Object recognition device, object recognition method, and object recognition program

Abstract

To provide an object recognition device, an object recognition method, and an object recognition program capable of estimating that there is a possibility that an object has gotten under a vehicle by using an external sensor for detecting the object around the vehicle.SOLUTION: The object recognition device 100 includes the object recognition unit 11 that recognizes whether an object is present in the detection range of the external camera 2 based on the detection result of the external camera 2 that detects an object around the vehicle, and the estimation unit 12 that estimates creeping of an object under the vehicle when the recognized object deviates from the detection range of the external camera 2 toward the lower side of the vehicle.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an object recognition device, an object recognition method, and an object recognition program. [Background technology]

[0002] Patent Document 1 discloses an object detection device that uses ultrasonic waves that travel vertically downward from a transmitter / receiver to detect the intrusion of an object between the transmitter / receiver and the road surface as an abnormality. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-122196 Summary of the Invention [Problem to be solved by the invention]

[0004] The above-mentioned device uses a transmitter / receiver that emits ultrasonic waves that travel vertically downward. However, vehicles equipped with such sensors that detect objects directly below the vehicle are not as common as vehicles equipped with external sensors that detect objects around the vehicle. [Means for solving the problem]

[0005] An object recognition device according to one aspect of the present disclosure includes an object recognition unit that recognizes objects present within the detection range of an external sensor based on the detection results of the external sensor that detects objects around the vehicle, and an estimation unit that estimates that the object will move under the vehicle when the recognized object deviates from the detection range of the external sensor and moves toward the bottom of the vehicle.

[0006] According to an object recognition device according to an aspect of the present disclosure, when an object recognized within the detection range of an external sensor deviates from the detection range of the external sensor so as to move toward below the vehicle, it is estimated that the object has slipped under the vehicle. This makes it possible to estimate that the object may have slipped under the vehicle by using the external sensor that detects objects around the vehicle.

[0007] In one embodiment, the estimation unit may estimate that the object has ceased to be under the vehicle when the object that was estimated to have been under the vehicle is detected again within the detection range of the external sensor. In this case, it is possible to estimate that the risk of the object having entered under the vehicle has been eliminated by using the external sensor that detects objects around the vehicle.

[0008] In one embodiment, the estimation unit may not estimate that the object has disappeared under the vehicle if the object estimated to have disappeared under the vehicle has not been detected again within the detection range of the external sensor for a predetermined time period for determining whether the object has disappeared under the vehicle. In this case, the influence of, for example, erroneous detection by the external sensor is reduced, thereby making it possible to more accurately estimate that the object may have disappeared under the vehicle.

[0009] An object recognition method according to another aspect of the present disclosure is an object recognition method for an object recognition device that recognizes objects around a vehicle, and includes an object recognition step in which the object recognition device recognizes an object that exists within the detection range of an external sensor based on the detection result of the external sensor that detects the object, and an estimation step in which the object recognition device estimates that the object will move under the vehicle if the recognized object deviates from the detection range of the external sensor toward under the vehicle.

[0010] According to an object recognition method according to another aspect of the present disclosure, when an object recognized within the detection range of an external sensor deviates from the detection range of the external sensor so as to move toward below the vehicle, it is estimated that the object has slipped under the vehicle. This makes it possible to estimate that the object may have slipped under the vehicle by using the external sensor that detects objects around the vehicle.

[0011] Yet another aspect of the present disclosure is an object recognition program that causes a computer to function as an object recognition device that recognizes objects based on the detection results of an external sensor that detects objects around a vehicle, and the object recognition program causes the computer to function as an object recognition unit that recognizes objects present within the detection range of the external sensor based on the detection results of the external sensor, and an estimation unit that estimates that the object will move under the vehicle when the recognized object deviates from the detection range of the external sensor toward the bottom of the vehicle.

[0012] According to an object recognition program according to yet another aspect of the present disclosure, when an object recognized within the detection range of an external sensor deviates from the detection range of the external sensor so as to move toward below the vehicle, it is estimated that the object has slipped under the vehicle. This makes it possible to estimate that the object may have slipped under the vehicle by using the external sensor that detects objects around the vehicle. [Effects of the Invention]

