JP2005165422A - Collision probability determination device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a collision probability determination device which adequately determines a probability of colliding with its own vehicle. <P>SOLUTION: The collision probability determination device recognizes objects of pedestrians, bicycles, other vehicles and so on ahead of its own vehicle (S10) and specifies the directions of lines of sight of the recognized pedestrians, drivers of the recognized bicycles, drivers of other recognized vehicles and so on (S20). Then, the above determination device gives the alarm from its own vehicle to arouse the attention of pedestrians, drivers of the bicycles, or drivers of other vehicles, who do not direct their lines of sight toward its own vehicle (S40) when it is determined that the specified lines of sight are not directed toward its own vehicle (S30). <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a collision possibility determination device that determines the possibility of collision with an object existing around a host vehicle.

Conventionally, a vehicle information providing device that provides information related to an obstacle according to the degree of recognition of the driver with respect to the obstacle has been proposed (see, for example, Patent Document 1). According to the vehicle information providing device disclosed in Patent Document 1, for example, a gaze point of a driver of the host vehicle is detected by a gaze point detection sensor, and a front obstacle is detected with respect to an obstacle existing ahead. When the positional relationship with the host vehicle is detected by the sensor, the obstacle is determined depending on whether the obstacle exists at a position corresponding to the visual field range set with the detected gazing point as a reference. Change the mode of information provision.
JP 2001-357498 A

  Since the conventional vehicle information providing apparatus described above selects an obstacle to be provided based on the visual field range of the driver, it also provides information on an object that may not collide with the host vehicle. As a result, unnecessary information is provided to the driver.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a collision possibility determination device that can accurately determine the possibility of collision with the host vehicle.

  The collision possibility determination device according to claim 1 recognizes at least one of a pedestrian, a bicycle, and another vehicle from an imaging unit that captures an image around the host vehicle and an image captured by the imaging unit. Recognizing means, gaze direction identifying means for identifying the gaze direction of at least one of a pedestrian, a bicycle driver, and a driver of another vehicle recognized by the recognizing means, and a gaze direction identified by the gaze direction identifying means Gazing direction determination means for determining whether the vehicle is in the direction in which the host vehicle is present, a pedestrian who has identified the sight line direction based on the determination result of the sight line direction determination means, a bicycle driven by the driver, and driving It is provided with the collision possibility determination means which determines the collision possibility of at least 1 of the other vehicles which a person drives, and the own vehicle.

  When a person recognizes an object visually, he or she looks in the direction that the object exists, but from the perspective of the vehicle where the person's line of sight is directed, the person who is looking at the person recognizes the presence of the vehicle Therefore, it can be determined that the user is looking at him. Then, for example, the direction in which the vehicle travels is changed or the vehicle is made aware of its presence so that it will not contact or collide with the vehicle that has recognized its presence.

  As described above, the present invention focuses on the fact that pedestrians, bicycle drivers, drivers of other vehicles, and the like who are looking at the own vehicle can be assumed to recognize the existence of the own vehicle. That is, it is determined whether the line-of-sight direction of a pedestrian, a bicycle driver, a driver of another vehicle, or the like is a direction in which the host vehicle exists. Thereby, the possibility of collision with the host vehicle can be determined.

  According to the collision possibility determination device according to claim 2, the line-of-sight direction specifying unit includes a feature point extracting unit that extracts a feature point characterizing a human face from an image captured by the imaging unit, and the feature point extracting unit. By identifying the direction of the face of at least one of the pedestrian and the bicycle driver based on the feature points of the face of at least one of the pedestrian and the bicycle driver extracted by the pedestrian, And the direction of the line of sight of at least one of the bicycle drivers.

  In general, the visual field of a person is roughly divided into a region of interest (central vision) when viewed with the eyes and its periphery (peripheral vision), and the cognitive ability differs between the central vision and the peripheral vision. In other words, humans can recognize what they are looking at in central vision, but it is difficult to recognize what they are in peripheral vision, and the tendency becomes stronger as they move away from central vision. . Therefore, when recognizing an object, a person turns his / her face toward the object to be recognized so that the object is located in the central vision. The present invention pays attention to this point, and by specifying the direction in which the face of the pedestrian or bicycle driver faces, the direction of the line of sight of the pedestrian or bicycle driver can be specified.

