Device and method for detecting distribution of driver of vehicle

Abstract

The invention relates to a device and a method for detecting a distraction of a vehicle driver (12) on a driver seat (14) of a vehicle, in which at least one image (100) of an image of a region having a vehicle interior comprising at least one arm (130) of the vehicle driver (12) is detected by means of an interior camera (52). Image data corresponding to the image 100 is generated and processed by a processing unit (20). From the image data, a position of at least one arm (132) and a hand position of a hand of the arm (130) are determined. At least one instrument position for arranging an end instrument (16, 28) introduced into the vehicle in a feasible location in the cockpit (10) is stored in a storage area of the processing unit (20). A test is made as to whether the determined hand position (118) of the vehicle driver (12) coincides with the instrument position.

Description

Technical Field

[0001] The invention relates to a device and a method for detecting distraction of a vehicle driver in a driver's seat of a vehicle. At least one camera of the vehicle is designed to capture at least one image with an image of a region of the vehicle interior including at least the upper body of the vehicle driver and to generate image data corresponding to the image. At least one processing unit is designed to process the image data. Background Art

[0002] Document DE 10 20 20 165 A1 discloses a monitoring device for monitoring a vehicle driver of a motor vehicle and for testing whether the vehicle driver is distracted according to predetermined criteria due to an intended operating action or an already performed operating action.

[0003] Document US2018 / 0238686A1 discloses a method for determining whether a vehicle driver is using a mobile phone or other portable electronic instrument. The information determined by means of a camera is compared with the information stored in a memory.

[0004] Document US2021 / 0174103A1 discloses a vehicle driver monitoring system having at least two cameras. For this purpose, the vehicle driver's posture in the recorded image can be used based on the image of the vehicle driver's face from the first and second cameras to detect possible distraction of the vehicle driver.

[0005] Vehicle drivers often introduce additional technical instruments, such as smartphones and tablets, into the vehicle and arrange them in the cockpit of the vehicle at a position that is advantageous from their perspective. Different brackets are available in the aftermarket for arranging such instruments. Likewise, such instruments are positioned at positions in the cockpit without additional brackets. When the vehicle driver operates these instruments, the vehicle driver is distracted. Thus, different legislators around the world require systems that measure the degree of attention of vehicle drivers. In China, such systems are called DAMS and are applicable to new vehicles starting in 2023. In Europe, such systems are called Advanced Driver Distraction Warning Systems (ADDW) and are applicable to new vehicles starting in 2024.

[0006] In order to detect the attention of the vehicle driver, a camera system is installed in the vehicle interior of the vehicle, which can be used for driver monitoring and interior monitoring at the same time. However, these interior cameras have a limited field of view, so that the instruments introduced additionally are often not in the field of view of the camera. As a result, the hands of the vehicle driver are also not visible or only partially visible in the image recorded by the interior camera when operating these instruments. The interior cameras in the vehicle for driver monitoring usually have a field of view of up to 150 degrees. In order to increase this field of view, complex optical systems must be provided, which require a relatively large structural space and cannot be integrated into the cockpit without problems. In addition, there is a relatively high cost associated with this. However, it is desirable that the interaction of the vehicle driver with the instruments outside the field of view of the interior camera is also detected as distraction. Summary of the invention

[0007] The object of the present invention is to specify a device and a method for detecting distraction of a vehicle driver who is seated in the vehicle by a terminal additionally introduced into the vehicle.

[0008] This object is achieved by a device having the features of claim 1 and by a method having the features of the independent method claim. Advantageous developments are specified in the dependent claims.

[0009] By means of the device according to claim 1 and the method with the features of the independent method claim, it is possible to determine that the vehicle driver is distracted by a terminal device such as a smartphone or a tablet computer introduced into the vehicle when the instrument position and the hand position are consistent. In particular, a plurality of possible instrument positions are stored, which are then all tested for consistency with the determined hand positions. Possible instrument positions are in particular a charging cradle installed in the vehicle, in the area in front of the vehicle's instrument cluster, in the area in front of the central information display, in the area in front of the head-up display, in or in front of a foldable storage compartment, on a rubber mat in the cockpit, in front of the air outlet nozzle, for example in the center of the cockpit or on the left next to the steering wheel, and possible arrangement positions of a suction cup holder, which can be fixed to the windshield of the vehicle and can have a so-called goose neck. In addition, a holder that can be fixed to the side window is known. Other possible instrument positions are the cup holder in the vehicle and the area in front of the steering wheel.

