Method for controlling a headlight assembly for a vehicle and headlight assembly

Abstract

Method for controlling a headlight arrangement for a vehicle (10), the headlight arrangement comprising two spaced headlights (1, 2), in which - road users (11, 12) are detected in the direction of travel in front of the vehicle (10) and - a first overall light distribution (39) can be generated, in which the luminous range on a first side of a central axis (36) is greater than on the other, second side of this central axis (36), the greater luminous range on the first side of the central axis (36) being generated by the light emission of both headlights (1, 2), and - a second overall light distribution (40) can be generated, in which the overall light distribution (40) can be regulated such that it has a central region (M) with a luminous range in the direction of at least the detected road user (12) that is less than the distance to the detected road user (12), and has side regions (S1, S2) on both sides next to this central region (M) with luminous ranges that are greater than the distance to the detected road user (12), characterized in that - when switching from the first overall light distribution (39) to the second overall light distribution (40), the light range of at least the second headlight (2) on the first side of the central axis (36) is first reduced at least to such an extent that it is smaller than the distance to the detected road user (12), then the light emission direction (L) of the second headlight (2) is pivoted about a vertical pivot axis (38) and then the second overall light distribution (40) is generated.

Description

[0001] The present invention relates to a method for controlling a headlight assembly for a vehicle and to a headlight assembly for a vehicle comprising a control device.

[0002] A vehicle's headlights are designed to illuminate the area in the direction of travel, especially the road, in poor visibility conditions, especially at night. They also serve as a distinguishing feature for other road users.

[0003] It is known to provide headlights that can provide a low beam and a high beam function for light emission in the direction of travel. The high beam function ensures very comprehensive illumination of the surroundings. However, it has the disadvantage that other road users, in particular drivers of vehicles ahead and oncoming vehicles, are dazzled. With the low beam, on the other hand, a light distribution can be created that does not dazzle other road users. However, the illumination of the surroundings is much less than with the high beam function. Due to the very high traffic densities nowadays, the high beam function can only be used very rarely. There is therefore a need to provide headlight arrangements that provide better illumination than the conventional low beam function, but which do not dazzle other road users like the high beam function.

[0004] DE 10 2007 045 150 A1 discloses a method for controlling a headlight assembly for a vehicle. In this case, the headlight assembly comprises two spaced-apart headlights, each emitting a light beam to generate an overall light distribution. Furthermore, the headlights each comprise a diaphragm assembly, the position of which can be adjusted to alter the overall light distribution. In the method, a road user is detected in the direction of the light emission of the headlight assembly. Once such a road user has been detected, the position of at least one diaphragm assembly is adjusted such that, in the overall light distribution, a central region with a shorter illumination range is formed in the direction of the detected road user, and side regions with a longer illumination range are formed on either side of this central region.The beam range in the central area, particularly in the direction of the detected road user, is controlled depending on the vertical angle, i.e., the distance of the road user. The light distribution is generated exclusively by changing the position of the apertures of the aperture assembly and, if necessary, by pivoting the light emission direction of the headlights around a vertical axis.

[0005] DE 10 2007 028 658 A1 describes another method for controlling a headlight assembly for a vehicle. In this case, the headlight assembly has a right and a left headlight unit, each comprising separate headlights for a low beam and a high beam. In the method, a road user is detected in the direction of the light emission of the headlight assembly. Once such a road user has been detected, the light distribution generated by the headlights for the high beam is modified with regard to the lateral illumination. The light distribution generated by the separate headlights for the low beam, however, remains unchanged.

[0006] Finally, DE 10 2007 040 042 A1 describes a system for generating a light beam in front of a motor vehicle. The system consists of a headlight with an LED array comprising several separately electrically controllable light-emitting diodes. The system further comprises an object detection device for detecting objects in the vicinity of the motor vehicle and a position determination device for determining the position of a detected object relative to the motor vehicle. In this system, the individual light-emitting diodes are controlled such that an illuminance limit is not exceeded in the area of ​​a detected object.

[0007] It is the object of the present invention to provide a method and a headlight arrangement of the type mentioned at the outset with which an overall light distribution can be generated which illuminates the surroundings of the vehicle as well as possible, but does not dazzle other road users.

[0008] According to the invention, this object is achieved by a method having the features of claim 1 and the headlight assembly specified in the patent claims. Advantageous embodiments and further developments are set forth in the dependent claims.

