JP2021144504A - Inattentive driving detection system - Google Patents

Abstract

To provide an inattentive driving detection system which can be retrofitted to a vehicle and is superior in detection accuracy.SOLUTION: An inattentive driving detection system 1 for detecting inattentive driving of a driver of a vehicle 10 comprises a speed detection device 2 which detects a speed of the vehicle 10, a vehicle compartment camera 3 which captures an image of the driver of the vehicle 10, and a control unit 5 connected to the speed detection device 2 and the vehicle compartment camera 3. When the driver takes an action to look back, the control unit 5 detects an angle and/or duration of driver's looking back on the basis of an image captured by the vehicle compartment camera 3 and determines whether the action to look back is considered as inattentive driving or not on the basis of the speed of the vehicle 10 detected by the speed detection device 2 and the angle and/or duration of driver's looking back.SELECTED DRAWING: Figure 1

Description

The present invention relates to an inattentive detection system.

Conventionally, in a vehicle, it has been required to prevent the driver from looking aside. Therefore, a detection device that detects the driving state of the driver of the vehicle and detects the inattentive driving has been proposed.

More specifically, an image input means for inputting image data of a face, an eye position detecting means for detecting an eye position from the input image data, a vehicle state detecting means for detecting a vehicle state when the vehicle is running, and a vehicle state detecting means. It was determined that there is a high probability that the driver is looking ahead of the vehicle based on the detection of the vehicle condition detecting means, and that the driver is likely to be looking ahead of the vehicle. An operation including a storage means for storing the sometimes detected eye position as a reference position and a comparison means for comparing whether or not the eye position detected while the vehicle is running matches a predetermined condition with respect to the reference position. A state detection device has been proposed (see, for example, Patent Document 1).

In the above-mentioned driving state detecting device, the image data of the driver's face is input by the image input means, and the eye position is detected by the eye position detecting means. On the other hand, the vehicle state detecting means detects the presence / absence of depression of the brake pedal, the steering angle of the steering device, the presence / absence of depression of the accelerator pedal, the acceleration / deceleration in the front-rear direction of the vehicle, and the forward gaze determination means obtains these detection information. Based on the above, it is determined whether or not the driver has a high probability of gazing at the front of the vehicle, and whether or not the driver is inattentive driving is determined.

Japanese Unexamined Patent Publication No. 11-195200

On the other hand, it may be required to retrofit the vehicle with a judgment system for inattentive driving. In such a case, the retrofitted inattentive driving determination system cannot be connected to the vehicle state detecting means mounted on the vehicle, and the detection information by the vehicle state detecting means cannot be acquired. Therefore, the retrofitted inattentive operation determination system has a problem that the presence or absence of inattentive operation cannot be accurately determined.

The present invention is an inattentive detection system that can be retrofitted to a vehicle and has excellent detection accuracy.

The present invention [1] is an inattentive detection system for detecting inattentive driving of a vehicle driver, a speed detecting means for detecting the speed of the vehicle, an in-vehicle imaging means for photographing the driver of the vehicle, and the speed detecting means. And a control means connected to the in-vehicle imaging means, the control means, based on the image taken by the in-vehicle imaging means when the driver makes a retrospective action, the driver's reflection angle and / Alternatively, the retrospective time is detected, and it is determined whether or not the retrospective action is inattentive driving based on the relationship between the speed of the vehicle detected by the speed detecting means and the retrospective angle and / or the retrospective time. Includes an inattentive detection system.

The inattentive detection system of the present invention includes a speed detecting means for detecting the speed of the vehicle, an in-vehicle imaging means for photographing the driver of the vehicle, and a control means connected to them.

Then, in the inattentive detection system of the present invention, the control means performs inattentive driving based on the speed of the vehicle detected by the speed detecting means and the driver's turning angle and / or turning time taken by the in-vehicle imaging means. Is detected.

Therefore, the inattentive detection system of the present invention can be retrofitted to the vehicle and can detect inattentiveness with high accuracy.

FIG. 1 is a schematic configuration diagram showing an embodiment of the inattentive detection system of the present invention. FIG. 2 is a flow chart showing a processing flow executed in the inattentive detection system shown in FIG.

1. 1. Inattentive detection system In FIG. 1, the inattentive detection system 1 is provided for detecting inattentive driving of a driver in an automobile 10 as a vehicle.

The inattentive detection system 1 may be, for example, an optional part retrofitted to the automobile 10 or a default part pre-installed in the automobile 10.

Preferably, the inattentive detection system 1 is an optional part that is retrofitted to the automobile 10. More preferably, the inattentive detection system 1 is a retrofit drive recorder as an optional part.

