DE102011084367A1 - Method for determining driver-side perception of e.g. pedestrian provided in surrounding of motor car by driver assistance system, involves determining perception based on frequency of view of driver striking same object per time unit - Google Patents

Abstract

The method involves determining (100) or receiving a position of an object and/or determining a direction of the object. A viewing direction of a driver is determined (110). A determination is made (120) whether view of the driver strikes the object by evaluating the viewing direction, the position, and the direction. A driver-side perception of an object is determined (140) based on a duration as long as the view of the driver strikes the object, a number of times that the driver view strikes the same object, and/or a frequency of the view of the driver striking the same object per time unit. An independent claim is also included for a device for determining a driver-side perception of an object provided in a surrounding of a motor car.

Description

The invention relates to the detection of a driver-side perception of an object located in the vicinity of a motor vehicle, for example another motor vehicle, a cyclist, a pedestrian or a stationary obstacle.

Current driver assistance systems for collision warning and / or collision prevention detect by sensors other road users and obstacles in the immediate vicinity of the vehicle. By means of a video sensor in the foot of the interior mirror, for example, lanes, the number of lanes, other vehicles or pedestrians can be detected in front of the vehicle. Additional radar sensors for the front as well as the rear area allow for a more robust design of the surroundings to the front or to be extended to the rear of the vehicle. Lateral obstacles can be detected, for example, by ultrasound or laser sensors known from the parking assistant, as well as by other video sensors in the exterior mirrors. Car2X communication enables appropriately equipped road users to communicate their position via radio.

The display and operating concept in today's vehicles with driver assistance systems for collision warning and / or collision prevention is typically designed as follows: A majority of the driver assistance systems warn the driver by time-escalated warning levels. Thus, in the first stage, the driver is usually warned visually of a possible danger via the vehicle display or the head-up display. In the subsequent second stage, the likelihood of an accident is higher, whereupon the visual output is enhanced by a more intrusive audible warning. These noises (typically narrow-band, high-frequency tones) have hitherto mostly been undirected and can only be located approximately from the direction of the vehicle display for the driver (eg in the case of a collision warning). In a third stage takes place in systems for collision prevention in the case of lack of or too little driver reaction, a dynamic driving intervention (for example, a braking). Both driver assistance systems of the longitudinal guidance and driver assistance systems of the transverse guidance often use the principle of time-escalated warning levels.

Visual warnings of the first warning level are often overlooked if the driver's gaze is not within the range of the warning at the time of the warning; This is especially the case with distraction. Furthermore, in complex traffic situations, an additional stress on the driver, if he fixes the visual warning too strong. Although the acoustic warning of the second warning level is eye-catching, it is often so late in terms of time that the accident may not be prevented due to the reaction time of the driver. In addition, the forgiveness - so the acceptance of false warnings - such a pushy audible warning is very low. With frequent false acoustic warnings, the acceptance of the driver in relation to such systems decreases sharply. In the worst case, the driver fills up so much of the system that the driver turns it off permanently. In addition, the number of assistance systems is increasing more and more. The driver's visual channel could soon be exhausted by the increasing number of visual warnings. In addition, in the previous warning concepts, the driver's momentary attention focus is usually not taken into account, ie. H. the warning strategy is independent of where the driver is looking.

From the publication DE 10 2009 041 187 A1 a method for adapting parameters of a driver assistance system is known, wherein the parameters to be set are a function of the state of the driver detected by the driver assistance system. For this purpose, the current viewing direction of the driver is detected by a gaze detection device. From the current viewing direction, the period of time during which the driver does not look at the road is determined. The parameters are adapted if the determined time duration exceeds a critical time value. For example, the warning can be suppressed or reduced and / or initiated later when the driver looks at the road. The warning can be enhanced when looking away from the road and automatic preventative vehicle responses can be initiated earlier.

