FCW device for leaning vehicle and FCW method for leaning vehicle

Abstract

This FCW device for a leaning vehicle is provided with, for the purpose of detecting a peripheral object and issuing an alarm while the leaning vehicle is turning: an alarm unit for issuing an alarm to a driver; and one or a plurality of control units for executing subject vehicle course prediction for predicting the course of a subject vehicle and subject vehicle peripheral object position detection for detecting the position of an object around the subject vehicle, calculating a possibility of collision between the subject vehicle and the object by using the result of the subject vehicle course prediction and the result of the subject vehicle peripheral object position detection, and executing alarm control for controlling whether to issue an alarm by the alarm unit on the basis of the possibility of collision. The FCW device for a leaning vehicle is used in a leaning vehicle tilting inward during turning. The FCW device for a leaning vehicle is further provided with: a monocular camera that is fixed to the subject vehicle and captures an image of the surroundings of the subject vehicle; and a turning state detection unit that detects a physical quantity relating to the yaw rate and roll rate of the subject vehicle. The control unit executes the subject vehicle peripheral object position detection on the basis of the image captured by the monocular camera, and executes the subject vehicle course prediction and / or the subject vehicle peripheral object position detection in consideration of the yaw rate and roll rate obtained from the detection result of the turning state detection unit, and uses the obtained result to execute alarm control as described in each of the following situations. Situation 1: alarm control is executed so as to issue an alarm if the possibility of collision between the subject vehicle and a first preceding vehicle becomes equal to or more than a predetermined collision possibility when the first preceding vehicle is turning in front of the subject vehicle at a vehicle speed smaller than that of the subject vehicle with the same turning radius as that of the subject vehicle while the subject vehicle is turning on a road without lane markings. Situation 2: alarm control is executed so as not to issue an alarm when a second preceding vehicle is turning in front of the subject vehicle at the same vehicle speed as that of the subject vehicle with the same turning radius as that of the subject vehicle while the subject vehicle is turning on a road without lane markings, and a third preceding vehicle is turning in front of the subject vehicle inside the turn of the second preceding vehicle at a vehicle speed smaller than that of the subject vehicle with a turning radius smaller than that of the subject vehicle.

Description

FCW device for lean vehicle and FCW method for lean vehicle

[0001] The present disclosure relates to a lean vehicle FCW device and a lean vehicle FCW method that issue a forward collision warning (FCW) to a driver of a lean vehicle.

[0002] Conventionally, FCW devices are known that issue a warning to a driver of a lean-to vehicle based on the possibility of a collision with a surrounding object. A lean-to vehicle turns by tilting its body. An example of a lean-to vehicle is a motorcycle. For example, Patent Document 1 discloses a device that issues a warning based on the positional relationship between a lean-to vehicle and a preceding vehicle traveling in front of the lean-to vehicle when multiple lean-to vehicles are traveling together. Patent Document 2 discloses a system that captures an image of the front using a camera attached to the lean-to vehicle, detects surrounding vehicles using the captured image, and issues a warning.

[0003] International Publication No. 2021 / 060357 Special Publication No. 2021-526681

[0004] The techniques disclosed in Patent Documents 1 and 2 do not take into consideration the operation of the FCW while a lean vehicle is turning.

[0005] An object of the present disclosure is to provide an FCW device for a lean vehicle and an FCW method for a lean vehicle that can detect surrounding objects and issue an alarm even while the lean vehicle is turning.

[0006] The present inventors have considered activating an FCW device while a lean-angle vehicle is turning. Warning control by the FCW device is performed using captured images of the vehicle's surroundings captured by an onboard camera fixed to the vehicle. A lean-angle vehicle turns with its body tilted. If the body tilts, the image captured by the onboard camera fixed to the vehicle also tilts. The body tilt of a lean-angle vehicle during a turn tends to be greater than the body tilt of a four-wheeled vehicle turning the same path. Therefore, the tilt of an image captured by a lean-angle vehicle also tends to be greater than the tilt of an image captured by a four-wheeled vehicle. The present inventors have discovered that when activating an FCW device while a lean-angle vehicle is turning, the influence of the tilt must be taken into consideration.

[0007] The FCW device executes various processes from the time an object is detected until an alert is issued to the driver. For example, the process executes processes such as detecting objects around the vehicle, selecting from the detected objects an object that is at risk of colliding with the vehicle, calculating the time to collision (TTC) between the selected object and the vehicle, determining whether to issue an alert based on the time to collision, and issuing an alert based on the determination result. The body of a lean vehicle is smaller than the body of a four-wheeled vehicle, and the space for installing the device is limited. For this reason, it is necessary to reduce the processing load of the processes executed by the FCW device and increase the design flexibility of hardware resources.

[0008] Thus, when an FCW device is activated while a lean vehicle is turning, it is necessary to suppress an increase in processing load while taking into consideration the effect of the lean vehicle's tilt. The inventors of the present application have discovered that it is possible to address both of the conflicting issues described above by configuring the FCW device to acquire captured images using a monocular camera and to detect vehicles that are the target of an alarm and issue an alarm without detecting white lines to recognize lane lines.

[0009] The FCW device for a lean vehicle according to the present disclosure employs the following configuration in order to achieve both of the above-mentioned issues related to lean vehicles.

