Method and device for detecting driver fatigue in a vehicle

Abstract

Method for detecting driver fatigue in a vehicle (100), the method comprising the following steps: Reading (201) an exceedance signal as an indication of an impending exceedance of a road edge marking (104, 106) by the vehicle (100); Evaluating (203; 503) a course (220; 640) of a vehicle's (100) trajectory prior to a time (t1, t2, t3, t4, t5; 642) of the reading of the speed limit signal, wherein a lateral distance (108, 110) between the vehicle (100) and the lane edge marking (104, 106) is evaluated, and setting a validity value attributable to the speed limit signal depending on a result of the evaluation to a value indicating validity of the speed limit signal or to a value indicating invalidity of the speed limit signal, wherein the validity value is set to the value indicating validity if the lateral distance (108, 110) is within a time interval between the time (t4) of the reading of the speed limit signal and a previous time (t3) of the reading a previous exceedance signal is greater than a threshold (222),and wherein the validity value is set to the invalidity value if the lateral distance (108, 110) within the time interval between the time (t2) of reading the exceedance signal and a previous time (t1) of reading a previous exceedance signal is permanently less than a threshold value (222); and , Providing (205) a fatigue warning signal using the exceedance signal when the validity value indicates the validity of the exceedance signal.

Description

State of the art

[0001] The invention relates to a device or a method according to the preamble of the independent claims. The present invention also relates to a computer program.

[0002] WO 2008 / 052827 A1 describes a method for driver state detection, whereby a signal indicating the driver state is derived from a quantity which indicates the frequency of occurring extreme values ​​in the time course of a quantity representing the lane behavior of the driver (TLC).

[0003] DE 10 2014 219 810 A1, subsequently published, relates to a method for determining the degree of fatigue of a driver of a vehicle, in which a driver-related determination of a change in the degree of fatigue is carried out using lane violation behavior by classifying lane violations as unconscious or conscious by the driver.

[0004] DE 10 2012 001 741 A1 discloses a monitoring system that is set up to provide an indicator whose value quantifies a driver's ability to operate the vehicle due to driver fatigue, based on the frequency of occurrence of an assistance action by the driver assistance system.

[0005] EP2143612A1 relates to a method for warning a driver of a motor vehicle of insufficient lane keeping ability, in which, during driving, the lateral position of the vehicle within a lane is determined, the activity of the driver with regard to at least one predetermined driving action is determined, a counter representing a measure of the driver's lane keeping ability is incremented depending on the determined vehicle position and the determined driver activity, and a warning is issued to the driver when the counter exceeds a predetermined threshold. Disclosure of the invention

[0006] Against this background, the approach presented here introduces a method and a device for detecting driver fatigue in a vehicle, as well as a corresponding computer program according to the main claims. Advantageous further developments and improvements of the device specified in the independent claim are possible through the measures listed in the dependent claims.

[0007] The described approach can be advantageously used, among other things, in conjunction with an algorithm for driver fatigue detection based on camera lane data. Frequent approaching of the lane is detected by the algorithm, which then calculates whether a driver has reached a critical level of fatigue. The fundamental building block of the algorithm is the Time-to-Lane Crossing (TLC) value, which indicates in seconds how long the vehicle would theoretically need to cross the right or left lane markings, given its current course. Fatigue is then determined by the frequency with which a driver falls below a predefined TLC threshold (right and left).

[0008] In the development of fatigue detection algorithms based on lane information (lane camera), it has been observed that multiple TLC events can occur in succession when the vehicle is driving near the lane boundary. A TLC event can indicate that a lane boundary crossing is likely to occur within a predetermined time or at the end of a predetermined time period. Such events can arise, for example, due to signal resolution, signal noise, sporty driving, etc., and do not necessarily indicate driver fatigue. Similarly, some drivers, for example, when cornering, do not want to stay precisely in the lane center. This also leads to TLC events, but these do not indicate driver fatigue. Advantageously, such events that do not indicate fatigue can be detected using the described approach. This allows for improved fatigue detection.