From the publication DE 10 2006 047 092 B4 an actual viewing direction is determined and compared with a reference direction of view determined by the deflection direction. The reference viewing direction typically corresponds to an evasion direction. If the actual viewing direction and the reference viewing direction deviate from one another, a signal is generated which signals the reference viewing direction. This signal may be an audible signal generated locally in the vehicle where the driver should look to take the reference direction.

In the publication DE 100 39 795 A1 the line of sight is determined and checked, if the line of sight coincides with the direction of a critical situation. Depending on the comparison, a corresponding warning is influenced.

From the publication "Road Safety Research Report No. 60 - Review Often Looked But Failed To See Accident Causation Factor ", ID Brown, November 2005, Department for Transport : London, the so-called "Look but Failed to See" phenomenon is known. According to this, judging the gaze in the direction of an object does not yet allow a reliable statement as to whether the driver has actually perceived this object, for example, has recognized the object as a source of danger.

It is an object of the invention to provide an improved method for evaluating the viewing direction.

The object is solved by the features of the independent claims. Advantageous embodiments are described in the dependent claims.

One aspect of the invention relates to a method for determining a driver-side perception of an object located in the surroundings of a motor vehicle. According to the method, on the vehicle side, the position or direction of an object, i. H. the direction from the point of view of the own vehicle. Alternatively or additionally, the position of the object can also be received by means of a radio device, for example via a Car2Car communication. In addition, the line of sight of the driver is determined. For this purpose, preferably the image of a camera is evaluated, which receives the driver.

The method checks whether the view meets the object, i. H. whether the driver directs the view of the object. For this purpose, the viewing direction and the position (and / or direction of the object) are evaluated.

• – die Dauer, wie lange der Blick das Objekt trifft,
• – die Anzahl, wie oft der Blick das gleiche Objekt trifft, oder
• – die Frequenz, wie häufig der Blick des Fahrers das gleiche Objekt pro Zeiteinheit trifft.

The invention is based on the finding that the statement that the driver's gaze hits the object does not yet permit a reliable statement as to whether the driver actually perceived the object. According to the invention, the determination as to whether the driver has perceived the object, ie noticed it, in dependence on one or more of the following variables:

• - the duration of how long the gaze meets the object,
• The number of times the view hits the same object, or
• The frequency of how often the driver's gaze hits the same object per unit of time.

The method according to the invention makes it possible to make a more reliable statement as to whether the driver who has been looking at an object actually notices this, ie H. perceived, has.

Duration is preferably the duration of how long the gaze has met the object without interruption, in particular without wandering to another object. Alternatively, it can also be the accumulated total duration, how long the gaze has hit the object, in which case interruptions are possible, in particular a wandering of the gaze to other objects.

• – ob die Dauer einen Schwellwert erreicht oder überschritten hat,
• – ob die Anzahl einen zweiten Schwellwert erreicht oder überschritten hat oder
• – ob die Frequenz einen dritten Schwellwert erreicht oder überschritten hat.

In order to decide whether the driver has perceived a particular object, one or more of the following criteria can be checked:

• - whether the duration has reached or exceeded a threshold,
• - whether the number has reached or exceeded a second threshold, or
• - whether the frequency has reached or exceeded a third threshold.

Preferably, the time course for the meeting of the object is stored by the view. To determine a driver-side perception of the object, this course is evaluated. In this saved history, one or more parameters such as duration, number and frequency are analyzed.

This history covers a specific sliding time window. Preferably, the history is stored within a time window having a length in the range of 1 to 5 seconds. The time window should not be too large because, after a certain period of time has elapsed, for example 5 seconds, a dangerous situation associated with the object (eg a turning process in a cyclist driving parallel to the motor vehicle) is already out of date.

Preferably, the time course relates to a plurality of objects. The time course thus indicates when the gaze hit which object from the majority of the objects.

To check whether the view hits the respective object, it is checked whether a straight line defined by the viewing direction intersects a two-dimensional or spatial area at the position of the object. For example, a sphere with a certain radius r can be assumed as the region, the center of the sphere corresponding to the three coordinates x, y, z of the position of the object. For example, the location is the coordinates of the center of gravity of the object. In a two-dimensional view with the coordinates x and y, a circle can be used instead of a sphere.