[0009] According to a first aspect of the present invention, a method is provided for controlling a headlight arrangement for a vehicle, in which the headlight arrangement has two spaced-apart headlights and in which road users are detected in the direction of travel in front of the vehicle. Furthermore, a first overall light distribution can be generated in which the luminous range is greater on a first side of a central axis than on the other, second side of this central axis. Furthermore, a second overall light distribution can be generated in which the overall light distribution can be controlled such that it has a luminous range in the direction of at least one detected road user that is less than the distance to the detected road user, and which has a luminous range in another direction that is greater than the distance to the detected road user.The method is characterized in that when switching from the first overall light distribution to the second overall light distribution, the luminous range of at least one headlight on the first side of the center axis is first reduced at least to such an extent that it is smaller than the distance to the detected road user and then the second overall light distribution is generated.

[0010] The center line is, in particular, the dividing line between the lanes on a straight road. In this case, the first side of this center line represents the right-hand lane, and the second side represents the left-hand oncoming lane. In left-hand traffic, the sides are reversed.

[0011] According to the invention, different overall light distributions are provided in the various aspects of the present invention. Furthermore, switching between these overall light distributions takes place. The overall light distributions are defined below: The first overall light distribution is asymmetrical, particularly with respect to a longitudinal axis, when considering the course of the cut-off line on the road. On the side of the vehicle's own lane, i.e. on the first side of the central axis, a greater luminous range is provided than on the side of the adjacent lane, which on single-lane roads represents the oncoming lane. On the adjacent lane, the luminous range only extends to a cut-off limit so that oncoming road users are not dazzled. The first overall light distribution represents, for example, a known low beam distribution. If the cut-off line is viewed on a screen arranged perpendicular to the vehicle's longitudinal axis, the low beam distribution on the driving side results in a characteristic rise in the cut-off line of 15° to the horizontal. This 15° rise corresponds to the greater luminous range on the driving side of the vehicle.

[0012] The second overall light distribution provided according to the invention, which is also referred to below as masked permanent high beam, is characterized in particular by a control of the headlight range as a function of one or more detected road users. In the direction of one or more such road users, i.e. in an area with an opening angle which is determined by the width and distance of the detected road user(s), the headlight range is less than the distance to the detected road user(s). The distance is defined such that the road user cannot be dazzled by the second overall light distribution. If the detected road user is another vehicle, for example, the distance can extend to the bumper of the other vehicle.If the other road user is a cyclist or pedestrian, the distance can be defined to reach the point where the other road user touches the ground.

[0013] The masked permanent high beam is further characterized in that in another direction, i.e. in particular in an area to the side of the aperture angle which includes the detected road user, a luminous range is provided which is greater than the distance to the detected road user. With the second overall light distribution, in particular a central area with a shorter luminous range is formed in the direction of the detected road user, and side areas with a longer luminous range are formed on either side of this central area. In this way, the masked permanent high beam provides optimal illumination of the area surrounding the vehicle in the direction of travel, but road users in front of the vehicle in the direction of travel are not dazzled.If the cut-off line of the masked permanent high beam is observed on a vertically arranged measuring screen, a horizontal cut-off line is formed, particularly in front of the detected road user, and a vertical cut-off line is formed next to the detected road user, which corresponds to the greater beam range in the area adjacent to the detected road user. The beam range in the central area, which in the second overall light distribution extends particularly to the detected road user, is preferably controlled by an existing beam range adjustment.

[0014] Finally, some aspects of the invention provide a third overall light distribution, which is also referred to below as a variable beam range. With the variable beam range, the maximum beam range is controlled so that it extends to a detected road user. In this case, the beam range is thus less than the distance to the detected road user, whereby the beam range is not static but is controlled depending on the distance of the detected road user. However, with the variable beam range, unlike with masked permanent high beam, no vertical cut-off line is formed in the center of the overall light distribution.

[0015] For the purposes of the invention, the term "light range" refers to an angle-dependent distance on the road at which the light intensity falls below a limit value. The limit value for the light intensity is defined in particular in the same way as for the cut-off point. At distances beyond the light range, the light intensity is so low that other road users are no longer dazzled. The angle is in particular a horizontal angle formed by a longitudinal axis through a headlight or headlight arrangement on the one hand and a connecting line from a point on the cut-off point and the intersection of the longitudinal axis with a transverse axis passing through the headlight or headlight arrangement on the other.