  According to the collision possibility determination device according to claim 3, the other vehicle is configured so that the other vehicle's line-of-sight direction detecting unit detects the line-of-sight direction of the driver of the other vehicle and the driver's detected by the other vehicle's line-of-sight direction detecting unit. Other vehicle communication means for transmitting information related to the line-of-sight direction to the own vehicle, the own vehicle includes own vehicle communication means for performing communication connection with the other vehicle communication means, and the line-of-sight direction specifying means is a driver of the other vehicle. The sight line direction is specified from information on the sight line direction of the driver of the other vehicle received by the own vehicle communication means. Thereby, it is possible to determine whether or not the driver's line-of-sight direction of the other vehicle is directed toward the host vehicle.

  According to the collision possibility determination device of the fourth aspect, when the collision possibility determination unit determines that the line-of-sight direction specified by the line-of-sight direction determination unit is not the direction in which the host vehicle exists. In addition, it is determined that there is a possibility that the host vehicle collides with at least one of a pedestrian who has specified the line-of-sight direction, a bicycle driven by the driver, and another vehicle driven by the driver. Thereby, a pedestrian, a bicycle, another vehicle, etc. with a possibility of colliding with the own vehicle can be specified.

  According to a fifth aspect of the present invention, there is provided a collision possibility determination device including at least one of a pedestrian whose direction of sight is identified by a collision possibility determination unit, a bicycle driven by the driver, and another vehicle driven by the driver, and the own vehicle. Alarm that generates an alarm for at least one of the pedestrian who specified the line-of-sight direction, the bicycle driven by the driver, and other vehicles driven by the driver. A generation means is provided. For example, by generating a horn of the host vehicle or turning on a headlight of the host vehicle, attention can be directed to the host vehicle with respect to an object that may collide with the host vehicle.

  The collision possibility determination device according to claim 6 is a self-vehicle and at least one of a pedestrian whose sight line direction is specified by a collision possibility determination unit, a bicycle driven by the driver, and another vehicle driven by the driver. When it is determined that there is a possibility that the vehicle will collide with the vehicle, a travel restriction unit that restricts the travel of the host vehicle is provided. Thereby, it is possible to reduce the possibility of collision with a pedestrian who is not looking at the own vehicle, a bicycle driven by the driver, or another vehicle.

  According to the collision possibility determination device of the seventh aspect, the travel restriction means restricts the operation to the open side in the opening / closing operation of the accelerator pedal of the host vehicle. Thereby, the time which approaches the pedestrian who may collide, the bicycle which a driver drives, and other vehicles can be delayed. As a result, it is possible to lengthen the time until the pedestrian or driver who has not turned his or her line of sight toward the own vehicle is made aware of the presence of the own vehicle.

  According to the collision possibility determination device of the eighth aspect, the host vehicle includes automatic braking means that automatically drives the braking device, and the travel restriction means performs automatic braking by the automatic braking means. To do. Thereby, it is possible to further reduce the possibility of collision with a pedestrian who is not looking at the own vehicle, a bicycle driven by the driver, or another vehicle.

  According to the collision possibility determination device according to claim 9, the collision possibility determination device according to any one of claims 1 to 8 is applied to a vehicle driven by using a motor as a power source. And That is, a hybrid vehicle, an electric vehicle, a fuel cell vehicle, and the like that are driven using a motor as a power source have lower noise during traveling than a vehicle that is driven using a conventional engine as a power source. Therefore, it is difficult for pedestrians and bicycle drivers to notice the approach of the vehicle.

  Therefore, by applying the collision possibility determination device of the present invention to a vehicle driven by using a motor as a power source, there is a possibility of collision with a pedestrian who is not looking at the own vehicle, a bicycle driven by the driver, or another vehicle. The presence or absence of can be determined.

  Hereinafter, the collision possibility determination device of the present invention will be described with reference to the drawings. In the present embodiment, an example in which the collision possibility determination device of the present invention is applied as one function of a driving support device that supports driving of a vehicle will be described.

  In FIG. 1, the whole structure of the driving assistance device in this embodiment is shown. As shown in the figure, the driving support device 200 includes a throttle opening sensor 10, a steering sensor 20, a laser radar sensor 30, a yaw rate sensor 40, a vehicle speed sensor 50, a visible camera 62, an infrared camera 64, and a navigation device 70. Each of these components is connected to a computer 80.