A warning of the object depends on the determination of whether the driver has perceived the object or not. For example, if the driver has not perceived the object (for example, a cyclist when turning) and thus overlooked, warned the driver whereas in the case of perceiving the object the warning does not occur at all, later in time or with less intensity, for example at lower volume.

Preferably, the driver is warned acoustically in front of the object. In this case, it is advantageous if the warning takes place by means of an Auditory Icon (also Audicon). An auditory icon is a sound characteristic of the object which is typically modeled on a sound of the object, such as a bicycle idle sound by a cyclist, children's voices or steps by a pedestrian, and a tire, wind, or engine sound by another automobile ,

The audible warning, in particular in the form of an auditory icon, should already take place early as part of the first warning stage, in particular more than 2 seconds before an assumed collision time. Since the audible warning often attracts more attention than a purely visual warning, the effectiveness of an early audible alert is higher than a purely visual alert.

However, the acoustic output should not be as obtrusive in its manifestation as, for example, a high-frequency warning sound, as this would not correspond to the criticality of the situation and would therefore be unacceptable to the driver. The warning sound, in particular the auditory icon, should preferably correspond to subtle, less critical, temporal acoustic information, thereby directing attention to potential hazards without disturbing the driver. Preferably, no time-critical or obtrusive auditory icons are used (eg, bicycle bells, brake squeals, horns, etc.), but rather ambiente auditory icons are used. These ambient auditory icons preferably correspond to natural locomotion noise (eg, engine noise) of the object and preferably do not correspond to the sound of a warning device (eg, horn). Such ambient auditory icons are perceived by the driver as background noise and can also be recorded over a longer period of time (for example over a period of more than 2 seconds).

It is advantageous if the warning sound is a spatially directed warning sound from a surround sound-capable audio system (at least one stereo audio system, but preferably a surround sound capable audio system). From the perspective of the driver, the warning sound comes from the direction of the object and indicates the direction of the object.

In this case, it may further be provided that the volume of the noise or the direction changes in the case of a spatially directed sound during the reproduction of the noise according to the movement of the object, so that the driver can roughly estimate the trajectory of the danger object on his auditory perception.

After a warning (or instead of a warning), a vehicle dynamic intervention (eg braking) of the vehicle can be triggered to prevent the collision with the object. This preferably takes place-in a manner similar to that already discussed above in connection with the warning-depending on the determination of a driver-side perception of the object.

It should be noted that some of the approaches described above for warning or triggering an intervention are in principle also applicable without the inventive teaching for determining a driver-side perception of an object. It would also be conceivable to merely check whether the driver's gaze hit the object or not, and make the triggering of the warning dependent on it.

It is also conceivable that the determination of a driver-side perception of a first object is dependent on whether it has been determined that the gaze has hit a second object or whether it has been determined that it is perceived. Namely, the likelihood of perception of the first object can be reduced by gazing at the second object or perceived perception of the second object because of the danger that the second object will be distracted from the first object. Similarly, the warning of the first object or the triggering of an intervention to prevent a collision with the first object may be made dependent on whether a second object has been detected, since the perception of the second object involves the risk that the second object distracts from the first object.

• – die Dauer, wie lange der Blick des Fahrers das Objekt trifft,
• – die Anzahl, wie oft der Blick des Fahrers das gleiche Objekt trifft, und/oder
• – die Frequenz, wie häufig der Blick des Fahrers das gleiche Objekt pro Zeiteinheit trifft.

A second aspect of the invention relates to a device for detecting a driver-side perception of an object located in the vicinity of a motor vehicle. The device may be part of a driver assistance system. The device comprises means for determining or receiving the position of an object and / or determining the direction of the object. In addition, means are provided for determining the viewing direction of the driver. Furthermore, the device is set up to check whether the view of the driver hits the object by evaluating the viewing direction and the position and / or direction of the object. Furthermore, the device is set up, a driver side Detecting the perception of the object as a function of one or more of the following variables:

• The duration of how long the driver's eyes meet the object
• The number of times the driver's gaze hits the same object, and / or
• The frequency of how often the driver's gaze hits the same object per unit of time.