[0013] According to some aspects of the present disclosure, it is possible to estimate whether an object may be under the vehicle by using an external sensor that detects objects around the vehicle. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a block diagram illustrating an object recognition device according to an embodiment. [Figure 2] FIG. 10 is a diagram for explaining an example of estimation of underwater movement. [Figure 3] FIG. 10 is a diagram for explaining another example of estimation of underwater movement. [Figure 4] FIG. 10 is a diagram for explaining an example of eliminating the underwater estimation. [Figure 5] FIG. 10 is a diagram for explaining an example of maintaining an estimation of underway. [Figure 6] FIG. 10 is a diagram for explaining another example of maintaining the estimation of underway. [Figure 7]10 is a flowchart illustrating an example of a process for estimating submersion. [Figure 8] 10 is a flowchart illustrating an example of a process for estimating the elimination of underpassing. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0016] Fig. 1 is a block diagram showing an object recognition device according to an embodiment. The object recognition device 100 shown in Fig. 1 is a device that uses an external camera 2 (external sensor) mounted on a vehicle such as a passenger car or freight vehicle to estimate whether an object may be slipping under the vehicle. The object recognition device 100 includes an ECU (Electronic Control Unit) 10, the external camera 2, a vehicle speed sensor 3, an ignition switch (IG switch) 4, an HMI (Human Machine Interface) 5, and a communication unit 6. The object recognition device 100 can also be applied to vehicles capable of executing driving assistance control or autonomous driving control.

[0017] [Configuration of object recognition device] As shown in FIG. 1 , the object recognition device 100 includes an ECU 10 that performs overall management of the device. The ECU 10 is an electronic control unit having a central processing unit (CPU) and storage units such as read-only memory (ROM), random access memory (RAM), and electrically erasable programmable read-only memory (EEPROM). The ECU 10 realizes various functions by, for example, executing programs stored in the storage units in the CPU. The ECU 10 may be composed of multiple electronic units. The ECU 10 may be configured as a unit integrated with the external camera 2, or may be configured as a unit separate from the external camera 2.

[0018] The ECU 10 is connected to an external camera 2, a vehicle speed sensor 3, an ignition switch 4, an HMI 5, and a communication unit 6. The external camera 2 is a camera for capturing images of the surroundings of the vehicle 1. The external camera 2 includes a front camera that captures images in front of the vehicle 1. The front camera may be installed inside the windshield of the vehicle 1. The external camera 2 may include a rear camera that captures images behind the vehicle 1, or may include side cameras (left side camera and right side camera) that capture images on the sides of the vehicle 1. The external camera 2 may be a panoramic view camera or the like installed on the grill, door mirror, or bumper of the vehicle 1. The external camera 2 has a distance measurement function. The external camera 2 may be a camera with a distance measurement function, such as a stereo camera. The external camera 2 may be a monocular camera that can measure the distance to an object in a captured image by performing predetermined image processing on the captured image. The external camera 2 transmits captured image information to the ECU 10.

[0019] The external camera 2 has a detection range 2R. The detection range 2R is the range within which the external camera 2 can capture an image of an object. For example, FIG. 2 shows a vehicle 1, the external camera 2, the detection range 2R (two-dot chain line) of the external camera 2, and an object 20. In the example of FIG. 2, the detection range 2R of the external camera 2 acting as a front camera is a sector-shaped space sandwiched between a pair of two-dot chain lines extending above and below the external camera 2 within a predetermined range in a side view of the vehicle 1. In this detection range 2R, a blind spot exists below the lower two-dot chain line in a side view of the vehicle 1. The external camera 2 acting as a front camera cannot detect an object present under the floor of the vehicle 1 or an object whose entirety is located below the lower two-dot chain line in a side view of the vehicle 1. Therefore, if an object that was within the detection range 2R moves downward from the detection range 2R of the external camera 2 toward the bottom of the vehicle 1, the object will move downward and no longer be included in the image captured by the external camera 2, resulting in a lost state.

[0020] The vehicle speed sensor 3 is a detector that detects the vehicle speed of the vehicle 1. As the vehicle speed sensor 3, for example, a wheel speed sensor that is provided on a wheel of the vehicle 1 or a drive shaft that rotates integrally with the wheel and detects the rotation speed of the wheel is used. The vehicle speed sensor transmits the detected vehicle speed information (wheel speed information) to the ECU 10.

[0021] The ignition switch 4 is a switch that functions to switch between energization and non-energization of the main power supply of the ECU 10. The ignition switch 4 has, for example, an ON state and an OFF state. When the ignition switch 4 is in the ON state, it energizes the main power supply to the ECU 10. When the ignition switch 4 is in the OFF state, it may intermittently energize the main power supply to the external camera 2 and the ECU 10 to operate the object recognition device 100. When the ignition switch 4 is in the OFF state, the ECU 10 may store the estimation results of the object recognition device 100, which will be described later, in a storage unit such as an EEPROM of the ECU 10.