  The driving support device 200 further includes a throttle driver 90, a brake driver 100, a steering driver 110, an automatic transmission controller 120, a display device 130, an input device 140, an alarm device 150, and a communication device 160. Each is connected to a computer 80.

  The computer 80 includes an input / output interface (I / O) (not shown) and various drive circuits. Since these hardware configurations are general, a detailed description of the configuration will be omitted. The computer 80 determines the possibility of collision with pedestrians, bicycles, and other vehicles existing around the host vehicle. If it is determined that there is a possibility of collision as a result of the determination, the computer 80 collides. An alarm generation process for generating an alarm for a possible pedestrian or other vehicle is executed.

  Further, the computer 80 drives the throttle driver 90, the brake driver 100, the steering driver 110, the automatic transmission controller 120, and the like based on information from each sensor, and maintains the traveling lane of the host vehicle. Driving support processing for supporting driving operations such as lane keeping running control for running the vehicle, and inter-vehicle distance control for running the vehicle while keeping the inter-vehicle time with the preceding vehicle at an appropriate inter-vehicle time.

  The throttle opening sensor 10 detects the opening of the throttle valve. The detected throttle valve opening signal is sent to the computer 80. The steering sensor 20 detects a change amount of the steering angle of the steering wheel, and a relative steering angle is detected from the value.

  The laser radar sensor 30 irradiates a predetermined range in front of the vehicle with the laser beam, thereby detecting the distance to the reflecting object that reflects the laser beam, the relative speed, the direction of the reflecting object with respect to the host vehicle, and the like. The object information composed of the detection result is converted into an electrical signal and then output to the computer 80. The laser radar sensor 30 detects an object using laser light, but may detect an object around the vehicle using radio waves such as millimeter waves and microwaves, ultrasonic waves, and the like. Good.

  The yaw rate sensor 40 detects an angular velocity around the vertical direction of the vehicle. The vehicle speed sensor 50 is a sensor that detects a signal corresponding to the rotational speed of the wheel. The visible camera 62 is an optical camera that is used as an imaging unit that captures a predetermined range in front of the host vehicle. The visible camera 62 converts the captured image into an electrical signal and outputs the electrical signal to the computer 80.

  The infrared camera 64 is a camera used as photographing means for photographing reflected light of infrared light projected within a predetermined range in front of the host vehicle from an unillustrated infrared projection lamp, and the photographed image is converted into an electrical signal. The data is converted and output to the computer 80. The infrared camera 64 captures the same range as the imaging range of the visible camera 62 described above.

  As is well known, the navigation device 70 executes various functions such as a map display function for displaying a map around the current position of the vehicle, and based on radio waves from a GPS (Global Positioning System) satellite. A GPS receiver for detecting the position (latitude, longitude) and a geomagnetic sensor for detecting the absolute direction of the traveling direction of the host vehicle are provided. Also provided is a map data input device for inputting map data for displaying a map.

  The throttle driver 90, the brake driver 100, the steering driver 110, and the automatic transmission controller 120 are all driven according to instructions from the computer 80. The throttle driver 90 controls the output of the internal combustion engine by adjusting the opening of the throttle valve. The brake driver 100 adjusts the brake pressure, and the steering driver 110 drives the steering by generating rotational torque in the steering. The automatic transmission controller 120 selects the gear position of the automatic transmission necessary for controlling the speed of the vehicle.

  The display device 130 is configured by a liquid crystal display, for example, and is installed near the center console in the vehicle interior. The display device 130 receives map display image data output from the navigation device 70, warning display image data for alerting the driver output from the computer 80, and the like. Display the image corresponding to the data.

  As the input device 140, for example, a touch switch or a mechanical switch integrated with the display device 130 is used, and is used for various inputs such as character input. The alarm device 150 is a device that generates an alarm sound for alerting the driver, and outputs an alarm according to an instruction from the computer 80.

  For example, in the lane keeping traveling control, an alarm is output when the own vehicle departs from the traveling lane. In the inter-vehicle distance control, the own vehicle suddenly approaches the preceding vehicle, and the inter-vehicle distance control is controlled. An alarm is output when the limit is exceeded. Communication device 160 is a mobile communication device that enables mobile communication with other vehicles.

  Next, a control block diagram of the computer 80 is shown in FIG. As shown in the figure, the computer 80 is divided into blocks of an input / output unit 81, an object recognition unit 82, a line-of-sight direction identification unit 83, a collision possibility determination unit 84, and an alarm generation unit 85.

