JP2015132966A - Posterolateral approaching body alarm apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate the inconvenience by an alarm.SOLUTION: A congestion detection unit 50 detects that the travelling road of a vehicle 5 is congested and an estimation unit 45 of a recognition state of the situation around a vehicle estimates that a driver 8 recognizes the state of the posterolateral road on an adjacency travelling lane Y1 (Y3). In this case, when the driver 8 operates the direction indicator in order to change the lane, an alarm suppression unit 60 suppresses the output of an alarm from an alarm output unit 70 even when an approaching vehicle detection unit 20 detects another vehicle 6 approaching a posterolateral and predetermined range on the adjacency travelling lane Y1 (Y3) to which the vehicle 5 intends to change the lane.

Description

  The present invention is installed in a vehicle and detects the presence of a vehicle approaching from the rear side on the deviating lane when there is a possibility of deviating from the lane at the time of lane change or the like. The present invention relates to a side approaching object warning device.

  In recent years, a camera is installed in a vehicle, and an approaching vehicle is detected from an image obtained by imaging the rear side of the vehicle that becomes a blind spot with respect to the driver, and this is alerted. BSW (Blind Spot Warning: hereinafter, rear side) A technique related to the approaching object warning device has been proposed.

  For example, in the rear side approaching object alarm device described in Patent Document 1, when a vehicle is present in the alarm area on the rear side of the vehicle, a lamp is turned on (hereinafter referred to as a primary alarm), and the driver In this state, when the driver expresses his intention to change lanes by noticing the presence of the vehicle, he blinks the lamp and blows a buzzer (hereinafter referred to as a secondary alarm) to indicate the direction. The fact that there is an approaching vehicle is transmitted to the lane on the side where the vessel is released.

JP 2011-141746 A

  However, the rear side approaching object warning device described in Patent Document 1 is a case where the driver has already noticed the presence of an approaching vehicle and has determined that the lane can be changed and then operated the direction indicator. When the approaching vehicle is in the alarm area at that time, a secondary alarm is output.

  Therefore, in such a conventional rear side approaching object warning device, even when the approaching vehicle is in the warning area, even if the driver knows the state of the rear side road Since the secondary alarm is output when the direction indicator is operated, the driver is bothered by outputting an alarm more than necessary.

  The present invention has been made in view of the above problems, and when it is estimated that the driver recognizes the state of the road including the state of the approaching vehicle in the adjacent traveling lane to which the lane is to be changed, the secondary is performed. An object of the present invention is to provide a rear side approaching object warning device that does not bother the driver by suppressing the warning.

  The rear side approaching object warning device according to the present invention suppresses the rear side warning when it is estimated that the driver is aware of the state of the road of the adjacent traveling lane to be changed.

  That is, the rear side approaching object warning device according to the present invention detects that an approaching vehicle exists in a predetermined distance range on the rear side of the vehicle in an adjacent traveling lane adjacent to the traveling lane of the vehicle. An approaching vehicle detection unit, a vehicle surrounding state recognition state estimation unit that estimates that the driver of the vehicle recognizes a state of a road behind the adjacent traveling lane, and the vehicle is traveling When the traffic detection unit detects that the road is congested and the approaching vehicle detection unit detects an approaching vehicle, an alarm is output when the driver tries to change the lane to the adjacent traveling lane When the warning output unit and the traffic jam detection unit detect traffic jam, and the vehicle surrounding situation recognition state estimation unit recognizes the state of the road on the rear side of the adjacent traveling lane, When estimated, the driver If an attempt is made to change lanes to the adjacent traffic lane, characterized in that it comprises, an alarm suppression unit suppresses output of alarm from the alarm output unit.

  According to the rear side approaching object alarm device according to the present invention configured as described above, the traffic congestion detection unit detects that the road on which the vehicle is traveling is congested, and recognizes the vehicle surrounding situation. When the state estimation unit estimates that the driver knows the state of the road behind the adjacent lane, the approaching vehicle is detected when the driver operates the direction indicator to change the lane. Even when the unit detects an approaching vehicle within a predetermined distance range on the rear side of the adjacent traveling lane in which the lane change is to be performed, the alarm suppression unit suppresses the output of the alarm from the alarm output unit. Therefore, when it is estimated that the driver is aware of the state of the road on the rear side of the adjacent traveling lane on the side where the lane is changed, it is possible to eliminate the troublesomeness caused by outputting an alarm more than necessary.

  According to the rear side approaching object warning device according to the present invention, an unnecessary warning is suppressed for the driver who is estimated to recognize the state of the rear side road. Can be eliminated.

It is a figure explaining the function of the rear side approaching object alarm device which concerns on this invention. It is a block diagram which shows schematic structure of the rear side approaching object alarm device which concerns on Example 1 of this invention. It is a flowchart of the main routine which shows the flow of the whole process performed with the rear side approaching object alarm device which concerns on Example 1 of this invention. It is a flowchart explaining the flow of the process which detects that the road currently drive | working is congested. It is a flowchart explaining the flow of another process which detects that the road currently drive | working is congested. It is a flowchart explaining the flow of the process which detects the approaching vehicle from a rear side. It is a figure explaining the principle of the process which detects a driver | operator's gaze direction, (a) is an example of the image of the face observed when a driver | operator is facing the front. (B) is an example of an image observed when the driver turns to the left. (C) is an example of an image observed when the driver turns his face to the left. (D) is an example of an image observed when the driver turns to the right. It is a figure explaining the specific method of detecting a driver | operator's gaze direction. It is a flowchart explaining the flow of the process which estimates whether the driver | operator is recognizing the state of the road behind the vehicle. It is a flowchart explaining the flow of the process which detects a driver | operator's gaze direction. It is a figure explaining the specific suppression content of the alarm suppression process performed with the rear side approaching object alarm device which concerns on Example 1 of this invention.

  Hereinafter, specific embodiments of the rear side approaching object alarm device according to the present invention will be described with reference to the drawings.

