JP2008077189A - Falling asleep while diving prevention device and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent the falling asleep of a driver by issuing warning in the optimal timing. <P>SOLUTION: A falling asleep while diving prevention device is provided with: a camera which picks up the face image of the driver of a vehicle; a closed eye judging part 32 for detecting the opened eyes or closed eyes of the driver, based on the face image of the driver picked up by the camera; a long closed eye judging part 33 for judging whether the driver has closed his or her eyes in a predetermined time or more, based on the detection result of an opened/closed eye detecting part 31; a reminding device 40 for reminding the driver since it is judged that the driver has closed the eyes in a predetermined time or longer by the long closed eye judging part 33 until the opened eyes are judged by the closed eye judging part 32; and a warning device 50 for issuing warning to the driver after the opened eyes are judged by the closed eye judging part 32. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a snooze driving prevention apparatus and program suitable for use in a driving support system.

  Conventionally, various techniques for detecting a driver's sleep and issuing an alarm have been proposed.

  Patent Document 1 discloses a dozing detection device that measures the eye closing time of a person to be detected and detects a dozing state based on the eye closing time. The time length of the eye closing time, the state of the person to be detected, the person to be detected According to at least one of the above environments, the weight of the doze evaluation criterion for one closed eye is changed.

  The dozing detection device of Patent Document 1 utilizes a phenomenon in which the eye-closing time increases when drowsiness increases. Specifically, the dozing detection device performs weighting according to the length of one eye-closing time, then accumulates the weighted eye-closing time, and when the integrated value becomes a certain level or more, It is judged and alarms.

  Patent Document 2 discloses a drowsiness driving alarm device that issues a warning in accordance with driver preference and driver characteristics. The dozing operation warning device of Patent Document 2 changes the dozing operation determination criterion of the dozing operation detection unit based on the length of the alarm operation time from when the alarm is issued until the alarm is stopped by manual operation.

  Patent Document 3 discloses a drowsiness prevention device that maintains the driver's arousal level and prevents the driver from falling asleep.

The dozing driving prevention device of Patent Document 3 is a technique for detecting a state before the dozing state and preventing the dozing operation. Specifically, the snooze driving prevention apparatus stores an operation amount of an object operated by a driver, sets a first reference operation amount based on the map data input from the navigation unit, and performs the operation It is determined whether or not the operation amount has decreased by comparing the amount and the first reference operation amount. When it is determined that the operation amount has decreased, the operation load is applied to the driver, and the content is displayed on the display unit.
JP 2002-279410 A JP 2002-183900 A JP 2005-141315 A

  However, even if the drowsiness detection device of Patent Document 1 is used, it is difficult to estimate the timing at which the driver wants an alarm even if the weight of the eye-closing time is changed because the way of feeling sleepiness varies from person to person. Also, if the alarm is issued early, it becomes annoying, and if the alarm is late, it becomes useless.

  The reason is as follows: (1) A warning is presented without the driver knowing how the drowsiness detection device of Patent Document 1 determines sleepiness; (2) Subjective sleepiness and Patent Document 1 This is considered to be because the time lag with sleepiness determined by the snoozing detection device is large. Furthermore, since the dozing detection device of Patent Document 1 issues a warning from the result of integrating past eye closure times, there is a high probability that the driver is closed at the time of warning, and the driver is surprised at the time of warning. There was also.

  The doze driving alarm device of Patent Document 2 synchronizes the alarm timing with the driver's feeling of drowsiness, so that the function as an alarm is impaired, and as a result, the alarm may be delayed.

  The snooze driving prevention device of Patent Document 3 has a problem that the driver's inability to understand intuitively is small and the effect of alerting is small because the decrease in the operation amount and drowsiness are not directly linked. Further, there is a problem that an operation load that is not essential for driving is troublesome for the driver.

  The present invention has been proposed to solve the above-described problems, and an object of the present invention is to provide a drowsiness prevention device and program for generating a warning at an optimal timing to prevent a driver from falling asleep.

  The snooze driving prevention apparatus according to the present invention determines whether the driver's eyes are open or closed based on an imaging unit that captures a face image of the driver of the vehicle and the driver's face image captured by the imaging unit. Based on the determination result of the open / closed eye determination unit, the long-closed eye determination unit that determines whether the driver is closed for a predetermined time or more based on the determination result of the open / closed eye determination unit; From the time when it is determined that the eyes are closed to the time when the opening / closing eye detection means detects the opening of the eyes, the attention raising means for calling attention to the driver, and the opening / closing eyes detection means detects the opening of the eyes. And alarm means for issuing an alarm to the driver.

