JP2010204847A - Driver condition detection device and collision early detection device using same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a highly reliable driver condition detection device for detecting such a detection abnormality generation state that a driver is not detected, and that even when the direction of the face of the driver is normal, it is erroneously detected that the driver is looking away. <P>SOLUTION: This driver condition detection device includes: a first driver condition detection means 10 for detecting the direction of a driver by image pickup; a detection abnormality determination means 20 for, when a person non-detection state or a detection state that the driver is looking away by the first driver condition detection means continues in a prescribed time, determining that the first driver condition detection means is put in the detection abnormality state that the detection of the direction of the face has not been normally performed; and a second driver condition detection means 40 for, when it is determined that the first driver condition detection means is put in the detection abnormality state by the detection abnormality determination means, detecting the posture of the driver from the position different from the position of the first driver condition detection means. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a driver state detection device and an early collision detection device using the same, and more particularly to a driver state detection device that detects a driver's face orientation by imaging and an early collision detection device using the same.

  Conventionally, a face image capturing camera that detects the driver's line-of-sight direction is provided on the steering column cover or in the combination meter, detects the driver's line-of-sight direction, and performs alarm control according to the detection result of the line-of-sight direction detection means. If the steering angle exceeds the specified steering angle, the turn signal switch is on, the shift position is not the forward position, or the lane is a curved road An in-vehicle alarm that disables the detection result of the line-of-sight detection means and that can perform an appropriate alarm in accordance with the driver's looking-ahead state without performing an alarm when it is inappropriate to detect the driver's looking-aside An apparatus is known (see, for example, Patent Document 1).

JP 2007-72629 A

  However, in the configuration described in Patent Document 1 described above, it is possible to prevent misjudgment of side-by-side detection based on the relationship between the vehicle traveling direction and the driver's line-of-sight direction, but the driving posture of the driver in the vehicle is lost. Side effects caused by the driver's condition and imaging environment, such as the presence of obstacles in front of the face imaging camera, strong sunlight, and inability to properly recognize the image. There has been a problem that false detection cannot be prevented.

  Accordingly, the present invention can detect a detection abnormality occurrence state in which a driver is not detected, or a false detection that the driver is in the lateral orientation is performed even if the driver's face orientation is normal. It is an object to provide a highly reliable driver state detection device.

In order to achieve the above object, a driver state detection device according to a first aspect of the present invention includes first driver state detection means for detecting a driver's face direction by imaging,
When the state in which no person is detected by the first driver state detection unit or the detection state in which the face direction is horizontal continues for a predetermined time, the face direction detection by the first driver state detection unit is normally performed. Detection abnormality determination means for determining that the detection abnormal state is not broken;
When the detection abnormality determination unit determines that the first driver state detection unit is in the detection abnormality state, a first position of the driver is detected from a position different from the first driving state detection unit. 2 driver state detecting means.

  Thus, when the driver's presence cannot be detected by the normal driving state detection means, or when it is determined that the sideways state is continuously detected and the sideways detection itself is a false detection, another driver state A detection means can be used, and a driver | operator's state can be detected correctly.

A second aspect of the present invention is the driver state detection apparatus according to the first aspect of the present invention.
The detection abnormality determination means determines the detection abnormality state based on the lateral detection state when the vehicle speed is equal to or higher than a predetermined speed.

  As a result, it is possible to distinguish whether the vehicle is in a normal running state, and because it is determined whether the vehicle is in a false detection state in a normal running state where the sideways state cannot continue. Therefore, it is possible to accurately determine the erroneous detection state.

A third aspect of the present invention is the driver state detection device according to the first or second aspect of the present invention.
The second driver state detection means includes a security camera for imaging the inside of the vehicle or a head position detection means for airbag deployment.

  As a result, it is not necessary to provide the second driver state detection means exclusively, and since the driver state detection means provided for other purposes is effectively used, the driver state can be detected appropriately in a small space and at a low cost. can do.

4th invention is the driver state detection apparatus which concerns on either 1st-3rd invention,
The second driver state detection unit includes a plurality of detection units, and determines the driver state based on evaluation values of the plurality of detection units.

  Thereby, since the evaluation value by a some detection means is used, a driver | operator state can be detected with high precision.

According to a fifth aspect of the present invention, in the driver state detection device according to any one of the first to fourth aspects of the invention,
The second driver state detection means includes communication means for transmitting the detected posture of the driver to a center outside the vehicle.

  Thereby, a center such as G-Book can be used, and an accurate driver state can be detected in real time. Also, you can notify the driver of the status from the center, accumulate the driver status data to the dealer, and inform the driver of the status trend when the driver visits the dealer, Can be used.

An early collision detection device according to a sixth aspect of the invention is an early collision detection device that detects that a vehicle may collide and issues a warning to a driver.
A driver state detection device according to a fifth invention is provided,
In the center where the driver's posture detected by the driver state detector is transmitted, the abnormal detection state is caused by a cause other than the driver's posture or the shielding of the first driving state detection means by the obstacle. When it is determined that the alarm has occurred, the timing for issuing the alarm is set later than in the normal state.

  Thereby, the false alarm of the early collision detection apparatus can be reduced, the driver's troublesomeness can be reduced, and the reliability of the collision alarm can be increased.

  According to the present invention, the reliability of the driver state detection can be increased.

