JP2020166479A - Drive support device - Google Patents

Abstract

To provide a drive support device capable of smoothly supporting drive of a moving body by making it possible to visually recognize a situation inside an object through a transmission member region of the object existing around the moving body.SOLUTION: A drive support device 10 includes: a camera 44 for detecting an object around an own vehicle 12; a polarization filter 76 provided with the camera 44; and an automatic adjustment processing unit 70 and a polarization filter adjusting unit 74 for determining one oncoming vehicle of vehicles existing around the own vehicle 12 detected by the camera 44 as an adjustment object of the polarization filter 76 to automatically adjust the polarization filter 76 relative to the transmission member area of the oncoming vehicle.SELECTED DRAWING: Figure 1

Description

The present invention relates to a driving support device that supports the driving of a moving body.

For example, Patent Document 1 discloses that the road surface on which the own vehicle travels is monitored by a camera (visual sensor) and the monitored road surface is analyzed.

Japanese Unexamined Patent Publication No. 2013-79937

By the way, for example, in an intersection, the own vehicle (moving body) usually turns left or right across the oncoming lane after passing by an oncoming vehicle (another moving body) traveling in the oncoming lane.

On the other hand, when the oncoming vehicle does not move forward in the direction of travel due to traffic congestion in the oncoming lane and stops in front of the intersection, the own vehicle can turn left or right across the oncoming lane while paying attention to the oncoming vehicle. At that time, the driver of the oncoming vehicle may perform a gesture action on the own vehicle with the intention of permitting a right or left turn at the intersection.

When the driver of the own vehicle performs a driving operation, the driver may make a right or left turn at the intersection after confirming the gesture behavior of the driver of the oncoming vehicle. In addition, when the driving support device mounted on the own vehicle assists the driver's driving operation, or when the driving support device automatically drives the own vehicle, turn left or right at the intersection from the gesture behavior of the driver of the oncoming vehicle. It is sufficient to judge whether or not it is possible and make a right or left turn based on the judgment result.

However, in recent years, in order to suppress the reflection of an object on a windshield or the like, an increasing number of vehicles have adopted a transparent member that has been cut against visible light or ultraviolet rays (UV) for the windshield or the like. As a result, it is difficult to visually recognize the gesture behavior of the driver of the oncoming vehicle from the outside through the transparent member (transparent member region).

The present invention has been made in consideration of such a problem, and by making it possible to visually recognize the situation in the object through the transmission member region of the object existing around the moving body, the operation of the moving body is performed. The purpose is to provide a driving support device that can smoothly support the above.

An aspect of the present invention is a driving support device that assists the driving of a moving body, in which a visual sensor that detects an object around the moving body, a polarizing filter provided in the visual sensor, and the visual sensor detect the object. One of the objects existing around the moving body is determined to be the adjustment target of the polarizing filter, and the transmitting member region capable of transmitting light is specified in the determined adjustment target, and the specified transmission is specified. It is provided with an automatic adjustment unit that automatically adjusts the polarizing filter with respect to the member region.

According to the present invention, by automatically adjusting the polarizing filter with respect to the transmission member region to be adjusted, the internal state of the adjustment target can be visually recognized through the transmission member region. As a result, based on the internal situation of the adjustment target (for example, the gesture behavior of the driver of the oncoming vehicle), the driving support of the moving body (for example, the running control for turning left or right in the intersection of the own vehicle) is smoothly performed. It becomes possible.

It is a block diagram of the vehicle which mounts the driving support device which concerns on this embodiment. It is explanatory drawing which schematically illustrated the arrangement of the camera and the like in the vehicle of FIG. It is explanatory drawing which schematically illustrated the right turn of the own vehicle at an intersection. It is a flowchart which shows the operation of the driving support device of FIG. It is explanatory drawing which schematically illustrated the detection of the driver of an oncoming vehicle by a camera and a radar.

Hereinafter, a preferred embodiment of the driving support device according to the present invention will be illustrated and described with reference to the accompanying drawings.

[1. Configuration of this embodiment]
<1.1 Overall configuration of driving support device 10>
FIG. 1 is a block diagram showing a configuration of a driving support device 10 according to an embodiment of the present invention. The driving support device 10 is incorporated in the vehicle 12 shown in FIG. 2 (hereinafter, also referred to as the own vehicle 12), and automatically or manually controls the traveling of the own vehicle 12 (moving body, own moving body). .. The “travel control” refers to control related to the travel of the own vehicle 12, including at least acceleration / deceleration control of the own vehicle 12. When "driving control" is performed by "automatic driving", "automatic driving" is not only "fully automatic driving" in which all driving control (driving support) of the own vehicle 12 is automatically performed, but also driving control (driving support). It is a concept that includes "partially automatic operation" in which

As shown in FIG. 1, the operation support device 10 is basically composed of an input system device group 14, a control system 16, and an output system device group 18. Each of the devices forming the input system device group 14 and the output system device group 18 is connected to the control system 16 via a communication line.

The input system device group 14 is connected to an external sensor 20, a communication device 22, a navigation device 24 (self-positioning device, self-positioning unit), a vehicle sensor 26, an automatic driving switch 28, and an operating device 30. It is provided with an operation detection sensor 32.

The output system device group 18 includes a driving force device 34 for driving a wheel (not shown), a steering device 36 for steering the wheel, a braking device 38 for braking the wheel, and a notification device 40 for notifying the driver through audiovisual means. It is provided with a direction indicator 42 that notifies the outside of the direction in which the vehicle 12 turns left or right.