[0010] Starting from the image data provided by the camera, the processing unit determines the position of the vehicle driver's arm. In particular, the processing unit determines the position of the forearm and preferably determines the position of the hand from this, even if the hand itself is not visible in the image. This can be achieved in particular with the help of a human posture estimation method, by which the position of the forearm can also be determined from a 2D image as a 3D vector, wherein the hand position can also be determined from this, if it is not visible in the image. If the hand is visible in the image, the position of each phalange of the hand can also be determined with the help of a known human posture estimation method and the movement of the finger for operating the terminal instrument (tapping and wiping) can also be reliably detected. In the known human posture estimation method, 3D key points are extracted from a sequence of images of a video image stream, wherein the actual movement of the person is analyzed with the help of the human posture estimation method and the individual limbs of the person can be determined. In common motion analysis, the known human posture estimation method usually uses 17 or 18 key points. However, in this case, more or fewer key points can also be used to determine the position and movement of the vehicle driver's arm.

[0011] It is particularly advantageous if a plurality of possible positions for arranging a terminal device introduced into the vehicle are stored in a memory area of ​​the processing unit. The processing unit then tests whether the determined hand position of the vehicle driver coincides with one of the instrument positions. This allows a plurality of instrument positions to be taken into account when detecting distraction of the vehicle driver.

[0012] Furthermore, the processing unit is advantageously designed to determine the course of a viewing direction of at least one eye of the vehicle driver based on the image data and to test whether the viewing direction extends through the one instrument position or one of the multiple instrument positions. This tests a further criterion for detecting distraction of the vehicle driver, which can be taken into account when deciding whether the vehicle driver is actually distracted by a terminal device installed in the vehicle.

[0013] Furthermore, the processing unit can determine the position of the vehicle driver's arm from the image data using a human pose estimation method and, as already explained, the hand position from the position of the vehicle driver's arm. Alternatively or additionally, the processing unit can determine the hand position from the image data using a human pose estimation method even if the hand is outside the recorded image, i.e., outside the field of view of the camera. It is particularly advantageous if the processing unit detects distraction if,

[0014] - the hand position is consistent with the instrument position within a preset time interval, and / or

[0015] - the hand position coincides with the instrument position multiple times within a preset time interval, and / or

[0016] The hand position coincides with the instrument position within a predefined time interval and the viewing direction of the vehicle driver extends through the instrument position.

[0017] In this way, distraction of the vehicle driver can be detected relatively reliably.

[0018] It is also advantageous if the cockpit position of an input unit or output unit of the vehicle is stored in the storage area of ​​the processing unit, the input unit or output unit being arranged in the instrument position or the input unit or output unit being pre-assigned to the instrument position or being arranged at a maximum distance of 5 cm from the instrument position. In other embodiments, the cockpit positions of a plurality of input units and / or output units of the vehicle are stored, the plurality of input units and / or output units being located at the instrument position or in the vicinity of the instrument position. The processing unit detects distraction only if, within a time interval in which the hand position coincides with the instrument position, the processing unit does not detect an operating input via the input unit arranged at the cockpit position and / or the processing unit determines that no warning signal is output via the output unit arranged at the cockpit position. This reliably prevents a correct interaction of the vehicle driver with the vehicle from being detected as a distraction caused by a terminal instrument introduced into the vehicle.