[0009] A procedure for detecting driver fatigue in a vehicle includes the following steps: Reading an overstepping signal as an indication of an impending overstepping of a road edge marking by the vehicle; Evaluating the vehicle's trajectory prior to the reading of the exceedance signal and setting a validity value attributable to the exceedance signal, depending on the result of the evaluation, to a first value indicating validity of the exceedance signal or to a second value indicating invalidity of the exceedance signal; and Providing a fatigue warning signal using the exceedance signal when the validity value indicates the validity of the exceedance signal.

[0010] The overstepping signal can indicate that the vehicle, given its current course, will soon cross a lane marking, such as a shoulder or center line, for example, within a shorter time than a predetermined threshold. The overstepping signal can indicate, for example, that an impending overstepping has been detected, or a time interval after which the overstepping is expected to occur, or that the overstepping will occur within a shorter time interval than a predetermined interval. For example, the overstepping signal can indicate the occurrence of a TLC event. Thus, the notification could, for example, be a warning of the occurrence of a TLC event.The speed limit signal can be determined, or may have been determined, using data from a vehicle's environmental sensing device, such as a camera. The vehicle's trajectory can be characterized by distance values, which indicate the time course of the distance between the vehicle and a lane marking, or by steering values, which indicate the time course of the vehicle's steering input. The distance values ​​can be determined, for example, using the vehicle's environmental sensing device. The steering values ​​can be determined, for example, using a sensor linked to the steering system. If the impending speed limit violation indicated by the speed limit signal was not intentionally caused by the driver, the speed limit signal may be an indication of driver fatigue.Whether the impending speed limit violation is intentional or unintentional can be indicated by the validity value. A valid speed limit signal can indicate an unintentional impending violation and thus driver fatigue. An invalid speed limit signal can indicate an intentional impending violation and thus the driver's attentiveness. The fatigue warning signal can be an indication of driver fatigue.

[0011] According to one embodiment, the lateral distance between the vehicle and the lane marking can be evaluated during the assessment step. Such a distance can be easily determined, for example, using a camera. A temporal profile of the lateral distance is suitable for detecting a driving trajectory influenced by driver fatigue.

[0012] The validity value can be set to the value indicating validity if the lateral distance within a time interval between the time the exceedance signal is read and a previous time the exceedance signal was read is greater than a threshold. It may be sufficient if the distance exceeds the threshold only once within that time interval. In this way, for example, the first exceedance signal that occurs after the threshold has been breached can be considered valid. A token, also called a marker, can be set if the distance is greater than the threshold. If an exceedance signal is read while the token is set, the exceedance signal can be considered valid. Reading the exceedance signal can trigger the token to be removed.If an overrun signal occurs when no token is set, the overrun signal can be considered invalid.

[0013] Thus, the validity value can be set to the invalid value if the lateral distance within a time interval between the time of reading the exceedance signal and a previous time of reading a previous exceedance signal is consistently less than a threshold. This prevents multiple fatigue warning signals from being generated by an impending exceedance, which could lead to a distortion of fatigue detection.

[0014] According to one embodiment, in addition to or as an alternative to evaluating the lateral distance, a change in the vehicle's steering angle can be evaluated during the evaluation step. A temporal profile of the steering angle is also suitable for detecting a driving trajectory influenced by driver fatigue.

[0015] During the evaluation step, the steering angle change pattern can be compared to a pattern to determine whether the validity value should be set to the value indicating validity or invalidity. For example, the validity value can be set to the value indicating validity if the steering angle change pattern within a time period prior to the time the override signal is read corresponds to a predetermined pattern. In this way, a simple pattern comparison may be sufficient to determine the validity of the override signal. The pattern may have been generated using characteristic steering angle curves that indicate driver fatigue.