[0016] According to one embodiment of the method according to the invention, the overall light distributions are generated by superimposing a first partial light distribution of the first headlight and a second partial light distribution of the second headlight. When switching from one overall light distribution to the other overall light distribution, in particular from the first overall light distribution to the second overall light distribution, the beam range of at least the second headlight, in particular of both headlights, is reduced on the first side of the central axis. The light emission direction of the second headlight is then pivoted about a vertical pivot axis to generate the second overall light distribution.

[0017] If the second overall light distribution, i.e. the masked permanent high beam, is generated in right-hand traffic by pivoting the light emission direction of the left headlight outwards away from the light emission direction of the right headlight, so that the angle between the light emission directions of the two headlights increases, and if the left headlight has a greater beam range on the right side of the central axis, i.e. on the road side, than on the left side of the central axis, i.e. on the oncoming lane, the problem arises that another road user may be dazzled when the light cone of the left headlight is pivoted. To prevent this, the method according to the invention first reduces the greater beam range on the first side of the central axis. Only then is the light emission direction pivoted outwards.

[0018] The reduction of the headlight range before generating the masked permanent high beam can be achieved, for example, by the headlight assembly generating a so-called city light function. The city light function is characterized by the creation of a symmetrical overall light distribution with a limited beam range, whereby the beam range is less than the maximum beam range of the first overall light distribution, e.g., the low beam. The second headlight can then be pivoted outward, as in a cornering light function. The second overall light distribution is then generated.

[0019] Conversely, if the driver switches from the second overall light distribution, i.e., the masked permanent high beam, to the first overall light distribution, e.g., the dipped beam, a city light distribution is first generated. The second headlight, and possibly also the first headlight, are then pivoted around a vertical pivot axis, and only then is the asymmetrical first overall light distribution generated. This prevents dazzling other road users even when switching back to the first overall light distribution.

[0020] Furthermore, according to the first aspect of the invention, a headlight assembly for a vehicle is provided. The headlight assembly according to the invention has at least two spaced-apart headlights for generating an overall light distribution.Furthermore, the headlight arrangement comprises a device for detecting road users in the direction of travel in front of the vehicle and a control device which is coupled to the road user detection device and with which a first overall light distribution can be generated, in which the luminous range is greater on a first side of a central axis than on the other, second side of this central axis, and a second overall light distribution can be generated, in which the overall light distribution can be controlled such that it has a luminous range in the direction of at least one detected road user which is less than the distance to the detected road user, and which has a luminous range in another direction which is greater than the distance to the detected road user.The headlight arrangement is characterized in that, when switching from the first overall light distribution to the second overall light distribution, the headlights can be controlled by the control device in such a way that the beam range of at least one headlight on the first side of the center axis is initially reduced at least to such an extent that it is smaller than the distance to the detected road user and then the second overall light distribution is generated.

[0021] The headlight arrangement according to the first aspect of the invention is particularly designed such that it can carry out the inventive method according to the first aspect of the invention completely or partially.

[0022] The first aspect of the invention relates in particular to the transition from an asymmetric light distribution, such as a low beam, to the so-called masked permanent high beam. The method and headlight arrangement according to the invention ensure, in particular, that another road user is not dazzled when switching between these two overall light distributions.

[0023] According to a second aspect of the invention, a method is provided for controlling a headlight arrangement for a vehicle, in which road users are detected in the direction of travel in front of the vehicle. In the method, a third overall light distribution can be generated by the headlight arrangement, in which the light distribution can be controlled such that it has a lighting range in the direction of a detected road user that is less than the distance to the detected road user and that is greater than or equal to the lighting range in other directions. Furthermore, a second overall light distribution can be generated, in which the overall light distribution can be controlled such that it has a lighting range in the direction of at least one detected road user that is less than the distance to the detected road user, and in another direction that has a lighting range that is greater than the distance to the detected road user.The method is characterized in that, during switching from one overall light distribution to another overall light distribution, the luminous range in the direction of the detected road user is regulated by pivoting the light emission direction of the headlight arrangement about a horizontal axis so that the luminous range reaches the detected road user.

