Method for adjusting sampling rate of viewing camera for viewing occupant of vehicle

Abstract

The invention relates to a method (30) for adjusting a sampling rate of an observation camera (28) for observing a vehicle occupant (24), comprising the following steps: providing observation data of the observation camera (28); identifying a blink event of the vehicle occupant (24) by evaluating the observation data, the start and end of the blink event being identified; according to the invention, an adjustment of the sampling rate of the observation camera (28) is determined as a function of the detected blink event, the sampling rate being increased if the start of the blink event of the vehicle occupant (24) is detected, and the sampling rate being reduced if the end of the blink event of the vehicle occupant (24) is detected.

Description

Technical Field

[0001] The invention relates to a method for adjusting the sampling rate of a viewing camera for viewing vehicle occupants and to a viewing device. Background Art

[0002] A method and a device for operating an occupant observation system are known from DE 10 2014 214 352 A1. Summary of the invention

[0003] A method for adjusting the sampling rate of an observation camera used to observe vehicle occupants is proposed, the method having the following steps:

[0004] Providing observation data of observation cameras;

[0005] By evaluating the observation data, identifying a blink event of the vehicle occupant, wherein a start and an end of the blink event are identified;

[0006] An adjustment of a sampling rate of the observation camera is determined as a function of the detected blink events, wherein the sampling rate is increased if the start of a blink event of the vehicle occupant is detected, and wherein the sampling rate is decreased if the end of a blink event of the vehicle occupant is detected.

[0007] The advantage of the method is that by adjusting the sampling rate, in particular by reducing the sampling rate, the energy input into the camera can be reduced. Advantageously, the sampling rate can be reduced if a high sampling rate is not required. Preferably, the sampling rate can be increased if this is desired and, for example, required to observe vehicle occupants. By reducing the energy input into the camera and other components, for example into the control device, the thermal properties of these components can be significantly improved. In other words, the heating of the camera and other components can be reduced, whereby they in particular do not overheat and therefore do not have temperature problems. Advantageously, the service life of these components can be increased or extended as a result. This is particularly advantageous for compact construction types and / or for high demands on the environment or surrounding components. As a result, the functionality of the observation device or the camera can be guaranteed or ensured.

[0008] By increasing the sampling rate when a blink event is identified, an optimal observation of the vehicle occupant can be achieved, whereby the performance of this observation can be maintained. Thus, the advantages are: the sampling rate is reduced, whereby no temperature issues occur, and the performance can be maintained by increasing the sampling rate at certain times, whereby no performance loss occurs.

[0009] In an exemplary embodiment, the method may also include a step of observing the vehicle occupant by means of an observation camera to determine observation data. In this way, in particular, the observation data can be generated safely and / or reliably. In this way, blink events of the vehicle occupant can also be determined safely and / or reliably.

[0010] Furthermore, the method can in particular include a step of adjusting the sampling rate of the observation camera as a function of the determined adjustment of the sampling rate of the observation camera. In this way, in particular the sampling rate of the observation camera can be adjusted safely and / or reliably. In this way, in particular the energy input, for example into the camera and / or into other components, can be reduced.

[0011] In an advantageous embodiment, the start of a blink event of the vehicle occupant can be detected if the difference in the degree of opening of the eyelid of the vehicle occupant between two points in time is greater than a threshold value. In addition, a first degree of opening of the eyelid of the vehicle occupant at the first point in time and a second degree of opening of the eyelid of the vehicle occupant at the second point in time can be determined, wherein the start of a blink event of the vehicle occupant is detected if the difference between the first degree of opening and the second degree of opening is greater than a threshold value. In this way, the start of a blink event of the vehicle occupant can be determined in a particularly safe and / or reliable manner, whereby the sampling rate can be adjusted in a particularly safe and / or reliable manner.

[0012] In an alternative or additional embodiment, a first sum of the visible sclera area, visible iris area and visible pupil area of ​​the eye of the vehicle occupant at a first time point and a second sum of the visible sclera area, visible iris area and visible pupil area of ​​the eye of the vehicle occupant at a second time point can be determined, wherein the start of a blink event of the vehicle occupant is detected if the difference between the first sum and the second sum is greater than a threshold value. In this way, the start of a blink event of the vehicle occupant can be determined in a safe and / or reliable manner, whereby the sampling rate can be adjusted in a safe and / or reliable manner.