[0016] In this case, the validity value can be set to the value indicating validity if the change in steering angle within an initial time period before the exceedance signal is read is less than a first threshold value, and within a second time period extending from the first time period until the exceedance signal is read, is greater than the first threshold value. This allows for the detection of a sudden, sharp steering input by the driver, which may indicate driver fatigue.

[0017] For example, the validity value can be set to the value indicating validity if a predetermined criterion is met by the ratio between a first maximum change in the steering angle within a first time interval preceding the time the exceedance signal is read and a second maximum change in the steering angle within a second time interval following the first time interval and extending to the time the exceedance signal is read. In this way, a steering intervention can be detected that is significantly more pronounced than the steering interventions otherwise being performed.

[0018] Such a pronounced steering intervention can, in turn, indicate driver fatigue.

[0019] The validity value can be set to the value indicating validity if, furthermore, the second maximum change is greater than a second threshold value. In this way, a minimum magnitude of a steering intervention indicative of fatigue can be defined.

[0020] Device for detecting fatigue in a driver of a vehicle, which is configured to perform the method according to one of the preceding claims.

[0021] This process can be implemented, for example, in software or hardware, or in a hybrid form of software and hardware, for example in a control unit.

[0022] The approach presented here thus also creates a device designed to carry out, control, and implement the steps of a variant of the method presented here in appropriate facilities. This embodiment of the invention in the form of a device also allows the problem underlying the invention to be solved quickly and efficiently.

[0023] In this context, a device can be understood as an electrical device that processes sensor signals and outputs control and / or data signals accordingly. The device may have an interface, which can be implemented in hardware and / or software. In the case of a hardware-based interface, the interfaces can, for example, be part of a so-called system ASIC, which incorporates various functions of the device. However, it is also possible that the interfaces are separate integrated circuits or consist at least partially of discrete components. In the case of a software-based interface, the interfaces can be software modules, which, for example, are present on a microcontroller alongside other software modules.

[0024] Also advantageous is a computer program product or computer program with program code that can be stored on a machine-readable carrier or storage medium such as a semiconductor memory, a hard disk memory or an optical memory and is used to carry out, implement and / or control the steps of the method according to one of the embodiments described above, in particular if the program product or program is executed on a computer or device.

[0025] Exemplary embodiments of the invention are shown in the drawings and explained in more detail in the following description. It shows: Fig. 1 a schematic representation of a vehicle with a device for detecting driver fatigue according to an exemplary embodiment; Fig. 2 a flowchart of a procedure for detecting driver fatigue according to an exemplary embodiment; Fig. 3 a diagram illustrating a method for detecting driver fatigue according to an exemplary embodiment; Fig. 4 a schematic representation of a vehicle with a device for detecting driver fatigue according to an exemplary embodiment; Fig. 5 a flowchart of a procedure for detecting driver fatigue according to an exemplary embodiment; Fig. 6 a diagram illustrating a method for detecting driver fatigue according to an exemplary embodiment; Fig. 7 a course of a steering angle velocity and an occurrence of a TLC event according to an embodiment; Fig. 8 a course of a steering angle velocity and an occurrence of a TLC event according to an exemplary embodiment.

[0026] In the following description of favorable embodiments of the present invention, the same or similar reference numerals are used for the elements shown in the various figures and acting similarly, without repeating these elements.

[0027] Fig. Figure 1 shows a schematic representation of a vehicle 100 with a device 102 for detecting driver fatigue in the vehicle 100 according to an exemplary embodiment. The vehicle 100 travels on a roadway bounded by a left lane marking 104 and a right lane marking 106. The vehicle 100 currently maintains a lateral distance 108 from the right lane marking 106 and a lateral distance 110 from the left lane marking 104. According to this exemplary embodiment, the vehicle 100 has an image acquisition device in the form of a camera 112. The lane markings 104 and 106 are within the detection range of the camera 112. The camera 112 is designed to provide image data captured by the camera 112 to an evaluation unit 114.