[0010] An FCW device for a lean vehicle according to one embodiment of the present disclosure includes an alarm unit that issues an alarm to a driver, and one or more control units that perform a host vehicle path prediction that predicts a path of the host vehicle and a host vehicle peripheral object position detection that detects the positions of objects around the host vehicle, calculate a possibility of a collision between the host vehicle and an object using the results of the host vehicle path prediction and the results of the host vehicle peripheral object position detection, and perform alarm control that controls whether or not an alarm is issued by the alarm unit based on the possibility of collision. The FCW device for a lean vehicle is used in a lean vehicle that leans to the inside of a turn when turning, and further includes a monocular camera that is fixed to the host vehicle and captures images of the area around the host vehicle, and a turning state detection unit that detects physical quantities related to the yaw rate and roll rate of the host vehicle, and the control unit performs the host vehicle path prediction and / or the host vehicle peripheral object position detection based on the obtained results, and performs the alarm control as described in each of the following situations. Situation 1: When the host vehicle is turning on a road without white lines and a first leading vehicle is turning ahead of the host vehicle at a speed slower than the host vehicle and with the same turning radius as the host vehicle, if the possibility of a collision between the host vehicle and the first leading vehicle reaches or exceeds a predetermined collision possibility, the control unit executes the warning control so that the warning unit issues a warning. Situation 2: When the host vehicle is turning on a road without white lines and a second leading vehicle is turning ahead of the host vehicle at the same speed and with the same turning radius as the host vehicle, and a third leading vehicle is turning ahead of the host vehicle on the inside of the turning of the second vehicle at a speed slower than the host vehicle and with a smaller turning radius than the host vehicle, the control unit executes the warning control so that the warning unit does not issue a warning.

[0011] In the above configuration, the FCW device for a lean vehicle may further include a vehicle speed detection unit that detects a vehicle speed of the host vehicle. In the host vehicle path prediction, the control unit may determine whether the host vehicle is turning based on a yaw rate obtained from the detection result of the turning state detection unit, and if it is determined that the host vehicle is turning, predict the host vehicle path based on the yaw rate obtained from the detection result of the turning state detection unit and the vehicle speed detected by the vehicle speed detection unit, and may perform the host vehicle path prediction taking into account the roll rate obtained from the detection result of the turning state detection unit, or perform the position detection of the object around the host vehicle taking into account the roll rate.

[0012] In the above configuration, the control unit may perform a prediction of the vehicle's course based on the yaw rate and vehicle speed of the turning vehicle to obtain a predicted course having a curved shape, and may perform any one of correcting the predicted course on the captured image, correcting the position of an object detected on the captured image, and rotating the captured image based on the roll rate.

[0013] An FCW method for a lean vehicle according to one embodiment of the present disclosure includes a warning unit that issues a warning to a driver, and one or more control units that perform host vehicle path prediction to predict a host vehicle path and host vehicle peripheral object position detection to detect the positions of objects around the host vehicle, calculate a possibility of a collision between the host vehicle and an object using the results of the host vehicle path prediction and the results of the host vehicle peripheral object position detection, and perform warning control to control whether or not to issue a warning by the warning unit based on the possibility of collision, and is an FCW method for a lean vehicle that leans to the inside of a turn when turning, the FCW method being executed by an FCW device used in a lean vehicle that leans to the inside of a turn, and the host vehicle is turning on a road without white lines, and a first leading vehicle is turning ahead of the host vehicle at a speed slower than the host vehicle, with the same turning radius as the host vehicle, and when the likelihood of a collision between the host vehicle and the first leading vehicle reaches or exceeds a predetermined likelihood of collision, the control unit executes the warning control so that the warning unit issues a warning. Situation 2: When the host vehicle is turning on a road without white lines, and a second preceding vehicle is turning in front of the host vehicle at the same speed and with the same turning radius as the host vehicle, and a third preceding vehicle is turning on the inside of the turning radius of the second vehicle in front of the host vehicle at a speed slower than the host vehicle and with a smaller turning radius than the host vehicle, the control unit executes the warning control so that the warning unit does not issue a warning.

[0014] In the above configuration, in the host vehicle path prediction, it is determined whether or not the host vehicle is turning based on the yaw rate obtained from the detection result of the turning state detection unit, and if it is determined that the host vehicle is turning, the path of the host vehicle is predicted based on the yaw rate obtained from the detection result of the turning state detection unit and the vehicle speed detected by a vehicle speed detection unit that detects the vehicle speed of the host vehicle, and in the host vehicle path prediction, a path prediction of the host vehicle is performed taking into account the roll rate obtained from the detection result of the turning state detection unit, or in the position detection of an object around the host vehicle, position detection of the object taking into account the roll rate may be performed.

[0015] The above configuration may include a step of predicting the vehicle's course based on the yaw rate and the vehicle speed to obtain a predicted course having a curved shape for the turning vehicle, and a step of correcting the predicted course on the captured image, correcting the position of an object detected on the captured image, or rotating the captured image based on the roll rate.

[0016] According to the FCW device for a lean vehicle and the FCW method for a lean vehicle of the present disclosure, it is possible to detect surrounding objects and issue an alarm even while the lean vehicle is turning.

[0017] [Lean Vehicle] A lean vehicle according to the present disclosure is a vehicle whose body leans to the right when turning right and to the left when turning left. The left and right directions are those seen from the driver of the lean vehicle. In this disclosure, the right-hand side, left-hand side, front, and rear, as seen from the driver of the lean vehicle, are referred to as the right, the left, the front, and the rear, respectively. The term "host vehicle" used in this disclosure refers to a lean vehicle. Lean vehicles include motorcycles. The tilt of a lean vehicle refers to the body of the lean vehicle being tilted to the left or right with respect to the vertical direction. The tilt of a lean vehicle can be expressed by the tilt angle between the body of the lean vehicle and the vertical direction. When the lean vehicle is traveling straight, the tilt angle is 0 degrees, and when the lean vehicle is turning, the tilt angle is greater than 0 degrees. A lean vehicle includes at least one front wheel and at least one rear wheel. For example, a lean vehicle may have two front wheels and one or two rear wheels, or one front wheel and one or two rear wheels. The lean vehicle may have steered front wheels or steered rear wheels. The lean vehicle runs by driving the drive wheels with a drive source. The drive source may be an internal combustion engine, an electric motor, or a hybrid drive source including an engine and an electric motor.

[0018] [FCW Device for Lean Vehicle] The FCW device for a lean vehicle described in the present disclosure includes an alarm unit, one or more control units, a monocular camera, and a turning state detection unit. The FCW device may further include a vehicle speed detection unit. The FCW device is mounted on a lean vehicle for use.