[0024] According to the second aspect of the invention, a headlight assembly for a vehicle is further provided, comprising at least two spaced-apart headlights for generating an overall light distribution and a device for detecting road users in the direction of travel in front of the vehicle. Furthermore, the headlight assembly comprises a control device coupled to the road user detection device and with which a third overall light distribution can be generated, in which the light distribution can be controlled such that it has a beam range in the direction of a detected road user that is less than the distance to the detected road user and that is greater than or equal to the beam range in other directions.Furthermore, a second overall light distribution can be generated, in which the overall light distribution can be controlled such that it has a luminous range in the direction of at least one detected road user that is less than the distance to the detected road user, and which has a luminous range in another direction that is greater than the distance to the detected road user. The headlight arrangement is characterized by a headlight range adjustment with which, during switching from one overall light distribution to another overall light distribution, the luminous range in the direction of the detected road user can be controlled by pivoting the light emission direction of the headlight arrangement about a horizontal axis such that the luminous range reaches as far as the detected road user.

[0025] The second overall light distribution is, in particular, the aforementioned masked permanent high beam. The third overall light distribution is, in particular, the aforementioned variable beam range. When switching between these two overall light distributions, it is ensured, in particular, that the beam range continuously extends to the detected road user. The method and headlight arrangement according to the invention can thus provide permanently good illumination up to a detected road user in a simple and cost-effective manner.

[0026] The headlight range adjustment of the headlight assembly according to the invention comprises, in particular, a first actuator for pivoting the light emission direction of at least one headlight about a horizontal axis. Furthermore, the headlight assembly may comprise a second actuator for pivoting the light emission direction of at least one headlight about a vertical axis. This actuator may, for example, be an actuator already present for a cornering light function.

[0027] According to a further embodiment of the headlight arrangement according to the invention, it comprises a diaphragm arrangement for each headlight, which has at least two vertically and / or horizontally movable flat diaphragms. In the second overall light distribution, i.e. in the masked permanent high beam, the side regions each form a vertical light-dark boundary to the central region, which can be generated by changing the vertical position and / or the horizontal position of at least one of the two diaphragms. For the vertical and horizontal displacement of the two diaphragms, the headlight arrangement in particular has a third actuator coupled jointly to both diaphragms. In this way, the various overall light distributions generated according to the invention can be generated by just three actuators. This ensures that the headlight arrangement according to the invention can be manufactured and operated cost-effectively.

[0028] According to a third aspect of the invention, a method for controlling a headlight assembly for a vehicle is provided, wherein at least two overall light distributions can be generated with the headlight assembly, and switching between the overall light distributions is possible. According to the method according to the invention, the driving behavior is detected, and when switching from one overall light distribution to another overall light distribution, the time interval for the transition from one overall light distribution to the other overall light distribution is determined as a function of the driving behavior.

[0029] In this method of the third aspect, in particular, a road user is detected in front of the vehicle in the direction of travel, and an overall light distribution generated by the headlight arrangement is controlled such that, in the direction of at least one detected road user, it has a beam range that is less than the distance to the detected road user, and in another direction, it has a beam range that is greater than the distance to the detected road users. This overall light distribution is, in particular, the aforementioned second overall light distribution, i.e., the masked permanent high beam.

[0030] The driving behavior is determined in particular by the driving dynamics and / or the magnitude of the vehicle's acceleration, in particular the magnitude of the vehicle's absolute value of acceleration. The time interval for the transition from one overall light distribution to another overall light distribution is shorter the greater the driving dynamics, i.e., in particular, the greater the acceleration or the absolute value of the acceleration. This design adapts the transition times to the driving dynamics, resulting in a pleasant transition between the overall light distributions for the user.

[0031] The time interval ΔT is calculated depending on the acceleration B in particular as follows: ΔT=−k1⋅abs(B)+k2, where for k1 applies 0.3s3 / m≤k1≤2.0s3 / m and for k2 applies: 2s≤k2≤10s

[0032] Preference applies to k1 0.5s3 / m≤k1≤0.9s3 / m and for k2: 4s≤k2≤6s

[0033] In particular, k1 = 0.7 s 3 / m and k2 = 5 s.

[0034] Furthermore, driving behavior can be determined based on the driver type. To determine the driver type, the driver can first be identified. Then, current and, if necessary, historical data about the driver is retrieved. Finally, the driver is assigned to a specific driver type.