[0022] The HMI 5 is an interface for inputting and outputting information between the object recognition device 100 and an occupant (user) of the vehicle 1. The HMI 5 has, for example, a display and a speaker. The HMI 5 outputs images on the display and / or sounds from the speaker in response to a control signal from the ECU 10. The display may be a MID (Multi Information Display) or a HUD (Head Up Display). The HMI 5 may have an indicator lamp that lights up or flashes when an object is estimated to be slipping under the vehicle 1, as described below.

[0023] The communication unit 6 is a communication device that controls wireless communication with the outside of the vehicle 1. The communication unit 6 communicates various types of information with, for example, a communication terminal such as a smartphone carried by a user who has exited the vehicle 1 via a communication network. The communication unit 6 is not particularly limited, and various known communication devices can be used.

[0024] Next, a description will be given of the functional configuration of the ECU 10. The ECU 10 has an object recognition unit 11, an estimation unit 12, and an attention calling unit 13.

[0025] The object recognition unit 11 recognizes objects present within the detection range 2R of the external camera 2 based on the detection results of the external camera 2 that detects objects around the vehicle 1. The object recognition unit 11 recognizes objects present within the detection range 2R of the external camera 2 based on the image captured by the external camera 2 using known image recognition processing. The objects recognized by the object recognition unit 11 may be, for example, other vehicles, pedestrians, bicycles, small animals, etc. The object recognition unit 11 may also recognize the relative position, relative speed, and movement direction of the object with respect to the vehicle 1.

[0026] When an object recognized by the object recognition unit 11 deviates from the detection range 2R of the external camera 2 so as to move downwards toward the vehicle 1, the estimation unit 12 estimates that the object has slipped under the vehicle 1. "Estimation of an object slipping under the vehicle" means that there is a possibility that the object has slipped under the vehicle. "There is a possibility that the object has slipped under the vehicle" means that there is a possibility that the object is lost and exists in the blind spot of the external camera 2 below the vehicle, such as under the floor of the vehicle or around the vehicle. The blind spot is a range outside the detection range 2R of the external camera 2. In the blind spot, it is difficult for the occupants of the vehicle 1 to see the object. The object here refers to an object that may be entirely hidden in the blind spot of the external camera 2 below the vehicle, and may be, for example, a pedestrian such as a child or a small animal.

[0027] "Departing from the detection range 2R of the external camera 2 as if moving downward from the vehicle 1" may mean "the object departs from the detection range 2R of the external camera 2 as it approaches the vehicle 1" or "the object departs from the detection range 2R of the external camera 2 as it moves away from the vehicle 1."

[0028] As an example, the estimation unit 12 calculates the relative distance from the vehicle 1 of an object present within the detection range 2R of the external camera 2 based on the image captured by the external camera 2. If the relative distance of the object within the detection range 2R of the external camera 2 decreases and the object is no longer recognized by the object recognition unit 11 through the lower boundary of the detection range 2R, the estimation unit 12 may determine that the detected object has deviated from the detection range 2R of the external camera 2 and is heading downwards toward the vehicle 1. If the relative distance of the object within the detection range 2R of the external camera 2 increases and the object is no longer recognized by the object recognition unit 11 through the lower boundary of the detection range 2R, the estimation unit 12 may determine that the detected object has deviated from the detection range 2R of the external camera 2 and is heading downwards toward the vehicle 1. For example, if the relative distance of an object does not change within the detection range 2R of the external camera 2, or if the object recognition unit 11 continues to recognize an object within the detection range 2R of the external camera 2, the estimation unit 12 may determine that the recognized object has not deviated from the detection range 2R of the external camera 2 so as to move downward toward the vehicle 1.

[0029] FIG. 2 is a diagram illustrating an example of intrusion estimation. In the situation shown in FIG. 2, when an object 20 that was within the detection range 2R of the external camera 2 mounted on the vehicle 1 approaches the vehicle 1 along the road surface, if the object 20 disappears from the detection range 2R, the object 20 moves downward in the image captured by the external camera 2 and is no longer included in the captured image, resulting in a lost state. This state corresponds to a state in which the object 20 may have intruded below the vehicle 1. The object 20 may be, for example, a small animal. In this way, the estimation unit 12 can infer that the object 20 has intruded below the vehicle 1 when the object 20 recognized by the object recognition unit 11 approaches the vehicle 1 and deviates from the detection range 2R of the external camera 2 so as to move downward toward the vehicle 1.