[0019] [White Lines] The white lines described in this disclosure refer to lines drawn on roads to indicate the vehicle's travel lane, and are not intended to be limited to a specific color. For example, the lines may be drawn in other colors, such as yellow. Such road lines may also be referred to as roadside lines, shoulder strips, diagonal boundary lines, or dividing lines. FCW devices can perform FCW control without detecting white lines. In this disclosure, when a vehicle is described as traveling on a road without white lines, this indicates that FCW control can be performed even on roads without white lines, and is not intended to limit roads on which lean vehicles travel to roads without white lines.

[0020] [Situation] In this disclosure, situations described with numbers, such as "Situation 1," indicate examples of situations in which the operation of the FCW device can be confirmed. One of the characteristics of the FCW device is that it can perform processes such as detecting the position of objects around the vehicle and predicting the vehicle's path, and can change whether or not to issue an alarm depending on the situation. Each situation described with a number indicates that the alarm control performed by the FCW device can be confirmed by recreating this situation, and it is not intended to limit the operation of the FCW device to the described situation.

[0021] [Warning Unit] The warning unit described in the present disclosure is a warning device that issues a warning to at least the driver of a lean vehicle. The warning unit issues a warning targeting at least one of the senses of hearing, seeing, and touch. A warning targeting the sense of hearing may be issued by emitting a voice or warning sound from a speaker. A warning targeting the sense of sight may be issued by displaying information on a display device or illuminating a light-emitting device. Warning devices such as speakers, display devices, and light-emitting devices may be installed, for example, on the handlebars or in their vicinity. If the driver wears a helmet, the warning device may be mounted on the helmet. Devices such as a speedometer and a navigation device provided in the lean vehicle may be used as warning devices. A warning targeting the sense of touch may be issued by vibrating at least one of the handlebars, helmet, and seat.

[0022] [Control Unit] The control unit described in the present disclosure controls each unit constituting the FCW device. The control unit controls each unit to realize the functions and operations of the FCW device described in the present disclosure. The control unit is a control device that executes FCW control. FCW control includes host vehicle path prediction, host vehicle peripheral object position detection, and warning control. The FCW device may include multiple control units, and multiple processes may be executed in parallel by separate control units. For example, a process related to host vehicle path prediction and a process related to host vehicle peripheral object position detection may be executed in parallel simultaneously. The control unit calculates a numerical value indicating the likelihood of a collision between the host vehicle and the object based on the host vehicle path prediction result and the host vehicle peripheral object position detection result. For example, the time to collision (TTC) may be used as the numerical value indicating the likelihood of a collision. Furthermore, the numerical value indicating the likelihood of a collision may be, for example, the distance between the leaning vehicle and the object or the relative movement speed. The control unit determines whether an alarm is necessary and when to issue the alarm based on the possibility of a collision, and executes alarm control to issue an alarm from the alarm device based on the determination result. The control unit can execute alarm control continuously while the lean vehicle is traveling, regardless of whether the lean vehicle is traveling straight or turning. A program corresponding to the control unit is prepared in advance in a non-volatile storage device, and the function and operation of the control unit are realized by executing this program by predetermined hardware. The control unit may be realized by an ECU (Electronic Control Unit) for vehicle control, or the FCW device may have a dedicated control unit.

[0023] [Monocular Camera] The monocular camera described in this disclosure is an imaging device that captures images of the periphery of a lean vehicle. The FCW device can perform FCW control using a single monocular camera, which is smaller than a stereo camera. The monocular camera is fixed to the lean vehicle and used. The monocular camera can capture images of the area ahead of the road on which the lean vehicle is traveling. The monocular camera captures images of leading vehicles traveling ahead of the lean vehicle while the lean vehicle is traveling straight or turning, vehicles stopped or parked in front of the lean vehicle, and other obstacles. The leading vehicles captured by the monocular camera include vehicles traveling on a predicted path obtained by host vehicle path prediction and vehicles traveling at positions outside the predicted path. The monocular camera captures moving images or continuously captures still images, thereby continuously capturing images of objects around the lean vehicle.

[0024] [Turning State Detection Unit] The turning state detection unit described in the present disclosure is a detection device that detects a physical quantity related to the yaw rate of a lean vehicle. The turning state detection unit can also detect a physical quantity related to the roll rate of a lean vehicle. The yaw rate, yaw angle, roll rate, and roll angle of the lean vehicle can be determined from the physical quantities detected by the turning state detection unit. For example, the turning state detection unit may detect either the yaw rate or the yaw angle, and calculate the other based on the detection result. The turning state detection unit may detect either the roll rate or the roll angle, and calculate the other based on the detection result. For example, an IMU (Inertial Measurement Unit) may be used as the turning state detection unit. Other devices such as a GPS (Global Positioning System) or a gyro sensor may be used as the turning state detection unit.

[0025] [Vehicle Speed ​​Detector] The vehicle speed detector described in the present disclosure is a detector for detecting the traveling speed of a lean vehicle. For example, a device that detects the vehicle speed based on the rotation speed of a front wheel or a rear wheel and the outer diameter of a tire acquired using a rotation sensor is used as the vehicle speed detector. Other devices, such as a device that detects the vehicle speed from position information obtained by GPS, may also be used as the vehicle speed detector.

[0026] [Vehicle Path Prediction] The vehicle path prediction described in the present disclosure is a process for predicting the path of the vehicle. Vehicle path prediction also includes predicting the path on an image captured by a monocular camera. Note that predicting the path on the captured image may be performed by mapping the predicted path of the vehicle in three-dimensional coordinates in the real world onto corresponding planar coordinates on the captured image. Planar coordinates on the captured image are, for example, an orthogonal coordinate system with the center of the captured image as the origin. When the lean vehicle is traveling straight, the FCW device can predict the path of the lean vehicle on the image captured by the monocular camera based on the mounting position, mounting angle, imaging direction, imaging range, etc. of the monocular camera on the lean vehicle. When the lean vehicle is turning, the FCW device can predict a curved path as the path of the turning lean vehicle based on the detection result by the turning state detection unit. When the lean vehicle is turning, the FCW device can also predict the path of the turning lean vehicle by curvature-correcting the predicted path on the captured image, which is obtained assuming that the lean vehicle is traveling straight, based on the detection results by the turning state detection unit.The FCW device can predict the path of the turning lean vehicle in approximately real time after acquiring an image using the monocular camera while the lean vehicle is traveling.