[0013] In a further improvement, the blinking behavior of the vehicle occupant can be learned by means of a machine learning method to predict the start and / or end of the blinking event of the vehicle occupant. Alternatively or additionally, the probability of the start of the blinking event of the vehicle occupant can be predicted by means of a learned algorithm, wherein if the probability of the blinking event is greater than a threshold value, the sampling rate is increased. Advantageously, alternatively or additionally, the probability of the start of the blinking event of the vehicle occupant can be predicted by means of a statistical method that takes into account the time until the last blinking event, wherein if the probability of the blinking event is greater than a threshold value, the sampling rate is increased. By predicting the probability of the start of the blinking event, the sampling rate can be increased, especially when the probability of the blinking event is relatively high. By this, it can be ensured that: in particular, the start of the blinking event is also covered by means of a high sampling rate. By this, the function of the observation device can be improved, in particular the observation of the vehicle occupant. By this, the vehicle occupant can be observed safely and / or reliably.

[0014] In an exemplary embodiment, if the beginning of a blink event of a vehicle occupant is detected, the sampling rate can be increased to a value between approximately 40 Hz and approximately 80 Hz, in particular between approximately 50 Hz and approximately 70 Hz, in particular between approximately 55 Hz and approximately 65 Hz, in particular approximately 60 Hz. In this way, a particularly reliable sampling rate can be achieved, whereby the vehicle occupant can be observed particularly safely and / or reliably. In this way, safety in road traffic can be increased in particular.

[0015] In an advantageous embodiment, if the end of a blink event of the vehicle occupant is detected, the sampling rate can be reduced to a value between approximately 10 Hz and approximately 50 Hz, in particular between approximately 20 Hz and approximately 40 Hz, in particular between approximately 25 Hz and approximately 35 Hz, in particular approximately 30 Hz. This can ensure, in particular, that the sampling rate is reduced such that the energy input into the camera or into other components is reduced. This can reduce the probability of overheating of these components, thereby ensuring the functionality and / or life of these components.

[0016] Furthermore, a viewing device is proposed, which is designed to carry out a method for adjusting a sampling rate of a viewing camera for viewing vehicle occupants.

[0017] In a further development, the observation device can have an observation camera for observing the vehicle occupants. In an alternative embodiment, the observation device can be designed as an observation camera. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Embodiments of the present invention are shown in the drawings and are explained in more detail in the following description. The same reference numerals are used for elements that are shown in different drawings and have similar functions, and a repeated description of these elements is omitted.

[0019] Figure 1 shows a schematic diagram of a vehicle having a viewing device for viewing a vehicle occupant according to an embodiment of the present invention;

[0020] Figure 2 A schematic diagram showing a method according to an embodiment of the present invention is shown;

[0021] Figure 3 A schematic diagram showing a method according to an embodiment of the present invention is shown;

[0022] Figure 4 A schematic diagram showing a method according to an embodiment of the present invention is shown;

[0023] Figure 5 a schematic diagram showing a graph of eyelid opening over time in accordance with an embodiment of the present invention;

[0024] Figure 6 A schematic diagram showing the eyes of a vehicle occupant according to an embodiment of the present invention;

[0025] Figure 7 A schematic diagram of the eyes of a vehicle occupant is shown according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0026] Figure 1 A schematic diagram of a vehicle 20 (e.g., a motor vehicle, such as a car) is shown, which has an observation device 22 for observing one or more vehicle occupants 24 or passengers 24 according to an embodiment of the present invention. The vehicle 20 has, in particular, an interior space 26 or a vehicle interior space 26, wherein, in particular, one or more seats 27 for one or more vehicle occupants 24 can be arranged in the vehicle interior space 26. The vehicle 20 also has an observation device 22 for observing the vehicle interior space 26 and / or one or more vehicle occupants 24, such as a driver and / or a co-driver and / or one or more other passengers. The observation device 22 can also be referred to as an observation system, a driver monitoring system (Driver Monitoring System), an occupant observation system and / or a monitoring system. The observation device 22 can be designed, for example, to detect the direction of sight, the body posture and / or the position of the head or face or eyes of the vehicle occupant 24, or to detect the fatigue state and / or other vital signs of the vehicle occupant 24. For example, the identity of the vehicle occupant 24 can also be detected.