  An input / output unit 81 shown in FIG. 2 inputs signals output from various sensors and the like, and outputs an output signal processed in the computer 80. The object recognition unit 82 performs predetermined image processing on a photographed image photographed by the visible camera 62 or the infrared camera 64, and characterizes a pedestrian, a bicycle, or a vehicle from the image data for each pixel subjected to the image processing. A feature extraction process for extracting feature points (amounts) is performed to recognize objects such as pedestrians, bicycles, and other vehicles in front of the host vehicle. For example, when an image in front of the host vehicle as shown in FIG. 3 is taken, pedestrians P1, P2, P3, bicycle B, and other vehicle V in the image are recognized.

  When recognizing objects such as pedestrians, bicycles, and other vehicles, feature amounts such as aspect ratio and shape are extracted from the captured image of the visible camera 62, and the captured image of the infrared camera 64 is extracted. In this case, feature quantities such as temperature level and temperature distribution are extracted. Then, based on the feature values extracted from each image, it is recognized whether the object is an object corresponding to a pedestrian, a bicycle, or another vehicle.

  The line-of-sight direction specifying unit 83 specifies the line-of-sight direction of a pedestrian recognized by the object recognition unit 82, the line-of-sight direction of a bicycle driver, and the line-of-sight direction of a driver of another vehicle. Here, with regard to the line-of-sight directions of pedestrians and bicycle drivers, pedestrians and drivers are obtained from image data for each pixel that has been image-processed with respect to the images captured by the visible camera 62 and the infrared camera 64. The face image portion is extracted, the positions of a plurality of feature points (for example, the eyes, corners of the eyes, nostrils, etc.) included in the extracted face image portion are detected, and the face on the image is determined based on the detected positions. Find the direction you are headed. Then, the obtained face direction is specified as the line-of-sight direction.

  That is, as described above, in general, the human visual field is roughly divided into an area of interest (central vision) when viewed with the eyes and its peripheral area (peripheral vision). Cognitive ability is different. In other words, humans can recognize what they are looking at in central vision, but it is difficult to recognize what they are in peripheral vision, and the tendency becomes stronger as they move away from central vision. .

  Therefore, when recognizing an object, a person turns his / her face toward the object to be recognized so that the object is located in the central vision. The present invention pays attention to this point and specifies the line-of-sight direction from the direction in which the face of the pedestrian faces. Thereby, the sight line direction of the pedestrian and the driver who drives the bicycle can be specified. Note that the direction of the face on the image is converted into an absolute direction by performing coordinate conversion or the like so that the position (latitude and longitude) where the host vehicle exists can be associated.

  In addition, for the driver's line-of-sight direction of the other vehicle, a transmission request signal is transmitted to the other vehicle around the host vehicle via the communication device 160 to transmit information on the driver's line-of-sight direction to the host vehicle. Information on the line-of-sight direction of the driver of the vehicle is acquired from the other vehicle that has received the transmission request signal. The other vehicle that has received the transmission request signal from the host vehicle acquires the driver's current line-of-sight direction and converts it to an absolute direction. And the information regarding the gaze direction converted into this absolute azimuth | direction is transmitted with respect to the own vehicle which transmitted the transmission request signal. Thereby, the gaze direction of the driver of the other vehicle can be acquired in the own vehicle.

  In other vehicles, it is necessary to include means for detecting the driver's line-of-sight direction and communication means for transmitting information on the detected line-of-sight direction to the host vehicle. For example, a gazing point detection sensor, an infrared floodlight lamp, an infrared region imaging camera, or the like disclosed in JP 2001-357498 A can be used.

  The collision possibility determination unit 84 uses the line-of-sight directions of the pedestrian, the bicycle driver, the driver of another vehicle, etc. from the line-of-sight direction specifying unit 83, and the pedestrian, bicycle, other vehicle, etc. existing in front of the host vehicle. And whether there is a possibility of collision with the host vehicle. Here, when it is determined that the line-of-sight direction of a pedestrian, a bicycle driver, a driver of another vehicle, etc. is not the direction in which the own vehicle exists, the driver who does not turn his / her line of sight in the direction in which the own vehicle exists It is determined that there is a possibility of collision between a bicycle or other vehicle driven by the vehicle, or a pedestrian and the host vehicle.