  In this embodiment, the present invention monitors the rear side of a traveling vehicle, and when there is an approaching vehicle in a predetermined distance range on the rear side of the adjacent traveling lane, the driver detects the presence of the approaching vehicle. When it is presumed to be recognized, it is applied to a rear side approaching object warning device having a function of suppressing a secondary warning.

  First, the operation of a conventional rear side approaching object alarm device that does not suppress the secondary alarm will be briefly described with reference to FIG. The vehicle 5 is traveling on the road lane Y2 of the road 2 in the left direction in the drawing. The rear camera 15a installed rearward at the rear end of the vehicle 5, the left rear camera 15b installed rearward in the vicinity of the left door mirror, and the right rear camera 15c installed rearward in the vicinity of the right door mirror The inside of the imaging range ω including the left and right adjacent traveling lanes is imaged. Then, from the captured images, the vehicle 5 on the adjacent traveling lane Y1 on the adjacent traveling lane Y1, inside the detection area X1 within a predetermined distance range from the vehicle 5 and on the adjacent traveling lane Y3 by image processing. , The other vehicle 6 approaching the vehicle 5 that is present in the detection area X2 in the predetermined distance range on the rear side is detected.

  The rear side approaching object warning device is activated, for example, when the vehicle 5 is traveling at a vehicle speed v (for example, 30 km / h or more) within a predetermined range. In addition, a dedicated start switch may be provided to start when the driver operates the start switch.

  When the other vehicle 6 approaching the vehicle 5 is detected within a predetermined distance range on the rear side of the vehicle 5 after the rear side approaching object warning device is activated, the left and right outside rear-view mirrors (door mirrors) are detected. ) By illuminating an indicator, such as an LED, installed in the vicinity, the presence of the other vehicle 6 is transmitted to the driver by visual information, and a warning is given when changing lanes (primary alarm).

Furthermore, if the driver is not aware of this primary warning and tries to change the lane by issuing a direction indicator on the adjacent traveling lane Y1 side or the adjacent traveling lane Y3 side where the other vehicle 6 exists, the indicator blinks. The visual information emphasized is output, and further, auditory information is output by sounding an alarm sound by a buzzer, etc., and the presence of the other vehicle 6 is more clearly communicated to the driver, and the lane change operation is interrupted (Secondary alarm).
(Description of configuration of rear side approaching object warning device)

  Next, the configuration of the rear side approaching object alarm device 10 according to Example 1 which is an embodiment of the present invention will be described with reference to FIG. FIG. 2 shows a configuration diagram when the rear side approaching object warning device 10 according to the present embodiment is mounted on the vehicle 5 (see FIG. 1).

  As shown in FIG. 2, the rear side approaching object alarm device 10 according to the first embodiment includes a rear side imaging unit 15 that images a road behind the vehicle 5 (see FIG. 1), and a rear side imaging. The approaching vehicle detection unit 20 that detects an approaching vehicle from the images captured by the unit 15, the driver imaging unit 30 that captures the range including the face of the driver 8 as needed, and the driver imaging unit 30. Based on the gaze direction detector 40 that detects the gaze direction of the driver 8 from the image including the face of the driver 8 and the detected gaze direction, the driver 8 A vehicle surrounding state recognition state estimation unit 45 that estimates whether or not a state is recognized, a traffic congestion detection unit 50 that detects whether or not a road on which the vehicle 5 is traveling is congested, and a driver 8 An alarm output unit 70 for outputting necessary alarms based on visual information and auditory information; When it is estimated by the state estimation unit 45 that the driver 8 is aware of the state of the road on the rear side of the adjacent traveling lane on the side where the vehicle 5 changes lanes, the approaching vehicle detection unit 20 detects the approaching vehicle. Even when the alarm is detected, the alarm output unit 70 instructs the alarm suppression unit 60 to suppress the output of the secondary alarm including the strong visual information and the auditory information, and to turn the vehicle 5 to the right or left. A blinker switch 80 that turns on, an LED 85 that lights or blinks according to an instruction from the alarm output unit 70, a buzzer 90 that rings according to an instruction from the alarm output unit 70, and a display 95 that displays necessary information according to an instruction from the alarm output unit 70 And.

  The approaching vehicle detection unit 20, the gaze direction detection unit 40, the vehicle surrounding situation recognition state estimation unit 45, the traffic jam detection unit 50, the alarm suppression unit 60, and the alarm output unit 70 are included in the camera controller 150. For example, the vehicle speed sensor 100 and the car navigation system 110 are connected via a CAN (Controller Area Network) communication line 130. In place of the CAN communication line 130, a communication line compliant with another communication standard used as an in-vehicle communication network, for example, LIN (Local Interconnect Network), Flex Ray, Ethernet (registered trademark), or the like may be used. it can.

  The rear side imaging unit 15 includes a rear camera 15a installed near the rear license plate toward the rear of the vehicle 5, and a left rear installed near the left door mirror of the vehicle 5 toward the left rear side of the vehicle 5. The camera 15b and the right rear camera 15c installed near the right door mirror of the vehicle 5 toward the right rear side of the vehicle 5. Then, the inside of the imaging range ω on the rear side of the vehicle 5 shown in FIG. 1 is monitored. Each camera constituting the rear side imaging unit 15 photoelectrically converts an optical system including a lens that collects light in a predetermined imaging range on the rear side of the vehicle 5 and an optical signal into an electrical signal. For example, a photoelectric conversion unit composed of a CMOS element or a CCD element is provided.

  The driver imaging unit 30 is provided near the meter cluster of the vehicle 5 and images a range including the face of the driver 8.

  The approaching vehicle detection unit 20 exists within a predetermined distance range from the vehicle 5 on the adjacent traveling lane Y1 or the adjacent traveling lane Y3 from the image captured by the rear side imaging unit 15, and is shown in FIG. As shown, another vehicle 6 approaching the vehicle 5 is detected. Specifically, from the rear side image of the vehicle 5 imaged by the rear side imaging unit 15, a difference (frame difference) between images captured continuously in time or analysis of luminance distribution, The other vehicle 6 is detected using an image processing method such as template matching. These image processing methods are widely used in general, and an approaching vehicle is detected by combining these image processing methods. A specific processing method will be described later.

