is arranged is defined as a signal direction, the feature acquisition unit arranges multiple generated amplitude images in a direction perpendicular to the signal direction to generate one concatenated amplitude image, and outputs the generated concatenated amplitude image as the amplitude feature information.

4. the receiving unit outputs the received signal for each of the superimposed reflected waves that are repeatedly received; the amplitude signal generation unit repeatedly outputs the amplitude signal to the feature acquisition unit based on the received signal output by the receiving unit each time the receiving unit receives the superimposed reflected wave; 3. The object detection device according to claim 2, wherein the feature acquisition unit generates the amplitude image for each of the amplitude signals repeatedly acquired from the amplitude signal generation unit, and, when a direction in which information indicating the magnitude of amplitude intensity for each reception timing in the amplitude image is arranged is defined as a signal direction, the feature acquisition unit arranges the amplitude images in a direction perpendicular to the signal direction in correspondence with the order in which the receiving unit received the superimposed reflected waves to generate one concatenated amplitude image, and outputs the generated concatenated amplitude image as the amplitude feature information.

5. a distance calculation unit (63) that calculates a distance measurement between the receiving unit and the object based on the amplitude signal and outputs distance measurement information that is information about the distance measurement, 5. The object detection device according to claim 2, wherein the object determination unit determines the shape of the object based on the amplitude feature information and the distance measurement information.

6. The object detection device according to any one of claims 2 to 4, wherein the object determination unit stores a learning model generated by machine learning the amplitude features according to the shape of the object, and determines the shape of the object based on the amplitude feature information and the learning model.

7. a feature acquisition unit (62) that detects amplitude features of the amplitude signal and outputs amplitude feature information that is information about the amplitude features; a distance calculation unit (63) that calculates a distance measurement between the receiving unit and the object based on the amplitude signal and outputs distance measurement information that is information about the distance measurement, the amplitude signal generation unit outputs the amplitude signal based on the superimposed reflected wave to the feature acquisition unit; the feature acquisition unit outputs waveform features indicating characteristic portions of an amplitude waveform of the amplitude signal as the amplitude feature information to the object determination unit based on the amplitude signal; The object detection device according to claim 1 , wherein the object determination unit determines the shape of the object based on the amplitude feature information and the distance measurement information.

8. The object detection device of claim 7, wherein the object determination unit stores a learning model generated by machine learning feature changes that indicate the waveform features according to the ranging distance, extracts the feature changes based on a plurality of pieces of amplitude feature information and a plurality of pieces of ranging distance information, and determines the shape of the object based on the extracted feature changes and the learning model.

9. The object detection device according to claim 7, wherein the object determination unit stores a feature map indicating a correspondence between the measured distance and the waveform feature for each object, and determines the shape of the object based on the amplitude feature information, the measured distance information, and the feature map.

10. The object detection device is applied to a vehicle, the transmitting unit and the receiving unit are provided in a bumper, 2. The object detection device according to claim 1, wherein the transmitter and receiver are attached to the bumper below the center of the bumper in the vehicle height direction.

Images