[0027] [Position Detection of Objects Surrounding the Vehicle] The position detection of objects surrounding the vehicle described in the present disclosure is a process of detecting objects around the vehicle and identifying the position of each object. The position detection of objects around the vehicle is performed using images captured by a monocular camera. Here, the position of an object may be a position on the image captured by the monocular camera. By performing image analysis to detect objects on the image captured by the monocular camera, the objects appearing in the captured image and the positions of each object are detected. Such image analysis is sometimes referred to as object detection, object recognition, etc. Objects detected from the captured image include a preceding vehicle traveling ahead of the lean-back vehicle while the lean-back vehicle is traveling straight or turning, a stopped or parked vehicle ahead, and other obstacles. The preceding vehicle detected from the captured image includes a vehicle traveling on a predicted path obtained by vehicle path prediction and a vehicle traveling at a position outside the predicted path. When a detected vehicle is referred to as a "leading vehicle," it refers to a vehicle traveling ahead of the lean-back vehicle in the same direction as the lean-back vehicle. The FCW device can detect the positions of objects around the vehicle in approximately real time after capturing an image using a monocular camera while the vehicle is traveling lean.

[0028] [Collision Possibility] The collision probability described in the present disclosure is a numerical value indicating the possibility of a collision between a lean vehicle and a surrounding object. For example, the relative movement speed of the object relative to the lean vehicle, the distance between the lean vehicle and the object, and the time until the lean vehicle collides with the object can be used as the collision probability. The collision probability is calculated based on the predicted path of the host vehicle obtained by performing host vehicle path prediction, the position of each object obtained by performing position detection of objects around the host vehicle, and the relative movement between each detected object and the lean vehicle. For example, the position detection of objects around the host vehicle determines the distance from the detected object to the lean vehicle in the captured image. The direction and movement speed of the detected object relative to the lean vehicle are determined by comparing multiple consecutively captured images. The time until collision is determined based on the distance from the object to the lean vehicle and the relative movement speed. Alarm control is performed to determine whether and when an alarm should be issued by the alarm unit based on the result of comparing the value indicating the collision probability with a threshold value prepared in advance according to the type of collision probability. For example, when the time to collision (TTC) is used as the collision probability, a predetermined time is set as a threshold, and when the time to collision reaches the threshold time, alarm control is performed to issue an alarm.

[0029] [Hardware Configuration] The FCW device according to the present disclosure is characterized by performing FCW control that has not been performed before. The hardware configuration of the FCW device according to the present disclosure and a lean vehicle equipped with the FCW device can be implemented by a person skilled in the art using the hardware configurations of devices and lean vehicles known in prior art documents, other publicly known documents, and conventional technology. Therefore, a detailed description of the hardware configuration of the FCW device and the lean vehicle will be omitted.

[0030] [Others] In this disclosure, when a process is described as being performed "based on A" or "taking A into consideration," the process is performed using A, but the information used is not limited to A, and information other than A may also be used. In this disclosure, when "at least one" is used in conjunction with an option, this means that all possible combinations of multiple options are included. In this disclosure, when "including," "comprising," "having," and their derivatives are used, this means that other content may be included in addition to the listed content. In this disclosure, when "may" or "may" is used, this means that this is a non-exclusive example and is not limited to the described example. In this disclosure, unless the number of components is clearly specified, the number of components may be one or more. Unless otherwise defined, the terms used in this disclosure have the general meaning understood by those skilled in the art of FCW control. The FCW device according to the present disclosure is not limited to the configurations described in the following embodiments or the configurations shown in the drawings. The FCW device according to the present disclosure can be implemented by combining multiple embodiments described below. The FCW device according to the present disclosure can also be implemented in forms other than the embodiments described below, or by modifying the embodiments described below.

[0031] FIG. 1 is a diagram for explaining an overview of an FCW device for a lean vehicle according to a first embodiment of the present disclosure. FIG. 2 is a diagram for explaining an overview of an FCW device for a lean vehicle according to a second embodiment of the present disclosure. FIG. 3 is a diagram for explaining an example of host vehicle path prediction taking into account turning. FIG. 4 is a diagram for explaining an example of host vehicle path prediction taking into account turning and roll rate. FIG. 5 is a diagram for explaining an example of position detection of an object around the host vehicle taking into account roll rate. FIG. 6 is a diagram for explaining another example of position detection of an object around the host vehicle taking into account roll rate.

[0032] [First Embodiment] An FCW device for a lean vehicle according to a first embodiment of the present disclosure will be described with reference to Fig. 1. As shown in Fig. 1, an FCW device 2 for a lean vehicle is provided on a lean vehicle 1 that turns by tilting the vehicle body toward the inside of the turn. The tilt angle θ shown in Fig. 1 indicates the tilt angle of the lean vehicle 1 with respect to the vertical direction. When the lean vehicle 1 is traveling straight, the tilt angle θ is 0 degrees, and when turning, the tilt angle θ is greater than 0 degrees.

[0033] The FCW device 2 includes an alarm unit 21 , one or more control units 22 , a monocular camera 23 , and a turning state detection unit 24 .