[0028] The evaluation unit 114 is configured to determine the trajectory of the vehicle 100 based on the image data. According to this embodiment, the evaluation unit 114 is configured to determine, based on the image data, a temporal profile of at least one of the distances 108, 110, and the trajectory, in turn, as a temporal profile of at least one of the distances 108, 110. Furthermore, the evaluation unit 114 is configured to determine a violation signal based on the image data and additionally or alternatively based on the temporal profile of at least one of the distances 108, 110. According to this embodiment, the violation signal indicates the occurrence of a TLC event, which in turn indicates the detection of an impending violation of one of the lane markings 104, 106 by the vehicle 100.The evaluation unit 114 is designed to provide the exceedance signal as well as data relating to the temporal course of the driving trajectory, according to this embodiment data relating to the temporal course of at least one of the distances 108, 110, to the device 102.

[0029] Device 102 has an interface for reading the exceedance signal and data relating to the temporal progression of the driving trajectory. Device 102 is designed to evaluate the temporal progression of the driving trajectory prior to the time of reading the exceedance signal in order to determine the validity of the exceedance signal. The time of reading the exceedance signal can be equated with the time of detection of the impending exceedance or the time of occurrence of the TLC event. If the evaluation of the driving trajectory shows that the exceedance signal is considered valid, the device is designed to provide a fatigue warning signal in response to the exceedance signal considered valid.If, on the other hand, the evaluation of the driving trajectory shows that the exceedance signal is considered invalid, the device is designed not to provide a fatigue warning signal in response to the exceedance signal considered invalid.

[0030] According to one embodiment, the device 102 is configured to evaluate the course of the vehicle trajectory within a predetermined time interval before the time the exceedance signal is read. According to this embodiment, the device 102 is configured to evaluate the temporal course of the vehicle trajectory by measuring the temporal progression of at least one of the distances 108, 110.

[0031] According to this embodiment, the vehicle 100 has a device 116 configured to provide a fatigue value representing the driver's fatigue based on the fatigue signal provided by the device 102 or based on a plurality of fatigue signals provided successively by the device 102. The fatigue value can be used, for example, to issue a warning to the driver.

[0032] To provide a fatigue warning signal, the device 102 has one or more devices configured to perform steps of a procedure as detailed in the following figures.

[0033] Fig. Figure 2 shows a flowchart of a method for detecting driver fatigue according to an exemplary embodiment. The method comprises steps 201, 203, and 205, which are carried out using suitable equipment based on Fig. The device described in section 1 can be implemented.

[0034] In step 201, a lane marking violation signal is read, indicating an impending violation of a lane marking by the vehicle, for example, the occurrence of a TLC event. In step 203, the vehicle's trajectory, prior to the reading of the lane marking violation signal, is evaluated. Depending on the result of this evaluation, a validity value indicating the validity of the lane marking signal is set to either a valid value or an invalid value. In step 205, a driver fatigue warning signal is generated in response to a lane marking signal deemed valid.

[0035] Step 203 can be performed continuously, so that when a TLC event occurs, an evaluated history of the driving trajectory may already be available.

[0036] In an optional subsequent step 207, successive fatigue warning signals can be evaluated to infer driver fatigue. For example, the temporal sequence or frequency of the occurrence of fatigue signals can be evaluated.

[0037] According to this embodiment, the course of the driving trajectory is evaluated based on the course of the lateral distance between the vehicle and the road edge marking.

[0038] According to one embodiment, the validity value is set to the value indicating validity only if the lateral distance within a predetermined time interval before the time the exceedance signal is read or the TLC event occurs is greater than a threshold value. This time interval can have a predefined value or extend back to the time of a previous exceedance signal being read or the occurrence of a previous TLC event. Conversely, if the distance is consistently less than the threshold value or another threshold value within the time interval of the exceedance signal, the validity value is set to the value indicating invalidity.

[0039] According to one embodiment, a token is used to decide whether the validity value is set to a value indicating validity or invalidity, as defined by… Fig. 3 is described in more detail.