[0030] FIG. 3 is a diagram illustrating another example of the estimation of underpassing. In the situation shown in FIG. 3 , when an object 21, a portion of which was within the detection range 2R of the external camera 2 mounted on the vehicle 1, moves away from the vehicle 1 along the road surface, the height of the object 21 decreases, causing the portion of the object 21 that was within the detection range 2R of the external camera 2 to no longer be within the detection range 2R. This results in the detection by the external camera 2 being lost downward. This state corresponds to a state in which the object 21 may have underpassed the vehicle 1. The object 21 may be, for example, a pedestrian such as a child. The reduction in the height of the object 21 may correspond, for example, to a standing pedestrian bending down. In this way, even for an object 21 moving away from the vehicle 1, the estimation unit 12 can estimate that the object 21 has underpassed the vehicle 1 if the object 21 recognized by the object recognition unit 11 deviates from the detection range 2R of the external camera 2 so as to move toward the vehicle 1.

[0031] The estimation unit 12 may estimate that the object has ceased to be under the vehicle 1 when the object estimated to have been under the vehicle 1 is redetected within the detection range 2R of the external camera 2. FIG. 4 is a diagram illustrating an example of the elimination of the estimation of the object being under the vehicle 1. In the situation illustrated in FIG. 4, if the object 22 that was under the vehicle 1 moves forward along the road surface and enters the detection range 2R, the object 22 moves from the bottom edge of the captured image of the external camera 2 so as to be included in the captured image. This state corresponds to a state in which the object 22 has been redetected by the external camera 2 and is no longer in a lost state, and the possibility that the object 22 has been under the vehicle 1 has been eliminated. The object 22 may be, for example, a small animal, similar to the object 20. In this way, the estimation unit 12 can estimate that the object 22 has ceased to be under the vehicle 1 when the object estimated to have been under the vehicle 1 is redetected within the detection range 2R of the external camera 2.

[0032] Here, when estimating the elimination of the obscuration of the object, there is a possibility that the external camera 2 may detect the object erroneously, such as due to, for example, fluctuations in the detection of the object by the external camera 2, or optical ghosts or reflections. In such a case, for example, the state in which the object is re-detected within the detection range 2R of the external camera 2 of the vehicle 1 may continue for only a very short time, and immediately thereafter, the object may deviate from the detection range 2R of the external camera 2. Therefore, if the state in which the object estimated to have obscured is re-detected within the detection range 2R of the external camera 2 does not continue for a predetermined elimination determination time, the estimation unit 12 may not estimate the elimination of the obscuration of the object (it may maintain the estimation of the obscuration of the object).

[0033] The resolution determination time is a threshold time for determining whether it is acceptable to estimate that the object has disappeared. The resolution determination time corresponds to the duration of redetection of the object that can be considered not to be a false detection by the external sensor. The resolution determination time may be, for example, about 0.5 seconds or 1 second. Alternatively, the resolution determination time may be a predetermined coefficient multiplied by the time required to estimate the object's disappearance. The required time may be the time from the time the external camera 2 begins to detect the object before it is lost to the time the object is lost and its disappearance is estimated. The predetermined coefficient may be, for example, 1 / 2 or 1 / 3.

[0034] The resolution determination time is used, for example, as a threshold for the time elapsed from a predetermined measurement start point. The timer measurement start point may be when an object re-detected within the detection range 2R of the external camera 2 falls within a predetermined measurement start distance relative to the vehicle 1 before the object becomes lost.

[0035] The measurement start distance is the value of the relative distance from the vehicle 1 at which the timer starts measuring. The measurement start distance is not particularly limited, but may be a distance corresponding to the lower limit of the detection range 2R of the external camera 2. The measurement start distance may be a distance from the vehicle 1 corresponding to the height relative to the mounting height of the external camera 2, such as the position of the lower two-dot chain line in Figure 2, for example.

[0036] FIG. 5 is a diagram illustrating an example of maintaining the estimation of underpassing. In the situation shown in FIG. 5, an object 23 that had deviated from the detection range 2R of the external camera 2 temporarily becomes partially present within the detection range 2R as it approaches the vehicle 1 along the road surface, but immediately thereafter deviates from the detection range 2R of the external camera 2. Here, the time during which the object 23 temporarily remains partially present within the detection range 2R is shorter than the resolution determination time. In this case, the temporary redetection of the object 23 may correspond to a detection fluctuation, such as when the object 23 is actually temporarily redetected due to a change in its height. In this way, if the state in which the object 23, which was estimated to have underpassed, is redetected within the detection range 2R of the external camera 2, does not continue for the predetermined resolution determination time, the estimation unit 12 does not need to estimate the elimination of the underpassing of the object 23 (it may maintain the estimation of the underpassing of the object 23).