The inattentive detection system 1 photographs the speed detecting device 2 as a speed detecting means for detecting the speed of the automobile 10, the vehicle interior camera 3 as an in-vehicle imaging means for photographing the driver of the automobile 10, and the front of the automobile 10. A front camera 4 as a front imaging means, a control unit 5 as a control means connected to these, and an alarm device 6 as an alarm means connected to the control unit 5 to warn the driver of the automobile 10 of inattentive driving. It has.

The speed detection device 2 is not particularly limited as long as it can detect the speed of the automobile 10, and examples thereof include a GPS device. The GPS device is, for example, based on a GPS sensor that acquires the position information of the automobile 10, a storage unit that stores the position information of the automobile 10, a timer that observes the passage of time, and a change in the position information per unit time of the automobile 10. It is equipped with a calculation unit for calculating the speed of the automobile 10, and can acquire the position information of the automobile 10 and calculate the speed of the automobile 10.

Further, as shown by the broken line in FIG. 1, the speed detection device 2 is connected by wire or wirelessly to the control unit 5 described later. As a result, the speed of the automobile 10 detected by the speed detection device 2 can be input to the control unit 5.

The vehicle interior camera 3 is not particularly limited as long as it can photograph the driver of the automobile 10, and a known small camera is used. The vehicle interior camera 3 is arranged toward the driver, for example, in the upper part of the driver's seat of the automobile 10. As a result, the vehicle interior camera 3 can photograph the driver's face (see arrow a in FIG. 1).

Further, as shown by the broken line in FIG. 1, the vehicle interior camera 3 is connected by wire or wirelessly to the control unit 5 described later. As a result, the driver's image taken by the vehicle interior camera 3 can be input to the control unit 5.

The front camera 4 is not particularly limited as long as it can photograph the road surface ahead in the traveling direction of the automobile 10, and a known small camera is used. For example, it is arranged in the upper part of the driver's seat of the automobile 10 toward the front of the windshield. As a result, the front camera 4 can photograph the road surface in front of the automobile 10 in the traveling direction (see arrow b in FIG. 1).

Further, as shown by the broken line in FIG. 1, the front camera 4 is connected by wire or wirelessly to the control unit 5 described later. As a result, the image of the front of the automobile 10 taken by the front camera 4 can be input to the control unit 5.

The control unit 5 is a control unit that executes electrical control of the inattentive detection system 1, and is composed of a microcomputer including a CPU, a ROM, a RAM, and the like.

Further, the control unit 5 has a determination program P for determining whether or not the driver is inattentive operation. The determination program P is not particularly limited, is designed by a known method, and is stored in the control unit 5.

The determination of the presence or absence of inattentive operation by the determination program P will be described in detail later.

The alarm device 6 is a device that warns the driver when the driver's inattentiveness is detected. Examples of the alarm device 6 include a speaker that warns the driver by voice, for example, a display that warns the driver by video, and a speaker is preferable.

As shown by the broken line in FIG. 1, the alarm device 6 is wired or wirelessly connected to the control unit 5 described later. As a result, when the driver's inattentive driving is detected in the determination program P of the control unit 5, the alarm device 6 can be operated by the control unit 5.

2. Inattentive detection by the inattentive detection system In the following, the processing flow for detecting the inattentive operation by the control program P will be described in detail with reference to FIG.

More specifically, first, at the start of operation of the automobile 10 (when the ignition is ON), the inattentive detection system 1 is started to operate, and the determination program P is executed (S1: start).

When the inattentive detection system 1 starts operating, first, while the automobile 10 is traveling, the driver is photographed by the vehicle interior camera 3, and the image is input to the control unit 5. Then, the determination program P of the control unit 5 detects the driver's retrospective behavior from the image taken by the vehicle interior camera 3 (S2).

The method of detecting the retrospective behavior by the determination program P is not particularly limited, and a known method such as pattern matching based on machine learning can be mentioned.

The determination program P monitors the driver until the driver's retrospective behavior is detected (S2: No).

Then, when the retrospective behavior is detected (S2: Yes), the determination program P determines the retrospective behavior based on the vehicle speed of the automobile 10 and the retrospective angle and the retrospective time of the driver by the method shown below. It is determined whether or not the driving is inattentive (S3 to S6).

That is, as the driver's retrospective behavior, for example, there are a retrospective behavior as a peripheral confirmation behavior when changing lanes and a retrospective behavior as an inattentive driving.

Therefore, if it is determined that all the retrospective actions are inattentive driving, it is determined that the driver is inattentive driving even during the peripheral confirmation action, and stress such as a warning (described later) may be given to the driver.

Therefore, the determination program P determines whether the driver's retrospective behavior is a peripheral confirmation behavior or an inattentive driving based on the vehicle speed, the retrospective angle, and the retrospective time.

More specifically, the determination program P first detects the driver's turning angle based on the image taken by the vehicle interior camera 3.