[0040] Fig. Figure 3 shows a diagram illustrating a method for detecting driver fatigue according to an exemplary embodiment. The abscissa of the diagram represents time t in seconds, and the ordinate represents the distance dy in meters between the vehicle and the road edge marking. The diagram shows the course 220 of the distance over time. The course 220 initially remains below a threshold value 222, exceeds the threshold value 222 at time t3, and falls below the threshold value 222 again after time t4. For example, the threshold value is set to 0.5 meters.

[0041] According to this embodiment, as shown by the Fig. 1 and Fig. 2 described, the course 220 of the lateral distance between the vehicle and the road edge marking was evaluated to assess the validity of an exceedance signal.

[0042] Time points t1, t2, t3, t4, and t5 mark the points in time at which successive exceedance signals are read, or equivalent points in time at which successive TLC events occur. Time points t1, t2, t3, t4, and t5 are assigned intervals x1, x2, x3, x4, and x5. Not all TLC events occurring at time points t1, t2, t3, t4, and t5 indicate driver fatigue. The described approach allows those TLC events that indicate fatigue to be classified as valid and those that do not to be classified as invalid.

[0043] According to this embodiment, those TLC events are considered valid if they were preceded by a period in which the distance was greater than the threshold value of 222. This applies to the TLC events at times t1, t4, and t5. Those TLC events preceded by a period in which the distance was consistently less than the threshold value of 222 are considered invalid. This applies to the TLC events at time t2. Time t3 represents a TLC event preceded by a period in which the distance was consistently less than the threshold value of 222. However, the distance reaches the threshold value of 222 at time t3, so the TLC event associated with time t3 can be considered either valid or invalid, depending on the implementation of the described approach.According to this embodiment, the time intervals range from the time of a current TLC event, for example time t2, to the time of a temporally preceding TLC event, in this case time t1.

[0044] If the trend 220 remains below the threshold 222 at subsequent time points after time t5, these time points are considered invalid. As soon as the trend 220 exceeds the threshold 222 again, the following time point is considered valid.

[0045] The following is a corresponding example of implementation based on: Fig. 3. As already explained, the curve 220 represents the lateral distance. Time points marked with a checkmark are considered valid TLC events, and time points marked with a cross are considered invalid TLC events. The validity or invalidity of a TLC event is set using a token. The areas designated with reference 224 indicate areas where the token is set. The areas designated with reference 226 indicate areas where the token is removed. Only those TLC events that occur at time points t1, t4, and t5, at which the token is set, can be considered valid. If a TLC event is considered valid, the token is removed, unless the distance during the time of the TLC event is greater than the threshold 222.

[0046] The described approach enables an increase in the robustness of trace-based fatigue detection. This is achieved by identifying TLC events not resulting from fatigue, in order to exclude them, for example, from subsequent frequency calculations.

[0047] According to this embodiment, a TLC event can only be triggered with a valid token. The token is revoked after a triggered TLC event, if necessary. The token only becomes valid again after a certain condition has been met. This approach is based on the premise that another valid TLC event should only occur if the vehicle has since moved a certain distance from the lane boundary.

[0048] In Fig. Figure 3 shows a corresponding example scenario. At time t1, a TLC event is triggered because the token has existed since the start time. Upon detection of the TLC event, the token is removed. At time t3, when the lateral distance x3 is greater than 0.5 m, the token is returned, and another TLC event can be triggered. The TLC event at time t2 is invalid because no token is present at that time. At time t4, another TLC event is triggered. Since the token is present at this time, another valid TLC event is triggered. Thus, consecutive TLC events at the track boundary are prevented.

[0049] Thus, according to one embodiment, a robust fatigue detection system is created, for example based on the frequency of exceeded TLC (Time-to-Lane-Crossing) values, by exclusively considering a further TLC value, whereby the further TLC value is only taken into account if the vehicle has previously left the critical area, i.e. an exceeded TLC value such as the threshold value 222.