[0034] The warning unit 21 issues a warning to the driver of the lean vehicle 1. The control unit 22 predicts the path of the lean vehicle 1, i.e., the host vehicle, and detects the positions of objects around the host vehicle. The control unit 22 calculates the possibility of a collision using the results of the host vehicle path prediction and the results of the position detection of objects around the host vehicle. The control unit 22 executes warning control, which controls whether or not the warning unit 21 issues a warning and the timing of issuing the warning, based on the possibility of a collision. The control unit 22 determines whether or not a warning is necessary based on the possibility of a collision, and controls the warning unit 21 so that a warning is issued from the warning unit 21 as necessary.

[0035] As shown within the dashed line frame in FIG. 1, the FCW device 2 executes FCW control including vehicle path prediction, position detection of objects around the vehicle, and warning control.

[0036] A monocular camera 23 fixed to the body of the lean vehicle 1 captures images of the vehicle periphery including the area ahead of the lean vehicle 1. The control unit 22 acquires the captured image captured by the monocular camera 23 (A). The turning state detection unit 24 detects physical quantities related to the yaw rate and roll rate of the body of the lean vehicle 1. The control unit 22 acquires the yaw rate and roll rate based on the physical quantities detected by the turning state detection unit 24 (B).

[0037] The control unit 22 executes the host vehicle path prediction (C). The control unit 22 can execute the host vehicle path prediction taking into account the yaw rate. The control unit 22 can execute the host vehicle path prediction taking into account the yaw rate and the roll rate.

[0038] The control unit 22 executes position detection of the object around the vehicle (D). The control unit 22 can execute position detection of the object around the vehicle taking the roll rate into consideration.

[0039] The control unit 22 uses the results obtained by predicting the vehicle's path and the results obtained by detecting the positions of objects around the vehicle to perform warning control (E) so that a warning is issued in situation 1 shown in Figure 1 and no warning is issued in situation 2.

[0040] Situations 1 and 2 shown in FIG. 1 show the positional relationship between the lean vehicle 1 while it is traveling and the preceding vehicles 101-103 detected by detecting the positions of objects around the vehicle, as viewed from above. The dashed and dotted lines shown in Situations 1 and 2 are the predicted paths of the lean vehicle 1 obtained by executing host vehicle path prediction. The dashed and dotted lines are the paths of the lean vehicle 1 obtained by executing host vehicle path prediction. The dashed lines are the paths that both ends of the lean vehicle 1 in the vehicle width direction will take when the lean vehicle 1 travels along the obtained path. If an object is present within the band-shaped area indicated by the two dashed lines, there is a possibility that the lean vehicle 1 will collide with this object while continuing to travel. For this reason, warning control is executed using the band-shaped area indicated by the dashed lines as the predicted path.

[0041] In situation 1, while the lean vehicle 1 is turning on a road without white lines, the first leading vehicle 101 is turning with the same turning radius as the lean vehicle 1. The first leading vehicle 101 is turning at a vehicle speed V2 that is slower than the vehicle speed V1 of the lean vehicle 1 (V2<V1). The first leading vehicle 101 is turning ahead of the lean vehicle 1. The first leading vehicle 101 is turning on the predicted path of the lean vehicle 1.

[0042] In situation 1, the control unit 22 executes warning control so that the warning unit 21 issues a warning when the possibility of a collision between the lean vehicle 1 and the first leading vehicle 101 becomes equal to or greater than a predetermined possibility of collision. For example, the control unit 22 controls the warning unit 21 to issue a warning when the distance between the lean vehicle 1 and the first leading vehicle 101 becomes a predetermined distance. For example, the control unit 22 controls the warning unit 21 to issue a warning when the relative speed between the lean vehicle 1 and the first leading vehicle 101 becomes a predetermined speed. For example, the control unit 22 controls the warning unit 21 to issue a warning when the time to collision becomes a predetermined time.

[0043] In situation 2, while the lean vehicle 1 is turning at vehicle speed V1 on a road without white lines, the second leading vehicle 102 is turning with the same turning radius and at the same vehicle speed V1 as the lean vehicle 1. In situation 2, in addition to the second leading vehicle 102, the third leading vehicle 103 is turning on the inside of the turning of the second leading vehicle 102 with a turning radius smaller than the turning radius of the lean vehicle 1 and at a vehicle speed V2 smaller than the vehicle speed V1 of the lean vehicle 1. Both the second leading vehicle 102 and the third leading vehicle 103 are turning ahead of the lean vehicle 1. The second leading vehicle 102 is turning on the predicted path of the lean vehicle 1. The third leading vehicle 103 is turning on a turning path that deviates from the predicted path of the lean vehicle 1.

[0044] In situation 2, the control unit 22 executes warning control so that the warning unit 21 does not issue a warning. The third leading vehicle 103, which, like the first leading vehicle 101 in situation 1, is traveling at a vehicle speed V2 that is slower than the vehicle speed V1 of the lean vehicle 1, moves relatively toward the lean vehicle 1, but the third leading vehicle 103 is located outside the predicted path of the lean vehicle 1. The lean vehicle 1 continues to turn on the predicted path, passing on the left side of the third leading vehicle 103. Therefore, the probability of a collision between the lean vehicle 1 and the third leading vehicle 103 is below a predetermined probability of collision that requires a warning, and the control unit 22 controls the warning unit 21 so as not to issue a warning for the third leading vehicle 103.

[0045] The second leading vehicle 102 is on the predicted path of the lean vehicle 1, but is traveling at the same vehicle speed V1 as the lean vehicle 1. The distance between the lean vehicle 1 and the second leading vehicle 102 is maintained, and the second leading vehicle 102 is not moving relative to the lean vehicle 1. Therefore, the possibility of a collision between the lean vehicle 1 and the second leading vehicle 102 is also below a predetermined collision possibility that requires an alarm, and the control unit 22 controls the alarm unit 21 not to issue an alarm for the second leading vehicle 102.