[0037] FIG. 6 is a diagram illustrating another example of maintaining the estimation of underpassing. In the situation shown in FIG. 6, an object 24 that had deviated from the detection range 2R of the external camera 2 is not moving relative to the vehicle 1, and a pseudo object 25 is temporarily detected as if it were within the detection range 2R. However, immediately thereafter, the pseudo object 25 is no longer detected. The pseudo object 25 is not an actual object, but an optical ghost or reflection image that is detected within the detection range 2R as if it were an object. Here, the time during which the pseudo object 25 is temporarily detected within the detection range 2R is shorter than the resolution determination time. In this case, although the object 24 and the pseudo object 25 are not the same object, the temporary detection of the pseudo object 25 can be considered similar to the detection fluctuation described above. Therefore, the temporary detection of the pseudo object 25 is treated as a pseudo equivalent to a temporary re-detection of the object 24. In this way, the estimation unit 12 does not need to estimate that the object 24 has ceased to be trapped under the object 24 (it may maintain the estimation of the object 24 being trapped under the object 24) if the state in which the pseudo object 25 is detected within the detection range 2R of the external camera 2, instead of the object 24 that was estimated to have been trapped under the object 24, does not continue for a predetermined time period for determining whether the pseudo object 25 has been detected within the detection range 2R of the external camera 2.

[0038] The attention-calling unit 13 may issue an attention warning regarding an object getting under the vehicle 1 when it is estimated that an object is getting under the vehicle 1. The attention-calling unit 13 may notify an occupant (user) of the vehicle 1 via the HMI 5 that an object may be getting under the vehicle 1.

[0039] For example, when the ignition switch 4 is in the ON state, the warning unit 13 may alert the occupant (user) of the vehicle 1 that an object may have slipped under the vehicle 1 by outputting an image on the display, outputting audio from the speaker, or lighting or flashing an indicator lamp.

[0040] For example, when the switch state of the ignition switch 4 is in the OFF state, the warning unit 13 may suspend the notification by the HMI 5, and when the switch state of the ignition switch 4 changes to the ON state, the warning unit 13 may notify the occupant (user) of the vehicle 1 by the HMI 5 that an object may have slipped under the vehicle 1 based on the estimated result of the object slipping under stored in a memory unit such as an EEPROM of the ECU 10.

[0041] For example, when the ignition switch 4 is in the OFF state, the warning unit 13 may alert the occupant (user) of the vehicle 1 that an object may have slipped under the vehicle 1 by communicating various information via a communication network via the communication unit 6 to a communication terminal such as a smartphone held by a user who has disembarked from the vehicle 1.

[0042] [Processing of object recognition device, object recognition method, and object recognition program] Next, an example of the processing of the object recognition device 100 will be described with reference to the flowcharts of Fig. 7 and Fig. 8. Fig. 7 is a flowchart showing an example of the process of estimating underpassing. The process (steps) shown in Fig. 7 is repeatedly executed at predetermined calculation intervals, for example, when the vehicle 1 is stopped and underpassing of an object is not estimated (for example, when the underpassing flag is set to 0). The switch state of the ignition switch 4 may be either an ON state or an OFF state.

[0043] 7 may be repeatedly executed at predetermined calculation intervals when the ignition switch 4 is in the ON state and the vehicle speed of the vehicle 1 is less than a vehicle stop determination threshold based on the detection result of the vehicle speed sensor 3. The vehicle stop determination threshold is a vehicle speed threshold for determining that the vehicle 1 is stopped. The vehicle stop determination threshold is not particularly limited, and may be set to 0.5 km / h, 1 km / h, or 5 km / h.

[0044] 7, in S11, the ECU 10 of the object recognition device 100 causes the object recognition unit 11 to recognize objects present within the detection range 2R of the external sensor (object recognition step). For example, in FIGS. 2 and 3, the object recognition unit 11 recognizes objects 20 and 21 present within the detection range 2R of the external camera 2 based on an image captured by the external camera 2.

[0045] In S12, the ECU 10 determines, using the estimation unit 12, whether the recognized object has deviated from the detection range 2R of the external sensor so as to move downward toward the vehicle. As an example, the estimation unit 12 calculates the relative distances from the vehicle 1 of the objects 20 and 21 present within the detection range 2R of the external camera 2 based on the image captured by the external camera 2. The estimation unit 12 may determine that the detected object 20 has deviated from the detection range 2R of the external camera 2 so as to move downward toward the vehicle 1 if the relative distance of the object 20 within the detection range 2R of the external camera 2 decreases and the object 20 is no longer recognized by the object recognition unit 11 through the lower boundary of the detection range 2R. The estimation unit 12 may determine that the detected object 21 has deviated from the detection range 2R of the external camera 2 so as to move downward toward the vehicle 1 if the relative distance of the object 21 within the detection range 2R of the external camera 2 increases and the object 21 is no longer recognized by the object recognition unit 11 through the lower boundary of the detection range 2R. For example, if the relative distance between the objects 20 and 21 does not change within the detection range 2R of the external camera 2, or if the object recognition unit 11 continues to recognize the objects 20 and 21 within the detection range 2R of the external camera 2, the estimation unit 12 may determine that the recognized objects 20 and 21 have not deviated from the detection range 2R of the external camera 2 so as to move downward from the vehicle 1.