  The doze driving prevention program according to the present invention is based on the driver's face image captured by the imaging means, and the opening / closing eye determination means for determining whether the driver's eyes are open or closed, and the determination by the opening / closing eye determination means Based on the result, the long-closed eye determining means for determining whether or not the driver has closed his eyes for a predetermined time or more, and when the driver has determined that the driver has closed his eyes for a predetermined time or longer, the opening and closing Until the time when eye opening is detected by the eye detection means, attention driver control means for controlling the alert means so as to alert the driver, and after the eye opening is detected by the opening / closing eye detection means, the driver It functions as an alarm control means for controlling the alarm means so as to issue an alarm.

  The open / close eye detection means detects the driver's eye opening or closing based on the driver's face image imaged by the imaging means. The long closed eye determination means determines whether the driver is closed for a predetermined time or more based on the determination result of the open / close eye determination means. The alerting means alerts the driver from the time when it is determined that the driver is closed for a predetermined time or more until the eye is detected. Then, the warning means issues a warning to the driver after alerting by the alerting means, that is, after eye opening is detected.

  Therefore, the above invention alerts the driver from when it is determined that the driver is closed for a predetermined time or more until the eye is detected, and issues an alarm to the driver after the eye is detected. Thus, an alarm can be issued at an optimal timing that does not surprise the driver.

  The snooze driving prevention apparatus and program according to the present invention can issue an alarm to a driver at an optimal timing.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

[First Embodiment]
FIG. 1 is a block diagram showing the configuration of the snooze driving prevention apparatus according to the first embodiment of the present invention. The snooze driving prevention device is installed, for example, diagonally forward of the driver, and the camera 10 that images the driver's face from diagonally above, and the alert device 40 and the alarm device 50 based on the face image captured by the camera 10. And a computer 20 for controlling the operation.

  The computer 20 includes a CPU 21 that performs image processing and arithmetic processing for controlling the alerting device 40 and the alarm device 50, a RAM 22 that is a data work area, and a ROM 23 that stores a control program for the CPU 21. I have. The computer 20 configured as described above is functionally configured as follows.

  FIG. 2 is a block diagram illustrating a functional configuration of the computer 20. The computer 20 includes an open / closed eye determining unit 31 that determines whether the driver is open or closed, an eye-closed determining unit 32 that determines whether the driver is in the closed state, a long-closed eye determining unit 33 that determines whether the driver is closed for a long time, A long eye-closing time storage unit 34 for storing a time for which eyes are closed continuously for a long time (long eye-closing time), a long eye-closing time integrating unit 35 for integrating the long-closing time, and an alarm determining unit 36 for determining whether to issue an alarm It is equipped with.

  Moreover, the alerting device 40 shown in FIG. 1 is controlled based on the determination result of the long-closed eye determination unit 33. The alarm device 50 is controlled based on the determination result of the alarm determination unit 36.

  FIG. 3 is a diagram illustrating an example of the alerting device 40 and the alarm device 50. The alerting device 40 is not particularly limited as long as it can alert the driver. For example, the alert device 40 may be provided in any one of a driver seat vibration device, a steering vibration device, a vehicle floor vibration device, an accelerator pedal vibration device, and a speaker device (audio system or navigation system) that outputs sound or sound. Yes, at least one such as a car air conditioner or a seat air conditioner that emits hot or cold air is applicable.

  The alarm device 50 is not particularly limited as long as it can issue an alarm to the driver. For example, the alarm device 50 may be provided in any one of a driver seat vibration device, a steering vibration device, a vehicle floor vibration device, an accelerator pedal vibration device, and a speaker device (audio system or navigation system) that outputs sound or sound. .), At least one of a car air conditioner or a seat air conditioner that emits hot or cold air, a light emitting element that lights or flashes, an information display device, and the like.

  Note that the alerting device 40 and the alarm device 50 may be different devices or the same device. For example, the speaker device may output a small sound when alerting the driver, and output a large sound when alerting the driver. Further, the speaker device may change the frequency and tone color between alerting and alarming.