[0046] In this way, the FCW device 2 issues an alarm if an object approaching the lean vehicle 1 due to relative movement is on the predicted path of the turning lean vehicle 1, but does not issue an alarm if the object is located outside the predicted path. Situations 1 and 2 are illustrative examples, and the alarm control by the FCW device 2 is not limited to situations 1 and 2. For example, when the first leading vehicle 101 is stopped on the road in situation 1, alarm control is executed in the same manner as in situation 1 described above. For example, when the second leading vehicle 102 is turning at a speed greater than the speed of the lean vehicle 1 in situation 2, or when the third leading vehicle 103 is stopped, alarm control is executed in the same manner as in situation 2 described above. The FCW device 2 can also execute alarm control in various other situations based on the predicted path of the lean vehicle 1 that takes into account the turning state obtained by executing host vehicle path prediction, the position of the object detected by position detection of the object around the host vehicle, and the relative movement between the lean vehicle 1 and the object.

[0047] Second Embodiment A lean vehicle FCW device 2 according to a second embodiment of the present disclosure will be described with reference to Fig. 2. Below, content that overlaps with the description of the first embodiment will be omitted, and only content necessary to describe the FCW device 2 of the second embodiment will be described.

[0048] As shown in FIG. 2A, the FCW device 2 according to the second embodiment further includes a vehicle speed detection unit 25 that detects the vehicle speed of the lean vehicle 1 in addition to the configuration shown in FIG.

[0049] The control unit 22 determines whether the lean vehicle 1 is traveling straight or turning from the yaw rate obtained using the turning state detection unit 24. The control unit 22 may determine that the lean vehicle 1 is turning when the magnitude of the yaw rate is equal to or greater than a predetermined threshold. If the sign of the yaw rate is defined as a positive yaw rate when the vehicle is turning to the right and a negative yaw rate when the vehicle is turning to the left, the control unit 22 may determine that the lean vehicle 1 is turning when the duration of the yaw rate sign is equal to or greater than a predetermined threshold. If the control unit 22 determines that the lean vehicle 1 is traveling straight, it predicts a path ahead in the traveling direction of the lean vehicle 1. If it is determined that the lean vehicle 1 is turning, the control unit 22 predicts a path of the lean vehicle 1 based on the yaw rate obtained using the turning state detection unit 24 and the vehicle speed detected by the vehicle speed detection unit 25. The control unit 22 performs host vehicle path prediction taking into account the roll rate obtained using the turning state detection unit 24, or performs object position detection taking into account the roll rate obtained using the turning state detection unit 24. Specifically, the control unit 22 performs the process shown in Fig. 2(b) or the process shown in Fig. 2(c).

[0050] 2B, the control unit 22 performs host vehicle path prediction that takes into account the turning of the lean vehicle 1 based on the yaw rate obtained using the turning state detection unit 24 and the vehicle speed detected by the vehicle speed detection unit 25 (S1). The control unit 22 corrects the obtained predicted path based on the roll rate obtained using the turning state detection unit 24. As a result, the host vehicle path prediction that takes into account the turning and roll rate of the lean vehicle 1 is performed (S2). Meanwhile, the control unit 22 performs position detection of objects around the host vehicle based only on the captured image obtained by the monocular camera 23 without using the vehicle speed, yaw rate, or roll rate (S3). The control unit 22 performs warning control based on the path prediction result that takes into account the turning and roll rate of the lean vehicle 1 and the positions of objects detected around the lean vehicle 1 (S4).

[0051] 2C, the control unit 22 predicts the vehicle path taking into account the turning of the lean vehicle 1 based on the yaw rate obtained using the turning state detection unit 24 and the vehicle speed detected by the vehicle speed detection unit 25 (S11). The control unit 22 detects the position of an object around the vehicle taking into account the roll rate of the lean vehicle 1 based on the roll rate obtained using the turning state detection unit 24 and the captured image obtained by the monocular camera 23 (S12). The control unit 22 executes warning control based on the predicted path taking into account the turning of the lean vehicle 1 and the position of the object around the lean vehicle 1 detected taking into account the roll rate of the lean vehicle 1 (S13).

[0052] Examples of processing performed in predicting the vehicle's path and detecting the positions of objects around the vehicle will be described with reference to Figures 3 to 6. Figures 3 to 6 schematically show the lean vehicle 1, predicted path, surrounding objects, etc., so that the details of each processing can be easily understood and compared.

[0053] When the lean vehicle 1 is traveling straight, a straight predicted path 301 is obtained, as shown on the left side of Fig. 3(a) . The predicted path 301 has a band shape of equal width formed by a left line 301L and a right line 301R. A center line 301C is a line passing through the center in the width direction of the left line 301L and the right line 301R.

[0054] On the image 51 captured by the monocular camera 23, the center line 301C of the predicted path 301 is a line perpendicular to the horizon 300. The left line 301L is a straight line tilted to the right, with its lower end farther to the left of the center line 301C than its upper end, and the right line 301R is a straight line tilted to the left, with its lower end farther to the right of the center line 301C than its upper end. The predicted path 301 on the captured image 51 is wide at the bottom of the captured image 51 and narrows as it approaches the horizon 300. The predicted path 301 on the captured image 51 is wider the closer it is to the lean vehicle 1 and narrows as the distance from the lean vehicle 1 increases.

[0055] The horizon line 300 on the captured image 51 is, for example, a line corresponding to the horizon ahead as seen from the lean-in vehicle 1 traveling straight on a level road. In the examples of Figures 3 to 6 that schematically show the captured images, the horizon line 300 is approximately parallel to the long side of the captured image, but in reality, the horizon line 300 is a line that is tilted according to the tilt angle of the lean-in vehicle 1. The horizon line 300 on each captured image may be set based on the mounting position, mounting angle, etc. of the monocular camera 23 on the lean-in vehicle 1.