The above statements on the method according to the invention according to the first aspect of the invention also apply correspondingly to the device according to the invention according to the second aspect of the invention.

The invention will be described below with reference to the accompanying drawings with reference to an embodiment. In these show:

1 an embodiment of an inventive method for detecting a driver-side perception of a hazard and to trigger a warning in response thereto;

2 an example for checking whether the straight line in the direction of the view vector intersects a space or surface area located at the coordinates of the object;

3 an exemplary time course of gaze fixations on two different sources of danger; and

4 an exemplary left turn, in which the inventive method is used.

1 shows an embodiment of an inventive method for detecting a driver-side perception of a source of danger and to trigger a warning in response thereto.

In the embodiment, the vehicle preferably includes a surround sound capable system having typically more than two speakers (for example, a 5.1 system or a 5.0 system without a subwoofer channel) capable of outputting a spatially directed sound from the direction of the object , Further, it is advantageous if the passenger compartment is largely isolated with respect to external noise, so that other vehicles or cyclists are not heard with the windows closed inside the passenger compartment.

Furthermore, on-board sensors (for example an ultrasound, radar, laser or camera system or a combination thereof) are provided for detecting objects in the surroundings of the own vehicle. Alternatively or additionally, it can be provided that the own vehicle receives information about other road users or hidden dangers (vehicles in intersections or pedestrians between parked vehicles) by a radio system in the vehicle. By means of on-board sensors or by means of the radio receiver, the vehicle has information about the position of objects in the environment, ie information about where in the environment relative to a vehicle reference point potential hazards (vehicles, cyclists, pedestrians, etc.) are located. For example, each object is a coordinate in the x, y, and z directions or only coordinates in the x and y directions. This is in step 100 shown.

Furthermore, the vehicle has a device for determining the viewing direction of the driver (see step 110 ). For this purpose, for example, detects an interior camera for each eye, the pupil position of the driver and calculates for each eye an eye-sighting vector with the coordinates (x, y, z) with respect to the origin in the pupil center. From the two eye direction vectors, a look vector is determined by averaging, which has, for example, its origin in the head center.

About linear vector geometry can now in step 120 be determined whether the driver has seen a source of danger, namely, whether the view has hit the object. For this purpose, it can be checked whether the straight line in the direction of the view vector intersects a space or surface area at the coordinates of the object. This is in 2 exemplified.

in Richtung des Blickvektors

ergibt sich anhand folgender Gleichung:

The point 0 corresponds to the origin of the 3-dimensional coordinate system, for example, at the location of the manufacturer's logo on the hood of the vehicle F. The straight line

in the direction of the look vector

results from the following equation:

die Koordinaten des Ursprungs des Blickvektors

beispielsweise den Kopfmittelpunkt. Die Größe I entspricht einem beliebigen Skalarwert. Die beispielsweise mittels eines Radarsensors oder Ultraschallsensors ermittelte Position der Gefahrenquelle wird durch den 3-dimensionalen Koordinatenvektor

beschrieben, hierbei handelt es sich beispielsweise um die Position des Schwerpunkts der Gefahrenquelle, beispielsweise eines Fahrradfahrers. An der Position

der Gefahrquelle wird ein flächiger oder räumlicher Körper gelegt, beispielsweise eine Toleranzkugel K mit dem Radius r und der Position

als Mittelpunkt. Bei einer Auswertung von zweidimensionalen Positionsinformationen kann stattdessen ein Toleranzkreis verwendet werden. Die Toleranzkugel K entspricht beispielsweise einer vereinfachten Abschätzung der Größe der Gefahrenquelle. Die Kugel K genügt folgender Kugelgleichung:

This describes the size

the coordinates of the origin of the view vector

for example, the head center. The size I corresponds to any scalar value. The position of the danger source determined, for example, by means of a radar sensor or ultrasound sensor is determined by the 3-dimensional coordinate vector

described, this is, for example, the position of the center of gravity of the hazard, such as a cyclist. At the position

the danger source is a surface or spatial body placed, for example, a tolerance sphere K with the radius r and the position

as the center. In an evaluation of two-dimensional position information, a tolerance circle can be used instead. The tolerance ball K corresponds, for example, to a simplified estimation of the size of the source of danger. The ball K satisfies the following ball equation:

den Körper K an der Position

der Gefahrenquelle schneidet, d. h. ob ein Schnittpunkt vorliegt. Die Gleichung der Blickgerade

lässt sich in die Gleichung des Körpers K (hier die Kugelgleichung) einsetzen:

To determine whether the view hits the object, it is checked if the line of sight

the body K at the position

the source of danger cuts, ie whether there is an intersection. The equation of the straight line

can be inserted into the equation of the body K (here the sphere equation):

und die Gefahrenquelle schneiden. Falls sich keine Lösung für I ergibt, trifft die Blickgerade

die Gefahrenquelle nicht.The above vector equation can be broken down into three partial equations for the three coordinates x, y and z, which in turn can be solved for 1. To check whether there is an intersection point, it can be checked whether the three scalar equations result in a joint solution for the size I, ie whether there is an intersection point. If a solution for I results, it means that the straight line is

and cut the source of the danger. If there is no solution for I, the line of sight hits

the source of danger is not.

According to the "Looked but Failed to See" phenomenon, however, a glance in the direction of an object does not yet provide any reliable information as to whether the driver also perceived this object. For this purpose, the number and duration of intersections from view and source of danger is evaluated. These intersections can also be termed gaze fixations to the corresponding source of danger. However, the term "gaze fixation" in this application does not necessarily mean that the gaze also remains on the source of danger for a certain duration.

If it is determined that the view is a source of danger (see step 120 ), this hit (ie, this gaze fixation) is stored in a temporal course of the gaze fixations (see step 130 ). The history indicates when there is an intersection with a source of danger. The temporal course of the fixations on sources of danger is stored over a certain sliding (ie running) time window (s. 3 ).

The sliding time window extends from the current time to a running time in the past. The sliding time window has a length in the range of 1 to 5 seconds and should not be chosen too large, because after a certain period of time the corresponding danger situation (eg the turning operation) is already out of date.

The course does not only concern a single source of danger but, if present, various detected sources of danger (for example a detected source of danger 1 and a detected hazard 2 ). During the period Δt ₁ and the time period Δt ₂ , the gaze meets the source of danger 1 ie there is an intersection between the gaze and the source of danger 1 in front. During the period of time Δt ₃ is an intersection with the source of danger 2 in front.

• – die Dauer (Fixationsdauer), wie lange der Blick des Fahrers das Objekt trifft,
• – die Anzahl (Fixationsanzahl), wie oft der Blick des Fahrers das gleiche Objekt trifft, und
• – die Frequenz (Fixationsfrequenz), wie häufig der Blick des Fahrers das gleiche Objekt pro Zeiteinheit trifft, beispielsweise die Anzahl der Fixationen pro Sekunde.

In this saved history, various parameters are evaluated, for example

• The duration (duration of fixation), how long the driver's eye meets the object,
• - the number (number of fixations), how often the driver's gaze hits the same object, and
• The frequency (fixation frequency), how often the driver's gaze hits the same object per unit of time, for example the number of fixations per second.