<1.2 Specific configuration of input system device group 14>
The outside world sensor 20 acquires information indicating the outside world state of the own vehicle 12 (hereinafter, referred to as outside world information) and outputs the outside world information to the control system 16. Specifically, the external sensor 20 includes a plurality of cameras 44 (visual sensors), a plurality of radars 46 (detection units using millimeter waves), and a plurality of LiDARs (Light Detection and Rangers, Laser Imaging Detection and Rangers). It includes 48 (a detection unit using light), an illuminance sensor 50, and a GPS (Global Positioning System) sensor 52 (self-positioning unit, self-positioning device).

The illuminance sensor 50 is a rain light sensor that detects the brightness (illuminance) around the own vehicle 12 due to irradiation with sunlight or the like. The GPS sensor 52 detects the current position (self-position information) of the own vehicle 12. The plurality of cameras 44, the plurality of radars 46, and the plurality of LiDAR 48s will be described later.

The communication device 22 is configured to be able to communicate with an external device including a roadside unit, another vehicle, and a server. For example, information related to traffic equipment, information related to other vehicles, probe information, or the latest map information 54 can be obtained. Send and receive. The map information 54 is stored in a predetermined memory area of the storage device 56 provided in the control system 16 or in the navigation device 24.

The navigation device 24 includes a satellite positioning device capable of detecting the current position (self-position information) of the own vehicle 12 and a user interface (for example, a touch panel display, a speaker, and a microphone). Therefore, the driving support device 10 may be provided with at least one of the navigation device 24 and the GPS sensor 52 with respect to the detection function of the current position of the own vehicle 12. Further, the navigation device 24 calculates a route to the designated destination based on the current position of the own vehicle 12 or the designated position by the user, and outputs the route to the control system 16. The route calculated by the navigation device 24 is stored as route information 58 in a predetermined memory area of the storage device 56.

The vehicle sensor 26 includes a speed sensor that detects the traveling speed (vehicle speed) of the own vehicle 12, an acceleration sensor that detects acceleration, a lateral G sensor that detects lateral G, a yaw rate sensor that detects angular speed around the vertical axis, and orientation / orientation. A directional sensor for detecting a gradient and a gradient sensor for detecting a gradient are included, and detection signals from each sensor are output to the control system 16. These detection signals are stored as own vehicle information 60 in a predetermined memory area of the storage device 56.

The automatic operation switch 28 is composed of, for example, a push button type hardware switch or a software switch using a navigation device 24. The automatic operation switch 28 is configured so that a plurality of operation modes can be switched by manual operation by a user including a driver.

The operation device 30 includes an accelerator pedal, a steering wheel, a brake pedal, a shift lever, and a turn signal lever. The operation device 30 is equipped with an operation detection sensor 32 that detects the presence / absence of operation by the driver, the amount of operation, and the operation position.

The operation detection sensor 32 constitutes the control system 16 for the accelerator depression amount (accelerator opening), steering operation amount (steering amount), brake depression amount, shift position, right / left turn direction, etc. as the detection result, which will be described later. Output to the control unit 62.

<1.3 Specific configuration of output system device group 18>
The driving force device 34 includes a driving force ECU (Electronic Control Unit) and a driving source including an engine and a driving motor. The driving force device 34 generates a traveling driving force (torque) of the own vehicle 12 according to a traveling control value input from the traveling control unit 62, and transmits the traveling driving force (torque) to the wheels via a transmission or directly.

The steering device 36 is composed of an EPS (electric power steering system) ECU and an EPS device. The steering device 36 changes the direction of the wheels (steering wheels) according to the travel control value input from the travel control unit 62.

The braking device 38 is, for example, an electric servo brake that also uses a hydraulic brake, and is composed of a brake ECU and a brake actuator. The braking device 38 brakes the wheels according to the travel control value input from the travel control unit 62.

The notification device 40 includes a notification ECU, a display device, and an audio device. The notification device 40 performs a notification operation related to automatic operation or manual operation in response to a notification command output from the notification control unit 64, which is described later, which constitutes the control system 16. The direction indicator 42 performs a notification operation related to turning left or right of the own vehicle 12 in response to a notification command output from the notification control unit 64.

<1.4 Configuration of control system 16>
The control system 16 is composed of one or a plurality of ECUs, and includes various function realization units in addition to the above-mentioned storage device 56, travel control unit 62, and notification control unit 64. In this embodiment, the function realization unit is a software functional unit whose functions are realized by one or a plurality of CPUs (Central Processing Units) executing a program stored in a non-transient storage device 56. Is. Instead of this, the function realization unit may be a hardware function unit composed of an integrated circuit such as an FPGA (Field-Programmable Gate Array).

In addition to the storage device 56, the travel control unit 62, and the notification control unit 64, the control system 16 includes an outside world recognition unit 66, an action planning unit 68, an automatic adjustment processing unit 70 (automatic adjustment unit), a clock unit 72, and a polarizing filter adjustment unit. It is configured to include 74 (automatic adjustment unit).

The outside world recognition unit 66 recognizes lane marks (white lines) on both sides of the own vehicle 12 by using various information (for example, outside world information from the outside world sensor 20) input by the input system device group 14, and stops lines. -Generates the position information of the traffic light or "static" external recognition information including the travelable area. In addition, the outside world recognition unit 66 uses various input information to recognize obstacles such as parked and stopped vehicles, traffic participants such as people and other vehicles, or "dynamic" outside world recognition including the light color of a traffic light. Generate information.