[0019] It is particularly advantageous that the processing unit determines the movement of the arm and / or hand of the vehicle driver based on image data of a plurality of images recorded in sequence by a camera and determines the operation of a terminal device introduced into the vehicle based on known movement patterns. In particular, typical movements such as tapping and wiping movements when operating a touch screen of a known terminal device are determined here by arm movements and / or hand movements. The processing unit preferably detects the distraction only when it determines the operation of the terminal device introduced into the vehicle within a time interval in which the hand position coincides with the device position. The known movement pattern can be stored in a storage area of ​​the processing unit. Alternatively or additionally, the data of the known movement pattern can be used to train an artificial neural network of the processing unit.

[0020] It is also advantageous that the time interval is stored in a preset manner in a storage area of ​​the processing unit and / or that the number of coincidences of hand positions and instrument positions determined within the preset time interval is preset in a storage area of ​​the processing area as a theoretical value, and when the theoretical value is exceeded, the processing unit detects distraction. This further increases the probability that the distraction is actually caused by the introduction of a terminal device. It is also advantageous that the introduction of a terminal device, such as a smartphone, to a known / suspicious location is also detected by means of a camera and detected by the processing unit. This can be taken into account by the processing unit as an additional determining feature when determining whether the vehicle driver is distracted. In particular, the detection of the introduction of a terminal device can act as a first indicator, in particular as a prerequisite, in order to detect distraction of the vehicle driver based on this.

[0021] It is particularly advantageous if the processing unit detects the hand of the vehicle driver in the image based on the image data of the image and tracks the hand based on further image data of a plurality of images of an image sequence recorded successively by the camera and detects the position at which the hand leaves the image. Such tracking is also referred to as tracking. Based on the subsequent movement of the arm, the position of the hand in the vehicle interior is determined. Preferably, the hand is determined as a 3D vector which moves with the arm based on the movement of the arm in space.

[0022] It is particularly advantageous if the field of view of the camera has a detection angle in the range of 100 to 150 degrees, wherein the camera detects only a portion of the entire cockpit. The camera can in particular be a monocular camera for recording 2D images. Preferably, the camera is an IR-RGB camera or a monochrome camera (with or without infrared light detection) that records grayscale images or a pure infrared camera.

[0023] A second aspect relates to a method having the features of the independent method claim, which can be developed in the same way as the device according to claim 1 , in particular having the features of the dependent claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Further features and advantages emerge from the following description, which explains exemplary embodiments in conjunction with the drawings.

[0025] in:

[0026] Figure 1 a perspective schematic diagram showing a cockpit of a vehicle including an additional separate display unit; and

[0027] Figure 2A schematic image recorded by an interior camera is shown, which includes key points determined by a processing unit of the vehicle using a human pose estimation method. DETAILED DESCRIPTION

[0028] Figure 1 The perspective schematic diagram of the cockpit 10 of the vehicle is shown. The vehicle driver 12 is sitting in the driver's seat 14 of the vehicle. A terminal device in the form of a tablet computer 16 has been introduced into the vehicle and can be arranged in the cockpit 10 at a more or less suitable position. Another terminal device in the form of a smartphone 28 is arranged on the dashboard of the cockpit 10 in front of the steering wheel 40. In particular, such terminal devices 16, 18 are arranged in the cockpit 10 by means of a holding device available in the aftermarket so that the vehicle driver 12 can see and operate them. By operating these terminal devices 16, 28, the vehicle driver 12 is distracted.

[0029] The cockpit 10 also has a central information display (CID) 26, a head-up display and a graphic instrument cluster 30, on which different display elements 32 and images can also be output. The cockpit 10 has a steering wheel 40 with operating elements 42, 44, a gear selector switch 46, pedals 48 and an input unit 34 with a rotary wheel and key functions and / or a touch input field. The input unit 34 is therefore also called an Ergo Commander.

[0030] An interior camera 52 is arranged in the upper region of the front windshield 50 of the vehicle, and is used to detect at least one image including an image of the interior region of the vehicle. The interior camera detects an image including an image of at least a part of the upper body and head of the vehicle driver 12. The interior camera 52 is used not only for monitoring the attention of the vehicle driver 12 but also for monitoring the interior. The interior camera 52 generates image data corresponding to the image and transmits the image data to the processing unit 20.