• 1. wenn die Fixationsdauer (s. die Dauern Δt₁, Δt₂, Δt₃ in 3) einer einzigen Fixation größer als eine erste Schwelle (beispielsweise 0,4 Sekunden) ist,
• 2. wenn die Fixationsanzahl der Fixationen des gleichen Objekts größer als eine zweite Schwelle (beispielsweise 2) ist (in 3 beträgt die Fixationsanzahl für die Gefahrenquelle 1 genau 2 und die Fixationsanzahl für die Gefahrenquelle 2 genau 1), oder
• 3. wenn die Fixationsfrequenz der Fixationen des gleichen Objekts größer als eine dritte Schwelle (beispielsweise 2 mal pro Sekunde) ist.

In order to be able to decide whether a driver has not only seen but also actually perceived a detected source of danger, it is checked on the basis of whether a perception criterion for the respective source of danger is fulfilled, for example, if one of the three aforementioned parameters exceeds a respective threshold (see step 140 ). For example, if only one of the following criteria is met, the corresponding hazard is considered perceived:

• 1. if the duration of fixation (see the durations Δt ₁ , Δt ₂ , Δt ₃ in 3 ) of a single fixation is greater than a first threshold (eg, 0.4 seconds),
• 2. if the fixation number of fixations of the same object is greater than a second threshold (eg 2) (in 3 is the number of fixations for the source of danger 1 exactly 2 and the number of fixations for the source of danger 2 exactly 1), or
• 3. if the fixation frequency of fixations of the same object is greater than a third threshold (eg, 2 times per second).

Additionally, software-based estimation methods may optionally be used to estimate whether or not the driver has perceived the object based on the driver-vehicle interaction (eg, the steering behavior or the operation of a vehicle controller). In addition, it can also be estimated on the basis of stored statistical knowledge (for example from studies) according to the traffic situation (position, movement speed of other road users), whether the corresponding sources of danger were with a certain probability not perceived.

Depending on whether or not the source of danger was perceived, a particular spatially directed acoustic sound is triggered to warn of the source of danger (see step 150 ). For example, an audible warning of a source of danger is triggered if the driver has not perceived the source of danger. For this purpose, for example, but additionally be required that the source of danger falls below a certain distance from the vehicle. The acoustic warning is preferably a spatially directed sound which, from the driver's point of view, comes from the direction of the corresponding source of danger. As a warning, an auditory icon is preferably used, which is modeled on the natural sound of the object. It is also advantageous if the Auditory Icon (for example, the volume or direction) changes in accordance with the movement trajectory of the object in relation to the own vehicle. The auditory icon should be triggered very early (for example, more than 2 seconds before a potential collision time). However, this type of acoustic output should not be so obtrusive as this would not correspond to the criticality of the situation and would therefore be unacceptable to the driver.

As described above, the knowledge of whether a source of danger has been detected can be used to trigger the acoustic staging described above. This has especially in complex traffic situations, such. Intersections in which multiple conflicts can occur in parallel, a huge advantage.

In 4 an example of a left turn situation is shown. As explained above, it is checked by the vehicle F whether the driver of the vehicle F has perceived the danger source G1 (here the cyclist).

mit dem Blickrichtungsvektor

im Wesentlichen auf die Gefahrenquelle G2 gerichtet; nur kurzfristig überstreift der Blick des Fahrers die Gefahrenquelle G1, wobei die Fixationsdauer kleiner als der Schwellwert bleibt. Der Fahrer wird von der Gefahrenquelle G2 abgelenkt, ohne dass der Fahrer die Gefahrenquelle G1 tatsächlich wahrgenommen hat. Das Fahrerassistenzsystem stellt fest, dass der Fahrer die Gefahrenquelle G1 nicht wahrgenommen hat, und löst ein Auditory Icon für die Gefahrenquelle G1 aus, beispielsweise ein Fahrrad-Leerlauf-Geräusch. Übersieht der Fahrer also beim Abbiegen beispielsweise einen Radfahrer G1, so kann über das Audiosystem unaufdringlich akustisch auf den Radfahrer G1 aufmerksam gemacht werden, so dass die Aufmerksamkeit des Fahrers intuitiv in die richtige Richtung gelenkt wird. Eine optische Warnung würde den visuellen Kanal des Fahrers nur zusätzlich belasten, da schon die Überwachung des Gegenverkehrs ein hohes Maß an Ressourcen auf diesem Kanal beansprucht. Da es sich um bereits gelernte Geräusche handelt, bekommt der Fahrer automatisch die Information, um welche Art von Gefahrenquelle es sich handelt, möglicherweise noch bevor er sie sehen kann. Die akustische Darbietung über das Audiosystem des Fahrzeugs sollte dabei ein so hohes Qualitätsmaß haben, dass der Fahrer die Richtung und Entfernung der jeweiligen Gefahrenquelle möglichst gut schätzen kann. Dadurch wird der visuelle Wahrnehmungskanal, der ohnehin schon beispielsweise durch den Abbiegevorgang ausgelastet ist (Überwachung Gegenverkehr, Lücke suchen usw.) entlastet und der relativ freie auditive Kanal besser genutzt.In a first situation is the view (see the straight line