  The gaze direction detection unit 40 detects the direction in which the driver 8 of the vehicle 5 is gazing.

  The vehicle surrounding state recognition state estimation unit 45 pays attention to the rear side area of the vehicle 5 so that the driver 8 is in the road state of the adjacent traveling lanes Y1, Y3 (see FIG. 1), for example, the adjacent traveling lane. It is estimated whether or not the existence of the other vehicle 6 traveling in Y1 and Y3 is recognized. Specifically, the driver 8 detects the direction of the face of the driver 8 and the direction of the line of sight from the image including the face of the driver 8 captured by the driver imaging unit 30. The frequency of looking toward the direction of the rear-view mirrors (left and right door mirrors) installed on the left and right, that is, the right rear side of the vehicle 5 or the left rear side of the vehicle 5 is detected. When it is detected that the driver 8 faces his face in the direction of the door mirror a predetermined number of times or more during a predetermined time interval set in advance, the driver 8 Presumed to be aware of the lateral state. A specific estimation method will be described later.

  The traffic jam detection unit 50 detects whether or not the road on which the vehicle 5 is traveling is jammed. The fact that the road on which the vehicle is traveling is in a traffic jam can be detected using various methods. For example, by receiving traffic information provided by radio broadcasting or VICS (registered trademark), the presence or absence of traffic jams can be detected. Can be detected. Alternatively, the vehicle speed v of the vehicle 5 is compared with the speed limit of the road on which the vehicle 5 is traveling, and when the vehicle speed v remains low for a predetermined time, it is determined that the road on which the vehicle 5 is traveling is congested. be able to. Details will be described later.

  In the alarm suppression unit 60, the traffic congestion detection unit 50 detects that the road on which the vehicle 5 is traveling is congested, and the vehicle surrounding state recognition state estimation unit 45 detects other vehicles on the rear side of the vehicle 5. When the driver 8 estimates that the state 6 is recognized, the secondary warning is suppressed.

  The alarm output unit 70 instructs the LED 85, the buzzer 90, and the display 95 installed in the vehicle 5 to output necessary alarms. That is, when the approaching vehicle detection unit 20 detects another vehicle 6 (see FIG. 1) approaching the vehicle 5 within a predetermined distance range from the vehicle 5 (see FIG. 1), the secondary alarm is suppressed. If not, a secondary warning that alerts the driver 8 to the approach of the other vehicle 6 is output to alert the driver 8 based on blinking of the LED 85 and ringing of the buzzer 90, that is, enhanced visual information and auditory information.

  The alarm output unit 70 is when the approaching vehicle detection unit 20 detects another vehicle 6 (see FIG. 1) approaching the vehicle 5 within a predetermined distance range from the vehicle 5 (see FIG. 1). Even when the secondary alarm is suppressed, the primary alarm for notifying the driver 8 of the approach of the other vehicle 6 is output to alert the driver 8 only by turning on the LED 85, that is, visual information. At this time, the output of the secondary alarm due to the blinking of the LED 85 and the sound of the buzzer 90 is suppressed.

  The turn signal switch 80 includes a left turn signal switch 80a and a right turn signal switch 80b. The turn signal switch 80 is linked to a direction indicator (a turn signal lever) operated by the driver 8 and the direction of the instructed turn signal. Is output.

The car navigation system 110 is connected to the traffic information receiving unit 120 and receives traffic information transmitted by radio or VICS (registered trademark).
(Description of the action of the rear side approaching object warning device)

  Next, the contents of a series of processes performed by the rear side approaching object warning device 10 according to the first embodiment will be described with reference to the flowchart of FIG. 3 and the system configuration diagram of FIG.

  (Step S100) In the vehicle surrounding state recognition state estimation unit 45, a left visual recognition time timer Tl and a right visual recognition time timer Tr are set. These timers are timers for setting a predetermined time interval in order to estimate whether or not the driver 8 is aware of the state of the rear side of the vehicle 5 by gazing at the left and right door mirrors. In detail, it demonstrates later in description of a driver | operator's recognition state estimation process performed in the vehicle surrounding condition recognition state estimation part 45. FIG.

  (Step S101) The traffic jam detection unit 50 performs traffic jam detection processing for detecting whether or not the road on which the vehicle 5 is traveling is jammed. Details will be described later.

  (Step S102) The approaching vehicle detection unit 20 determines whether or not the activation condition of the rear side approaching object warning device 10 (hereinafter referred to as BSW10) is satisfied. Specifically, the BSW 10 is automatically activated when the vehicle speed v of the vehicle 5 is equal to or higher than a predetermined vehicle speed (for example, 30 km / h) set in advance.

  (Step S <b> 104) The approaching vehicle detection unit 20 performs approaching vehicle detection processing for detecting the other vehicle 6 approaching from the rear side of the vehicle 5. Details will be described later.

  (Step S106) The vehicle surrounding state recognition state estimation unit 45 performs a recognition state estimation process for the driver 8 for estimating whether the driver 8 recognizes the state of the road on the rear side. Details will be described later.

  (Step S108) It is determined whether or not the left turn signal switch 80a is turned on. When the left turn signal switch 80a is ON, the process proceeds to step S110, and when the left turn signal switch 80a is OFF, the process proceeds to step S114.

  (Step S110) It is determined whether the traffic jam flag Fj is ON. When the traffic jam flag Fj is ON, the process proceeds to step S112, and when the traffic jam flag Fj is OFF, the process proceeds to step S122. Note that the traffic jam flag Fj is turned on when it is determined that the road on which the vehicle 5 is traveling is congested in the traffic jam detection processing performed at step S101, and when it is determined that there is no traffic jam. This flag is turned off.