1 is a diagram illustrating an example of the overall configuration of a driver state detection device 150 and an early collision detection device 30 according to a first embodiment. It is the figure which showed an example of the installation state of a face detection camera. FIG. 2A is a side view of the installation state of the face detection camera. FIG. 2B is a front view of the installation state of the face detection camera. It is the figure which showed the example of the face image. FIG. 3A is a diagram illustrating an example of a front image in a normal detection state. FIG. 3B is a diagram illustrating an example of a face image in an erroneous detection state. FIG. 3C is a diagram illustrating an example of a face image in a false detection state due to sunlight. It is the reference figure which showed the image of the state in which the horizontal orientation determination was performed appropriately. It is a figure showing an example of a state where a driver is not detected. It is the figure which showed the example in which the false detection generate | occur | produced due to the driving | running posture of a driver | operator. FIG. 6A is a diagram illustrating an example of an erroneous detection state from the side. FIG. 6B is a diagram illustrating an example of an erroneous detection state from an oblique direction. FIG. 6 is a diagram illustrating an example of a driving posture detection unit 41 of a second driver state detection unit 40. It is the figure which showed an example of the driving posture detection means 41 different from FIG. FIG. 6 is an operation flowchart of the driver state detection device 150 according to the first embodiment. It is the figure which showed an example of the whole structure of the driver | operator state detection apparatus 150a and the collision early detection apparatus 30 which concern on Example 2. FIG. It is the figure which showed an example of the communication form of the driver | operator information detection apparatus 150a and the collision early detection apparatus 30 which concern on Example 2. FIG. FIG. 12 is a diagram illustrating an example of communication modes of a driver state detection device 150a and an early collision detection device 30 according to a second embodiment different from FIG. It is the figure which showed the correspondence example of the center 120 of the collision early detection apparatus 30 of Example 2. FIG. FIG. 13A is a diagram showing a processing flow corresponding to the center 120 side. FIG. 13B is a diagram showing a relationship example between the driver state detection device 150a and the early collision detection device 30.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram illustrating an example of the overall configuration of a driver state detection device 150 and an early collision detection device 30 using the same according to a first embodiment to which the present invention is applied. In FIG. 1, the driver state detection device 150 according to the first embodiment includes a first driver state detection unit 10, a detection abnormality determination unit 20, and a second driver state detection unit 40. The driver state detection device 150 may include a vehicle speed detection unit 21 and an early collision detection system 30 as necessary. Further, in FIG. 1, a center 120 outside the vehicle is shown as a related component of the driver state detection device 150.

  The first driver state detection means 10 is means for detecting the driver's face orientation by imaging. The first driver state detection means 10 includes a face detection camera 11 and an image processing ECU (Electronic Control Unit) 12.

  The face detection camera 11 is means for imaging the driver's face, and is mounted, for example, on the steering column so that the driver's face can be imaged from the front. From the image captured by the face detection camera 11, it is possible to detect whether the driver's face is in the landscape orientation, whether the eyes are closed, and the like, and the driver's state can be detected. . Note that various cameras can be applied to the face detection camera 11, for example, a CCD (Charge Coupled Device), a CMOS (Complementary Metal Oxide Semiconductor) camera, or the like.

  The image processing ECU 12 performs image processing on the face image captured by the face detection camera 11 and performs image recognition such as whether the face is in a sideways state, that is, whether it is a side-by-side state or whether the face is in a closed state. Thereby, the state of the driver can be detected. The image processing ECU 12 performs image processing such as feature point extraction necessary for image recognition. For example, a predetermined electronic circuit such as an ASIC (Application Specific Integrated Circuit) that performs such arithmetic processing or a CPU (Central It may be configured as a microcomputer provided with a processing unit (central processing unit) and the like.

  The detection abnormality determination unit 20 is a unit that determines whether or not the driver state detected by the first driver state detection unit 10 is a normally detected driver state. The detection abnormality determination unit 20 can reliably detect the detected driver state when the driver state detected by the first driver state detection unit 10 corresponds to a predetermined condition that is not considered to be a normal detection result. First, it is determined that the detection is abnormal. Specifically, the detection abnormality determination unit 20 is the first driver state detection unit 10 when the state where no person is detected continues for a predetermined time or more, or when the lateral detection state continues for a predetermined time or more. In addition, the detected driver state is a detection result under a state in which a defect occurs in detection, and the first driver state detection means 10 is in a state where normal detection cannot be performed. Make a decision. Although the details of the specific contents of the detection abnormality determination method will be described later, the detection abnormality determination means 20 is configured as a predetermined electronic circuit such as an ASIC capable of performing such arithmetic processing or a microcomputer equipped with a CPU. May be. Moreover, the detection abnormality determination means 20 may be comprised independently, and may be integrated and mounted in collision judgment ECU31 as shown in FIG. Furthermore, the detection abnormality determination means 20 may be incorporated and mounted in the face image processing ECU 12. As described above, the detection abnormality determination unit 20 can be provided at any position as long as it has a function capable of determining whether or not the detection result of the driver state by the first driver state detection unit 10 is normal.

  The vehicle speed detection means 21 is a means for detecting the traveling speed of the vehicle on which the driver state detection device 150 according to this embodiment is mounted. Since the vehicle speed may be used as a part of the determination condition in the erroneous detection determination by the detection abnormality determination unit 20, the vehicle speed detection unit 21 provides the vehicle speed to the detection abnormality determination unit 20 in that case. Therefore, the vehicle speed detection means 21 may be provided as needed. For example, a wheel speed sensor may be applied to the vehicle speed detection means 21.

  The early collision detection device 30 is a device that detects a vehicle collision at an early stage and notifies the driver. The driver state detection device 150 according to the present embodiment may be configured to be incorporated as a part of the early collision detection device 30, for example. The early collision detection apparatus 30 includes a collision determination ECU 31, an obstacle detection means 32, and an alarm means 36 in addition to the components of the driver state detection apparatus 150 according to the present embodiment. The alarm means 36 may include, for example, a buzzer 33, a meter 34, a brake ECU 35, and the like depending on the application.

  The obstacle detection means 32 is means for detecting the presence of an obstacle in front of the vehicle. The obstacle detection means 32 may be various detection means as long as it can detect the presence of an obstacle and can immediately measure the distance between the obstacle and the vehicle. For example, a millimeter wave radar is applicable. May be. Millimeter wave radar receives radio waves reflected from obstacles by emitting millimeter waves, and measures the position of the object and the relative speed with the vehicle based on the propagation time and frequency difference caused by the Doppler effect. .