[0031] The processing unit 20 has a data output 22 and a data input 24 which are used for connection to other units of the vehicle, in particular to other input and output units and control units of assistance systems.

[0032] Figure 2 A schematic image 100 recorded by interior camera 52 is shown, which includes key points 102 to 130 determined by processing unit 20 of the vehicle using a method for estimating a human pose.

[0033] The interior camera 52 has a field of view of 120 degrees. In other embodiments, the camera can also have a field of view in the range of 100 to 150 degrees. Usually, the interior camera 52 is arranged and oriented so that it does not detect some areas of the cockpit 10, so that the input instruments 16, 20 introduced into the vehicle arranged there are not in the field of view of the interior camera 52 and the operation of the terminal instruments by the vehicle driver is therefore not visible in the image. The instrument position of the introduced terminal instrument 16, 28 is stored in the storage area of ​​the processing unit 20 as a possible instrument position of the introduced terminal instrument in addition to other possible instrument positions of the terminal instrument. These other positions can be in particular:

[0034] - a charging station provided in the vehicle,

[0035] - the area in front of the instrument cluster 30,

[0036] - before the central information display 26,

[0037] - In or before the retractable storage compartment,

[0038] - on rubberized surfaces present in the cockpit 10 , in particular on rubber mats arranged in the cockpit 10 ,

[0039] to the right and left of the vehicle driver 12 before the air in the cockpit 10 exits the nozzles or ventilation grilles,

[0040] the usual position of the suction cup holder, which can have a goose neck and can be arranged usually on the front window 50 or on the side windows to the right and left of the vehicle driver 12 ,

[0041] - a clampable retaining device in the side window,

[0042] - in the cup holder, on the dashboard in front of the steering wheel 40 and

[0043] - Other partially vehicle-specific instrument locations.

[0044] All possible instrument positions for the vehicle are stored in a memory area of ​​the processing unit 20, since the terminal instruments 16, 28 can be arranged at each of these instrument positions. The processing unit 20 implements a human body pose estimation method, by which key points 102 to 130 on the human body of the vehicle driver 12 are identified and classified. Starting from the image of the vehicle driver 12 recorded by the interior camera 52, the following key points are determined with the aid of the human body pose estimation method:

[0045] Right eye 102 , left eye 104 , base of neck 106 , upper torso 108 , right shoulder 110 , left shoulder 112 , right elbow 114 , left elbow 116 , right hand 118 , left hand 120 , right hip 122 , left hip 124 , lower torso 126 , right knee 128 , and left knee 130 .

[0046] Starting from these key points 102 to 130, vectors between the key points 102 to 130 can be determined and tracked during movements of, in particular, right arm 132 or left arm 134 of vehicle driver 12. The position of a hand outside of image 100 in the interior of the vehicle can thus also be determined as hand position 118.

[0047] Furthermore, methods are known by which the direction of the sight of the eyes 102, 104 of the vehicle driver 12 can be precisely determined, so that the area of ​​the cockpit 10 towards which the sight of the vehicle driver 12 is directed can be determined. If the hand position 118 is at the instrument position and / or the sight is directed towards the instrument position within a certain time interval, it can be detected that the vehicle driver 12 is distracted and the introduced terminal is at the instrument position. In addition, it can be tested whether an input and / or output unit, such as a central information display 26 or an instrument cluster 30, is at a potential instrument position, whether a message 32 that can attract the attention of the vehicle driver is output, or whether an input is made via a corresponding input unit at a potential instrument position or within a circumference of a maximum of 5 cm. In this case, the processing unit 20 does not detect that the vehicle driver is distracted by the introduced terminal, but rather detects a common operating behavior in the vehicle.