[0046] If it is determined that the recognized objects 20, 21 have deviated from the detection range 2R of the external camera 2 so as to move downwards of the vehicle 1 (S12: YES), the ECU 10 proceeds to the process of S13. If it is determined that the recognized objects 20, 21 have not deviated from the detection range 2R of the external camera 2 so as to move downwards of the vehicle 1 (S12: NO), the ECU 10 ends the process of FIG.

[0047] In S13, the ECU 10 causes the estimation unit 12 to estimate whether an object has slipped under the vehicle (estimation step). The estimation unit 12 estimates that there is a possibility that the object 20, 21 has slipped under the vehicle 1, and sets the slip-under flag to 1, for example.

[0048] In S14, the ECU 10 causes the attention warning unit 13 to issue a warning about underpassing. The attention warning unit 13 issues a warning about underpassing, for example, by outputting a notification sound from the speaker of the HMI 5. The attention warning unit 13 may issue a warning about underpassing by displaying a notification image on the display of the HMI 5. The attention warning unit 13 may issue a warning about underpassing by transmitting notification information to a terminal such as a smartphone of the user of the vehicle 1 via the communication unit 6. Note that the processing of S14 is not essential. After that, the ECU 10 ends the processing of FIG. 7.

[0049] Fig. 8 is a flowchart showing an example of a process for estimating the elimination of underpassing. The process (steps) shown in Fig. 8 is repeatedly executed at predetermined calculation intervals, for example, when the vehicle 1 is stopped in a state where underpassing of an object is estimated (for example, a state where the underpassing flag is set to 1). The switch state of the ignition switch 4 may be either on or off.

[0050] As shown in FIG. 8 , in S21, the ECU 10 of the object recognition device 100 causes the estimation unit 12 to determine whether an object estimated to have dived under has been redetected within the detection range of the external sensor. For example, in FIG. 4 showing a state in which diving under of an object 22 has been estimated, the estimation unit 12 determines that the object 22 estimated to have dived under has been redetected within the detection range 2R of the external camera 2 if, based on the image captured by the external camera 2, the object 22 that was not present in the detection range 2R and not recognized by the object recognition unit 11 is re-recognized by the object recognition unit 11 to reappear in the detection range 2R through the lower boundary of the detection range 2R. For example, in FIG. 4 showing a state in which diving under of an object 22 has been estimated, based on the image captured by the external camera 2, the estimation unit 12 determines that the object 22 estimated to have dived under has not been redetected within the detection range 2R of the external camera 2 if, based on the image captured by the external camera 2, the object 22 that was not present in the detection range 2R and not recognized by the object recognition unit 11 remains unrecognized.

[0051] If it is determined that the object 22 that was presumed to have entered under the vehicle has been re-detected within the detection range 2R of the external camera 2 (S21: YES), the ECU 10 proceeds to processing of S22. If it is determined that the object 22 that was presumed to have entered under the vehicle has not been re-detected within the detection range 2R of the external camera 2 (S21: NO), the ECU 10 proceeds to processing of S24.

[0052] In S22, the ECU 10 determines whether the state (redetection state) in which the object presumed to have submerged has been redetected within the detection range of the external sensor by the estimation unit 12 has continued for a predetermined resolution determination time. For example, in FIG. 4 showing a state in which submergence of the object 22 has been presumed, the estimation unit 12 determines that the redetection state has continued for the predetermined resolution determination time if, based on the image captured by the external camera 2, the object 22 redetected within the detection range 2R has been detected for the predetermined resolution determination time. For example, in FIG. 4 showing a state in which submergence of the object 22 has been presumed, the estimation unit 12 determines that the redetection state has not continued for the predetermined resolution determination time if, based on the image captured by the external camera 2, the object 22 redetected within the detection range 2R has not been detected within the predetermined resolution determination time.

[0053] If it is determined that the re-detection state has continued for the predetermined resolution determination time (S22: YES), the ECU 10 proceeds to the process of S23. If it is determined that the re-detection state has not continued for the predetermined resolution determination time (S22: NO), the ECU 10 proceeds to the process of S24.