The action planning unit 68 creates an action plan (time series of events) for each traveling section based on the recognition result by the outside world recognition unit 66, and updates the action plan as necessary. As a type of event, for example, in the case of an intersection 90 (see FIG. 3) described later, the event is an event in which the own vehicle 12 smoothly travels at the intersection 90.

Further, the action planning unit 68 uses the map information 54, the route information 58, and the own vehicle information 60 read from the storage device 56 to generate a traveling track (time series of target behavior) according to the created action plan. Specifically, this traveling track is a time-series data set whose data units are position, attitude angle, speed, acceleration, curvature, yaw rate, and steering angle.

The automatic adjustment processing unit 70 performs processing for automatically adjusting the polarizing filter 76 provided in the camera 44 by using various information from the outside world recognition unit 66, the current time measured by the clock unit 72, and the like. .. The automatic adjustment processing unit 70 outputs a command signal for performing the above processing to the polarizing filter adjustment unit 74. Further, the automatic adjustment processing unit 70 outputs information related to the result of the automatic adjustment processing to the polarizing filter 76 to the action planning unit 68.

In this case, the automatic adjustment processing unit 70 includes the vehicle identification unit 70a, the adjustment target identification unit 70b, the transmission member area identification unit 70c, the filter adjustment determination unit 70d, the body movement identification unit 70e, the body movement content estimation unit 70f, and the travel control permission. It functions as a determination unit 70 g. The functions of these components of the automatic adjustment processing unit 70 will be described later.

The travel control unit 62 has each travel control value for controlling the travel of the own vehicle 12, including at least acceleration / deceleration control of the own vehicle 12, according to the travel trajectory (time series of target behavior) generated by the action planning unit 68. To determine. Then, the travel control unit 62 outputs each of the obtained travel control values to the driving force device 34, the steering device 36, and the braking device 38. The notification control unit 64 drives and controls the notification device 40 or the direction indicator 42 according to the traveling track generated by the action planning unit 68. The polarizing filter adjusting unit 74 automatically adjusts the polarizing filter 76 based on a command signal from the automatic adjusting processing unit 70. Hereinafter, the travel control unit 62, the notification control unit 64, and the polarizing filter adjustment unit 74 may be collectively referred to as the “operation control unit 78”.

<1.5 Arrangement of external sensor 20 in own vehicle 12>
FIG. 2 is an explanatory diagram schematically illustrating the arrangement of a plurality of cameras 44, a plurality of radars 46, and a plurality of LiDAR 48s in the vehicle 12 of the external world sensors 20. Note that FIG. 2 illustrates a case where the vehicle 12 is a vehicle traveling on the left side as an example.

The plurality of cameras 44 arranged in the own vehicle 12 include a front camera 44a provided at the front of the roof, a right front side camera 44b arranged on the driver's seat side (right front side) of the vehicle body 80, and the vehicle body 80. The left front side camera 44c arranged on the passenger side (left front side), the right rear side camera 44d arranged on the right rear seat side of the vehicle body 80, and the left rear seat side of the vehicle body 80. There are a left rear side camera 44e, a rear camera 44f provided at the rear of the roof, and a front camera 44g which is two stereo cameras provided at the front of the roof. In the following description, for convenience of explanation, the front cameras 44a and 44g, the right front side camera 44b, the left front side camera 44c, the right rear side camera 44d, the left rear side camera 44e, and the rear camera 44f are referred to. It may be referred to as a camera 44a to 44g.

The front camera 44a captures an image within a predetermined angle range (imaging area 82a) in front of the own vehicle 12, and outputs the captured image to the control system 16 (see FIG. 1). The right front side camera 44b captures an image within a predetermined angle range (imaging area 82b) on the right side of the own vehicle 12, and outputs the captured image to the control system 16. The left front side camera 44c captures an image within a predetermined angle range (imaging area 82c) on the left side of the own vehicle 12, and outputs the captured image to the control system 16. In this case, a part of the imaging region 82a, which is the front region, and the imaging region 82b, which is the right region, overlap. Further, a part of the imaging region 82a and the imaging region 82c, which is the left side region, overlap.

Similarly, with respect to the right rear side camera 44d, the left rear side camera 44e, the rear camera 44f, and the front camera 44g, a predetermined angle region outside the own vehicle 12 to which the lens of the camera is directed is partially covered. Is imaged so as to overlap, and the captured image is output to the control system 16. Note that FIG. 2 also shows the imaging regions 82d to 82f of the right rear side camera 44d, the left rear side camera 44e, and the rear camera 44f.

Further, a polarizing filter 76 (polarizing filter 76a to 76g) is attached to the lens of each camera 44a to 44g.

Further, a plurality of radars 46 and a plurality of LiDAR 48s are also arranged in the own vehicle 12. The plurality of radars 46 include three radars 46a to 46c arranged in front of the vehicle body 80 and two radars 46d and 46e arranged behind the vehicle body 80. The LiDAR 48 includes two LiDAR 48a and 48b arranged in front of the vehicle body 80 and three LiDAR 48c to 48e arranged behind the vehicle body 80.

[2. Operation of driving support device 10]
The driving support device 10 in the present embodiment is configured as described above. Subsequently, the operation of the driving support device 10 when turning left or right at the intersection 90 will be described with reference to FIGS. 3 to 5. Here, a case where the own vehicle 12 equipped with the driving support device 10 turns right in the intersection 90 shown in FIG. 3 by automatic driving will be described.