[0050] Fig. Figure 4 shows a schematic representation of a vehicle 100 with a device 102 for detecting driver fatigue in the vehicle 100 according to an exemplary embodiment. The vehicle 100 corresponds to the one shown in the following: Fig. The vehicle described in Figure 1. Additionally, a device 430 for detecting the steering angle of a steering system of the vehicle 100 is shown. The device 430 is designed to provide steering angle data representing the steering angle to the evaluation unit 114.

[0051] In addition to or alternatively to the one based on Fig. As described in section 1, the evaluation unit 114 is designed to determine the vehicle trajectory based on the image data from camera 112. According to this embodiment, the evaluation unit 114 is configured to determine a change in the steering angle based on the steering angle data and, in turn, to determine the vehicle trajectory as a temporal progression of the change in the steering angle. Furthermore, the evaluation unit 114 is configured, as already described, to determine the vehicle trajectory based on the steering angle data. Fig. As described in Figure 1, the violation signal is determined based on the image data. According to this embodiment, the violation signal indicates the occurrence of a TLC event, which in turn indicates the detection of an impending violation of one of the lane markings 104, 106 by the vehicle 100. The evaluation unit 114 is designed to provide the violation signal, as well as data regarding the temporal progression of the driving trajectory, and, according to this embodiment, data regarding the temporal progression of the change in the steering angle, to the device 102.

[0052] The device 102 has an interface for reading the exceedance signal as well as the data relating to the temporal progression of the driving trajectory. The device 102 is designed to, as shown by Fig. 1 described how to evaluate a temporal progression of the driving trajectory prior to the time of reading the exceedance signal in order to determine the validity of the exceedance signal.

[0053] According to one embodiment, the device 102 is configured to evaluate the course of the vehicle trajectory within a predetermined time period before the time the exceedance signal is read. According to this embodiment, the device 102 is configured to evaluate the temporal course of the vehicle trajectory by measuring the temporal course of the change in the steering angle.

[0054] To provide the fatigue warning signal, the device 102 has one or more devices configured to perform steps of a procedure as detailed in the following figures.

[0055] Fig. Figure 5 shows a flowchart of a method for detecting driver fatigue according to an exemplary embodiment. The method comprises steps 201, 503, 205, and an optional step 207. Steps 201, 503, 205, and 207 can be performed using suitable equipment based on Fig. The device described in section 4 will be implemented.

[0056] In step 201, as already shown in the Fig. As described in section 2, a violation signal is read in, indicating an impending violation of a lane marking by the vehicle, for example, the occurrence of a TLC event. In step 503, a vehicle trajectory is evaluated that precedes the reading of the violation signal. Depending on the result of the evaluation, a validity value indicating the validity of the violation signal is set to either a valid value or an invalid value. In step 205, as already described in section 203, the vehicle's trajectory is evaluated. Fig. 2 described, responding to a valid exceedance signal, a fatigue warning signal is provided.

[0057] According to this embodiment, the course of the vehicle trajectory is evaluated based on the change in the vehicle's steering angle. For this purpose, the change in the steering angle is evaluated using at least one reference pattern.

[0058] According to one embodiment, the validity value is only set to the value indicating validity if the change in the steering angle within a time period prior to the reading of the exceedance signal or the equivalent occurrence of the TLC event corresponds to a predetermined pattern. This time period can have a predefined value or extend back to the reading of a previously received exceedance signal or the occurrence of a previously received TLC event.

[0059] According to one embodiment, in step 503 two sub-time periods are considered, and the validity value is only set to the value indicating validity if the change in the steering angle within a first sub-time period is less than a first threshold value and within a second sub-time period is greater than the first threshold value. In this embodiment, the sub-time periods are directly adjacent to each other, and the second sub-time period lies between the first sub-time period and the time the exceedance signal is read.According to one embodiment, a ratio is further calculated between a first maximum change in the steering angle within the first sub-time interval and a second maximum change in the steering angle within the second sub-time interval. The validity value is set to the value indicating validity only if the ratio meets a predetermined criterion, for example, if it is greater than or less than a reference value. According to another embodiment, the validity value is set to the value indicating validity only if the second maximum change is greater than a second threshold value.