  (Step S112) It is determined whether the left rear side recognition flag Flr is ON. When the left rear side recognition flag Flr is ON, the process proceeds to step S120, and when the left rear side recognition flag Flr is OFF, the process proceeds to step S122. The left rear side recognition flag Flr is estimated that the driver 8 recognizes the state of the road on the left rear side of the vehicle 5 in the recognition state estimation process of the driver 8 performed in step S106. It is a flag that is turned on at times and turned off when it is estimated that the state of the road on the left rear side is not recognized.

  (Step S114) It is determined whether or not the right turn signal switch 80b is turned on. When the right turn signal switch 80b is ON, the process proceeds to step S116, and when the left turn signal switch 80a is OFF, the process returns to step S101.

  (Step S116) It is determined whether the traffic jam flag Fj is ON. When the traffic jam flag Fj is ON, the process proceeds to step S118, and when the traffic jam flag Fj is OFF, the process proceeds to step S122.

  (Step S118) It is determined whether the right rear side recognition flag Frr is ON. When the right rear side recognition flag Frr is ON, the process proceeds to step S120, and when the right rear side recognition flag Frr is OFF, the process proceeds to step S122. The right rear side recognition flag Frr is estimated that the driver 8 recognizes the state of the road on the right rear side of the vehicle 5 in the recognition state estimation process of the driver 8 performed in step S106. It is a flag that is turned ON when it is estimated that the state of the road on the right rear side is not recognized.

  (Step S120) The alarm suppression unit 60 suppresses the secondary alarm. Specifically, as shown in FIG. 11, the LED 85 is turned on as visual information. At this time, the fact that the alarm is suppressed is displayed on the display 95 using character information or the like according to an instruction from the alarm output unit 70.

  (Step S122) The alarm suppression unit 60 performs non-suppression of the secondary alarm. Specifically, as shown in FIG. 11, the LED 85 blinks as the enhanced visual information and the buzzer 90 sounds as the auditory information.

(Step S124) The approaching vehicle detection unit 20 determines whether an end condition with the BSW 10 is satisfied. Specifically, the BSW is configured to automatically end when the vehicle speed v of the vehicle 5 reaches a predetermined vehicle speed (for example, 0 km / h) set in advance.
(Explanation of traffic jam detection process)

  Next, the flow of the traffic jam detection process performed by the traffic jam detection unit 50 will be described with reference to the flowchart of FIG.

  (Step S <b> 200) In response to an instruction from the traffic jam detection unit 50, the car navigation system 110 determines whether traffic information has been received from the traffic information reception unit 120. When the traffic information is received, the process proceeds to step S202, and when the traffic information is not received, step S200 is repeated.

  (Step S202) With reference to the traffic information received by the car navigation system 110 by an instruction from the traffic congestion detection unit 50, whether or not the road on which the vehicle 5 is currently traveling is congested in the traveling direction is determined. Determine. If it is determined that there is traffic, the process proceeds to step S204. If it is determined that there is no traffic, the process proceeds to step S206.

  (Step S204) The traffic jam detection unit 50 turns on the traffic jam flag Fj. Then, the process returns to the main routine (FIG. 3). The traffic jam flag Fj is a flag that is given when it is determined that the road on which the vehicle 5 is currently traveling is congested in the traveling direction.

  (Step S206) The traffic jam detection unit 50 turns off the traffic jam flag Fj. Then, the process returns to the main routine (FIG. 3).

  Various methods other than the method shown in FIG. 4 are conceivable as a method for detecting the traffic jam.

  For example, vehicle speed data v (t) indicating that the vehicle speed v of the vehicle 5 obtained from the vehicle speed sensor 100 changes with time t is measured over a predetermined time, and the measured vehicle speed data The average vehicle speed ave (v (t)) is calculated by averaging v (t). Then, a difference value obtained by subtracting the average vehicle speed ave (v (t)) from the speed limit v0 of the currently traveling road acquired from the car navigation system 110 (see FIG. 2) is calculated. When this difference value is larger than a preset speed difference threshold value vth, it is determined that the currently running road is congested. The speed difference threshold value vth is a value set based on an evaluation experiment or the like.

  FIG. 5 is a flowchart for performing this process.

  (Step S210) In accordance with an instruction from the traffic jam detection unit 50, the speed limit v0 of the currently traveling road stored in the car navigation system 110 is acquired.

  (Step S212) In the traffic jam detection unit 50, the vehicle speed data v (t) obtained over a predetermined time is averaged to calculate the average vehicle speed ave (v (t)) of the vehicle 5.

  (Step S214) A difference value v0−ave (v (t)) obtained by subtracting the average vehicle speed ave (v (t)) calculated in Step S212 from the speed limit v0 acquired in Step S210 is calculated, and the calculated difference is calculated. It is determined whether or not the value is larger than the speed difference threshold value vth. When the difference value is larger than the speed difference threshold value vth, the process proceeds to step S216, and otherwise, the process proceeds to step S218.

  (Step S216) The traffic jam detection unit 50 turns on the traffic jam flag Fj. Then, the process returns to the main routine (FIG. 3).

  (Step S218) The traffic jam detector 50 turns off the traffic jam flag Fj. Then, the process returns to the main routine (FIG. 3).

The traffic jam detection unit 50 performs traffic jam detection by executing one of the processes shown in FIGS. In order to perform the processing shown in FIG. 4, it is necessary to obtain traffic information from the outside of the vehicle 5. For this reason, it is difficult to detect traffic jams in places where infrastructure for obtaining traffic information is not yet developed or where it is difficult to receive traffic information. For this reason, the two processes shown in FIGS. 4 and 5 may be performed in parallel to improve the reliability of traffic jam detection.
(Explanation of approaching vehicle detection process)

  Next, the flow of the approaching vehicle detection process performed by the approaching vehicle detection unit 20 will be described with reference to the flowchart of FIG.

  (Step S300) The images captured by the rear camera 15a, the left rear camera 15b, and the right rear camera 15c constituting the rear side imaging unit 15 are input to the approaching vehicle detection unit 20.