with the line of sight vector

essentially directed to the source of danger G2; Only in the short term does the driver's gaze sweep over the source of danger G1, whereby the fixation duration remains smaller than the threshold value. The driver is distracted from the source of danger G2, without the driver having actually perceived the source of danger G1. The driver assistance system determines that the driver has not perceived the hazard source G1 and triggers an auditory icon for the hazard source G1, such as a bicycle idle noise. If, for example, the driver overlooks a cyclist G1 when turning, the cyclist G1 can be sensitively informed via the audio system in an unobtrusive manner, so that the driver's attention is intuitively directed in the right direction. An optical warning would only add additional weight to the driver's visual channel, since the monitoring of oncoming traffic already requires a high level of resources on this channel. Because they are already learned sounds, the driver automatically gets the information about what type of hazard it is, possibly even before he can see them. The acoustic performance via the audio system of the vehicle should have such a high quality that the driver can estimate the direction and distance of the respective source of danger as well as possible. As a result, the visual perception channel, which is already busy, for example, by the turning process (monitoring oncoming traffic, looking for a gap, etc.) relieved and better used the relatively free auditory channel.

mit dem Blickrichtungsvektor

für eine ausreichende Fixationsdauer, die größer als der entsprechende Schwellwert ist, auf die Gefahrenquelle G1 gerichtet. Das System erkennt, dass der Fahrer die Gefahrenquelle G1 tatsächlich wahrgenommen hat, und löst kein Auditory Icon für die Gefahrenquelle G1 aus (oder löst das Auditory Icon später aus). Für den Fall, dass die Fixationsdauer für die Gefahrenquelle G2 zu kurz ist oder überhaupt keine Fixation der Gefahrenquelle G2 erfolgt ist, wird ein Warnsignal für die Gefahrenquelle G2 ausgelöst.In a second situation, the gaze (see the straight line of sight)

with the line of sight vector

for a sufficient fixation period that is greater than the corresponding threshold, directed to the hazard G1. The system recognizes that the driver has actually detected the hazard G1 and does not trigger an auditory icon for the hazard G1 (or triggers the auditory icon later). In the event that the fixation period for the source of danger G2 is too short or no fixation of the source of danger G2 has taken place at all, a warning signal for the source of danger G2 is triggered.

It would also be conceivable, depending on the driver's perception of the danger, to trigger the Auditory Icon earlier (if the source of the danger is not perceived) or later (if the source of danger is perceived). Alternatively or additionally, this could also affect the volume of the auditory icon.

Parameters such as activation time or volume can also be varied depending on the situation. For example, depending on the distance and / or the approach speed of the hazard, the volume can be adjusted (for example, as a vehicle approaches, the noise becomes louder).

Preferably, the passenger compartment is largely isolated with respect to outside noise so that other vehicles or cyclists are not heard with the windows closed inside the passenger compartment. Thus, the acoustically ambitious information triggered by the driver assistance system can be understood as a filtering of the outside world, ie. H. only relevant acoustic information is transmitted artificially in the driver's compartment.