  When the obstacle detection means 32 detects the presence of the obstacle including the position and relative speed, the collision determination ECU 31 determines whether or not the vehicle may collide with the obstacle based on these detected values. It is a means to judge. Further, when the collision determination ECU 31 determines that the vehicle may collide with the collision object, the collision determination ECU 31 notifies the driver that there is a possibility of collision using the warning means 36. The alarm means 36 may be provided in accordance with a necessary one and use among the buzzer 33, the meter 34, the brake ECU 35, and the like. The timing of issuing an alarm by the alarm means 36 may be set a predetermined several seconds before the collision. For example, control may be performed so as to issue an alarm 2 seconds before the predicted collision timing.

  The buzzer 33 is a means for issuing a warning to the driver by voice (including a buzzer sound), and the meter 34 is a means for performing a warning display and giving a warning to the driver visually. The brake ECU 35 is a means for performing control to operate the brake and avoiding or relieving the collision by deceleration.

  The collision determination ECU 31 uses such alarm means 36 to make the driver recognize the danger of collision and perform control for avoiding or mitigating the collision. Therefore, the collision determination ECU 31 performs collision determination and calculation processing for control based on the determination, and may be configured as a predetermined electronic circuit or microcomputer.

  When the driver state detection device 150 according to the present embodiment is associated with the early collision detection device 30 and incorporated and used as a part thereof, the determination result by the face image processing ECU 12 and the detection abnormality determination means 20 is also a collision. It is preferable that the collision determination ECU 31 performs the collision determination based on the results input to the determination ECU 31. For example, based on the detection from the obstacle detection means 32, in the setting for notifying the driver of the collision 2 seconds before the collision prediction timing, the driver state detection result by the first driver state detection means 10 is the sideways state. , The control can be performed so that the alarm timing of the collision notification is advanced and the alarm is issued early. In that case, according to the driving | running state detection apparatus 150 which concerns on a present Example, since the misdetection of a driver | operator's sideways state can be recognized, a false alarm can be reduced. In addition, when the erroneous detection state continues, the control for advancing the above-mentioned alarm timing can be stopped, or the control for reducing the time width for the advance of the alarm timing can be performed. A highly reliable early collision detection device 30 can be obtained.

  The second driver state detection unit 40 detects the driver's posture when the detection abnormality determination unit 20 determines that the driver state detection by the first driver state detection unit 10 is not a normal detection result. It is means to do. If the second driver state detection unit 40 is provided at the same position as the first driver state detection unit 10, there is a risk of causing the same detection abnormality as the first driver state detection unit 10. Therefore, it is installed at a position different from the first driver state detection means. The second driver state detection unit 40 includes a driving posture detection unit 41 and a driving posture information processing unit 45. The driving posture information processing unit 45 may include the communication unit 42 and / or the driving posture determination unit 43 depending on the application.

  The driving posture detection means 41 is a means for detecting the driver's posture. As the driving posture detecting means 41, various detecting means may be used as long as the posture of the driver can be detected. For example, an indoor camera for anti-theft security, a head position detecting camera at the time of airbag deployment, etc. May be used. In addition to the image pickup means such as a camera, a means such as a laser that can detect the position (posture) of the driver may be used. The driving posture detection means 41 may be provided exclusively for the second driver state detection means 40, but as described above, it is prepared for the security camera for theft prevention and the airbag deployment. Detection means having other purposes, such as a camera or a laser, may also be used. As a result, it is not necessary to newly provide the dedicated driving posture detecting means 41, and the driver state detecting device 150 according to the present embodiment can be realized in a space-saving and cost-saving manner.

  The driving posture information processing means 45 is a means for processing information related to the driver posture detected by the driving posture detection means 41 and recognizing the driver posture. The driving posture information processing unit 45 includes a communication unit 42 and / or a driving posture determination unit 43 as necessary.

  The communication means 42 is means for transmitting information on the driving posture of the driver detected by the driving posture detection means 41 to the center 120 outside the vehicle. For example, if a person is resident in the center 120, the driving posture information sent from the communication means 42 is analyzed on the center 120 side to grasp the cause of the detection abnormality in the first driver state detecting means 10. can do. Then, if the cause of the detection abnormality is that the driver's posture is bad or an obstacle is placed in front of the face detection camera 11 and is caused by the driver's posture or the arrangement of articles, driving The driver can be fed back to improve the driving posture. Further, even if feedback is not provided directly to the driver, it is possible to take measures such as sending the information to the dealer and notifying the driver of the opportunity. In addition, it is possible to take a countermeasure such that the control instruction for delaying the timing of the collision warning as described above is performed on the vehicle side.

  The driving posture determination unit 43 determines whether or not the erroneous detection by the first driver state detection unit 10 is caused by the driving posture of the driver from the driving posture information detected by the driving posture detection unit 41. Means. For example, since the driving posture detection unit 41 is provided at a position different from that of the first driver state detection unit 10, the driving posture detected from the driving posture detection unit 41 is the first driver state detection unit 10. If the driving posture determining means 43 stores in advance the posture patterns that prevent the normal detection of the driver state by the above, and pattern matching between these patterns and the detected driving posture information is performed, It can be determined whether or not the cause of the detection abnormality is due to the driver's posture. In such a case, a pattern is stored in advance in the storage unit 44 built in the driving posture determination unit 43, and the stored pattern is compared with the driving posture detected by the driving posture detection unit 41. be able to.