  (Step S302) Hereinafter, the processing from step S302 to step S308 is performed in the approaching vehicle detection unit 20. First, in this step S302, a pixel whose brightness differs greatly between adjacent pixels is detected as a luminance change point from the input image.

  (Step S304) Next, the images taken at different times are compared to calculate an optical flow. The optical flow is information representing movement between two images observed at different times, and is vector information connecting two points associated with each other as representing the same point between the two images. expressed. In this step S304, two images captured by the same camera at different times are compared, the brightness change points detected in step S302 are associated, and an optical flow is calculated. Since a method for calculating corresponding points between images taken at different times is widely used in searching for corresponding points of a stereo camera and the like, detailed description thereof is omitted.

  (Step S306) Only the optical flow toward the vehicle 5 is detected from the optical flows calculated in Step S304. Specifically, the direction of the optical flow calculated in step S304 is analyzed, and only the optical flow approaching the vehicle 5 is extracted. Here, since the mounting positions and mounting directions of the rear camera 15a, the left rear camera 15b, and the right rear camera 15c are known in advance, the other vehicle 6 traveling in the adjacent traveling lanes Y1, Y3 approaches the vehicle 5. The direction of the optical flow observed when the vehicle is in the vehicle can be predicted in advance. Therefore, it is only necessary to detect an optical flow having a predicted direction.

  (Step S308) Next, an optical flow group detected at positions close to each other from the optical flows toward the vehicle 5 and having substantially the same size is detected. The optical flow group detected in this way constitutes an approaching vehicle.

  (Step S310) It is determined whether an approaching vehicle is detected. When an approaching vehicle is detected, the process proceeds to step S312. When an approaching vehicle is not detected, the process returns to the main routine (FIG. 3).

(Step S312) A primary warning is output by the action of the warning output unit 70 to notify that an approaching vehicle has been detected. Specifically, the LED 85 installed near the door mirror on the side where the approaching vehicle is detected is turned on. Thereafter, the process returns to the main routine (FIG. 3).
(Explanation of driver's gaze direction detection principle)

  Next, the gaze direction detection process of the driver 8 performed by the gaze direction detection unit 40 will be described with reference to FIGS.

  When the range including the face of the driver 8 is imaged by the driver imaging unit 30, the images shown in FIGS. 7A to 7D are obtained. FIG. 7A is an example of an image of the face of the driver 8 facing the front. FIG. 7B is an example of an image of the face of the driver 8 facing slightly to the left. FIG. 7C is an example of an image of the face of the driver 8 facing further left. FIG. 7D is an example of an image of the face of the driver 8 facing slightly to the right.

  As can be seen from FIGS. 7A to 7D, the distance between both eyes changes depending on the direction in which the user is gazing. If the face rotation angle is increased, one eye is hidden as shown in FIG. 7C, so that both eyes are not observed. In such a case, the distance between the observed eyeball and the edge of the face is regarded as the distance between both eyes.

  Now, as shown in FIG. 7A, the distance between the eyes when the driver 8 is facing the front, that is, when the face orientation θ = 0, is Dee. When the face direction is gradually moved to the left from the state, the face direction θ = θ2 (> 0) is obtained through the state of FIG. 7B in which the face direction θ = θ1 (> 0). The state shown in FIG.

  When in the state of FIG. 7B, the distance between the eyes is dee (<Dee). Further, at this time, the distance p (p> 0) between the center line m1 of the face and the center line m2 of both eyes is uniquely determined.

  Further, when in the state of FIG. 7C, the distance between both eyes is dee (<Dee). Further, as in FIG. 7B, the distance p (p> 0) between the center line m1 of the face and the center line m2 of both eyes is uniquely determined.

The face direction θ of the driver 8 can be calculated by θ = cos ⁻¹ (dey / Deye).

  When the driver 8 turns his face to the right, the state of FIG. 7D where the face orientation θ = θ3 (<0) is observed. At this time, as shown in FIG. 7D, the center line m2 of both eyes is observed on the left side of the center line m1 of the face. In this state, it is considered that the distance p between the center line m1 of the face and the center line m2 of both eyes is negative. In this way, it is possible to determine whether it is facing right or left by the sign of the distance p between the center line m1 of the face and the center line m2 of both eyes.

  Here, there are individual differences in the distance between both eyes, and there is also an individual difference in the seating position of the driver 8, so it is assumed that the distance Eye between the eyes differs depending on the driver 8.

  Therefore, the distance Deye between both eyes may be measured in advance when the driver 8 is seated before starting driving, and the measured value may be stored.

  During driving, the distance between the eyes of the driver 8 is measured at any time.

  FIG. 8 is a scatter diagram in which the face orientation θ calculated from the binocular distance deye measured when a specific driver 8 gets on the right-hand drive vehicle and gazes at the left and right door mirrors. .

  Since the driver 8 can move the eyeball and the face independently, there are various variations in the gaze method when the left and right door mirrors are watched. That is, there are cases where the movement of the face is reduced and the door mirror is gazed with a horizontal eye, or when the face is moved greatly and the eyeball is confronted with the door mirror.

  Therefore, when the driver 8 gazes the left and right door mirrors several times, as shown in FIG. 8, the face orientation θ calculated from the binocular distance deye varies.

  That is, sample points eli (i = 1, 2,...) Representing the face orientation calculated when the left door mirror is watched form a cluster S1 as shown in FIG. Further, sample points eri (i = 1, 2,...) Representing the face orientation calculated when the right door mirror is watched form a cluster S2 as shown in FIG. Here, the cluster S1 represents a range of ± 2σ (σ is a standard deviation) predicted when the distribution state of the sample points eli (i = 1, 2,...) Is estimated to be a normal distribution. . The cluster S2 represents a range of ± 2σ (σ is a standard deviation) predicted when the distribution state of the sample point eri (i = 1, 2,...) Is estimated to be a normal distribution.