<2.1 Explanation of intersection 90>
FIG. 3 is an explanatory diagram showing a state in which the own vehicle 12 enters the intersection 90 and tries to make a right turn. The own vehicle 12 attempts to pass through the intersection 90 where the first road 94 and the second road 96 intersect along the planned travel route 92 indicated by the broken line arrow. In FIG. 3, each of the first road 94 and the second road 96 is a road having four lanes. Further, FIG. 3 illustrates a case where a vehicle such as the own vehicle 12 travels on the left side.

That is, the first road 94 is adjacent to the first traveling lane 94a and the first traveling lane 94a of two lanes in which a vehicle such as the own vehicle 12 can go straight, and the oncoming vehicle 98 (another vehicle, another moving body). ) And the like, including the first oncoming lane 94b having two lanes on which the vehicle can go straight. On the other hand, the second road 96 is adjacent to the second traveling lane 96a and the second traveling lane 96a having two lanes on which the vehicle can travel straight, and the vehicle such as the vehicle 102 can travel straight on the second road 96, similarly to the first road 94. It is configured to include a second oncoming lane 96b having two lanes.

<2.2 Explanation of the operation of FIGS. 3 to 5>
Next, regarding the operation of the driving support device 10 (see FIG. 1) shown in the flowchart of FIG. 4, it corresponds to the situation where the own vehicle 12 (see FIGS. 1 to 3 and 5) travels along the planned travel route 92. I will explain while letting you. Here, in the first oncoming lane 94b, when the oncoming vehicle 98 is temporarily stopped before the intersection 90 due to traffic congestion or the like, the situation of the driver 100 in the oncoming vehicle 98 (for example, gesture behavior) is taken into consideration. The operation of the driving support device 10 from the first traveling lane 94a into the intersection 90, turning right, and reaching the second traveling lane 96a will be described. In FIG. 3, the illustration of vehicles other than the oncoming vehicle 98 existing in the first oncoming lane 94b is omitted.

First, when the own vehicle 12 travels in the first traveling lane 94a and enters the intersection 90, in step S1 of FIG. 4, the automatic adjustment processing unit 70 (see FIG. 1) uses the camera 44 (for example, FIG. 2, FIG. The image data captured by the front camera 44a) of FIGS. 3 and 5 is acquired. In addition, the automatic adjustment processing unit 70 also acquires the detection result of the radar 46 (for example, the radar 46a) and the detection result of the LiDAR 48. Further, the automatic adjustment processing unit 70 includes the current position of the own vehicle 12 specified by the GPS sensor 52 or the navigation device 24, the illuminance (illuminance information) around the own vehicle 12 detected by the illuminance sensor 50, and the current time measured by the clock unit 72. Also get the time.

In step S2, the vehicle identification unit 70a identifies an object reflected in the image data, that is, an object existing around the own vehicle 12, from the image data acquired in step S1. Next, the vehicle identification unit 70a identifies a moving body (another moving body) existing around the own vehicle 12 from the identified object. Specifically, in the example of FIG. 3, the vehicle identification unit 70a is an oncoming vehicle 98 (see FIGS. 3 and 5) existing at an intersection 90, a first road 94 and a second road 96, and a second oncoming lane 96b. The vehicle 102 traveling on the vehicle 102 is specified. Next, the vehicle identification unit 70a identifies the types of the specified oncoming vehicle 98 and the vehicle 102. As this type, for example, there are types such as (1) whether the identified vehicle is a two-wheeled vehicle, a three-wheeled vehicle, or a four-wheeled vehicle, and (2) whether it is an ordinary passenger car, a bus, or a truck.

As described above, the operation of FIGS. 3 to 5 is an operation when the own vehicle 12 travels by automatic driving in consideration of the situation (for example, gesture behavior) of the driver 100 in the oncoming vehicle 98. In this case, if the driver 100 makes a gesture that the own vehicle 12 may turn right at the intersection 90, for example, the camera 44 passes through the transmission member region 106 such as the windshield 104 of the oncoming vehicle 98. By imaging (detecting) the driver 100, the detected gesture behavior of the driver 100 can be recognized.

However, in recent vehicles, in order to suppress the reflection of an object on a transparent member such as glass or resin forming the transparent member region 106, a cut treatment against visible light and ultraviolet rays (UV) is performed. As a result, when the transparent member region 106 is visually recognized from the outside, it is difficult to recognize the gesture behavior of the driver 100 of the oncoming vehicle 98 through the transparent member region 106 due to glare or reflection in the transparent member region 106. Is.

In the operation support device 10, even if the transmission member region 106 is subjected to such a cut process, the polarizing filter 76 is automatically adjusted to reduce glare or reflection so as to face the transmission member region 106 via the transmission member region 106. The driver 100 of the car 98 can be imaged (detected).

Therefore, in the next step S3, the adjustment target identification unit 70b identifies the vehicle closest to the own vehicle 12 among the vehicles identified in step S2 as the adjustment target of the polarizing filter 76. Specifically, among the image data captured by the front camera 44a, the oncoming vehicle 98, which is the closest vehicle, is set as the adjustment target of the polarizing filter 76.

That is, the own vehicle 12 turns right in the intersection 90 along the planned travel route 92 so as to cross the front in the traveling direction of the oncoming vehicle 98. Therefore, the adjustment target identification unit 70b determines that it is necessary to identify the gesture behavior of the driver 100 in the oncoming vehicle 98 ahead in order to turn right in the intersection 90, and adjusts the oncoming vehicle 98 to the polarizing filter 76. Determine as a target.