  In the snooze driving prevention apparatus configured as described above, when the camera 10 captures the driver's face, the computer 20 executes the next snoozing driving prevention processing routine.

  FIG. 4 is a flowchart illustrating a dozing operation prevention process routine according to the first embodiment.

  In step S1, the following threshold values are set.

Long closed eye threshold i _th = a
Long closed eye cumulative time threshold (Σj) _th = b
Alarm time threshold k _th = c
Closed eye threshold L _th = d

In step S2, each parameter is initialized.
Closed eye count value (closed eye time) i = 0
Long closed eye duration count value (long closed eye duration) j = 0
Alarm time k = 0

  In step S <b> 3, the open / close eye detection unit 31 detects the open / close state of the driver's eyes by performing image processing on the driver's face image captured by the camera 10. For example, the open / closed eye detection unit 31 calculates the ratio of the current pupil length to the driver's pupil length stored in advance at the time of awakening (opening of the eye), and calculates the degree of eye closure L from the opening of the eye. (= 1-opening degree) is calculated. In addition, the calculation method of the eye closing degree L is well-known, and is not specifically limited.

In step S <b> 4, the eye closing determination unit 32 determines whether the eye is closed or in an open state based on the eye closing degree L detected by the opening / closing eye detection unit 31. Specifically, the eye closing determination unit 32 compares the eye closing degree L with the eye closing threshold L _th .

FIG. 5 is a diagram showing the closing degree L with respect to the closing eye threshold L _th . Closed-eye determination section 32 determines that the closed-eye state when the closing degree L exceeds the closed-eye threshold L _th, determines the open-eye state when the closing degree L does not exceed the closed-eye threshold L _th. And when it determines with an eye-closed state, it progresses to step S5, and when it determines with an eye-open state, it progresses to step S9.

In step S5, the long closed eye determination unit 33 performs long closed eye determination. Specifically, the long-eye-closed determination unit 33 determines whether or not the closed-eye count value i exceeds the long-closed threshold value i _th. It is determined that it is not. When it is determined that the eye is in the long closed state, the process proceeds to step S7, and when it is determined that the eye is not in the long closed state, the process proceeds to step S6.

In step S6, the closed eye count value i is incremented (i = i + 1), and the process returns to step S3. Accordingly, steps S3 to S6 are repeatedly executed, and when the closed eye count value i exceeds the long eye closing threshold value i _th , that is, when the eye closing time exceeds a predetermined time, the process proceeds to step S7.

  In step S7, the long closed eye duration count value j is incremented (j = j + 1). Here, since the long closed eye duration count value j is counted up from the time when it is determined that the eye is long closed (in the case of an affirmative determination in step S5), the long closed eye duration time (long eye duration) is indicated. ing.

  In step S8, the long closed eye determination unit 33 operates the alerting device 40 to alert the driver and returns to step S3. Thus, steps S3 to S5 and S7 to S8 are repeatedly executed until a negative determination is made in step S4, that is, until the driver opens his eyes, and the alerting device 40 operates continuously.

  On the other hand, in step S <b> 9, the long closed eye determination unit 33 turns off the alerting operation by the alerting device 40. Therefore, the alerting device 40 stops alerting when the driver opens his eyes.

  In step S10, the long closed eye storage unit 34 stores the last long closed eye count value j, that is, the long closed eye duration.

  In step S11, the long closed eye time accumulating unit 35 accumulates the long closed eye duration j from the present to a certain period before the long closed eye duration j stored in the long closed eye storage unit 34, The accumulated time Σj is calculated.

  FIG. 6 is a diagram illustrating the integration range. The integration range is expressed as a range from the present to a certain time ago. Therefore, if the current time changes, the integration range is also shifted by the change, and the integration time Σj also changes.

  In step S12, the warning determination unit 36 determines whether the long closed eye duration j is greater than zero. If the determination is affirmative, the process proceeds to step S13. If the determination is negative, the process returns to step S2.

In step S13, the warning determination unit 36 performs a warning determination as to whether or not to issue a warning. Specifically, the alarm determination unit 36 determines whether or not the integration time Σj calculated by the long eye closing time integration unit 35 has exceeded the integration time threshold value (Σj) _th . If the determination is affirmative, the process proceeds to step S14. If the determination is negative, the process returns to step S2. Therefore, the warning determination unit 36 performs a warning determination after a negative determination in step S4, that is, when the driver is in the open eye state. As a result, an alarm is not issued when the driver is in an eye-closed state, so that the driver can be prevented from being surprised by a sudden alarm.