[0035] Furthermore, it is possible for the user to freely select the driving behavior. For example, the driver can specify the desired driving behavior by entering a value before starting a journey.

[0036] According to the third aspect of the invention, a headlight assembly for a vehicle is further provided, comprising at least two spaced-apart headlights for generating an overall light distribution and a control device with which at least two overall light distributions can be generated and with which it is possible to switch between the overall light distributions. The headlight assembly further comprises a detection device for detecting driving behavior. The headlight assembly is characterized by a timer with which, when changing from one overall light distribution to another overall light distribution, the time interval for the transition from one overall light distribution to the other overall light distribution can be determined as a function of driving behavior.

[0037] The two total light distributions generated according to the third aspect of the invention can be the first and the second, the first and the third, and the second and the third aforementioned total light distribution.

[0038] The third aspect of the invention ensures that the time interval for the transition from one overall light distribution to another overall light distribution is carried out in such a way that, depending on the driving behavior, optimal illumination of the surroundings of the vehicle is achieved.

[0039] According to a fourth aspect of the invention, a method is provided for controlling a headlight arrangement of a vehicle, in which preceding and oncoming road users are detected in the direction of travel in front of the vehicle and the light distribution generated by the headlight arrangement is regulated such that it has a lighting range in the direction of a detected preceding road user that is less than the distance to the detected preceding road user, and whose lighting range in the direction of the adjacent lane is switched back and forth between at least a first lighting state with an increased illumination of the adjacent lane and a second lighting state with a reduced illumination of the adjacent lane depending on the detection of another, in particular oncoming, road user.The method according to the invention is characterized in that the switching between the two lighting states for the lighting range in the direction of the adjacent lane is delayed depending on the detection rate of the other, in particular oncoming, road users.

[0040] The light distribution generated by the headlight arrangement is in particular the above-mentioned second overall light distribution, i.e. the masked permanent high beam. In the two lighting states of this overall light distribution, the method according to the invention adjusts the lighting range of a side area next to a first detected road user as a function of the detection of a further road user. If the adjustment goes so far that the lighting range in the side areas corresponds to the central area of ​​the masked permanent high beam, the result is that the first lighting state of the light distribution corresponds to the masked permanent high beam and the second lighting state of the light distribution corresponds to the above-mentioned third overall light distribution, i.e. the variable lighting range.

[0041] The delay time increases the more frequently additional road users, especially oncoming ones, are detected. For example, with each additional road user detected, the delay time is increased by one extension interval and decreased at a defined rate. However, a minimum and maximum value can be specified for the delay time to prevent the delay time from becoming arbitrarily long in very dense traffic. The specific selection of the delay time ensures that the driver receives a pleasant light pattern that does not appear jarring due to excessive changes.

[0042] Furthermore, the extension interval can be determined depending on the horizontal angular position in which the additional, particularly oncoming, road user was detected. In particular, it can be taken into account whether the road user appears on the left or right side or in the middle in front of the vehicle. Finally, the extension interval and / or the decay rate can also be determined depending on the vehicle speed. For example, in a speed range typical of country roads, the delay time can only be increased slowly when oncoming road users are detected, but can be decreased quickly. On the other hand, at high speeds, such as those typical on a motorway, a newly detected additional road user can lead to a significant extension of the delay time, with the reduction in the delay time occurring slowly due to the decay rate.

[0043] For example, the delay time is in a range from one second to 400 seconds, in particular in a range from 2 seconds to 200 seconds.

[0044] According to a further embodiment of the method according to the invention, when the light is distributed in the direction of the detected road user, a central region with a shorter luminous range is formed and, on either side of this central region, a first and a second side region with a longer luminous range is formed if no oncoming road user has been detected. In this case, a masked permanent high beam is provided, in which the second side region illuminates the adjacent carriageway. In this case, the decay rate is greater if the first, i.e. right-hand, side region has a greater luminous range than the central region than if the first side region has a smaller luminous range than or the same as the central region. In this case, the adjacent carriageway is therefore illuminated again more quickly if the other side region of the masked permanent high beam is illuminated.In this case, the illumination of the first side area is also switched on or off, particularly depending on the detection of a road user. Overall, the method according to the invention can thus achieve a significant calming of the light pattern during periods of higher traffic volume.