  In addition, since the driving posture detection unit 41 of the second driving posture detection unit 40 is not necessarily provided exclusively for detecting the driver's posture, the information is insufficient for determining the driving posture. There is also. In such a case, the driving posture determination unit 43 may include only the storage unit 44 and store in the storage unit 44 a state in which an erroneous detection has occurred. For example, if the data stored in the storage means 44 is collectively analyzed by a dealer or the like and the tendency of the driver's posture or the like that is likely to cause a detection abnormality can be grasped thereby, a detection abnormality occurs in the subsequent driving. Therefore, the driver can take care not to take an easy posture, and the appropriate detection rate of the driver state by the first driver state detection means 10 can be increased.

  Thus, the driving posture information processing means 45 can use an appropriate driving posture information processing means 45 according to the application and the mode of the driving posture detection means 41.

  The center 120 is an engine provided independently outside the vehicle, and can send various information to the vehicle and manage the vehicle. Therefore, if the center 120 can analyze the driving posture information detected by the driving posture detection means 41, the center 120 analyzes whether the cause of the detection abnormality is caused by the driver's posture, The above-mentioned correspondence can be performed according to the application.

  Next, an example of the installation position of the face detection camera 11 of the first driver state detection unit 10 will be described with reference to FIG. FIG. 2 is a diagram illustrating an example of an installation state of the face detection camera. 2A is a side view of the state where the face detection camera is installed, and FIG. 2B is a front view of the driver when the face detection camera is installed.

  In FIG. 2A, the face detection camera 11 is mounted on the upper surface of the steering column cover 51 to which the steering wheel 50 is attached with the imaging surface facing the driver. In this state, when the face detection camera 11 images the driver, the driver's face is imaged from the lower front of the driver. Therefore, when the driver takes a normal driving posture without twisting the body or leaning on the steering wheel 50, the driver's face can be imaged from below the front.

  FIG. 2B shows a state where the driver is in the field of view of the face detection camera 11 and the meter 34 from the upper semicircle region of the steering wheel 50. It can be seen that the face detection camera 11 is provided at a position where the driver's face can be imaged without a shield from the hollow portion of the steering wheel 50. Further, the driver can visually recognize the meter 34 and can recognize that the alarm can be recognized if a collision warning display is displayed at any position around the meter 34. Therefore, as described in FIG. 1, the meter 34 can be used as a visual alarm means to notify the driver of the probability of collision.

  Thus, the face detection camera 11 of the first driver state detection means 10 is provided at a position where the driver's face can be imaged from the front, such as the upper surface of the steering column cover 51, for example.

  Next, an example in which a detection abnormality occurs in the first driver state detection unit 10 will be described with reference to FIGS. 3 to 6.

  FIG. 3 is a diagram showing an example of a face image detected by the first driver state detection means 10 using the face detection camera 11 mounted on the upper surface of the steering column cover 51 as shown in FIG. It is.

  FIG. 3A is a diagram illustrating an example of an image in a normal state in which the driver faces the front as a comparative example. Of the driver's face imaged by the face detection camera 11 shown in the upper diagram, eyes, nose, chin, and the like are extracted as feature points and edges by the face image processing ECU 12, and image processing as shown in the lower diagram is performed. It is determined that the driver is facing the front.

  FIG. 3B is a diagram illustrating a state in which the driver himself faces the front, but the driver's arm covers the face detection camera 11, and the driver's face can only be captured below. . In such a case, in the upper captured image, it can be recognized that the driver is facing the front. However, when the face image processing ECU 12 performs image processing, the image is as shown in the lower figure. It is possible to obtain an image that is hardly recognizable to be facing the front. In such a case, the face image processing ECU 12 performs the lateral orientation determination even though the driving difference is facing the front. That is, erroneous detection is performed.

  FIG. 3C is a diagram showing a state in which strong sunlight is incident although the driver is facing the front. In such a case, the face image acquired by the upper face detection camera 11 can recognize that the driver is facing the front, but when image processing is performed by the image processing ECU 12, an image like the lower side is obtained. It cannot be recognized that the feature points such as the chin and eyes are facing the front, and it is determined that the feature point is in the horizontal state. That is, due to the influence of sunlight, the image processing ECU 12 misidentifies the face width, and a side-view determination is made even though the driver is facing the front, and erroneous detection is performed.

  In this way, even if the driver's driving posture is not a shielding object such as an arm, or even if such a driving posture is not taken, erroneous detection of face orientation determination occurs due to external factors such as sunlight. In addition, there are individual differences depending on the shape of the face, and there are faces that are difficult to recognize when facing the front even when facing the front. In other words, due to the presence of the shielding object and the limit of the image recognition capability, there are cases in which the occurrence of erroneous detection cannot be avoided in the driver's lateral detection.

  FIG. 4 is a diagram illustrating an image in a state where the horizontal orientation determination is appropriately performed as a reference example. As shown in the upper diagram of FIG. 4, an image that can be recognized that the driver's face orientation detected by the face detection camera 11 is the landscape orientation can be captured. Also in the lower image processed by the image processing ECU 12, it is shown that feature points and edges such as eyes, nose and chin can be detected appropriately and the driver can recognize that the driver is in the sideways state. Comparing FIG. 4 with FIG. 3 (A), when the face orientation can be determined normally, the lowest image that has undergone image processing is also at an image level that allows the face orientation to be recognized. I understand.

  FIG. 5 is a diagram showing an example of a state in which the first driver state detection means 10 cannot detect the presence of a person and the driver is not detected. As shown in FIG. 5, the face detection camera 11 is mounted on the upper surface of the steering column cover 51 that supports the steering wheel 50, but an obstacle 60 is placed in front of the imaging surface of the face detection camera 11, The face detection camera 11 is shielded, and the driver cannot be detected. In this case, even if the driver himself is seated in a normal posture on the seat 70 and is facing the front, the imaging surface of the face detection camera 11 is shielded by the shielding object 60, so nothing other than the shielding object 60 is imaged. It becomes impossible. Therefore, no feature points such as edges are extracted from the acquired image, and the image processing ECU 12 is in a state where the driver is not detected.