[0056] In the host vehicle path prediction (S1 and S11) taking turning into consideration shown in FIGS. 2(a) and 2(b), a predicted path is obtained based on the yaw rate and vehicle speed of the lean vehicle 1 during turning. Specifically, when the lean vehicle 1 tilts its body and starts turning, a curved predicted path 311 is obtained based on the yaw rate of the lean vehicle 1, as shown in FIG. 3(b). As shown on the left side of FIG. 3(b), the predicted path 311 during turning is a curved band-like region formed by a left line 311L and a right line 311R. Similar to the predicted path 301 during straight driving, the predicted path 311 is a band-like region of equal width. A center line 311C is a line passing through the center of the width of the predicted path 311. As shown on the right side of FIG. 3(b), the predicted path 311 on the captured image 52 is wide at the bottom of the captured image 52 and narrows as it approaches the horizon 300. The predicted path 311 on the captured image 52 has a curved shape that is wider the closer it is to the lean vehicle 1 and narrows as the distance from the lean vehicle 1 increases.

[0057] The vehicle path prediction (S2) taking into account the turning and roll rate shown in Figure 2(a) is performed by correcting the predicted path obtained based on, for example, the yaw rate and vehicle speed based on the roll rate of the lean vehicle 1.

[0058] Specifically, for example, when the control unit 22 performs a host vehicle path prediction (S1) that takes turning into consideration, i.e., a host vehicle path prediction based on the yaw rate and vehicle speed, the predicted path 411 shown in FIG. 4A should be obtained. On the other hand, the predicted path on the captured image is the predicted path 401 shown in FIG. 4B. When the lean vehicle 1 is turning, the monocular camera 23 fixed to the vehicle body is tilted with respect to the road. Therefore, if the path is predicted on the captured image without considering the effect of the tilt, the predicted path will differ from the actual path. Therefore, the control unit 22 performs a host vehicle path prediction (S2) that takes turning and roll rate into consideration, i.e., corrects the predicted path 401 based on the roll rate. By correcting the predicted path 401 based on the roll rate, the predicted path 401 on the captured image is corrected as shown by the arrow in FIG. 4B, and the predicted path 411 shown in FIG. 4C is obtained. The control unit 22 calculates the roll angle from the roll rate obtained from the detection result of the turning state detection unit 24, and corrects the predicted course 401 based on the roll rate by rotating the points on the captured image that make up the predicted course 401 in the reverse direction by the calculated roll angle, with the center of the captured image as the axis.

[0059] 4(a) and 4(b) are objects detected by detecting the positions of objects around the vehicle (S3). If the lean vehicle 1 shown in FIG. 4(a) actually continues turning, the turning path will be along the corrected predicted path 411, and the position of the object 201 will be on the turning path.

[0060] 4(b), in the predicted path 401 obtained by the host vehicle path prediction (S1) that takes only turning into consideration, unlike the actual situation, the object 201 is located off the predicted path 401, and the object 202 is located on the predicted path 401. In order to execute warning control with higher accuracy, the control unit 22 corrects the predicted path 401 and executes warning control based on the corrected predicted path 411.

[0061] 4(c), in the predicted path 411 obtained by the host vehicle path prediction (S2) taking into account the turning and roll rate, the object 201 is on the predicted path 401, and the object 202 is off the predicted path 401, just like in reality. This allows the control unit 22 to execute more accurate warning control (S4).

[0062] In the processing example shown in Fig. 2(a), the predicted path obtained by taking the turn into consideration is corrected by taking the roll rate into consideration, whereas in the processing example shown in Fig. 2(b), the position of an object around the vehicle is detected by taking the roll rate into consideration, and the position of the object is corrected.

[0063] Fig. 5 shows an example in which, under the same circumstances as Fig. 4, the position of an object is corrected instead of the predicted path. When the host vehicle path prediction (S11) taking turning into consideration is executed, a predicted path 401 is obtained on the captured image as shown in Fig. 5(a). Fig. 5(a) shows the positions of two objects 201 and 202 detected when the position of an object around the host vehicle is detected without considering the roll rate. As described above, in reality, the object 201 is on the predicted path and the object 202 is located off the predicted path. However, in Fig. 5(a), the object 201 is located off the predicted path 401, and the object 202 is on the predicted path 401.

[0064] When the control unit 22 performs position detection of objects around the vehicle taking the roll rate into consideration (S12), the positions of the objects 201 and 202 are corrected as shown by the arrows in Figure 5(a) and move to the positions shown in Figure 5(b).

[0065] 5B, the object 201 is on the predicted path 401, and the object 202 is off the predicted path 401, resulting in a result that matches the case where the lean vehicle 1 actually continues to turn. The control unit 22 can execute more accurate warning control (S13) based on the predicted path 401 of the lean vehicle 1 shown in FIG. 5B and the detected positions of the objects 201 and 202.

[0066] Correction of the object positions in the captured image can be performed by correcting the object positions as shown in FIG. 5( a) or by rotating the captured image as shown by the arrows in FIG. 6( a). When rotating the captured image, the positions of the objects 201 and 202 in the captured image in FIG. 6( a) are corrected to the positions shown in FIG. 6( b). This results in a result that matches the actual state of the leaning vehicle 1 continuing to turn, allowing the control unit 22 to execute more accurate warning control. The corrections indicated by the arrows in FIGS. 5( a) and 6( a) are performed using the calculated roll angle, similar to the correction of the predicted path. In the example of FIG. 5( a), the control unit 22 performs correction so that the planar coordinates of the two detected objects 201 and 202 in the captured image are rotated by the calculated roll angle around the center of the captured image. In the example of FIG. 6( a), the control unit 22 performs correction so that the entire captured image is rotated by the calculated roll angle around the center of the captured image.

[0067] In this way, the control unit 22 can perform any one of correcting the predicted path in the captured image based on the roll rate, correcting the position of an object detected in the captured image, and rotating the captured image, thereby enabling more accurate warning control. It goes without saying that the control unit 22 is not limited to correcting the predicted path and the position of an object after acquiring them, but may also perform processing so as to obtain the corrected predicted path and the position of an object from the beginning.