The method described above allows an intuitive ambiente attention in potential dangers. As a result, accidents can be avoided without patronizing or disturbing the driver. The driver assistance can be experienced more often by triggering an auditory icon as early as possible.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102009041187 A1 [0005]
• DE 102006047092 B4 [0006]
• DE 10039795 A1 [0007]

Cited non-patent literature

• "Road Safety Research Report No. 60 - Review Often Looked But Failed to See Accident Causation Factor ", ID Brown, November 2005, Department for Transport [0008]

Claims (14)

Method for determining a driver-side perception of an object (G1, G2) located in the vicinity of a motor vehicle (F), comprising the steps of: - determining ( 100 ) or receiving the position

an object (G1, G2) and / or determining the direction of the object (G1, G2); - Determine ( 110 ) the viewing direction

the driver; - Check ( 120 ), whether the view of the driver hits the object, by evaluating - the line of sight

and - the position

and / or direction of the object (G1, G2); and - determining ( 140 ) a driver-side perception of the object depending on the duration (Δt ₁ , Δt ₂ , Δt ₃ ), how long the driver's gaze hits the object, The frequency of how often the driver's gaze hits the same object per unit of time. The method of claim 1, wherein the perception of the object is detected when - the duration (Δt ₁ , Δt ₂ , Δt ₃ ) has reached or exceeded a threshold, - the number has reached or exceeded a second threshold, or - the frequency is a third Threshold reached or exceeded. Method according to one of the preceding claims, wherein a time course for the meeting of the object (G1, G2) is stored by the view and the determination of a driver-side perception by evaluating the course is carried out. The method of claim 3, wherein the history is stored over a sliding time window having a length in the range of 1 to 5 seconds. Method according to one of claims 3-4, wherein the time course relates to a plurality of objects. A method according to any one of the preceding claims, wherein the step of checking comprises: - checking if one by the line of sight

defined straight line

a flat or spatial area (K), which is located at the position

of the object is cutting. Method according to one of the preceding claims, comprising the additional step of: - triggering ( 150 ) A warning of the driver in front of the object (G1, G2) depending on the detection or non-detection of a driver's perception of the object, wherein the triggering of the warning at all or the manner of the warning, in particular the activation point of the warning or the intensity of the warning , depends on it. The method of claim 7, wherein the warning in front of the object (G1, G2) is not triggered or triggered later in time, if it was determined that the driver has perceived the object. Method according to one of claims 7-8, wherein the warning of the driver is acoustic, in particular by an object associated Auditory Icon. A method according to claim 9, wherein, for warning, a spatially directed sound is generated with a surround sound-capable audio system, in particular with a surround sound capable audio system, which indicates to the driver the direction of the object (G1, G2). Method according to one of claims 9 or 10, wherein the warning sound changes during the reproduction of the sound according to the movement of the object (G1, G2), in particular the volume of the sound and / or direction in the case of a spatially directed noise. Method according to one of the preceding claims 9-11, wherein the audible warning occurs at an early time which is more than 2 seconds before an assumed collision time. Method according to one of the preceding claims, comprising the additional step: Initiating engagement of the vehicle to prevent the collision with the object (G1, G2) in response to detection or non-detection of a driver-side perception of the object, the triggering of the intervention depending on or the manner of engagement therefrom. Device for detecting a driver-side perception of an object (G1, G2) located in the vicinity of a motor vehicle (F), comprising: - means for determining or receiving the position

of an object (G1, G2) and / or determining the direction of the object (G1, G2) and - means for determining the viewing direction

the driver, the device is set up - to check whether the driver's view hits the object, by evaluating - the line of sight

and - the position

and / or direction of the object, and - determine a driver-side perception of the object depending on - the duration (.DELTA.t ₁ , .DELTA.t ₂ , .DELTA.t ₃ ), how long the driver's gaze hits the object, - the number of times the view of the Driver encounters the same object, and / or the frequency of how often the driver's gaze hits the same object per unit of time.

Images