  The shielding object 60 may include various objects. For example, a cloth for wiping a windshield or a book such as a guide book may be placed.

  FIG. 6 is a diagram illustrating an example in which the first driver state detection unit 10 performs erroneous detection due to the driving posture of the driver. 6A is a diagram illustrating an example of a state of erroneous detection from the side, and FIG. 6B is a diagram illustrating an example of a state of erroneous detection from an oblique direction. .

  In FIG. 6 (A), the driving posture is such that the driver reclines the seat 70 largely rearward, and the driver greatly slides backward. When the driver takes such a driving posture, the entire face of the driver does not enter the imaging range of the face detection camera 11, and only the forehead and the upper part of the head are imaged. In such a state, even if the driver is facing the front, depending on the edge processing of the image processing of the image processing ECU 12, the driver may determine that the driver is in the sideways state. Therefore, erroneous detection may occur depending on the driver's posture.

  FIG. 6B shows a state where the driver is driving while leaning on the door side. In such a state, the driver's face is captured only at the upper left, and the edge of the headrest 71 of the seat 70 behind the driver may be captured. Then, the edge of the headrest 71 is subjected to edge processing, and the driver is determined to be in the sideways state as the entire acquired image, and may be erroneously detected.

  As described above, even if the driver's driving posture is set, all or part of the driver's face is out of the imaging area of the face detection camera 11, and the driver is facing the front, but is erroneously detected as being sideways. May be done.

  As described above with reference to FIGS. 3 to 6, in various cases, the first driver state detection means 10 that detects the driver state by imaging can generate a detection abnormality due to erroneous detection or non-detection. Based on such a situation, it is necessary to determine the erroneous detection determination condition of the detection abnormality determination means 20. The erroneous detection by the first driver state detection means 10 is roughly divided into the erroneous detection described in FIG. 3 and FIG. 6 that determines that the driver is in the sideways state although facing the front, As described in FIG. 5, there is a case where the driver itself is not detected and is not detected. The undetected state described in FIG. 6 is a detection abnormality that continues as long as the shielding object 60 is not removed, and the erroneous detection described in FIGS. 3 and 6 is performed as long as the driver's posture or the surrounding environment does not change. It is a detection abnormality that continues. Therefore, the detection abnormality determination means 20 determines that a detection abnormality has occurred when the undetected state or the state in which the lateral orientation is detected continues for a predetermined time or longer. That is, it can be determined that the undetected state continues in a state where the shielding object 60 described in FIG. 5 covers and shields the front of the face detection camera 11. Further, when the sideways detection state continues, it is difficult for the driver to drive the vehicle normally with the sideways state being continued, so for example in FIG. 3 or FIG. It is highly likely that the erroneous detection state as described has continued. Therefore, in the driver state detection device 150 according to the present embodiment, in the first driver state detection means 10, when the undetected state continues for a predetermined time or more, or when the lateral detection state continues for a predetermined time or more. The first driver state detection unit 10 determines that the detection is abnormal. As a result, it is possible to detect a state in which a detection abnormality as shown in FIGS. 3 to 6 occurs.

  Note that the predetermined time that is the determination criterion for the undetected state and the predetermined time that is the determination criterion for the landscape state may be the same or different. For example, the predetermined time serving as the determination criterion for the undetected state may be several minutes or more, and the determination criterion for continuing the horizontal state may be several tens of seconds or several seconds within one minute. The time used as these criteria can be set appropriately according to the application.

  In the driver state detection device 150 according to the present embodiment, when the detection abnormality determination unit 20 determines that the detection of the driver state by the first driver state detection unit 10 is a detection abnormality, The second driver state detection means 40 detects the driver state.

  FIG. 7 is a diagram illustrating an example of the driving posture detection unit 41 of the second driver state detection unit 40. FIG. 7 shows a case where a security camera 41 a for preventing someone from entering the vehicle and stealing the vehicle is applied as the driving posture detection means 41. The security camera 41a is often provided at a substantially central portion of the ceiling 80 in the vehicle so that the entire vehicle interior can be imaged. Therefore, it is possible to image the driver from above with the security camera 41a, and to detect and detect the driver's posture from above. Thus, for example, the security camera 41a may be applied to the driving posture detection means 41.

  In many cases, the security camera 41a is usually configured to transmit an image to the center 120 when a vehicle theft or the like occurs and a person on the center 120 side responds. Therefore, when the security camera 41a is used as the driving posture detection means 41 of the second driver state detection means 40, an image is transmitted to the remote center 120 using the existing security system, and the center 120 may analyze whether the erroneous detection is based on the driver's posture or non-detection.

  FIG. 8 is a diagram showing an example of the driving posture detection means 41 of the second driver state detection means 40 different from FIG. FIG. 8 shows an example in which a head position detecting camera 41b for airbag deployment is applied to the driving posture detecting means 41 of the second driver state detecting means 40. In FIG. 8, the airbag 100 is accommodated in the instrument panel 90. The airbag 100 is deployed from the deployment surface 91 of the instrument panel 90. The position of the driver's head is provided in front of the ceiling 80, and is provided with a head position detection camera 41b in which an imaging surface is installed toward the rear and downward of the driver. The head position detection camera 41b detects the driver's head position, and when the airbag 100 is deployed, the airbag 100 is deployed when the driver's head is in a position close to the front deployment surface 11. Without the control, control is performed so that the airbag 100 is deployed when the driver's head is behind at least the head position detection camera 41b.

  Thus, for example, the airbag deployment head position detection camera 41b provided on the ceiling 80 in front of the vehicle may be applied to the posture position detection means 41 of the second driver state detection means 40. It is possible to detect the driver's posture from the front upper side of the driver, and whether or not the erroneous detection by the first driver state detecting means 10 is based on the driver's posture based on the detected image. Can be analyzed. At that time, the posture of the driver can be detected with higher accuracy by utilizing a three-dimensional recognition technology or the like used as a head position detection technology for airbag deployment.