[0027] The viewing device 22 can be arranged in particular in a dashboard, in an instrument panel, in a steering wheel, on the roof, on a window, on a rearview mirror or on a column, for example on an A-pillar and / or a B-pillar of the vehicle 20 , or at another location in the vehicle 20 .

[0028] In order to observe the vehicle occupant 24, the observation device 22 can have a camera 28 or a recording unit. In a further development, the observation device 22 can have an illumination unit for emitting light, in particular infrared light. Alternatively, the observation device 22 can have no separate illumination unit and can illuminate the vehicle interior or the vehicle occupant with the aid of ambient light.

[0029] The lighting unit can be directed in particular in the direction of the vehicle occupant 24 and thus in the direction of the vehicle seat 27 in order to illuminate the vehicle occupant 24 with light, in particular with infrared light. In other words, light, in particular infrared light, can be emitted by means of the lighting unit in the direction of the vehicle occupant 24. The lighting unit can be designed, for example, as a lighting unit, a lighting element, a light emitting diode, an LED, an OLED and / or a laser diode and / or can have a lighting unit, a lighting element, a light emitting diode, an LED, an OLED and / or a laser diode.

[0030] The recording unit can be designed, for example, as an image recording unit, such as a sensor or a camera, in particular as an infrared camera module, wherein the recording unit is directed in the direction of the vehicle occupant 24 and thus in the direction of the vehicle seat 27 in order to visually detect the vehicle occupant 24. By being designed as an infrared camera module, this observation can also be carried out at night without the vehicle occupant 24 being brightly illuminated and thus in particular not being disturbed or dazzled.

[0031] In other words, the observation device 22 is designed to observe the vehicle interior of the vehicle. The observation device 22 may include: a lighting unit for emitting light, such as infrared light, in the direction of the vehicle interior; and a recording unit for recording light, such as infrared light. Preferably, the light emitted by the lighting unit may be reflective at or in the vehicle interior, wherein the reflected light may be deflected in the direction of the recording unit.

[0032] The observation device 22 can also have a control unit 29 or an evaluation unit 29 or a computing unit 29 for controlling the illumination unit and / or the recording unit and / or for processing the data recorded by means of the recording unit.

[0033] The monitoring device 22 is designed in particular to carry out Figure 2 , Figure 3 and / or Figure 4For example, the method described in the method can be implemented on the control unit 29 or the computing unit 29 or by means of the control unit or the computing unit. Figure 2 , Figure 3 and / or Figure 4 method.

[0034] Figure 2 A schematic diagram of a method 30 for adjusting the sampling rate of a viewing camera for viewing vehicle occupants according to an embodiment of the present invention is shown. The viewing camera may be part of a viewing device or may be designed as a viewing device. The viewing device may be configured as follows: Figure 1 The observation device is designed and / or arranged in the vehicle.

[0035] In a first step 32 of the method 30, observation data of the observation camera are provided. These observation data can be received, for example, by means of a receiving unit of the control device.

[0036] In a second step 34 of method 30 , a blink event of the vehicle occupant is detected by evaluating the observation data, wherein the start and end of the blink event are detected.

[0037] Here, the following approach can be envisaged for detecting a blink event.

[0038] The start of a blink event of the vehicle occupant can advantageously be identified if the difference in the degree of opening of the eyelid of the vehicle occupant between two points in time is greater than a threshold value. In other words, a first degree of opening of the eyelid of the vehicle occupant at a first point in time and a second degree of opening of the eyelid of the vehicle occupant at a second point in time can be determined, wherein the start of a blink event of the vehicle occupant is identified if the difference between the first degree of opening and the second degree of opening is greater than a threshold value. In other words, the difference between the degrees of opening of two different eyelids at different points in time is determined, wherein the start of a blink event is identified if the difference is greater than a threshold value.

[0039] In other words, the onset of a blink event of the vehicle occupant can advantageously be identified by determining a decrease in eyelid opening. The algorithm can determine the eyelid opening δ at each time point t t If the difference in eyelid opening between two time points {t0, t1}|t0<t1 Above a threshold value ∈, the sampling rate of the camera is advantageously increased or should be advantageously increased. In this case, the eyelid opening can be interpreted not only in terms of metrics but also at the image level. At the image level, compensation of the estimated head movement between the time points t0 and t1 can be performed. In this case, conditions related to the monotonicity of the eyelid signal are conceivable.