  In step S14, the warning determination unit 36 operates the warning device 50 to issue a warning to the driver.

In step S15, the alarm determination unit 36 determines whether or not the alarm time k exceeds the alarm time threshold k _th so as to determine the alarm time. If the determination is negative, the process proceeds to step S16, where an affirmative determination is made. In this case, the process proceeds to step S17.

In step S16, the alarm time k is incremented (k = k + 1), and the process returns to step S15. Thus, until the alarm time k exceeds the alarm time threshold k _th, steps S15 and S16 is repeatedly executed.

  In step S17, the alarm determination unit 36 turns off the operation of the alarm device 50 and returns to step S2. Therefore, after the alarm is issued for a predetermined time, the process is executed again from step S2.

  As described above, the snooze driving prevention apparatus according to the first embodiment alerts the driver when the driver is closed for a long time, the accumulated time of the long eye-closing duration exceeds a predetermined time, and the driver Alerts the driver when the eyes open. As a result, the snooze driving prevention device not only alerts the driver to snooze, but can also issue an alarm without surprise the driver.

  In other words, the snooze driving prevention device gives the driver information on the eye closing time, so that the driver can intuitively understand that it is dangerous if the eyes are closed for a long time. In addition, the snooze driving prevention device provides a sensory stimulus to the driver at the timing when the eye is opened from the closed eye state when the accumulated time of the long eye-closed duration exceeds a predetermined time, and sensorially Can inform the danger. On the other hand, since the driver can easily recognize what the snooze driving prevention device detects, the degree of understanding of the snoozing driving prevention device increases and can be trusted.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. In addition, the same code | symbol is attached | subjected to the site | part same as 1st Embodiment, and a different part is mainly demonstrated.

  FIG. 7 is a block diagram showing a functional configuration of the computer 20 of the snooze driving prevention apparatus according to the second embodiment. The computer 20 is obtained by adding a threshold value changing unit 37 to the configuration shown in FIG.

  In the snooze driving prevention apparatus configured as described above, when the camera 10 captures the driver's face, the computer 20 executes the next snoozing driving prevention processing routine.

  FIG. 8 is a flowchart illustrating a dozing operation prevention processing routine according to the second embodiment. This routine executes step S20 between step S11 and step S12 of FIG.

In step S20, the threshold value changing unit 37 calculates the long eye closing threshold value i _th using the integrated time Σj calculated by the long eye closing time integrating unit 35.

FIG. 9 is a diagram illustrating a function f (Σj) for calculating i _th . According to the function f, the length closed-eye threshold i _th when from Σj zero to b ₁ is the maximum value a _max. When Σj is from b ₁ to b ₂ , the long closed eye threshold i _th decreases monotonously. The length closed-eye threshold i _th when Σj is b ₂ or more of a minimum value a _min.

The long eye closing threshold value i _th calculated by the threshold value changing unit 37 is used in the long eye closing determining unit 33 (step S5).

FIG. 10 is a diagram illustrating the frequency of alerting when the drowsiness level increases with time. Σj shown in FIG. 9 corresponds to the sleepiness level. Therefore, when the drowsiness level is small, the long-closed eye threshold i _th becomes the maximum value a _max , so the frequency of alerting is low and the driver does not feel annoying. On the other hand, when drowsiness increases, Σj increases, and thus the long-closed eye threshold i _th decreases toward the minimum value a _min . As a result, Σj is likely to increase, the frequency of alerting is increased at an accelerated rate, and the alarm timing is eventually advanced. On the other hand, when sleepiness is awakened and the eye-opening state continues, Σj decreases and the long-closed eye threshold i _th increases, so the frequency of alerting decreases and the alarm timing also delays.

  Thus, the frequency of alerting using the long closed eye duration is important information for the driver's sleepiness recognition.

  FIG. 11 is a diagram showing the frequency of eye closing with respect to a general eye closing time according to the state of sleepiness. When there is no drowsiness, blinking is quick (closed eye time is short), but when there is drowsiness, blinking is slow (closed eye time is long), and the occurrence frequency of a state where the closed eye time is long increases.