  That is, as described above, when a person visually recognizes an object, the person turns his / her line of sight in the direction in which the object exists. It can be determined that the user is looking in order to recognize the presence of the vehicle. Then, for example, the direction in which the vehicle travels is changed or the vehicle is made aware of its presence so that it will not contact or collide with the vehicle that has recognized its presence.

  As described above, the present invention focuses on the fact that pedestrians, bicycle drivers, drivers of other vehicles, and the like who are looking at the own vehicle can be assumed to recognize the existence of the own vehicle. That is, whether or not there is a possibility of collision with the own vehicle by determining whether or not the line of sight of the pedestrian, the driver of the bicycle, the driver of the other vehicle, etc. is the direction in which the own vehicle exists. judge. Note that the determination result of the possibility of collision is sent to the alarm generation unit 85.

  2 receives a determination result that there is a possibility of a collision with a pedestrian, a bicycle, or another vehicle by the collision possibility determination unit 84, the warning generation unit 85 warns these objects. Is generated. For example, the horn of the own vehicle is generated or the headlight of the own vehicle is turned on. Thereby, attention can be directed to the own vehicle with respect to an object such as a pedestrian, a bicycle, or another vehicle that may collide with the own vehicle.

  Next, alarm generation processing by the driving support apparatus 200 related to the characteristic part of the present embodiment will be described with reference to the flowchart shown in FIG. First, in step (hereinafter referred to as “S”) 10 shown in FIG. 7, an object such as a pedestrian, bicycle, or other vehicle existing in front of the host vehicle is recognized. In S20, the line-of-sight directions of the pedestrian, the bicycle driver, the driver of another vehicle, etc. recognized in S10 are specified.

  In S30, it is determined whether or not the specified line-of-sight direction is directed in the direction in which the host vehicle exists. If the determination is affirmative, the process proceeds to S10 and the above-described process is repeated. On the other hand, if a negative determination is made, it is determined that there is a possibility that the bicycle or other vehicle driven by the driver who does not turn his / her line of sight in the direction where the own vehicle exists, or the pedestrian and the own vehicle may collide. Then, the process proceeds to S40.

  In S40, an alarm is generated from the host vehicle to alert a pedestrian, a bicycle driver, or a driver of another vehicle who is not looking in the direction in which the host vehicle is located.

  As described above, the driving support device 200 according to the present embodiment specifies the line-of-sight direction of a pedestrian, a driver of a light vehicle such as a bicycle, and the driver of another vehicle existing around the own vehicle, and the line-of-sight direction is the own vehicle. If it is not directed in the direction in which the vehicle exists, it is determined that there is a possibility of collision with a bicycle or other vehicle or a pedestrian driven by a driver who does not turn his / her line of sight toward the host vehicle. Thereby, the possibility of collision with the host vehicle can be accurately determined.

(Modification 1)
In this embodiment, in S30 shown in FIG. 4, there is a possibility that a bicycle or other vehicle driven by a driver who does not turn his / her line of sight in the direction in which the host vehicle is located, or a pedestrian and the host vehicle collide. When the determination is made, a process for generating an alarm is executed in S40, but a limit may be imposed on the traveling of the host vehicle.

  For example, in S50 shown in FIG. 5, the opening / closing operation of the accelerator pedal of the host vehicle may be restricted so that the operation to the opening side becomes invalid. Or you may make it add a restriction | limiting so that the own vehicle may not accelerate. Thereby, the time which approaches the pedestrian who may collide, the bicycle which a driver drives, and other vehicles can be delayed. As a result, it is possible to lengthen the time until the pedestrian or driver who is not looking at the own vehicle notices the presence of the own vehicle. Note that the processing of S10 to S40 shown in FIG. 5 is the same as the processing described in the present embodiment, and thus the description thereof is omitted.

(Modification 2)
As a means for limiting the travel of the host vehicle, for example, the brake driver 100 may be driven to perform automatic braking. Thereby, it is possible to reduce the possibility of collision with a pedestrian who is not looking at the own vehicle, a bicycle driven by the driver, or another vehicle.

(Modification 3)
The driving support device 200 of the present embodiment may be applied to a vehicle that drives using a motor as a power source. That is, a hybrid vehicle, an electric vehicle, a fuel cell vehicle, and the like that drive using a motor as a power source have lower noise during traveling than a vehicle that uses a conventional engine as a power source. Therefore, it is difficult for pedestrians and bicycle drivers to notice the approach of the vehicle.