  7 and 8, an example in which the posture of the driver is detected using the imaging unit of the security camera 41a or the head position detection camera 41b has been described. For example, a laser, an ultrasonic radar, or the like Can also be applied to the posture position detecting means 41.

  Further, a plurality of these posture position detecting means 41 may be used to detect the driving posture when an erroneous detection occurs. For example, when the vehicle includes both the security camera 41a and the head position detection camera 41b, or includes both the security camera 41a and the laser, driving from a plurality of posture position detection means 41 is performed. The posture of the person may be detected, and the driving posture may be comprehensively determined using these evaluation values.

  Next, the operation flow of the driver state detection device 150 according to the first embodiment will be described with reference to FIG. FIG. 9 is an operation flowchart of the driver state detection device 150 according to the first embodiment.

  In step 100, in the first driver state detection means 10, the driver's face is imaged by the face detection camera 11, and it is determined by the image processing ECU 12 whether or not the driver's face is in front. When the image processing ECU 12 detects that the face direction is the front direction, it is considered that the driver state is normal and the first driver state detection means 10 is also functioning normally. End the flow. On the other hand, if the image processing ECU 12 determines that the driver's face is not facing the front, a determination result signal is output from the image processing ECU 12 to the detection abnormality determination unit 20, and the process proceeds to step 110.

  In step 110, the detection abnormality determination unit 20 determines whether or not a person undetected state is output for a certain time or more, or whether a landscape state is output for a certain time or more. If any one of these conditions is satisfied, it is determined that the detection of the driver state by the first driver state detecting means 10 is an abnormal detection state, and the process proceeds to step 120. On the other hand, if none of the above conditions is satisfied, the process proceeds to step 160.

  In step 120, based on the detection abnormality determination result of the detection abnormality determination unit 20, the driving posture detection unit 41 of the second driving state detection unit 40 starts detection, and the posture of the driver is detected. For example, the driving posture detection means 41 is an imaging means provided at a position different from the face detection camera 11 provided in the vehicle compartment, such as a security camera 41a and a head position detection camera 41b for airbag deployment. In this case, the driving posture detection unit 41 starts imaging. When the driving posture detection means 41 is a laser or ultrasonic radar, the laser or ultrasonic radar is emitted. The driving posture information of the driver detected by the driving posture detection means 41 is output to the driving posture information processing means 45.

  In step 130, the driving posture information is transmitted from the communication unit 42 to the center 120, or the driving posture determination unit 43 determines the driving posture. For example, when the driving posture information is sent from the communication unit 42 to the center 120, for example, when the driving posture image can be acquired by the driving posture detection unit 41 by imaging, the acquired image information is the communication unit. 42 to the center 120 side.

  In step 140, the center 120 performs a cause analysis of the detection abnormality based on the received driving posture information. If the received driving posture information is an image, a driving posture in which the driver does not properly detect whether or not the shielding object 60 is placed in front of the face detection camera 11 and is erroneously detected as being in landscape orientation. The cause analysis is performed as to whether or not it has been taken or whether it is a false detection due to the influence of sunlight.

  In the case where the driving posture determination unit 43 is used without using the communication unit 42, the determination result of the driving posture performed in step 130 is stored in the storage unit 44, and the factor when the dealer is visited Analysis will be performed.

  In step 150, the corresponding processing for the cause of the detected abnormality found in the analysis in step 140 is performed, and the processing flow ends. For example, the response processing can give a feedback to the driver as an example, and can issue an alarm such as “please correct the driving posture” or “please do not place an object in front of the camera”. Also, for example, when information is transmitted to the dealer and a complaint is made about the driving information detection device 150 or the early collision detection device 30 by the user, the dealer shows the driving posture image to the user and takes the correct driving posture. It is also possible to take measures such as prompting them to receive them.

  As described above, various measures can be taken according to the purpose of use and the user's desire, and the detection abnormality of the driver information detection device 150 can be reduced and the reliability can be increased.

  Next, the process returns to step 110, and if the detection abnormality condition is not satisfied, the process proceeds to step 160.

  In step 160, it is not determined in step 110 that the driver's face is front-facing, and in step 110, the detection abnormality condition of the first driver state detection means 10 is not satisfied. It is determined that the face orientation is landscape.

  In step 170, the early collision detection device 30 determines whether or not the vehicle is likely to collide. The obstacle detection means 32 to which a millimeter wave radar or the like is applied may detect the position and relative speed of an obstacle in front of the vehicle, and the collision determination ECU 31 may determine the possibility of a collision.

  If the collision determination ECU 31 determines in step 170 that there is no possibility of a collision, the process flow is ended as it is. On the other hand, if it is determined that there is a possibility of a collision, the process proceeds to step 180.

  In step 180, the collision determination ECU 31 issues a warning to the driver by the warning means 36 to notify that there is a possibility of a collision. As a result, the driver can avoid or mitigate the impact.

  In FIG. 9, Step 140 and Step 150 are processes on the center 120 side, and Step 170 and Step 180 are processes in the early collision detection apparatus 30.

  FIG. 10 is a diagram illustrating an example of the overall configuration of the driver state detection device 150a and the early collision detection device 30 according to the second embodiment to which the present invention is applied. In FIG. 10, the driver state detection device 150a according to the second embodiment includes a first driver state detection unit 10, an erroneous determination detection determination unit 20, and a second driver state detection unit 40a. In common with the driver state detection device 150 according to the first embodiment, it is configured as a part of the early detection device 30. In the driver state detection device 150a according to the second embodiment, the internal configuration of the second driver state detection means 40a is configured by a security camera 41a and a data communication module 42a. This is different from the driver state detection device 150 according to the first embodiment in that it is used for the above. That is, in the second embodiment, an example in which the present invention is applied to a vehicle that supports remote security will be described as a more specific embodiment of the driver state detection device 150 according to the first embodiment. In addition, about the components other than the 2nd driver | operator state detection means 40a which are the same components as Example 1, the same referential mark as Example 1 is attached | subjected and the description is abbreviate | omitted.