[0048] The interior camera 52 continuously records images as an image sequence, i.e., a video sequence. As a result, the movement of the arm 132, 164 of the vehicle driver 12 can be detected, from which the hand movement can be inferred. Based on this, operating inputs on possible terminal devices arranged at the instrument position, such as scrolling and tapping with the index finger or thumb, can be detected. By means of feedback from an input device (such as the central information display 26) to the processing unit 20, the processing unit 20 can easily distinguish between the operation of possible terminal devices arranged at the instrument position and the correct vehicle operation. In particular, the operation of the air conditioning operating unit or the audio operating unit can be determined by the processing unit 20 and correspondingly taken into account in the case of detected distraction. Thus, the interaction with the introduced terminal device and the interaction with the vehicle operating elements can be easily distinguished. In the case of detection of distraction by the terminal device, a warning or prompt can be output according to legal requirements or the corresponding information can be given to the assistance system, so that, for example, the degree of autonomous driving of the vehicle is increased, the distance from other vehicles is increased, the speed is reduced and / or the brake system is preactivated.

[0049] Several examples for detecting distraction of the vehicle operator 12 are described next.

[0050] Example 1:

[0051] Interior camera 52 detects an image of the interior, which includes parts of the body of vehicle driver 12 and parts of the interior. Vehicle driver 12 looks at the smartphone, which is not within the field of view of interior camera 52, for example, the smartphone is arranged on a holder on the ventilation grille of the air outlet located in the center of cockpit 10, that is, arranged to the right in front of vehicle driver 12 from the perspective of vehicle driver 12.

[0052] The processing unit 20 processes the image data of the images captured by means of the interior camera 52 and additionally detects sight lines in the direction of the ventilation grille. To this end, the processing unit 20 detects the number of sight lines of the vehicle driver 12 within a preset time (in the range of 5 seconds to 60 seconds) and / or accumulates the time during which the vehicle driver 12 looks at the area of ​​the ventilation grille. The ventilation grille or the area in front of the ventilation grille is stored in the processing unit 20 as a possible instrument position.

[0053] The processing unit 20 does not detect any interaction with operating elements in the area, such as a warning flashing button arranged in the area of ​​the ventilation grille, during the time interval of the increased viewing angle. The output of warning signals to output units in this area is also monitored by the processing unit 20, since such warning signals also provide an explanation for the viewing angle of the vehicle driver 12 in the direction of the instrument position in front of the ventilation grille.

[0054] If a preset threshold value is exceeded, the processing unit 20 can classify that the vehicle driver 12 is interacting with the smartphone. As a threshold value, a time interval can be predefined within which an increased gaze is permitted and distraction is then detected if the threshold value is exceeded. Alternatively or additionally, the number of gazes within a preset time interval and / or the cumulative duration of gazes within a preset time interval can also be detected and compared with a theoretical value. If the theoretical value is exceeded, the processing unit 20 detects distraction of the vehicle driver 12. Therefore, if the smartphone is not visible in the image 100 recorded with the aid of the interior camera 52, that is, if the smartphone is outside the field of view of the interior camera 52, then according to the present invention, an extension of the evaluation algorithm of the processing unit 20 is implemented to detect smartphone interactions by the vehicle driver 12.

[0055] Example 2:

[0056] The vehicle driver 12 looks at a smartphone, which is not arranged in the field of view of the interior camera 52. For example, the smartphone is arranged in the cockpit 10 in the area before the air outlet nozzle or the ventilation grille, as in Example 1. With the aid of the continuously recorded camera images, the movement of the arm 130 of the vehicle driver 12 or, if the fingers of the right arm 130 of the vehicle driver 12 are visible in the image 100, the movement of the fingers is detected and evaluated in order to infer the operation of the terminal or the smartphone from the movement of the arm 132, 134 and / or the fingers. In addition, the processing unit 20 detects the line of sight in the direction of the ventilation grille in addition to the image captured by the interior camera 52. The processing unit 20 detects the movement of the arm 130 and the hand 118 away from the steering wheel 40 by means of the fingers of the hand extending in the direction of the ventilation grille before the fingers and the hand disappear from the field of view of the interior camera 52 due to the movement of the hand in the direction of the ventilation grille.

[0057] The processing unit 20 detects no operating actions or interactions with the input and output units in this area of ​​the cockpit 10 which would necessitate sight lines and hand movements in this area.