In step S4, the transparent member region specifying unit 70c identifies the transparent member region 106 from the image data in which the oncoming vehicle 98 is reflected in the oncoming vehicle 98 identified in step S3. In this case, the transmission member region identification unit 70c may specify the transmission member region 106 based on the detection result of the radar 46 or LiDAR 48 and the current position of the own vehicle 12. That is, the distance and angle of the oncoming vehicle 98 with respect to the own vehicle 12 can be specified from the detection result of the oncoming vehicle 98 by the radar 46 or the LiDAR 48 and the current position of the own vehicle 12. 3 and 5 show a case where the angle θ of the transmission member region 106 (driver riding region 108) with respect to the own vehicle 12 (position of the radar 46 or LiDAR 48) is specified from the detection result. Further, by specifying the type of the oncoming vehicle 98 from the identification result in step S2, it is possible to specify the installation location of the transmission member region 106 such as the windshield 104 in the oncoming vehicle 98. As a result, the transparent member region 106 can be easily specified. Alternatively, the transparent member region specifying unit 70c may specify the transparent member region 106 using only the image data in which the oncoming vehicle 98 is reflected. 3 and 5 show, as an example, a case where the windshield 104 of the oncoming vehicle 98 is specified in the transmission member region 106 in the image data captured by the front camera 44a.

In the oncoming vehicle 98, in addition to the windshield 104, the left and right front door glasses, the left and right rear door glasses, and the back glass are also transparent members. Among these transparent members, the transparent member region specifying unit 70c specifies the windshield 104 closest to the own vehicle 12 as the transparent member region 106. Further, the windshield of the own vehicle 12 is inclined toward the front, and the front camera 44a images the front through the inclined windshield. Further, since the windshield 104 of the oncoming vehicle 98 is also inclined, in the radar 46 or LiDAR 48, the light or millimeter wave reflected in the transmission member region 106 may be incident from an oblique direction. Therefore, it is desirable that the transmission member region specifying portion 70c specifies the transmission member region 106 in consideration of the inclination angle of the windshield of the own vehicle 12 and the inclination angle of the windshield 104 of the oncoming vehicle 98. In the following description, for convenience of explanation, the image of the transparent member region 106 reflected in the image data may also be referred to as “transparent member region 106”.

Further, in step S4, the transparent member region specifying unit 70c identifies a region (driver riding region 108) in which the camera 44 can detect the driver 100 in the transparent member region 106.

In this case, in step S2, the vehicle identification unit 70a identifies the type of vehicle. Therefore, the transmission member region identification unit 70c operates in the transmission member region 106 in the image data based on the identification result of the vehicle identification unit 70a, the detection result of the radar 46 or LiDAR48, and the current position of the own vehicle 12. The area where the person 100 is presumed to be reflected is specified as the driver boarding area 108. That is, (1) the oncoming vehicle 98 is a vehicle traveling on the left side, (2) the type of the oncoming vehicle 98 based on the identification result in step S2, and (3) the detection result of the oncoming vehicle 98 by the radar 46 or LiDAR48. , And the distance and angle of the oncoming vehicle 98 with respect to the own vehicle 12 based on the current position of the own vehicle 12 (angle θ in FIGS. 3 and 5), the region on the left side of the transmission member region 106 when viewed from the own vehicle 12. Can be specified as the driver boarding area 108. Alternatively, the transparent member region specifying unit 70c may specify the driver boarding region 108 using only the image data in which the oncoming vehicle 98 is reflected by using the identification result of the vehicle identification unit 70a.

Even when the driver's riding area 108 is specified, the transmission member area specifying portion 70c is operated in consideration of the inclination angle of the windshield of the own vehicle 12 and the inclination angle of the windshield 104 of the oncoming vehicle 98. The passenger boarding area 108 may be specified. In the following description, for convenience of explanation, the image of the region in which the driver 100 is reflected in the transparent member regions 106 of FIGS. 3 and 5 and the region in which the driver 100 is reflected in the image data is referred to as ". It may be referred to as "driver boarding area 108".

In step S5, the filter adjustment determination unit 70d sets the polarizing filter 76 (in FIG. 3 and FIG. 5, the polarizing filter of the front camera 44a) with respect to the transmission member region 106 specified in step S4, more specifically, the driver riding region 108. It is determined whether or not the automatic adjustment of 76a) is necessary.

When the driver 100 cannot be recognized from the driver riding area 108 reflected in the image data because the transparent member region 106 (windshield 104) is cut, the filter adjustment determination unit 70d It is determined that the automatic adjustment of the polarizing filter 76 with respect to the driver riding area 108 is necessary (step S5: YES), and the process proceeds to the next step S6.

In step S6, the filter adjustment determination unit 70d outputs a command signal to the polarization filter adjustment unit 74 to instruct the execution of automatic adjustment based on the angle θ or the like. In this case, the filter adjustment determination unit 70d may generate a command signal instructing the automatic adjustment of the polarizing filter 76 in consideration of the current position of the own vehicle 12, the illuminance, and the current time.