  The second driver state detection means 40a is a system for imaging the vehicle interior for detecting and recording theft and intruders that are normally used as a security system for theft prevention, and for security cameras 41a, And a data communication module 42a.

  The security camera 41a is an imaging means for imaging a situation of theft and an intruder, and can capture the entire interior of the vehicle. As described in FIG. 7 of the first embodiment, for example, the security camera 41a is installed at a substantially central portion of the ceiling 80 in the vehicle, and is arranged at a position where the entire vehicle interior can be imaged.

  The operation of the security camera 41a is often controlled by a dedicated remote security ECU (not shown). In that case, the detection abnormality determination unit 20 and the remote security ECU may cooperate to drive the security camera 41a as the driving posture detection unit 41.

  As the security camera 41a, various cameras such as a CCD camera, a CMOS camera, or an analog camera may be used depending on the application. The driver's posture image captured by the security camera 41a is output to the data communication module 42a.

  The data communication module 42 a is a communication unit that performs communication with the center 120. The data communication module 42a may be, for example, a dedicated wireless device or a mobile phone connected to the network of the center 120. The data communication module 42a may have a function to access any website on the Internet, but if it can access the network of the center 120, the function for recognizing the driving posture according to the present embodiment. Therefore, it is sufficient to provide a communication function with the center 120. In general, in a security system, a dedicated communication module connected only to the network of the center 120 is applied to the data communication module 42a.

  The center 120 is an engine outside the vehicle that manages the vehicle and provides information. For example, in the security system, the center 120 stores vehicle status information and makes it possible to obtain vehicle status information by accessing it using a mobile phone or a personal computer held by the user. . In addition, when the center 120 obtains the vehicle status information, the center 120 may further perform a service for notifying that the information is obtained to the user's mobile phone or personal computer by e-mail. The security system having such a communication function is detected by the first driver state detecting means 10 in the driver state detecting device 150a and the early collision detection device 30 using the same. Used to analyze the cause of anomalies.

  FIG. 11 is a diagram illustrating an example of communication modes of the driver information detection device 150a and the early collision detection device 30 according to the second embodiment using a security system. In FIG. 11, the relationship among the vehicle 110, the center 120, and the dealer 130 is shown.

  First, in the vehicle 110, if it is determined that a detection abnormality has occurred in the first driver information detection means 10 and the normal face orientation cannot be determined, the second driver state detection means 40a. Then, imaging of the driver is started using the security camera 41a. Then, the acquired image is transmitted from the vehicle 110 to the center 120 existing outside by using the data communication module 42a.

  At the center 120, if a person is resident, the factor analysis of the first driver information detecting means 10 can be performed from the driver's posture and the situation inside the vehicle. If no person is resident in the center 120, the received image is saved and stored.

  Next, the center 120 transmits image information about the driver's posture to the dealer 130. When a person is resident in the center 120 and there is an analysis result, the driving posture information is transmitted to the dealer 130 together with the analysis result. On the other hand, when a person is not resident in the center 120, only the driver attitude information recorded when a malfunction of detection abnormality occurs in the driver information detection device 150a is transmitted to the dealer 130.

  In the dealer 130 that has received the driver attitude information, when the user has complained about the malfunction of the driver information detection device 150a or when visiting the dealer 130 for other purposes, Advice can be given on the proper use of the person information detection device 150a. At that time, if no person is resident in the center 120 and only the image at the time of detection abnormality is transmitted, the same service is provided to the user if the analysis is performed on the dealer 130 side. can do.

  Thus, according to the driver state detection device 150a according to the present embodiment, for example, when the detection by the first driver information detection means 10 is not normal, the situation image at the time of detection abnormality is transmitted to the dealer 130. Thus, it is possible to propose a driving posture or the like to the user, and the reliability of the driver state detection device 150a and the early collision detection device 30 using the same can be improved.

  FIG. 12 is a diagram illustrating an example of a communication mode in which a security system different from that in FIG. 11 is applied to the driver state detection device 150a and the early collision detection device 30 according to the second embodiment. For specific vehicles such as predetermined luxury cars, there is a case where a specific vehicle owner's desk is provided in the center 120 and a person is stationed for dealing with the owner of the specific vehicle from the viewpoint of ensuring security. 12, the example which comprised the driver state detection apparatus 150a and the early collision detection apparatus 30 which concern on Example 2 using such a form is shown.

  Under the above-described situation, when a malfunction of detection abnormality occurs in the driver state detection device 150a, the driver posture information is transmitted from the data communication module 42a to the center 120, as in the case of FIG. . When the specific vehicle owner's desk in the center 120 receives the driver attitude information, it analyzes the cause of the malfunction of the driver state detection device 150a. If it is determined that the cause of the malfunction is due to the driver's posture or the shielding object 60, the fact is notified to the driver of the vehicle 110 using the communication means. The notification may be a voice notification such as an in-car telephone, or a notification displayed on the meter 34 or a navigation display. As a result, the driver can recognize that the driver state detection device 150a is in a state in which the driver state cannot be normally detected due to his / her own driving posture and the presence of the shield 60, and the cause thereof can be determined. By correcting, the driver state detection device 150a can function normally.

  As described above, according to the driver state detection device 150a according to the aspect shown in FIG. 12, it is possible to know the detection abnormal state of the driver state detection device 150a and the corresponding solution in real time. It can be used effectively with high reliability.

  FIG. 13 is a diagram showing an example of the correspondence on the center 120 side when the form of the driver state detection device 150 a according to the second embodiment described in FIG. 12 is incorporated in the early collision detection device 30.