[0058] If the processing unit 20 determines that the threshold value is exceeded, the processing unit classifies that the vehicle driver 12 is interacting with the smartphone, i.e., the processing unit 20 detects that the vehicle driver 12 is distracted. For a description of the threshold value and for further possibilities for determining that the vehicle driver 12 is interacting with the smartphone or that the vehicle driver 12 is distracted, see Example 1.

[0059] Example 3:

[0060] In the same way as in examples 1 and 2, the vehicle driver 12 interacts with the smartphone, wherein movements of the vehicle driver's 12 fingers and / or arms are visible in the image 100, but the smartphone is not arranged in the field of view of the interior camera 52. The smartphone can be arranged in the same way as in examples 1 and 2 in the center of the cockpit 10 before the air outlet nozzle or the ventilation grille.

[0061] Starting from images 100 continuously captured by interior camera 52 , whose image data are processed by processing unit 20 , processing unit 20 detects a movement of a hand or key point 118 of vehicle driver 12 with a finger extended in the direction of an air outlet nozzle not visible in image 100 .

[0062] Furthermore, the processing unit 20 does not detect any operating actions on operating elements in this area, such as key actuation or output of warning signals, which would require a hand movement or a gaze direction toward the potential device position of the smartphone. In particular, distraction of the vehicle driver 12 can be reliably detected by determining a hand movement and / or an extended finger before the hand leaves the field of view of the interior camera 52.

[0063] In principle, the positions of key points 102 to 130, such as the position of hand 118 or the position of fingers that are no longer visible in the image of the person, can be extrapolated using methods for estimating the person's posture. By determining the vectors between the individual key points, the 3D vectors of the person's limbs and the 3D positions of key points 102 to 130 can be determined. This allows both the position of the hand and the position of the fingertips to be determined, even if this body part is no longer visible in image 100. Distraction can be detected as a time threshold if the hand position and the instrument position persist for longer than a predefined time, which has a value in the range of 0.5 seconds to 5 seconds, in particular a value of 1 second.

[0064] Operating inputs (such as tapping or swiping) based on movements of arm 130 are also possible with the aid of gesture recognition methods if the hand carrying out the operating input or the finger of the hand carrying out the operating input is not visible in detected image 100 .

[0065] As a further possible measure after the distraction has been detected, a threshold value may be adapted (in particular by an assistance system). Additionally or alternatively, a graphical or acoustic warning may also be output to the vehicle driver 12 . Reference numerals list

[0066] 10 Cockpit

[0067] 12 Vehicle Driver

[0068] 14 Driver's Seat

[0069] 16 tablets

[0070] 20Control Unit

[0071] 22 Data output terminal

[0072] 24 data input terminals

[0073] 26 Central information display

[0074] 28 Smartphone

[0075] 30 Instrument cluster

[0076] 32 display elements

[0077] 34 Ergo Controller

[0078] 40 Steering wheel

[0079] 42, 44 operating elements

[0080] 46 gear selector switch

[0081] 48 pedals

[0082] 50 front window glass

[0083] 52 Internal Space Camera

[0084] 100 camera images

[0085] 102 to 130 key points

[0086] 132, 134 Vehicle driver's arm

Claims

1. A device for detecting distraction of a vehicle driver in a driver's seat of a vehicle, The device comprises at least one camera (52) of the vehicle, the at least one camera being designed to detect at least one image (100) having an image of an area of ​​the vehicle interior including at least one arm (132, 134) of a vehicle driver (12) and to generate image data corresponding to the image (100). The device comprises at least one processing unit (20) which is designed at least for processing the image data. in, The processing unit (20) is further designed to determine, based on the image data, a position of at least one arm (132, 134) of the vehicle driver (12) and a hand position (118) of a hand of the arm (132, 134), At least one device position of a possible location for arranging a terminal device introduced into the vehicle in the cockpit (10) of the vehicle is stored in a memory area of ​​the processing unit (20), The processing unit (20) is further designed to test whether the determined hand position (118) of the vehicle driver (12) matches the instrument position.