As a result, the polarizing filter adjusting unit 74 automatically adjusts the polarizing filter 76 (76a). As a result, when the automatic adjustment processing unit 70 acquires the image data captured by the camera 44 (44a) again after the automatic adjustment, the driver 100 reflected in the driver boarding area 108 is recognized in the acquired image data. be able to. The configuration of the polarizing filter 76 and the adjusting method of the polarizing filter 76 are known. For example, (1) a configuration in which a polarizing lens is rotated and adjusted, and (2) a plurality of polarizers are installed on a CMOS image sensor, and the degree of polarization is adjusted according to the ambient conditions acquired selectively or in combination. There is a configuration. Therefore, detailed description of the configuration and adjustment method of the polarizing filter 76 will be omitted.

On the other hand, in step S5, when the transmission member region 106 (windshield 104) is not cut, the radar 46 or LiDAR 48 can detect the driver 100 via the driver riding region 108. In this case, the filter adjustment determination unit 70d determines from the detection result of the radar 46 or LiDAR 48 that the automatic adjustment of the polarizing filter 76 is unnecessary (step S5: NO), and skips step S6. Alternatively, if the driver 100 can be recognized in the driver riding area 108 reflected in the image data, the filter adjustment determination unit 70d does not need to automatically adjust the polarizing filter 76 with respect to the driver riding area 108. May be determined (step S5: NO).

In the next step S7, the body movement identification unit 70e determines whether or not the driver 100 reflected in the driver riding area 108 is exhibiting some body movement (for example, gesture behavior). Then, when it is determined that some kind of body movement is exhibited, the body movement identification unit 70e identifies the type of body movement of the driver 100.

In this case, the body movement identification unit 70e determines (1) the viewing direction of the driver 100 (whether or not the own vehicle 12 is visually recognized), and (2) regarding the determination of the presence or absence of gesture behavior and the identification of the type of body movement. The degree of caution of the driver 100 (whether or not he is paying attention to his own vehicle 12, whether or not he is looking at the screen of a portable device in the oncoming vehicle 98, whether or not he is sleeping), (3) The driver 100 It is determined whether or not the vehicle 12 is performing a gesture action, and (4) whether or not the vehicle is a gesture action that signals permission for a right turn of the vehicle 12. In this discrimination process, if the body movement identification unit 70e discriminates (1) to (4) based on the detection results of various sensors provided in the driving support device 10, computer vision technology, or machine learning. Good.

In step S8, the body movement content estimation unit 70f estimates the content of the gesture action from the processing result in step S7 if it is a gesture action that allows the vehicle 12 to turn right. That is, even if the driver 100 performs some gesture action, it may not be an action with respect to the own vehicle 12, or may be performing some action without looking at the own vehicle 12. Therefore, the body movement content estimation unit 70f performs the estimation process of step S8 based on the processing result in step S7, using the gesture action for the own vehicle 12 as a trigger. Here, for example, when the driver 100 is waving his / her hand toward the own vehicle 12 in the left-right direction, the body movement content estimation unit 70f allows the driver 100 to make a right turn to the own vehicle 12. Presumed to be a body movement. In this case as well, the body movement content estimation unit 70f may estimate the body movement content by machine learning.

A side camera (right front side camera 44b, left front side camera 44c, right rear side camera 44d, left rear side camera 44e) is arranged as a visual sensor in the own vehicle 12, and the image acquired by the side camera. Therefore, the gesture behavior of the driver of another vehicle located on the side of the own vehicle 12 may be recognized. Specifically, a gesture action of permission from the driver of another vehicle existing at a position orthogonal to the traveling direction of the own vehicle 12 in the intersection 90, and a stop request from another vehicle such as an adjacent emergency vehicle. It becomes possible to recognize the gesture behavior of.

In step S9, the travel control permission determination unit 70g determines whether or not the estimation result in step S8 should be reflected in the action plan. For example, if the estimation result is that the driver 100 is waving his / her hand in the left-right direction toward the own vehicle 12, the driving control permission determination unit 70g determines that the intersection 90 can turn right. Is notified to the Action Planning Department 68. The action planning unit 68 updates the action plan in response to the determination result of the travel control permission determination unit 70g.

As a result, in step S10, the travel control unit 62 executes a right turn of the own vehicle 12 at the intersection 90 by controlling each device forming the output system device group 18 based on the updated action plan.

In the next step S11, the body movement identification unit 70e acquires the image data of the camera 44 again, and from the image of the driver 100 in the acquired image data, whether or not the driver 100 continues the same body movement. Is determined. If it is determined that the same body movement is continued (step S11: YES), the driving support device 10 returns to step S7 and executes the processes of steps S7 to S11 again. As a result, when the driver 100 is performing a gesture action that permits the own vehicle 12 to make a right turn, the own vehicle 12 repeatedly executes steps S7 to S11 to enter the second traveling lane 96a in the intersection 90. You can turn right towards.

In step S11, when the body movement identification unit 70e recognizes that the driver 100 has stopped moving from the image of the driver 100 in the image data (step S11: NO), the body movement content estimation unit 70f For example, because the own vehicle 12 crosses the front of the oncoming vehicle 98 in the traveling direction and turns right, the gesture action for permitting the right turn is no longer necessary, or even if the own vehicle 12 is turning right, the oncoming vehicle 98 is turned. It is presumed that the driver 100 has stopped the gesture action in order to move forward. Then, the driving support device 10 proceeds to step S12.

In the next step S12, the automatic adjustment processing unit 70 determines whether or not the own vehicle 12 has turned right at the intersection 90 based on the image data or the like acquired from the camera 44. Alternatively, the driving support device 10 determines whether or not to stop the automatic driving.