[0054] In S23, the ECU 10 causes the estimation unit 12 to estimate that the object 22 has disappeared from under the vehicle (elimination step). The estimation unit 12 estimates that the possibility that the object 22 has disappeared from under the vehicle 1 has been eliminated, and sets the obscuration flag to 0, for example. Thereafter, the ECU 10 ends the processing of FIG. 8.

[0055] In S24, the ECU 10 does not estimate, by the estimation unit 12, that the object 22 has disappeared under the vehicle (maintenance step). The estimation unit 12 estimates that the possibility that the object 22 has disappeared under the vehicle 1 has not been eliminated (maintenance of the estimation of the object 22 disappearing under the vehicle 1). The estimation unit 12 estimates that the possibility that the object 22 has disappeared under the vehicle 1 continues, and maintains the object 22 disappearance flag at 1, for example. In S24, the ECU 10 may continue to issue a warning regarding the object 22 disappearing under the vehicle 1 by the warning unit 13. After that, the ECU 10 ends the processing of FIG. 8.

[0056] [Object Recognition Program] The object recognition program causes the ECU 10 (computer) to function (operate) as the above-mentioned object recognition unit 11, estimation unit 12, and attention calling unit 13. The object recognition program is provided by a non-transitory recording medium such as a ROM or a semiconductor memory. Alternatively, the object recognition program may be provided via communication such as a network.

[0057] According to the object recognition device 100, object recognition method, and object recognition program described above, when an object 20, 21 recognized within the detection range 2R of the external camera 2 deviates from the detection range 2R so as to move toward below the vehicle 1, it is estimated that the object 20, 21 will slip below the vehicle 1. This makes it possible to estimate that the object 20, 21 may be slipping below the vehicle 1 by using the external camera 2 that detects the objects 20, 21 around the vehicle 1.

[0058] In the object recognition device 100, the object recognition method, and the object recognition program, if the object 22 that was predicted to have slipped under the vehicle 1 is detected again within the detection range 2R of the external camera 2, it is predicted that the object 22 has stopped slipping under the vehicle 1. This makes it possible to predict that the risk of the object 22 slipping under the vehicle 1 has been eliminated by using the external camera 2 that detects the objects 22 around the vehicle 1.

[0059] In the object recognition device 100, the object recognition method, and the object recognition program, if the object 23 that was predicted to have entered under the vehicle 1 is not detected again within the detection range 2R of the external camera 2 for a predetermined time period for determining whether the object 23 has entered under the vehicle 1, the object recognition device 100 does not predict whether the object 23 has entered under the vehicle 1. This reduces the influence of, for example, erroneous detection by the external camera 2, and makes it possible to more accurately predict whether the object 23 may have entered under the vehicle 1.

[0060] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments.

[0061] In the above embodiment, the external camera 2 is exemplified as an external sensor, but the external sensor is not limited to this example. A radar sensor may be used as the external sensor. The radar sensor is an external sensor for detecting objects outside the vehicle 1. Examples of radar sensors include millimeter-wave radar and LiDAR (Light Detection and Ranging). The radar sensor uses radio waves (e.g., millimeter waves) or light to detect objects around the vehicle 1 and detect the distance between the vehicle 1 and the object. In other words, the radar sensor also functions as a ranging sensor that measures the distance between the vehicle 1 and the external object. The external sensor may be a sonar sensor. The object recognition unit 11 may recognize an object present within the detection range of the external sensor based on the detection result of the external sensor. The detection range of the external sensor may be a range that does not include the space below the vehicle 1. The detection range of the external sensor may include part of the space below the vehicle 1.

[0062] The external sensor may be a combination of the external camera 2 and a radar sensor. When multiple sensors or cameras are used as the external sensors, some of the external sensors may not be used when the ignition switch 4 is in the OFF state, thereby reducing power consumption.

[0063] In the above embodiment, the estimation unit 12 did not estimate that the object had ceased to be present under the vehicle (maintained the estimation that the object had ceased to be present under the vehicle) if the state in which the object estimated to have entered under the vehicle was re-detected within the detection range 2R of the external camera 2 did not continue for a predetermined time period for determining whether the object had entered under the vehicle (maintained the estimation that the object had entered under the vehicle) but is not limited to this example. For example, it may determine whether to estimate that the object had ceased to be present under the vehicle or maintain the estimation that the object had entered under the vehicle based on the result of averaging or filtering the state in which the object estimated to have entered under the vehicle was re-detected within the detection range 2R of the external camera 2 over a predetermined time period.

[0064] In the above embodiment, the estimation unit 12 estimates that the object has ceased to be present when the object estimated to have entered the vehicle is redetected within the detection range 2R of the external camera 2, but the present invention is not limited to this example. For example, the object may be eliminated from the vehicle by having the user of the vehicle 1 visually check an area outside the detection range 2R of the external camera 2 and inputting the result into the HMI 5. Furthermore, the estimation unit 12 may estimate that the object has entered the vehicle even when the vehicle 1 is moving.