2. The device according to claim 1, characterized in that The device positions of a plurality of feasible locations are stored in a storage area of ​​the processing unit (20), and The processing unit (20) is further designed to test whether the determined hand position (118) of the vehicle driver (12) coincides with one of the instrument positions.

3. The device according to any one of the preceding claims, characterized in that The processing unit (20) is further designed to determine the course of a viewing direction of at least one eye of the vehicle driver (12) based on the image data and to test whether the viewing direction extends through the one instrument position or one of the multiple instrument positions.

4. The device according to any one of the preceding claims, characterized in that The processing unit (20) determines the position of the arm (132, 134) of the vehicle driver (12) based on the image data by means of a human posture estimation method and determines the hand position and / or The processing unit (20) determines the hand position (118) based on the image data by means of a human pose estimation method.

5. The device according to any one of the preceding claims, characterized in that The processing unit (20) detects distraction when: The hand position (118) is consistent with the instrument position within a preset time interval, and / or The hand position (118) coincides with the instrument position multiple times within a preset time interval, and / or The hand position (118) coincides with the instrument position within a predefined time interval and the viewing direction of the vehicle driver (12) extends through the instrument position.

6. The device according to claim 5, characterized in that In a memory area of ​​the processing unit, a cockpit position of at least one input unit and / or output unit (26, 30) of the vehicle is stored, the input unit and / or output unit being arranged at the instrument position or at a distance of up to 5 cm from the instrument position or being assigned to the instrument position in a preset manner, The processing unit (20) detects the distraction only if, during a time interval in which the hand position (118) coincides with the instrument position, the processing unit does not detect an operating input via an input unit arranged at the cockpit location and / or the processing unit determines that no warning signal (32) is output via an output unit (26, 30) arranged at the cockpit location.

7. The device according to claim 5 or 6, characterized in that The processing unit (20) determines the movement of the arm (132, 134) and / or hand of the vehicle driver (12) based on image data of a plurality of images (100) recorded successively by the camera (52) from an image sequence and determines the operation of a terminal device introduced into the vehicle based on known movement patterns, The processing unit (20) detects the distraction only if the processing unit (20) determines an operation of a terminal device introduced into the vehicle within a time interval in which the hand position (118) coincides with the position of the device. The known movement pattern is in particular stored in a memory area of ​​the processing unit (20) and / or is used to train an artificial neural network of the processing unit.

8. The device according to any one of the preceding claims 5 to 7, characterized in that The time interval and / or In a memory area of ​​the processing unit (20), the number of times the hand position (118) and the instrument position coincide within the preset time interval is preset as a target value, and when the target value is exceeded, the processing unit detects distraction.

9. The device according to any one of the preceding claims, characterized in that The processing unit (20) detects the hand of the vehicle driver (12) in the image based on the image data of the image and tracks the hand based on further image data of a plurality of images (100) of an image sequence recorded successively by the camera (52) and detects the position at which the hand leaves the image (100) and determines the hand position (118) in the vehicle interior based on subsequent movements of the arm (132, 134).

10. The device according to any one of the preceding claims, characterized in that The field of view of the camera (52) has a detection angle in the range of 100° to 150°, the camera (12) detects only a part of the entire cockpit (10), and the camera (52) is preferably a monocular camera and / or an RGB camera.

11. A method for detecting distraction of a vehicle driver in a driver's seat of a vehicle, in, detecting at least one image (100) with an image of a region of a vehicle interior including at least the upper body of a vehicle driver (12) and generating image data corresponding to the image and processing it by means of a processing unit (20), determining the position of at least one arm (132, 134) of the vehicle driver (12) based on the image data and determining the hand position (118) of the hand of the vehicle driver (12) based thereon, storing at least one position as an instrument position for arranging a terminal device introduced into the vehicle at a possible location in the cockpit (10) in a storage area of ​​the processing unit (20), A test is performed to determine whether the determined hand position (118) of the vehicle driver (12) corresponds to an instrument position for arranging a terminal instrument introduced into the vehicle at a possible location in the cockpit (10).

Images