[0040] For example, you can use Figure 5 diagrams and / or with the help of Figure 6 eyes to present the recognition of blink events with the help of eyelid opening.

[0041] Alternatively or additionally, the start of a blink event can also be determined by determining a sum. For example, a first sum of the visible sclera area, visible iris area and visible pupil area of ​​the eye of the vehicle occupant at a first time point and a second sum of the visible sclera area, visible iris area and visible pupil area of ​​the eye of the vehicle occupant at a second time point can be determined, wherein the start of a blink event of the vehicle occupant is detected if the difference between the first sum and the second sum is greater than a threshold value.

[0042] In other words, the onset of a blink event of a vehicle occupant can advantageously be identified by determining the areas of the visible sclera, pupil and iris. With the aid of an algorithm, the sum of the areas A of the visible sclera, visible iris and visible pupil at each time point t can be determined: t If the difference between the sum of the areas at two time points {t0, t1}|t0<t1 Above a threshold value ∈, the sampling rate of the camera is increased or should be increased. The threshold value can be designed, for example, as an area value. In this case, the area sum can be interpreted not only metrically but also at the image level. At the image level, compensation of the estimated head movement between the time points t0 and t1 can be performed. In this case, conditions related to the monotonicity of the area signal are conceivable.

[0043] For example, you can use Figure 7 The eyes are presented to identify blink events with the help of the area.

[0044] In a further development, the optical flow in the eyelid region, which compensates for the head movement, can be used to switch the sampling rate of the camera. This optical flow represents the movement between the two images. The difference between the area around the eyelid at time t0 and the eyelid area at time t1 can be stored in a vector field via this optical flow. By means of this vector field, in particular, it is known how the pixels move from one image to the next. Based on this movement, it can be advantageously known how the eyelid moves and whether the eyelid is closed.

[0045] In an advantageous embodiment, the blinking behavior of a vehicle occupant, in particular a specific vehicle occupant, can be learned by means of a machine learning method in order to predict the start and / or end of a blinking event of the vehicle occupant. For example, a vehicle occupant can be identified, wherein a blinking behavior can be learned accordingly for the identified vehicle occupant and can be called up later during the method. For example, the probability of the start of a blinking event of a vehicle occupant can be predicted by means of a learned algorithm, wherein the sampling rate is increased if the probability of a blinking event is greater than a threshold value. Alternatively or additionally, the probability of the start of a blinking event of a vehicle occupant can be predicted by means of a statistical method that takes into account the time until the last blinking event, wherein the sampling rate is increased if the probability of a blinking event is greater than a threshold value.

[0046] In other words, blink events of the vehicle occupant can be predicted with the aid of a predictor. In other words, blink events can be predicted by a predictor. If the probability of a blink event is above a threshold, the sampling rate of the camera can be increased. In this case, predictors based on AI methods and / or statistical methods can be used, which take into account the time until the last blink event.

[0047] In particular, it is conceivable that the start of a blink event can be recorded only with a lower sampling rate. In these cases, the start of the blink event process can be estimated, for example, with the aid of a parameterizable estimation model. For calibration and determination of the model parameters, for example, past blink events, the subsequent course of blink events with a switched sampling frequency, and assumptions can be used for this purpose.

[0048] In particular, false alarms, while they may increase the energy input into system components, have no further impact on the overall system.

[0049] In a third step 36 of the method 30 , an adjustment of the sampling rate of the observation camera is determined as a function of the detected blink events, wherein the sampling rate is increased if the start of a blink event of the vehicle occupant is detected, and wherein the sampling rate is reduced if the end of a blink event of the vehicle occupant is detected.

[0050] Advantageously, if the start of a blink event of the vehicle occupant is detected, the sampling rate can be increased to a value between about 40 Hz and about 80 Hz, in particular between about 45 Hz and about 80 Hz, in particular between about 45 Hz and about 75 Hz, in particular between about 50 Hz and about 70 Hz, in particular between about 55 Hz and about 65 Hz, in particular about 60 Hz. Preferably, if the end of a blink event of the vehicle occupant is detected, the sampling rate can be reduced to a value between about 10 Hz and about 50 Hz, between about 10 Hz and about 45 Hz, between about 15 Hz and about 45 Hz, in particular between about 20 Hz and about 40 Hz, in particular between about 25 Hz and about 35 Hz, in particular about 30 Hz.