The turning angle is detected by the driver's line-of-sight direction, face direction, and the like. The method of detecting the reflection angle by the determination program P is not particularly limited, and known methods such as pattern matching based on machine learning can be mentioned.

Further, the speed of the automobile 10 detected by the speed detection device 2 (GPS device or the like) is input to the determination program P.

Then, the determination program P determines whether or not the driver's retrospective behavior is inattentive driving based on the relationship between the speed of the automobile 10 and the retrospective angle.

More specifically, in driving the automobile 10, when the retrospective behavior is a peripheral confirmation behavior, the retrospective angle is relatively small, while when the retrospective behavior is inattentive driving, the retrospective angle is relatively large. ..

The turning angle is also affected by the speed of the automobile 10.

That is, when the speed of the automobile 10 is relatively high, the reflection angle in the peripheral confirmation action becomes even smaller. In other words, when the speed of the automobile 10 is relatively high, it can be determined that the retrospective behavior is inattentive driving even if the retrospective angle is medium. On the other hand, when the speed of the automobile 10 is relatively low, the opposite is true, and even if the reflection angle is medium, the reflection behavior can be determined to be a peripheral confirmation behavior.

Therefore, in the determination program P, a correlation map between the turning angle in inattentive driving and the speed of the automobile 10 is registered.

Further, in the determination program P, a threshold value (angle threshold value) indicating the boundary between the reflection angle in the peripheral confirmation action and the reflection angle in the inattentive operation is registered.

Then, the determination program P determines whether or not the driver's reflection angle exceeds the threshold value (angle threshold value) (S3).

When the driver's reflection angle is equal to or less than the threshold value (S3: No), the determination program P determines that the driver's reflection behavior is not an inattentive driving but a peripheral confirmation behavior (S6).

In this case, the determination program P continues to monitor the driver without warning the driver (described later) and detects the next retrospective action (S10: return).

On the other hand, when the driver's reflection angle exceeds the threshold value (S3: Yes), the determination program P determines that the driver's reflection behavior may be inattentive driving.

In such a case, the determination program P subsequently detects the driver's retrospective time based on the image taken by the vehicle interior camera 3.

The retrospective time is the maintenance time of the retrospective behavior. The method of detecting the retrospective time by the determination program P is not particularly limited, and a known method can be mentioned.

Then, the determination program P determines whether or not the driver's retrospective behavior is inattentive driving based on the relationship between the speed of the automobile 10 and the retrospective time.

More specifically, in driving the automobile 10, when the retrospective behavior is a peripheral confirmation behavior, the retrospective time is relatively short, while when the retrospective behavior is inattentive driving, the retrospective time is relatively long. ..

The retrospective time is also affected by the speed of the automobile 10.

That is, when the speed of the automobile 10 is relatively high, the reflection time in the peripheral confirmation action becomes even shorter. In other words, when the speed of the automobile 10 is relatively high, it can be determined that the retrospective behavior is inattentive driving even if the retrospective time is medium. On the other hand, when the speed of the automobile 10 is relatively low, the opposite is true, and even if the retrospective time is medium, the retrospective action can be determined to be a peripheral confirmation action.

Therefore, in the determination program P, a correlation map between the retrospective time in the inattentive driving and the speed of the automobile 10 is registered.

Further, in the determination program P, a threshold value (time threshold value) indicating the boundary between the retrospective time in the peripheral confirmation action and the retrospective time in the inattentive operation is registered.

Then, the determination program P determines whether or not the driver's reflection time exceeds the above threshold value (time threshold value) (S4).

When the driver's reflection time is equal to or less than the threshold value (S4: No), the determination program P determines that the driver's reflection behavior is not an inattentive driving but a peripheral confirmation behavior (S6).

In this case, the determination program P continues to monitor the driver without warning the driver (described later) and detects the next retrospective action (S10: return).

On the other hand, when the driver's reflection time exceeds the threshold value (S4: Yes), the determination program P determines that the driver's reflection behavior is inattentive driving (S5).

Further, when the driver looks aside, the automobile 10 may deviate from the traveling lane and cause a danger.

Therefore, the determination program P detects the presence or absence of lane deviation of the automobile 10 by the following method based on the image taken by the front camera 4 (S7 to S9).

More specifically, in the determination program P, when it is determined that the driver's retrospective behavior is inattentive driving, the traveling lane of the automobile 10 and the traveling lane of the automobile 10 are based on the image taken by the front camera 4. Detects the running position.

The traveling lane is detected by a white line on the road surface in front, a road sign, a median strip, a roadside tree, etc., taken by the front camera 4. Further, the traveling position of the automobile 10 is detected by the relative relationship between the traveling lane and the automobile 10 photographed by the front camera 4. The method of detecting the traveling lane and the traveling position by the determination program P is not particularly limited, and a known method such as pattern matching based on machine learning can be mentioned.