When the own vehicle 12 is turning right at the intersection 90 or when the automatic driving is continuously performed (step S12: NO), the driving support device 10 returns to step S1 and executes the processes of steps S1 to S12 again. To do. On the other hand, when the own vehicle 12 turns right in the intersection 90, passes through the intersection 90 and enters the second traveling lane 96a, or when the automatic driving is terminated by operating the automatic driving switch 28 (step S12: YES). The driving support device 10 ends the operation shown in FIG.

<2.3 Deformation example>
In the above description, the case where the own vehicle 12 traveling on the left side turns right in the intersection 90 using the front camera 44a has been described. In the operation of FIGS. 3 to 5, in addition to the front camera 44a or instead of the front camera 44a, another camera 44 (for example, the right front side camera 44b or the left front side camera 44c, or the rear camera 44f) is used. By using this, it is also applicable to the case where the own vehicle 12 traveling on the left side turns right in the intersection 90 by using the right front side camera 44b or the left front side camera 44c or the rear camera 44f.

Further, the operations of FIGS. 3 to 5 can be applied to the case where the vehicle 12 traveling on the left side turns left in the intersection 90 and the vehicle 12 traveling on the right side turns right or left in the intersection 90. In the case of traveling on the right side, the region on the right side of the own vehicle 12 in the transparent member region 106 of the oncoming vehicle 98 is specified as the driver riding region 108.

Further, the operations of FIGS. 3 to 5 are applicable not only to the oncoming vehicle 98 but also to other vehicles such as the vehicle 102. For example, on a highway or the like, when the own vehicle 12 merges into a lane, the driver of another vehicle traveling in the lane recognizes the gesture behavior of permitting the own vehicle 12 to merge. It is possible.

Further, in the operations of FIGS. 3 to 5, it is also possible to detect the oncoming vehicle 98 via the transparent member of another vehicle.

Further, in step S8 of FIG. 4, the polarizing filter 76 may be automatically adjusted based on the processing result of step S7. As a result, the gesture behavior of the driver 100 can be estimated more accurately. Furthermore, in the present embodiment, instead of the hardware-like automatic adjustment processing of the polarizing filter 76 described above, it is also possible to make the driver 100 recognizable by software-based image processing on the image data. As a result, processing equivalent to the automatic adjustment of the polarizing filter 76 can be realized by image processing.

[3. Effect of this embodiment]
As described above, the driving support device 10 according to the present embodiment is a device that supports the driving of the own vehicle 12 (moving body), and the oncoming vehicle 98 and the vehicle 102 (object) around the own vehicle 12 are supported. One of the detection camera 44 (visual sensor), the polarizing filter 76 (76a to 76 g) provided on the camera 44 (camera 44a to 44 g), and the objects existing around the own vehicle 12 detected by the camera 44. One oncoming vehicle 98 is determined to be the adjustment target of the polarizing filter 76, the transmitting member region 106 capable of transmitting light is specified among the determined adjustment targets, and the polarizing filter 76 is automatically adjusted for the specified transmitting member region 106. An automatic adjustment processing unit 70 and a polarizing filter adjustment unit 74 (automatic adjustment unit) are provided.

As a result, the polarizing filter 76 is automatically adjusted with respect to the transmission member region 106 of the oncoming vehicle 98 to be adjusted, so that the internal condition of the oncoming vehicle 98 can be visually recognized through the transmission member region 106. Become. As a result, based on the internal situation of the oncoming vehicle 98 (gesture behavior of the driver 100 of the oncoming vehicle 98), the driving support of the own vehicle 12 (for example, the support for turning left or right in the intersection 90 of the own vehicle 12) is smoothly performed. It becomes possible to do it.

In this case, the oncoming vehicle 98 to be adjusted is another moving body including the transmission member region 106 among the objects existing around the own vehicle 12 (self-moving body) equipped with the driving support device 10. The automatic adjustment processing unit 70 and the polarizing filter adjusting unit 74 request the polarizing filter 76 with respect to the driver riding area 108 in which the camera 44 can detect the driver 100 in the oncoming vehicle 98 in the transmission member area 106. Is automatically adjusted. This makes it possible to easily detect the gesture behavior of the driver 100.

Further, the transparent member area specifying unit 70c of the automatic adjustment processing unit 70 uses the current position (self-position information) of the own vehicle 12 specified by the navigation device 24 provided in the own vehicle 12 or the GPS sensor 52, and the driver boarding area. Estimate 108. As a result, the driver's boarding area 108 can be accurately specified.

Further, the automatic adjustment processing unit 70 and the polarizing filter adjusting unit 74 automatically adjust the polarizing filter 76 so that the camera 44 can detect the situation of the driver 100 via the driver riding area 108. .. As a result, the gesture behavior of the driver 100 can be detected with high accuracy.

In this case, the situation of the driver 100 includes at least a situation in which the driver 100 is visually recognizing the own vehicle 12 and a gesture action intended to permit the driver 100 to perform a predetermined action such as turning left or right of the own vehicle 12. The situation of whether or not the operation is performed is included. The driving support device 10 executes a predetermined action when the driver 100 visually recognizes the own vehicle 12 and confirms that the driver 100 is performing a gesture action based on the detection result of the camera 44. Further prepare. As a result, it is possible to perform appropriate traveling control for the own vehicle 12 according to the gesture behavior.