  That is, when the left door mirror is watched, it is estimated that the face orientation θ is observed in a range of θlmin to θlmax. Further, when the right door mirror is watched, it is estimated that the face orientation θ is observed in a range of θrmin to θrmax.

  Therefore, when the face orientation θ calculated from the binocular distance deye is in the range of θlmin to θlmax, it is estimated that the driver 8 is gazing at the left door mirror. Further, when the face orientation θ calculated from the binocular distance deye is in the range of θrmin to θrmax, it is estimated that the driver 8 is gazing at the right door mirror.

  Note that, when applied to a real scene, the driver 8 has to gaze the left and right door mirrors in order several times in advance, and each time the face orientation θ is calculated, θlmin, θlmax, θrmin described in FIG. , Θrmax is preferably subjected to calibration for calculating a range.

Since FIG. 8 shows the result of measurement with a right-hand drive vehicle, the right door mirror is closer to the driver 8. Therefore, the absolute value of the face direction θ when the right door mirror is watched is observed to be smaller than the absolute value of the face direction θ when the left door mirror is watched. In the case of a left-hand drive vehicle, the opposite result is obtained.
(Explanation of driver's gaze direction detection principle)

  In order to measure the face orientation θ of the driver 8, it is necessary to specify the position of the face of the driver 8 and the position of the eyeball. These processes are performed in the gaze direction detection unit 40.

  Details of the processing will be described with reference to FIG. Note that various methods have been proposed for image processing for detecting the face area and the position of the eyeball from the captured image, such as face detection in a camera. Therefore, any method may be used. That is, the method described below is only an example.

  The position of the face of the driver 8 is detected from the image picked up by the driver image pickup unit 30 by template matching using a face model. Then, after the face area is detected, the position of the face center line m1 shown in FIG. 7 is specified.

As for the position of the eyeball of the driver 8, an area that matches the eyeball model prepared in advance is searched from the image captured by the driver imaging unit 30. At this time, using conditions such as the eyeball being symmetrical with respect to the center line of the face, the black eye (pupil) region being sandwiched by the white eye (sclera) region, etc. Detection is performed by narrowing down the eyeball candidate region. Then, after the eyeball position is detected, the distance between both eyes shown in FIG. 7 specifies dee and the position of the center line m2 of both eyes.
(Explanation of driver's rear side recognition state estimation method)

  Next, the principle of a process for estimating whether or not the driver 8 recognizes the state of the road on the rear side of the vehicle 5 performed by the vehicle surrounding state recognition state estimation unit 45 will be described.

  In general, before changing the lane, the driver 8 looks at the adjacent traveling lane on the side where the lane is changed, and confirms that there is no approaching vehicle or that the lane can be changed even if an approaching vehicle exists. Check. Specifically, the presence of the approaching vehicle and the state are recognized by gazing at the door mirror.

  At this time, if you focus on the confirmation of the rear side, you will not be able to obtain the information ahead of the road, so generally, while checking the road conditions ahead, look at the door mirror several times, The state of the rear side is checked.

  Therefore, the vehicle surrounding situation recognition state estimation unit 45 counts the number of times the door mirror is watched during a preset predetermined time, and when it is detected that the door mirror is watched more than a predetermined number of times, the driver 8 It is presumed that the state of the road behind the door mirror is recognized.

  As the predetermined time, for example, a time of about 3 seconds is set. For example, when the door mirror is watched three times or more during the predetermined time, it is estimated that the driver 8 recognizes the state of the road behind the door mirror. The number of times of gazing at the door mirror is reset every time a predetermined time elapses, and information on how many times the door mirror has been gazed during the latest predetermined time is always held.

Note that the values described as the predetermined time and the number of times of gazing at the door mirror are examples, and actually, appropriate values are set according to the traveling state of the vehicle 5 at that time. For example, when the vehicle speed is high, it is more difficult to look away from the front than when the vehicle speed is low. Therefore, it is estimated that the state of the road on the rear side can be recognized with a shorter predetermined time and fewer gazes. Also good.
(Explanation of driver's rear side recognition state estimation process)

  Next, the flow of the driver's recognition state estimation process performed by the vehicle surrounding state recognition state estimation unit 45 will be described with reference to the flowchart of FIG.

  (Step S400) The gaze direction detecting unit 40 detects the face direction of the driver 8. Details will be described later.

  (Step S402) A left viewing time timer Tl for measuring a predetermined time for determining how many times the left door mirror is watched during a predetermined time interval set in Step S100 (see FIG. 3), and a predetermined The right-side visual recognition time timer Tr for measuring a predetermined time for determining how many times the right door mirror is watched during the time interval is decremented.

  (Step S404) It is determined whether or not the left visual recognition time timer Tl has become zero. When the left visual recognition time timer Tl is 0, the process proceeds to step S408, and otherwise, the process proceeds to step S406.

  (Step S406) It is determined whether or not the right side visual recognition time timer Tr has become zero. When the right side visual recognition time timer Tr is 0, it progresses to step S414, and when that is not right, it progresses to step S420.

  (Step S408) The left rear side gaze counter Clg is reset.

  (Step S410) The left rear side gaze flag Flg is turned off.

  (Step S412) The left visual recognition time timer Tl is set.

  (Step S414) The right rear side gaze counter Crg is reset.

  (Step S416) The right rear side gaze flag Frg is turned OFF.

  (Step S418) The right side visual recognition time timer Tr is set.

  (Step S420) It is determined whether the left rear side gaze flag Flg is ON. When the left rear side gaze flag Flg is ON, the process proceeds to step S422. Otherwise, the process proceeds to step S424.

  (Step S422) The left rear side gaze counter Clg is incremented.

  (Step S424) It is determined whether or not the right rear side gaze flag Frg is ON. When the right rear side gaze flag Frg is ON, the process proceeds to step S426, and otherwise, the process returns to the main routine (FIG. 3).

  (Step S426) The right rear side gaze counter Crg is incremented.