[0060] This procedure for deciding on the validity or invalidity of a TLC event indicated by the exceedance signal is explained below using the following: Fig. Sections 6 to 8 are described in detail.

[0061] Fig. Figure 6 shows a diagram illustrating a method for detecting driver fatigue according to an embodiment. The abscissa of the diagram represents time t in seconds, and the ordinate represents the steering angle velocity, indicating a change in the steering angle. The diagram shows a curve 640 of the change in the steering angle over time. During a first time interval t11, also referred to above as the first sub-interval, the curve 640 initially oscillates around the origin with small fluctuations, reaching a maximum value y1. During a second time interval t12 following the first time interval t11, also referred to above as the second sub-interval, the curve 640 exhibits a sharp increase, reaching a value y2 at the end of time interval t12 that is a multiple of the value y1.The time 642 of the occurrence of the value y2 can correspond to a time of occurrence of a TLC event and thus to a time 642 of reading an exceedance signal.

[0062] According to this embodiment, as shown by the Fig. 4 and Fig. As described in section 5, the curve 640, responding to the occurrence of a TLC event, is evaluated to determine whether the curve 640 of the change in steering angle in the time interval t11, t12 prior to time 642 of the occurrence of the TLC event exhibits a predetermined pattern. Based on this evaluation, the validity of the TLC event and thus of the exceedance signal is assessed.

[0063] The one based on Fig. The approach described in section 6 enables fatigue detection by coupling track-based and steering-based components. This is achieved by recognizing TLC events not resulting from fatigue, in order to exclude them, for example, in a subsequent frequency calculation.

[0064] According to one embodiment, the TLC events (based on lane information) are validated by the driver's steering movements. This approach is based on the premise that a TLC event should always be accompanied by a pattern in which the driver has not actively steered the steering wheel for a certain period of time and then suddenly applies a stronger correction to the steering wheel. Fig. Figure 6 illustrates one possible configuration. During period t11, the instantaneous steering angular velocity (dSA) has its maximum value y1. In the immediately following period t12, the steering angular velocity has its maximum value y2. For the exemplary pattern that leads to a valid evaluation, the following conditions must be met, according to one embodiment: the first time interval t11 has a first predetermined value, the second time interval t12 has a second predetermined value, the quotient of y1 and y2 is greater than a predetermined value, and y2 is greater than a predetermined steering angular velocity. For example, the following conditions are set: the first time interval t11 is equal to two seconds, the second time interval t12 is equal to one second, the quotient y1 / y2 is greater than four, and y2 is greater than 2° / s. According to one embodiment, only those TLC events that exhibit such a steering pattern are valid.

[0065] For the in Fig. In example 6, the steering wheel pattern is recognized at time t=4s (642) because "y2 / y1 > 4" and "y2 > 2". Therefore, the TLC event that occurred at time t=4s is considered valid.

[0066] This approach primarily solves problems in curves where TLC events often occur without relevance to fatigue.

[0067] Fig. Figure 7 shows a curve 640 of a steering angle velocity and the occurrence of a TLC event 750 according to an exemplary embodiment. In a time interval 752 before the occurrence of the TLC event 750, the curve 640 is evaluated to determine whether a steering pattern indicative of driver fatigue can be detected, as can be seen, for example, from Fig. 6 is described. According to the illustrated embodiment, the steering pattern is detected within the time interval 752, so that the TLC event 750 is a steering pattern TLC event which is considered a valid TLC event because it is regarded as an indication of fatigue.