[0065] In the above embodiment, the object recognition device 100 includes the attention calling unit 13, but the attention calling unit 13 is not essential. The result of the estimation of the object's penetration by the object recognition device 100 may be used for other control purposes other than calling the user's attention.

[0066] If the vehicle 1 is equipped with an autonomous driving ECU capable of executing autonomous driving control, the result of estimating the obscuration of an object may be used for the autonomous driving control. When the object recognition device 100 estimates the obscuration of an object while the vehicle 1 is being driven autonomously, the object recognition device 100 may cause the autonomous driving ECU to execute autonomous driving control of the vehicle 1 so that the tires of the vehicle 1 do not run over the obscured object.

[0067] For example, when the object recognition device 100 estimates that an object recognized by the object recognition device 100 as being present near the center of the vehicle 1 in the width direction will run under the vehicle 1 while the vehicle 1 is traveling straight, the autonomous driving ECU may control the steering angle of the vehicle 1 so that the tires of the vehicle 1 do not run up on the object that has run under the vehicle 1. In this case, if it is estimated based on the relative speed of the recognized object in the vehicle width direction that the run-under object is present near the center of the vehicle 1 in the width direction below the vehicle 1, the autonomous driving ECU may suppress or prohibit changes in the steering angle of the vehicle 1 while traveling straight, thereby reducing the possibility that the rear tires will run up on the object that has run under the vehicle 1. If it is estimated based on the relative speed of the recognized object in the vehicle width direction that the run-under object is present near both ends of the vehicle 1 in the width direction below the vehicle 1, the autonomous driving ECU may change the steering angle of the vehicle 1 so that the trajectory of the rear tires does not overlap the position of the estimated object, thereby reducing the possibility that the rear tires will run up on the object that has run under the vehicle 1.

[0068] In addition, when vehicle 1 is moving slowly in a straight line and the object recognition device 100 predicts that an object that is recognized as being present near either end of the vehicle 1 in the width direction will slip under the vehicle, the autonomous driving ECU may activate the brakes of vehicle 1 to prevent the tires of vehicle 1 from riding over the object that has slipped under the vehicle 1.

[0069] Note that, when the vehicle 1 is equipped with a driving assistance ECU capable of executing driving assistance control instead of an autonomous driving ECU capable of executing autonomous driving control, the result of estimating the obscuring of an object may be used for the driving assistance control. When the object recognition device 100 estimates that an object will obscure the road while the vehicle 1 is being manually driven, the object recognition device 100 may cause the driving assistance ECU to execute driving assistance control such as the above-mentioned control of the steering angle or brakes so that the tires of the vehicle 1 do not run up on the obscured object. [Explanation of symbols]

[0070] 1...vehicle, 2...external camera (external sensor), 2R...detection range, 10...ECU (computer), 11...object recognition unit, 12...estimation unit, 20, 21, 22, 23, 24...object, 100...object recognition device

Claims

1. an object recognition unit that recognizes an object present within a detection range of an external sensor based on a detection result of the external sensor that detects an object around the vehicle; an estimation unit that, when the recognized object deviates from the detection range of the external sensor and moves toward below the vehicle, estimates that the object will slip under the vehicle.

2. The object recognition device according to claim 1 , wherein the estimation unit estimates that the object has ceased to be trapped when the object estimated to have been trapped is redetected within the detection range of the external sensor.

3. 3. The object recognition device according to claim 2, wherein the estimation unit does not estimate that the object has ceased to be trapped if the object whose trapping has been estimated is re-detected within the detection range of the external sensor for a predetermined period of time to determine whether the object has been trapped.

4. An object recognition method for an object recognition device that recognizes objects around a vehicle, comprising: an object recognition step of recognizing, by the object recognition device, whether or not the object is present within a detection range of the external sensor based on a detection result of the external sensor that detects the object; an estimation step of estimating, by the object recognition device, that the object will move under the vehicle when the recognized object deviates from the detection range of the external sensor and moves toward under the vehicle.

5. An object recognition program that causes a computer to function as an object recognition device that recognizes an object based on a detection result of an external sensor that detects an object around a vehicle, the object recognition program comprising: an object recognition unit that recognizes whether or not an object exists within a detection range of the external sensor based on a detection result of the external sensor; and An object recognition program that functions as an estimation unit that estimates that an object will slip under the vehicle when the recognized object deviates from the detection range of the external sensor and moves toward the bottom of the vehicle.

Images