  FIG. 13A is a diagram showing a processing flow corresponding to the center 120 side. In the processing flow of FIG. 13A, the flow starts from the stage at which the driving attitude information is received at the center 120 from the data communication module on the vehicle 110 side.

  In step 141, the center 120 analyzes the received image of the driving posture information, and the cause of the detection abnormality is due to the driver's posture or the driver such as blocking the face detection camera 11 due to the placement of the shielding object 60. It is determined whether it is a factor. If it is determined in step 141 that the detected abnormality is caused by the driver, the process proceeds to step 151. If it is determined that the detected abnormality is not caused by the driver, the process proceeds to step 152.

  In step 151, since the detected malfunction is caused by the driver, the specific vehicle owner's desk notifies the driver so as to remove the specific malfunction factor. The notification at this time may be performed using a telephone or display means as described above. If the detection failure is resolved by the notification to the driver, the processing of the driver state detection device 150a according to the present embodiment ends.

  In step 152, since the detection failure is not caused by the driver, the driver cannot be advised. In this case, for example, a response to feedback to the collision determination ECU 31 of the early collision detection device 30 is performed.

  In step 153, as a feedback to the collision determination ECU 31, the degree of giving an alarm early is reduced. The specific contents will be described with reference to FIG.

  FIG. 13B is a diagram illustrating an example of the relationship between the driver state detection device 150 a and the early collision detection device 30. In FIG. 13B, TTC (Time to Collision) obtained by dividing the inter-vehicle distance from the preceding vehicle by the relative speed is shown as a safety evaluation index indicating a follow-up behavior assuming avoidance of a rear-end collision accident. Assuming that the point at which TTC becomes 0 is the collision timing, the collision early detection device 30 according to the present embodiment sets the normal collision warning timing several seconds before the collision timing, for example, 2 seconds before. Here, in the early collision detection device 30 according to the present embodiment, the driver state detected by the driver state detection device 150a is also a control condition of the alarm timing, and when the driver's sideways state or closed eye state is detected, Since the collision is more likely to occur, control is performed to advance the alarm timing by, for example, about 1 to 2 seconds from the normal alarm timing. When such a control function is installed in the collision determination ECU 31, if an erroneous detection of the driver state detection device 150a occurs, the alarm timing becomes earlier than the normal alarm timing. It is emitted and it becomes a very troublesome state for the driver.

  Thus, if the setting for reducing the early timing of the alarm timing is made in the collision determination ECU 31 as in step 153 of FIG. 13A, the erroneous detection of the driver state detection device 150a occurs. It is possible to avoid a state in which alarms occur frequently. If the cause of the detection abnormality is not caused by the driver, the driver state detection device 150a and the collision can be performed without causing trouble to the driver by taking the measures described in step 152 and step 153. The early detection device 30 can function.

  As for the alarm timing of the early collision detection device 30, such control can be realized if the alarm timing set value of the collision determination ECU 31 can be remotely operated from the center 120.

  Thus, according to the driver state detection device 150a and the early collision detection device 30 according to the second embodiment, it is a case where an abnormality in detection of the driver state occurs by effectively utilizing the functions of the security system. However, the center 120 can be used to make a flexible response, and the driver state detection device 150a and the early collision detection device 30 can be made reliable and troublesome for the user.

  The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the above-described embodiments, and various modifications and substitutions can be made to the above-described embodiments without departing from the scope of the present invention. Can be added.

  INDUSTRIAL APPLICABILITY The present invention can be used for a driver state detection device that detects a driver's state, an early collision detection device that uses this to improve driving safety, and the like.

DESCRIPTION OF SYMBOLS 10 1st driver state detection means 11 Face detection camera 12 Face image processing ECU
20 Detection abnormality determination means 21 Vehicle speed detection means 30 Early collision detection device 31 Collision determination ECU
32 Obstacle detection means 33 Buzzer 34 Meter 35 Brake ECU
36 Warning means 40, 40a Second driver state detection means 41, 41a, 41b Driving posture detection means 42, 42a Communication means 43 Driving posture determination means 44 Storage means 45 Driving posture information processing means 50 Steering wheel 51 Steering column cover 60 Shield 70 Seat 71 Headrest 80 Ceiling 90 Instrument panel 91 Deployment surface 100 Air bag 110 Vehicle 120 Center 130 Dealer 150, 150a Driver state detection device

Claims (6)

First driver state detecting means for detecting the driver's face orientation by imaging;
When the state in which no person is detected by the first driver state detection unit or the detection state in which the face direction is horizontal continues for a predetermined time, the face direction detection by the first driver state detection unit is normally performed. Detection abnormality determination means for determining that the detection abnormal state is not broken;
When the detection abnormality determination unit determines that the first driver state detection unit is in the detection abnormality state, a first position of the driver is detected from a position different from the first driving state detection unit. And a driver state detecting means. 2. A driver state detecting device comprising:   2. The driver state detection device according to claim 1, wherein the detection abnormality determination unit determines the detection abnormality state based on the lateral detection state when a vehicle speed is equal to or higher than a predetermined speed.   3. The driver state detection apparatus according to claim 1, wherein the second driver state detection unit includes a security camera that captures an image of the inside of the vehicle or a head position detection unit for airbag deployment. 4.   The second driver state detection unit includes a plurality of detection units, and determines the driver state based on evaluation values of the plurality of detection units. The driver state detection device according to item.   5. The driver according to claim 1, wherein the second driver state detection unit includes a communication unit that transmits the detected posture of the driver to a center outside the vehicle. State detection device. A collision early detection device that detects that a vehicle may collide and issues a warning to the driver,
The driver state detection device according to claim 5 is provided,
In the center where the driver's posture detected by the driver state detector is transmitted, the abnormal detection state is caused by a cause other than the driver's posture or the shielding of the first driving state detection means by the obstacle. If it is determined that the alarm has occurred, the timing for issuing the alarm is set later than in the normal state.