[0068] Fig. 8 indicates accordingly Fig. 7 a curve 640 of a steering angle velocity and the occurrence of a TLC event 750 according to an exemplary embodiment. In the time interval 752 before the occurrence of the TLC event 750, the curve 640 is evaluated to determine whether a steering pattern can be identified, such as that which may be based on Fig. 6 is described. According to the illustrated embodiment, the steering pattern is not detected within the time interval 752, so that the TLC event 750 is a TLC event without a steering pattern, which is considered an invalid TLC event, since it is not seen as an indication of fatigue, but for example as an indication of the driver of the vehicle intentionally cutting a corner.

[0069] The one based on the Fig.The approach described in sections 5 to 8 enables the validation of TLC results using steering patterns. This allows for a distinction between unintentional driving maneuvers, which indicate fatigue, and intentional driving maneuvers, such as corner cutting. According to one embodiment, this is achieved by validating TLC detection using steering parameters. In particular, a characteristic steering pattern is used for this purpose. Such a specific pattern is employed to identify intentional and unintentional steering maneuvers.

[0070] If an embodiment includes an “and / or” connection between a first feature and a second feature, this is to be read as meaning that the embodiment according to one embodiment has both the first feature and the second feature, and according to another embodiment either only the first feature or only the second feature.

Claims

[1] Method for detecting fatigue in a driver of a vehicle (100) wherein the method comprises the following steps: Reading (201) an exceedance signal as an indication of an impending exceedance of a road edge marking (104, 106) by the vehicle (100); Evaluating (203; 503) a course (220; 640) of a vehicle's (100) trajectory prior to a time (t1, t2, t3, t4, t5; 642) of the reading of the speed limit signal, wherein a lateral distance (108, 110) between the vehicle (100) and the lane edge marking (104, 106) is evaluated, and setting a validity value attributable to the speed limit signal depending on a result of the evaluation to a value indicating validity of the speed limit signal or to a value indicating invalidity of the speed limit signal, wherein the validity value is set to the value indicating validity if the lateral distance (108, 110) is within a time interval between the time (t4) of the reading of the speed limit signal and a previous time (t3) of the reading a previous exceedance signal is greater than a threshold (222),and wherein the validity value is set to the invalidity value if the lateral distance (108, 110) within the time interval between the time (t2) of reading the exceedance signal and a previous time (t1) of reading a previous exceedance signal is permanently less than a threshold value (222); and, Providing (205) a fatigue warning signal using the exceedance signal when the validity value indicates the validity of the exceedance signal. [2] Method according to claim 1, wherein in the evaluation step a change (640) of a steering angle of the vehicle (100) is evaluated. [3] Method according to claim 2, wherein the validity value is set to the value indicating validity when a progression (640) of the change in the steering angle within a time interval (t11, t12) before the time (642) of reading the exceedance signal corresponds to a predetermined pattern. [4] Method according to claim 2 or 3, wherein the validity value is set to the value indicating validity if the course (642) of the change of the steering angle within a first time interval (t11) before the time (642) of reading the exceedance signal is less than a first threshold value and within a second time interval (t12) which extends from the first time interval (t11) to the time (642) of reading the exceedance signal is greater than the first threshold value. [5] Method according to one of claims 2 to 4, wherein the validity value is set to the value indicating validity when a ratio between a first maximum change (y1) of the steering angle within a first time interval (t11) prior to the time of reading the exceedance signal and a second maximum change (y2) of the steering angle within a second time interval (t12) following the first time interval (t11) and extending to the time (640) of reading the exceedance signal satisfies a predetermined criterion. [6] Method according to claim 5, wherein the validity value is set to the value indicating validity, if furthermore the second maximum change (y2) is greater than a second threshold value. [7] Device (102) for detecting fatigue of a driver of a vehicle (100), which is configured to perform the method according to one of the preceding claims. [8] Computer program configured to perform the method according to any one of claims 1 to 6. [9] Machine-readable storage medium on which the computer program according to claim 8 is stored.

Images