[0051] In other words, the method 30 can be designed as a method for adjusting the image sampling rate according to the behavior of the eyelids. The goal is in particular to reduce the average image sampling rate in order to reduce the energy input into the system components. In order to observe the driver, one or more driver observation cameras can be operated with a variable sampling rate, for example in order to determine parameters such as the duration, start or end of a blink event. Advantageously, a camera with a lower frame rate can be operated by means of the method 30, wherein the frame rate is increased only during the determination or recognition of a blink event. In other words, blink events can be classified.

[0052] Figure 3 A schematic diagram of a method 30 for adjusting the sampling rate of a viewing camera for viewing vehicle occupants according to an embodiment of the present invention is shown. The viewing camera may be part of a viewing device or may be designed as a viewing device. The viewing device may be configured as follows: Figure 1 The observation device is designed and / or arranged in the vehicle.

[0053] The method 30 has in particular the following Figure 2 The steps of the method 30 thus have a first step 32, a second step 34 and a third step 36. Figure 3 The method 30 has a further fourth step 38. In the fourth step 38 of the method 30, the sampling rate of the observation camera is adjusted according to the determined adjustment of the sampling rate of the observation camera. Advantageously, the sampling rate of the observation camera is increased if the start of a blink event of the vehicle occupant is detected, and the sampling rate of the observation camera is reduced if the end of a blink event of the vehicle occupant is detected. The fourth step 38 can in particular take place after the third step 36 of determining the adjustment of the sampling rate.

[0054] Figure 4A schematic diagram of a method 30 for adjusting the sampling rate of a viewing camera for viewing vehicle occupants according to an embodiment of the present invention is shown. The viewing camera may be part of a viewing device or may be designed as a viewing device. The viewing device may be configured as follows: Figure 1 The observation device is designed and / or arranged in the vehicle.

[0055] The method 30 has in particular the following Figure 2 The method 30 thus has a first step 32, a second step 34 and a third step 36. In a further improvement, the method 30 may additionally have Figure 3 The fourth step 38 of the method 30 is as follows. Figure 4 The method 30 has a further fifth step 40 . In the fifth step 40 of the method 30 , the vehicle occupants are observed by means of an observation camera for determining observation data. The fifth step 40 can in particular take place before the first step 32 .

[0056] Figure 5 A schematic diagram of a diagram 42 of eyelid opening over time according to an embodiment of the present invention is shown. The diagram 42 has an x-axis 44, on which time is shown. The diagram 42 also has a y-axis 46, on which the eyelid opening in millimeters is shown. In the diagram 42, two functions 48, 50 of eyelid opening over time are shown. The two functions 48, 50 initially run at least approximately parallel to the x-axis 44. In this case, the sampling rate according to bracket 52 is maintained at approximately 30 Hz, for example. Preferably, the sampling rate can be maintained at a value between approximately 10 Hz and approximately 45 Hz, in particular between approximately 20 Hz and approximately 40 Hz, in particular between approximately 25 Hz and approximately 35 Hz, in particular approximately 30 Hz.

[0057] In the area of ​​circle 54, the start of a blink event is detected. This detection can be performed, for example, according to Figure 2 , Figure 3 and / or Figure 4 The eyelid opening and therefore the functions 48, 50 fall to lower values. As a result, the eyelid opening decreases. In the region of this decrease, the sampling rate according to bracket 56 is increased or raised, for example, to approximately 60 Hz. Advantageously, if the beginning of a blink event of the vehicle occupant is detected, the sampling rate can be raised to a value between approximately 45 Hz and approximately 80 Hz, in particular between approximately 50 Hz and approximately 70 Hz, in particular between approximately 55 Hz and approximately 65 Hz, in particular approximately 60 Hz.

[0058] In other words, the graph 42 shows the eyelid opening in mm over time during a blink event. In other words, the graph shows two examples of the course of the eyelid opening during a blink event in the development of a reference frame for the eyelid opening. The method can identify the start of a blink event, wherein the frame rate of the camera is then increased.

[0059] Figure 6 A schematic diagram of an eye 58 of a vehicle occupant according to an exemplary embodiment of the present invention is shown. The method or algorithm identifies a lower eyelid 60 and an upper eyelid 62. The maximum distance between the lower eyelid 60 and the upper eyelid 62 is measured. From this, an eyelid opening 64 is determined, in particular at a first point in time. The eyelid opening 64 can be specified, for example, in millimeters. This determination can be performed at multiple points in time. For example, it can be performed by means of Figure 2 , Figure 3 and / or Figure 4 The method identifies the onset of a blink event by determining the degree of eyelid opening.