  (Step S428) It is determined whether or not the value of the left rear side gaze counter Clg exceeds a preset threshold value Lth. When the value of the left rear side gaze counter Clg exceeds the threshold value Lth, the process proceeds to step S430, and otherwise, the process returns to the main routine (FIG. 3).

  (Step S430) The left rear side recognition flag Flr is turned ON and the process returns to the main routine (FIG. 3).

  (Step S432) It is determined whether or not the value of the right rear side gaze counter Crg has exceeded a preset threshold value Rth. When the value of the right rear side gaze counter Crg exceeds the threshold value Rth, the process proceeds to step S434. Otherwise, the process returns to the main routine (FIG. 3).

(Step S434) The right rear side recognition flag Frr is turned ON and the process returns to the main routine (FIG. 3).
(Explanation of driver's gaze direction detection process)

  Next, the flow of the driver's gaze direction detection process performed by the gaze direction detection unit 40 will be described with reference to the flowchart of FIG.

  (Step S500) The range including the face of the driver 8 is imaged.

  (Step S502) The gaze direction detecting unit 40 detects the face of the driver 8 from the captured image.

  (Step S504) Based on the result of detecting the face of the driver 8, the position of the eyeball of the driver 8 is further detected.

  (Step S506) Based on the detection result of the eyeball of the driver 8, the distance deye between both eyes is measured.

  (Step S508) A gaze direction that is the direction of the face of the driver 8 is calculated.

  (Step S510) Based on the gaze direction calculated in step S508, the driver 8 determines whether or not the left door mirror is gaze. If it is determined that the left door mirror is being watched, the process proceeds to step S512. Otherwise, the process proceeds to step S514.

  (Step S512) The left rear side gaze flag Flg is turned ON and the process returns to the main routine (FIG. 3).

  (Step S514) Based on the gaze direction calculated in step S508, the driver 8 determines whether the right door mirror is being gaze. If it is determined that the right door mirror is being watched, the process proceeds to step S516, and otherwise, the process returns to step S500.

  (Step S516) The right rear side gaze flag Frg is turned ON, and the process returns to the main routine (FIG. 3).

  As described above, according to the rear side approaching object alarm device 10 according to the present invention configured as described above, the traffic congestion detection unit 50 detects that the road on which the vehicle 5 is traveling is congested. In addition, when the vehicle surrounding situation recognition state estimation unit 45 estimates that the driver 8 recognizes the state of the other vehicle 6 traveling behind the adjacent traveling lane Y1 (Y3), When the person 8 operates the turn indicator to change the lane, the approaching vehicle detection unit 20 determines a predetermined distance behind the adjacent traveling lane Y1 (Y3) where the vehicle 5 is changing the lane. Even when another vehicle 6 approaching the vehicle 5 is detected in the range, the alarm suppression unit 60 suppresses the alarm output from the alarm output unit 70. Therefore, when it is estimated that the driver 8 is aware of the state of the road on the rear side of the adjacent traveling lane Y1 (Y3) on the side where the lane is changed, the troublesomeness caused by the output of an unnecessary warning is reduced. Can be eliminated.

  Moreover, according to the rear side approaching object alarm device 10 according to the present invention configured as described above, the alarm output unit 70 outputs visual information including light and display and auditory information including sound and voice. Therefore, since the alarm suppression unit 60 suppresses the output of information other than the visual information, the troublesomeness at the time of alarm output can be reduced.

  Furthermore, according to the rear side approaching object alarm device 10 according to the present invention configured as described above, when the alarm output from the alarm output unit 70 is suppressed by the alarm suppression unit 60, the display 95 (alarm suppression The status display section) indicates that the alarm output from the alarm output section 70 is suppressed, so even if the driver 8 notices that the secondary alarm is not output, the cause However, since it can be understood that it is not a failure of the rear side approaching object warning device 10, a sense of security can be given.

  And according to the rear side approaching object warning device 10 which concerns on this invention comprised in this way, the vehicle surrounding condition recognition state estimation part 45 is the rear side in the predetermined time interval of the driver | operator 8 of the vehicle 5. When the direction confirmation frequency is detected and the rear side confirmation frequency is higher than a predetermined frequency, it is estimated that the driver 8 recognizes the state of the road behind the adjacent traveling lane Y1 (Y3). Therefore, the recognition state of the driver 8 can be reliably estimated.

  Moreover, according to the rear side approaching object warning device 10 according to the present invention configured as described above, the frequency of the rear side confirmation within a predetermined time interval is determined so that the face of the driver 8 travels adjacently within the predetermined time. Since it is calculated based on the frequency of turning to the direction of the rear-view mirror on the lane Y1 (Y3) side, the driver's rear side confirmation frequency can be calculated by simple image processing.

  Moreover, according to the rear side approaching object warning device 10 according to the present invention configured as described above, the traffic jam detection unit 50 has received traffic jam information indicating that there is traffic jam around the current position of the vehicle 5. By detecting this, it is determined that the running road is congested, so that the traffic jam situation around the vehicle 5 can be detected by simple processing.

  Furthermore, according to the rear side approaching object warning device 10 according to the present invention configured as described above, the traffic jam detection unit 50 calculates the average of a predetermined time of the vehicle 5 from the speed limit v0 of the road on which the vehicle 5 is traveling. A difference value obtained by subtracting the vehicle speed ave (v (t)) is calculated. When the difference value is larger than the speed difference threshold value vth (predetermined value), it is determined that the road is congested. Even in a situation where reception is not possible, the traffic situation around the vehicle 5 can be detected.

  In the first embodiment, the other vehicle 6 is detected from the image captured by the rear side imaging unit 15, but the detection of the other vehicle 6 is, for example, as described in Patent Document 1, It can also be performed using a distance measuring sensor such as a laser radar. In addition, the detection accuracy can be improved by using a rear side imaging unit 15 formed of a camera and a distance measuring sensor such as a laser radar in combination.