Then, the determination program P detects whether or not the automobile 10 deviates from the lane based on the relationship between the traveling lane of the automobile 10 and the traveling position of the automobile 10 (S7).

When the lane deviation of the vehicle 10 is detected (S7: Yes), the control unit 5 activates the alarm device 6, the driver is driving aside, and the vehicle 10 deviates from the lane due to the inattentive driving. It warns the driver that it is doing (S8 (large warning)).

The alarm method is not particularly limited, but for example, a relatively loud voice is generated from the speaker as the alarm device 6 to warn the driver.

On the other hand, when the lane deviation of the automobile 10 is not detected (S7: No), the control unit 5 activates the alarm device 6 to warn the driver only that the driver is driving aside (S9 (small). alarm)).

The warning method is not particularly limited, but for example, a voice smaller than the above-mentioned large alarm is generated from the speaker as the alarm device 6 to warn the driver.

Then, when the control unit 5 confirms the driver's inattentive driving and recovery from the lane deviation, the control unit 5 continues to monitor the driver and detects the next retrospective action (S10: return). Further, the above processing is stopped when the traveling of the automobile 10 is stopped. s
3. 3. Action / Effect The above-mentioned inattentive detection system 1 includes a speed detection device 2 for detecting the speed of the automobile 10, a vehicle interior camera 3 for photographing the driver of the automobile 10, and a control unit 5 connected to them. There is.

Then, in the inattentive detection system 1, the control unit 5 is based on the speed of the automobile 10 detected by the speed detection device 2 and the driver's turn-back angle and / or turn-back time taken by the vehicle interior camera 3. And detect inattentive driving.

Therefore, the inattentive detection system 1 can be retrofitted to the automobile 10 and can accurately detect inattentiveness.

Further, the inattentive detection system 1 is provided with a front camera 4 for photographing the front of the automobile 10, and when the inattentive driving of the driver 10 is detected, the control unit 5 is based on the image taken by the front camera 4. , Detects the presence or absence of lane deviation of the automobile 10.

That is, according to the inattentive detection system 1 described above, lane deviation due to inattentive driving can be detected.

In addition, the inattentive detection system 1 includes an alarm device 6 connected to the control unit 5, and the control unit 5 activates the alarm device 6 when the inattentive operation is detected. Further, the control unit 5 warns the driver with a relatively loud voice (large warning) when a lane deviation occurs, and the driver with a relatively small voice for inattentive driving that does not cause a lane deviation. Can be warned (small alarm).

Therefore, according to the inattentive detection system 1 described above, the accident risk of the automobile 10 can be reduced.

4. Modification example In the above description, the inattentive detection system 1 detects the presence or absence of inattentive driving based on the relationship between the driver's turning angle and the driver's turning time and the speed of the automobile 10. It is also possible to detect the presence or absence of inattentive driving based on the relationship between either the driver's turning angle or the driver's turning time and the speed of the automobile 10. In this case, in the processing flow by the determination program P, either the step of determining the relationship between the reflection angle and the threshold value (S3 in FIG. 2) or the step of determining the relationship between the reflection time and the threshold value (S4 in FIG. 2). Is omitted.

Further, in the above description, the inattentive detection system 1 includes the front camera 4 and determines whether or not the vehicle deviates from the lane. For example, the inattentive detection system 1 does not have to include the front camera 4 and deviates from the lane. It is not necessary to determine the presence or absence of. In this case, in the processing flow by the determination program P, the step of determining the presence or absence of lane deviation (S7 in FIG. 2) and the warning regarding inattentive driving and lane departure (S8 (large warning in FIG. 2)) are omitted, and inattentive operation is omitted. (S9 in FIG. 2) is executed.

Further, in the above description, the inattentive detection system 1 includes an alarm device 6 and warns the driver when the inattentive driving is detected. For example, the inattentive detection system 1 warns the driver. It is not necessary to have. In such a case, the inattentive detection system 1 can be connected to, for example, an automatic driving system, and when the inattentive driving is detected, the automatic driving system can be automatically operated to interrupt the driving by the driver.

1 Inattentive judgment system 2 Speed detection device 3 Vehicle interior camera 4 Front camera 5 Control unit 6 Alarm device 10 Automobile

Claims (1)

It is an inattentive detection system that detects inattentive driving of the driver of the vehicle.
Speed detection means to detect the speed of the vehicle and
In-vehicle imaging means for photographing the driver of the vehicle,
The speed detecting means and the control means connected to the in-vehicle imaging means are provided.
When the driver looks back, the control means
Based on the imaging image taken by the in-vehicle imaging means, the driver's reflection angle and / or reflection time is detected.
It is characterized in that it is determined whether or not the turning action is inattentive driving based on the relationship between the speed of the vehicle detected by the speed detecting means and the turning angle and / or the turning time. Inattentive detection system.

Images