  Therefore, as shown in FIG. 10, the snooze driving prevention apparatus shortens the frequency of alerting as sleepiness progresses, and thus can inform the driver of the progress of sensory sleepiness. Then, the driver can check sleepiness by taking correspondence with his / her own sleepiness awareness. This means that the drowsy driving prevention device and the driver share the closed eye information.

  Therefore, when the drowsy driving prevention device gives a warning using the past eye-closing time, the driver can recognize the warning early and can respond to an appropriate vehicle operation.

  As described above, the snooze driving prevention apparatus according to the second embodiment can prevent delays in alerting and warning since Σj increases when sleepiness rapidly progresses. In addition, when the drowsiness awakens rapidly, the doze driving prevention device can delay the timing of alerting and warning so that the driver does not feel bothered.

  Furthermore, even when drowsiness progresses and issues an alarm, the snooze driving prevention device issues an alarm at the timing of opening the eye as in the first embodiment, so that it is possible to prevent the driver from being surprised by a sudden alarm. it can.

  Note that the present invention is not limited to the above-described embodiment, and it is needless to say that the present invention can also be applied to a design modified within the scope of the claims.

It is a block diagram which shows the structure of the dozing driving prevention apparatus which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the functional structure of a computer. It is a figure which shows an example of an alerting device and an alarm device. It is a flowchart which shows the dozing driving | operation prevention processing routine of 1st Embodiment. It is a diagram showing a closing degree L against the closed-eye threshold L _th. It is a figure which shows an integration range. It is a block diagram which shows the functional structure of the computer of the dozing driving prevention apparatus which concerns on 2nd Embodiment. It is a flowchart which shows the dozing driving | operation prevention processing routine of 2nd Embodiment. It is a figure which shows the function f (Σj) for calculating i _th . It is a figure which shows the frequency of alerting when a sleepiness degree becomes large with time. It is a figure which shows the closed eye frequency with respect to the general closed eye time according to the state of sleepiness.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Camera 20 Computer 31 Opening / closing eye detection part 32 Eye closing determination part 33 Long eye closing determination part 34 Long eye closing time memory | storage part 35 Long eye closing time integration part 36 Alarm determination part 37 Threshold change part 40 Warning device 50 Alarm apparatus

Claims (5)

Imaging means for imaging a face image of a driver of the vehicle;
Open / closed eye determination means for determining whether the driver's eyes are open or closed based on the driver's face image captured by the imaging means;
Based on the determination result of the opening / closing eye determination means, a long-closed eye determination means for determining whether the driver has closed his eyes for a predetermined time or more;
Attention alerting means for alerting the driver from when it is determined by the long eye closure determination means that the driver is closed for a predetermined time or more until when eye opening is detected by the opening / closing eye detection means; ,
Alarm means for issuing an alarm to the driver after opening of the eye is detected by the open / close eye detection means;
A snooze driving prevention device with From the present to the past certain period of time, further comprising an integrated time calculating means for calculating the integrated time by integrating the long closed eye duration when the driver closed the eyes for a predetermined time or more,
The alarm means issues an alarm to the driver after the accumulated time calculated by the accumulated time calculating means exceeds a predetermined accumulated time threshold value and the opening / closing eye detecting means detects an open eye. Item 1. The snooze driving prevention apparatus according to item 1. A closing threshold value setting means for setting a closing threshold value based on the accumulated time calculated by the accumulated time calculating means;
The opening / closing eye detection means calculates an eye closing degree based on the driver's face image, and the eye closing degree exceeds a closing degree threshold set by the closing degree threshold setting means. The driver's eyes are sometimes detected, and the eyes of the driver are detected when the degree of closing of the eyes does not exceed the closing threshold value. apparatus. The doze driving prevention device according to claim 3, wherein the closing threshold value setting unit decreases the closing threshold value as the integration time becomes longer. Computer
Open / closed eye determination means for determining whether the driver's eyes are open or closed based on the driver's face image captured by the imaging means,
A long-closed eye determination unit that determines whether the driver has closed his eyes for a predetermined time or more based on a determination result of the open / close eye determination unit;
From the time when it is determined that the driver has closed the eyes for a predetermined time or more by the long eye closing determination means to the time when the opening / closing eye detection means detects the opening of eyes, the driver is alerted Attention control means for controlling the means,
Alarm control means for controlling the alarm means so as to issue an alarm to the driver after the opening / closing eye detection means has detected eye opening;
A snooze driving prevention program that functions as

Images