  In the first embodiment, the rear side imaging unit 15 is configured using the three cameras of the rear camera 15a, the left rear camera 15b, and the right rear camera 15c, and the other vehicle 6 is detected. This is not limited to the configuration of three cameras. That is, the rear side imaging unit 15 may be configured by using only one rear camera 15a or two cameras, the left rear camera 15b and the right rear camera 15c.

  Further, in Example 1, the BSW 10 was activated by the speed of the vehicle 5 or the switch operation of the driver 8. However, the BSW 10 is a system that is originally used while traveling on a road with two or more lanes on one side. Accordingly, the map information stored in the car navigation system 110 is referred to the currently traveling road type information so that it is activated only when the currently traveling road is two or more lanes on one side. Also good.

  Further, in the first embodiment, an alarm is output based on visual information and auditory information. However, antennal information can be used instead of auditory information. For example, an actuator that outputs vibration information is installed on the seat on which the driver 8 is seated, and the actuator installed on the seat is vibrated in accordance with the blinking of the LED when the secondary alarm is output. Information can also be transmitted to the back and legs of the driver 8. Also in this case, the secondary alarm can be suppressed by the action of the alarm suppression unit 60.

  In the first embodiment, it is estimated that the driver 8 is gazing at the door mirror of the vehicle 5 and recognizes the state of the road on the rear side. In a vehicle equipped with a monitoring system that captures the state of the road on the rear side with a camera and displays it on a monitor in the vehicle, the driver 8 watches the monitor more than a predetermined number of times within a predetermined time. It can be estimated that the state of the road on the rear side of the vehicle 5 is recognized.

  Furthermore, it is possible to estimate that the driver 8 recognizes the state of the road on the rear side of the vehicle 5 by detecting the action of the driver 8 looking directly at the rear side of the vehicle 5 instead of gazing at the door mirror. . In this case, in place of the gaze direction detection process described in the embodiment, a process for detecting a larger facial movement is performed. For example, a process of detecting a state in which the face is greatly rotated left and right by associating images including the face of the driver 8 captured continuously at different times, or directly viewing the left and right rear sides prepared in advance This can be executed by performing template matching between the template image being processed and the actually captured image.

  Further, the congestion detection method described in the first embodiment is not limited to the described method. That is, when the number of vehicles in the adjacent lane is detected by the surrounding monitoring camera or radar installed in the rear side approaching object warning device 10, and the detected vehicles exceed a certain number of detections within a certain time. In addition, it can be determined that the running road is congested. Further, the vehicle speed of surrounding vehicles can be detected using the inter-vehicle communication function, and when the vehicle speed of the surrounding vehicles is as low as that of the vehicle 5, it can be determined that the running road is congested. .

  As mentioned above, although the Example of this invention was explained in full detail with drawing, since an Example is only an illustration of this invention, this invention is not limited only to the structure of an Example. Of course, changes in design and the like within a range not departing from the gist are included in the present invention.

5 Vehicle 8 Driver 10 Rear side approaching object warning device (BSW)
15 rear side imaging unit 15a rear camera 15b left rear camera 15c right rear camera 20 approaching vehicle detection unit 30 driver imaging unit 40 gaze direction detection unit 45 vehicle surrounding state recognition state estimation unit 50 traffic congestion detection unit 60 alarm suppression unit 70 alarm Output part 80 Turn signal switch 80a Left turn signal switch 80b Right turn signal switch 85 LED
90 Buzzer 95 Display (alarm suppression status display)
100 vehicle speed sensor 110 car navigation system 120 traffic information receiving unit 130 CAN communication line 150 camera controller

Claims (7)

An approaching vehicle detection unit that detects that an approaching vehicle exists in a predetermined distance range on the rear side of the vehicle in an adjacent traveling lane adjacent to the traveling lane of the vehicle;
A vehicle surrounding state recognition state estimation unit that estimates that the driver of the vehicle recognizes the state of the road on the rear side of the adjacent traveling lane;
A traffic jam detector for detecting that the road on which the vehicle is running is jammed;
When the approaching vehicle detection unit detects an approaching vehicle, an alarm output unit that outputs an alarm when the driver tries to change the lane to the adjacent traveling lane;
When the traffic congestion detection unit detects traffic congestion and the vehicle surrounding situation recognition state estimation unit estimates that the driver recognizes the state of the road behind the adjacent traveling lane, A rear side approaching object alarm device comprising: an alarm suppression unit that suppresses output of an alarm from the alarm output unit when a driver tries to change a lane to the adjacent traveling lane.   The alarm output unit outputs visual information including light and display, and auditory information including sound and voice or contact information including vibration, and the alarm suppression unit includes the visual information. The rear side approaching object warning device according to claim 1, wherein output of information other than that is suppressed.   When an alarm output from the alarm output unit is suppressed by the alarm suppression unit, an alarm suppression state display unit indicating that an alarm output from the alarm output unit is suppressed is provided. The rear side approaching object warning device according to claim 1 or 2.   The vehicle surrounding state recognition state estimation unit detects a rear side confirmation frequency within a predetermined time interval of the driver of the vehicle, and the driving is performed when the rear side confirmation frequency is higher than a predetermined frequency. The rear side approaching object warning device according to any one of claims 1 to 3, wherein the person estimates that the state of the road on the rear side of the adjacent lane is recognized.   The rear side confirmation frequency is calculated based on a frequency at which the driver's face faces the direction of the rear-view mirror on the side of the adjacent traveling lane within the predetermined time interval. The rear side approaching object alarm device according to claim 4.   The traffic jam detection unit determines that the running road is jammed by detecting that the traffic jam information indicating that the current position of the vehicle is jammed is reported. The rear side approaching object alarm device according to any one of claims 1 to 5.   The traffic jam detection unit calculates a difference value obtained by subtracting an average vehicle speed for a predetermined time of the vehicle from a speed limit of the road on which the vehicle is traveling, and when the difference value is larger than a predetermined value, the road The rear side approaching object warning device according to any one of claims 1 to 5, wherein it is determined that the traffic is congested.