[0068] REFERENCE SIGNS LIST 1 lean vehicle 2 FCW device 21 alarm unit 22 control unit 23 monocular camera 24 turning state detection unit 25 vehicle speed detection unit

Claims

1. An FCW device for a lean vehicle, comprising: an alarm unit that issues an alarm to a driver; and one or more control units that execute a host vehicle path prediction that predicts a path of the host vehicle and a host vehicle peripheral object position detection that detects the positions of objects around the host vehicle, calculate a possibility of a collision between the host vehicle and the object using the results of the host vehicle path prediction and the results of the host vehicle peripheral object position detection, and execute an alarm control that controls whether or not an alarm is issued by the alarm unit based on the possibility of collision. The FCW device for a lean vehicle is used in a lean vehicle that leans to the inside of a turn when turning, and further comprises: a monocular camera that is fixed to the host vehicle and captures an image of the surroundings of the host vehicle; and a turning state detection unit that detects physical quantities related to a yaw rate and a roll rate of the host vehicle. The control unit executes the host vehicle path prediction and / or the host vehicle peripheral object position detection based on the captured image. A FCW device for a lean vehicle, comprising: a first leading vehicle turning ahead of the host vehicle at a vehicle speed slower than the vehicle speed of the host vehicle and with the same turning radius as the host vehicle while the host vehicle is turning on a road without white lines, the control unit executes the warning control so that the warning unit issues a warning when a collision probability between the host vehicle and the first leading vehicle reaches or exceeds a predetermined collision probability; a second leading vehicle turning ahead of the host vehicle at the same vehicle speed and with the same turning radius as the host vehicle while the host vehicle is turning on a road without white lines, and a third leading vehicle turning ahead of the host vehicle at a vehicle speed slower than the vehicle speed of the host vehicle and with a turning radius smaller than the turning radius of the host vehicle on the inside of the turning of the second vehicle, the control unit executes the warning control so that the warning unit does not issue a warning.

2. The FCW device for a lean vehicle as described in claim 1, further comprising a vehicle speed detection unit which detects the vehicle speed of the host vehicle, wherein the control unit, in the host vehicle course prediction, determines whether the host vehicle is turning based on the yaw rate obtained from the detection result of the turning state detection unit, and if it is determined that the host vehicle is turning, predicts the course of the host vehicle based on the yaw rate obtained from the detection result of the turning state detection unit and the vehicle speed detected by the vehicle speed detection unit, and, in the host vehicle course prediction, performs a course prediction of the host vehicle taking into account a roll rate obtained from the detection result of the turning state detection unit, or, in the position detection of an object around the host vehicle, performs position detection of the object taking into account the roll rate.

3. The FCW device for a lean vehicle as described in claim 2, characterized in that the control unit performs the vehicle path prediction based on the yaw rate and vehicle speed of the turning vehicle to obtain a predicted path having a curved shape, and performs one of correcting the predicted path on the captured image, correcting the position of an object detected on the captured image, and rotating the captured image based on the roll rate.

4. An FCW method for a lean vehicle that is executed by an FCW device for use in a lean vehicle that leans to the inside of a turn when turning, the FCW method comprising: a warning unit that issues a warning to a driver; and one or more control units that execute a host vehicle path prediction that predicts a path of the host vehicle and a host vehicle peripheral object position detection that detects the positions of objects around the host vehicle, calculate a possibility of a collision between the host vehicle and the object using the results of the host vehicle path prediction and the results of the host vehicle peripheral object position detection, and execute warning control that controls whether or not a warning is issued by the warning unit based on the possibility of collision, the FCW method comprising: a step of capturing an image of the surroundings of the host vehicle with a monocular camera fixed to the host vehicle; a step of detecting the positions of the objects around the host vehicle based on the captured image obtained by the monocular camera; a step of acquiring a yaw rate and a roll rate of the host vehicle from the detection results of a turning state detection unit that detects physical quantities related to the yaw rate and roll rate of the host vehicle; and a step of executing the host vehicle path prediction and / or the position detection of the objects around the host vehicle, taking into account the yaw rate and the roll rate obtained from the detection results of the turning state detection unit. and executing the warning control as described in each of the following situations using the result of the host vehicle trajectory prediction and the result of the position detection of the object around the host vehicle. Situation 1: When the host vehicle is turning on a road without white lines and a first leading vehicle is turning in front of the host vehicle at a vehicle speed slower than the host vehicle and with the same turning radius as the host vehicle, the control unit executes the warning control so that the warning unit issues a warning when the possibility of a collision between the host vehicle and the first leading vehicle becomes equal to or greater than a predetermined possibility of a collision. Situation 2: When the host vehicle is turning on a road without white lines and a second leading vehicle is turning in front of the host vehicle at the same vehicle speed and with the same turning radius as the host vehicle, and a third leading vehicle is turning on the inside of the turning of the second vehicle in front of the host vehicle at a vehicle speed slower than the host vehicle and with a turning radius smaller than the host vehicle, the control unit executes the warning control so that the warning unit does not issue a warning.

5. The FCW method for a lean vehicle as described in claim 4, characterized in that in the host vehicle path prediction, the path of the host vehicle is predicted based on a yaw rate obtained from the detection result of the turning state detection unit and a vehicle speed detected by a vehicle speed detection unit that detects the vehicle speed of the host vehicle, and in the host vehicle path prediction, it is determined whether the host vehicle is turning based on the yaw rate obtained from the detection result of the turning state detection unit, and if it is determined that the host vehicle is turning, a path prediction of the host vehicle is performed taking into account a roll rate obtained from the detection result of the turning state detection unit, or in the position detection of an object around the host vehicle, the position detection of the object is performed taking into account the roll rate.

6. The FCW method for lean vehicles described in claim 5, characterized in that it includes a step of: executing a vehicle path prediction based on the yaw rate and the vehicle speed to obtain a predicted path having a curved shape of the turning vehicle; and a step of executing one of the following steps based on the roll rate: correcting the predicted path on the captured image, correcting the position of an object detected on the captured image, and rotating the captured image.

Images