  Therefore, by applying the driving support device 200 of the present embodiment to a vehicle that drives using a motor as a power source, there is a possibility of collision with a pedestrian who is not looking at the own vehicle, a bicycle driven by the driver, or another vehicle. The presence or absence of can be determined.

It is a block diagram which shows the whole structure of the driving assistance apparatus 200 concerning embodiment of this invention. It is a control block diagram of the computer 80 concerning the embodiment of the present invention. It is an image figure of the image which image | photographed the surroundings of the own vehicle based on embodiment of this invention. It is a flowchart which shows the flow of the alarm generation process concerning embodiment of this invention. It is a flowchart which shows the flow of the alarm generation process concerning the modification 1 of embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 81 Input / output part 82 Object recognition part 83 Collision determination target extraction part 84 Gaze direction specific part 85 Collision possibility judgment part 86 Alarm generation part 200 Driving support device

Claims (9)

Imaging means for capturing an image of the surroundings of the host vehicle;
Recognizing means for recognizing at least one of a pedestrian, a bicycle, and another vehicle from an image taken by the imaging means;
Gaze direction identifying means for identifying the gaze direction of at least one of a pedestrian, a bicycle driver, and a driver of another vehicle recognized by the recognition means;
Gaze direction determining means for determining whether or not the gaze direction specified by the gaze direction specifying means is a direction in which the host vehicle exists;
Based on the determination result of the line-of-sight direction determination means, the possibility of a collision between the own vehicle and at least one of a pedestrian who has specified the line-of-sight direction, a bicycle driven by the driver, and another vehicle driven by the driver. A collision possibility judging device comprising: a collision possibility judging means for judging.   The line-of-sight direction specifying unit includes a feature point extracting unit that extracts a feature point that characterizes a human face from an image captured by the imaging unit, and the pedestrian extracted by the feature point extracting unit and driving the bicycle. The direction of the face of at least one of the pedestrian and the driver of the bicycle is specified based on the feature points of the face of at least one of the person, so that the pedestrian and the driver of the bicycle The collision possibility determination device according to claim 1, wherein a gaze direction of at least one person is specified. The other vehicle is
Other vehicle gaze direction detection means for detecting the gaze direction of the driver of the other vehicle;
And other vehicle communication means for transmitting information related to the driver's line-of-sight direction detected by the other-vehicle line-of-sight detection means to the host vehicle,
The vehicle is
Own vehicle communication means for performing communication connection with the other vehicle communication means,
The line-of-sight direction specifying means specifies the line-of-sight direction of the driver of the other vehicle from information on the line-of-sight direction of the driver of the other vehicle received by the own vehicle communication means. The collision possibility determination device described.   The collision possibility determining means, when the line-of-sight direction determining means determines that the line-of-sight direction specified by the line-of-sight direction specifying means is not the direction in which the host vehicle exists, the pedestrian who specified the line-of-sight direction, 4. The vehicle according to claim 1, wherein the vehicle is determined to collide with at least one of a bicycle driven by the driver and another vehicle driven by the driver. 5. The collision possibility determination device described.   It is determined by the collision possibility determination means that there is a possibility that the host vehicle collides with at least one of a pedestrian who has specified the line-of-sight direction, a bicycle driven by the driver, and another vehicle driven by the driver. In this case, it is provided with a warning generation means for generating a warning for at least one of the pedestrian who specified the line-of-sight direction, the bicycle driven by the driver, and the other vehicle driven by the driver. The collision possibility determination device according to claim 4.   It is determined by the collision possibility determination means that there is a possibility that the host vehicle collides with at least one of a pedestrian who has identified the line-of-sight direction, a bicycle driven by the driver, and another vehicle driven by the driver. 6. The collision possibility determination device according to claim 4, further comprising travel restriction means for restricting travel of the host vehicle.   The collision possibility determination device according to claim 6, wherein the travel restriction unit restricts an operation toward an open side in an opening / closing operation of an accelerator pedal of the host vehicle. The host vehicle includes automatic braking means for automatically driving a braking device,
7. The collision possibility determination device according to claim 6, wherein the travel restriction unit performs automatic braking by the automatic braking unit.   The collision possibility determination device according to any one of claims 1 to 8 is applied to a vehicle driven by using a motor as a power source.

Images