The driving support device 10 further includes a radar 46 and a LiDAR 48 (detection unit) that detect an object around the own vehicle 12 by using light or millimeter waves. The transmission member region identification unit 70c of the automatic adjustment processing unit 70 identifies the transmission member region 106 based on the detection result of the oncoming vehicle 98 by the radar 46 or LiDAR 48. As a result, the driver 100 can be accurately detected via the transparent member region 106.

In this case, the radar 46 or LiDAR 48 detects the angle θ of the transmission member region 106 with respect to the own vehicle 12. The automatic adjustment processing unit 70 and the polarizing filter adjusting unit 74 automatically adjust the polarizing filter 76 based on the angle θ. As a result, the driver 100 can be detected with high accuracy.

Further, among the objects existing around the own vehicle 12 on which the driving support device 10 is mounted, the oncoming vehicle 98 having the transmission member region 106 is the adjustment target, and the radar 46 or the LiDAR 48 is the oncoming vehicle via the transmission member region 106. When the driver 100 in 98 is detected, the filter adjustment determination unit 70d of the automatic adjustment processing unit 70 determines that the automatic adjustment of the polarizing filter 76 is unnecessary. As a result, the driver 100 can be easily detected even in the transparent member region 106 that has not been cut.

Further, the driving support device 10 further includes a GPS sensor 52 for specifying the current position of the driving support device 10, an illuminance sensor 50, and a clock unit 72. In this case, the filter adjustment determination unit 70d and the polarization filter adjustment unit 74 of the automatic adjustment processing unit 70 measure the current position specified by the GPS sensor 52, the illuminance (illuminance information) detected by the illuminance sensor 50, and the clock unit 72. The polarizing filter 76 is automatically adjusted based on the current time. As a result, the polarizing filter 76 can be accurately adjusted according to the degree of sunlight shining.

The adjustment target of the polarizing filter 76 may be another moving body (oncoming vehicle 98) that exists within a predetermined angle range (imaging region 82) with respect to the camera 44 and is closest to the own vehicle 12. As a result, the driver 100 can be easily and accurately detected.

It should be noted that the present invention is not limited to the above-described embodiment, and it goes without saying that various configurations can be adopted based on the contents described in this specification.

10 ... Driving support device 12 ... Own vehicle (moving body, self-moving body)
44, 44a-44g ... Camera (visual sensor)
70 ... Automatic adjustment processing unit (automatic adjustment unit)
74 ... Polarizing filter adjustment unit (automatic adjustment unit)
76, 76a-76g ... Polarizing filter 98 ... Oncoming vehicle (object, adjustment target, other moving object)
102 ... Vehicle (object) 106 ... Transparent member area

Claims (10)

In a driving support device that supports the driving of a moving object
A visual sensor that detects an object around the moving object,
The polarizing filter provided in the visual sensor and
Among the objects existing around the moving body detected by the visual sensor, one object is determined to be the adjustment target of the polarizing filter, and the transmitting member region capable of transmitting light is specified in the determined adjustment target. An automatic adjustment unit that automatically adjusts the polarizing filter with respect to the specified transmission member region.
A driving support device equipped with. In the driving support device according to claim 1,
The adjustment target is, among the objects existing around the self-moving body on which the driving support device is mounted, another moving body having the transparent member region.
The automatic adjusting unit automatically adjusts the polarizing filter with respect to the driver riding area in which the visual sensor can detect the driver in the other moving body in the transmitting member region. Support device. In the driving support device according to claim 2,
The automatic adjusting unit is a driving support device that estimates the driver's boarding area by using the self-position information of the self-moving body specified by the self-positioning device provided in the self-moving body. In the driving support device according to claim 2 or 3.
The automatic adjustment unit is a driving support device that automatically adjusts the polarizing filter so that the visual sensor can detect the situation of the driver via the driver's riding area. In the driving support device according to claim 4,
The driver's situation includes at least a situation in which the driver is visually recognizing the self-moving body and whether or not the driver is performing a gesture action intended to permit the driver to perform a predetermined action of the self-moving body. Includes the situation and
The driving support device executes the predetermined action when it is confirmed that the driver visually recognizes the self-moving body and performs the gesture action based on the detection result of the visual sensor. A driving support device further equipped with a driving control unit. In the driving support device according to any one of claims 1 to 5,
Further equipped with a detection unit that detects an object around the moving body using light or millimeter waves.
The automatic adjustment unit is a driving support device that identifies the transmission member region based on the detection result of the adjustment target by the detection unit. In the driving support device according to claim 6,
The detection unit detects the angle of the transmission member region with respect to the moving body, and detects the angle of the transmission member region.
The automatic adjustment unit is a driving support device that automatically adjusts the polarizing filter based on the angle. In the driving support device according to claim 6 or 7.
Among the objects existing around the self-moving body on which the driving support device is mounted, the other moving body having the transmissive member region is the adjustment target, and the detection unit is the other through the transmissive member region. A driving support device that determines that automatic adjustment of the polarizing filter is unnecessary when the automatic adjustment unit detects a driver in a moving body. In the driving support device according to any one of claims 1 to 8.
A self-positioning unit that specifies the position of the driving support device as self-position information, an illuminance sensor, and a clock unit are further provided.
The automatic adjustment unit automatically adjusts the polarizing filter based on the self-position information specified by the self-positioning unit, the illuminance information detected by the illuminance sensor, and the current time measured by the clock unit. Driving support device to do. In the driving support device according to any one of claims 1 to 9.
The adjustment target is a driving support device that exists within a predetermined angle range with respect to the visual sensor and is another moving body that is closest to the moving body.

Images