[0060] Figure 7 A schematic diagram of an eye 58 of a vehicle occupant according to an embodiment of the present invention is shown. The eye 58 has the area of ​​the visible sclera 66, the area of ​​the visible iris 68 and the area of ​​the visible pupil 70. These areas can be added, for example, to form a total area or area sum A. For example, the total area can be calculated by means of Figure 2 , Figure 3 and / or Figure 4 The method identifies the onset of a blink event by determining the sum of the areas.

Claims

1. A method (30) for adjusting a sampling rate of an observation camera (28) for observing a vehicle occupant (24), the method comprising the following steps: Providing observation data of the observation camera (28); By evaluating the observation data, blink events of the vehicle occupant (24) are identified, wherein: identifying the beginning and the end of the blink event; An adjustment of the sampling rate of the observation camera (28) is determined based on the recognized blink event, wherein the sampling rate is increased if the start of the blink event of the vehicle occupant (24) is recognized, and wherein the sampling rate is reduced if the end of the blink event of the vehicle occupant (24) is recognized.

2. The method (30) according to claim 1, comprising the step of observing the vehicle occupant (24) by means of the observation camera (28) to determine observation data.

3. The method (30) according to any one of the preceding claims, comprising the step of adjusting the sampling rate of the observation camera (28) as a function of the determined adjustment of the sampling rate of the observation camera (28).

4. The method (30) according to any one of the preceding claims, wherein: If the difference in the degree of opening of the eyelids of the vehicle occupant (24) between two points in time is greater than a threshold value, the beginning of a blink event of the vehicle occupant (24) is detected.

5. The method (30) according to any one of the preceding claims, wherein: A first degree of opening of the eyelids of the vehicle occupant (24) at a first point in time and a second degree of opening of the eyelids of the vehicle occupant (24) at a second point in time are determined, wherein if a difference between the first degree of opening and the second degree of opening is greater than a threshold value, the start of a blink event of the vehicle occupant (24) is identified.

6. The method (30) according to any one of the preceding claims, wherein: A first sum of the visible sclera area, visible iris area and visible pupil area of ​​the eye of the vehicle occupant (24) at a first point in time and a second sum of the visible sclera area, visible iris area and visible pupil area of ​​the eye of the vehicle occupant (24) at a second point in time are determined, wherein if the difference between the first sum and the second sum is greater than a threshold value, the start of a blink event of the vehicle occupant (24) is identified.

7. The method (30) according to any one of the preceding claims, wherein: The blinking behavior of the vehicle occupant (24) is learned by means of a machine learning method in order to predict the start and / or end of a blink event of the vehicle occupant (24).

8. The method (30) according to any one of the preceding claims, wherein: The probability of the onset of a blink event of the vehicle occupant (24) is predicted by means of a learned algorithm, wherein the sampling rate is increased if the probability of a blink event is greater than a threshold value.

9. The method (30) according to any one of the preceding claims, wherein: The probability of the onset of a blink event of the vehicle occupant (24) is predicted by means of a statistical method taking into account the time until the last blink event, wherein the sampling rate is increased if the probability of a blink event is greater than a threshold value.

10. The method (30) according to any one of the preceding claims, wherein: If the onset of a blink event of the vehicle occupant (24) is detected, the sampling rate is increased to a value between approximately 40 Hz and approximately 80 Hz, in particular between approximately 50 Hz and approximately 70 Hz, in particular between approximately 55 Hz and approximately 65 Hz, in particular approximately 60 Hz.

11. The method (30) according to any one of the preceding claims, wherein: If the end of a blink event of the vehicle occupant (24) is detected, the sampling rate is reduced to a value between approximately 10 Hz and approximately 50 Hz, in particular between approximately 20 Hz and approximately 40 Hz, in particular between approximately 25 Hz and approximately 35 Hz, in particular approximately 30 Hz.

12. An observation device (22) designed to carry out the method (30) according to any one of the preceding claims.

13. The observation device (22) according to claim 12, comprising an observation camera (28) for observing a vehicle occupant (24).

Images