[0045] According to a further embodiment of the method according to the invention, the curvature of the road is determined, and the system switches to the second lighting state with a lower illumination of the adjacent lane if the curvature exceeds a threshold value. The curvature of the road can be determined from data obtained from vehicle sensors, such as the temporal change in the steering angle, or based on the current position of the vehicle and a digital geographical map, for example, available in the navigation system.

[0046] According to the fourth aspect of the invention, a headlight arrangement for a vehicle is further provided, which has at least two spaced headlights for generating an overall light distribution and a device for detecting preceding and oncoming road users in the direction of travel in front of the vehicle.Furthermore, the headlight arrangement comprises a control device which is coupled to the road user detection device and with which an overall light distribution can be generated which can be controlled such that it has a lighting range in the direction of a detected road user traveling ahead which is less than the distance to the detected road user traveling ahead, and whose lighting range in the direction of the adjacent lane is switched back and forth between at least a first lighting state with increased illumination of the adjacent lane and a second lighting state with less illumination of the adjacent lane depending on the detection of an oncoming road user.The headlight arrangement is characterized in that the control device has a delay unit with which the switching between the two lighting states for the lighting range in the direction of the adjacent lane can be delayed depending on the detection rate of oncoming road users.

[0047] The fourth aspect of the invention, in particular, makes it possible to avoid excessive switching back and forth between different light distributions. This stabilizes the generated light distribution and prevents driver distraction due to changes in the light distribution.

[0048] According to a fifth aspect of the invention, a method is provided for controlling a headlight assembly of a vehicle, in which road users are detected in the direction in front of the vehicle and the light distribution generated by the headlight assembly is regulated such that it has a luminous range in the direction of at least one detected road user that is less than the distance to the detected road user, and which has a luminous range in another direction that is greater than the distance to the detected road user. The method is characterized in that the luminous range in the other direction is regulated as a function of the horizontal angle between the direction of travel of the vehicle and the connecting line from the vehicle to the detected road user or another detected road user.The light distribution produced by the headlight arrangement is in particular the second overall light distribution mentioned above.

[0049] The beam range, especially in the other direction, is greater the smaller the horizontal angle. According to one embodiment of the method according to the invention, the following applies to the beam range L in the other direction: LW=Lmax(−mΦ+n), where L max is the maximum beam range, Φ is the horizontal angle, for m applies: 0.167Degree−1≤m≤0.4Degree−1, in particular 0.2 degrees−1≤m≤0.3 degrees−1 and especially preferred m=0.25 degrees−1 and for n: 1≤n≤1,2, in particular 1.1≤n≤1.15 and especially preferred n=1,125.

[0050] Furthermore, it is preferably specified that up to a certain angle, the luminous range corresponds to the maximum luminous range, and above a certain angle, the luminous range is minimum. For horizontal angles Φ ≤ Φ1, the luminous range in the other direction corresponds to the maximum luminous range, where Φ1 is in a range from 0° to 2°, in particular from 0.2° to 0.8°. Furthermore, for horizontal angles Φ ≥ Φ2, the luminous range in the other direction corresponds to the minimum luminous range, where Φ2 is in a range from 3° to 6°, in particular from 4° to 5°.

[0051] The parameters mentioned above can ensure that the reduction in headlight range is not too abrupt when approaching or overtaking a road user. This measure also contributes to a calming of the light pattern.

[0052] The beam range in the other direction is controlled primarily depending on the horizontal angle between the vehicle's direction of travel and the line connecting the vehicle to the other detected road user. In this case, the light distribution illuminates, among other things, the oncoming lane.

[0053] Furthermore, the beam range in the other direction can be adjusted depending on the horizontal angle between the vehicle's direction of travel and the line connecting the vehicle to the detected road user. In this case, the light distribution in the other direction illuminates the area adjacent to the lane of the detected road user during an overtaking maneuver, i.e., in right-hand traffic, the area to the right of the road user being overtaken. The beam range in the other direction can be adjusted, in particular, after a signal for overtaking has been given in the direction of the oncoming lane, for example, after the direction indicator has been activated in the direction of the oncoming lane.

[0054] According to one embodiment of the method according to the invention, in the light distribution in the direction of at least the detected road user, a central region with a shorter illumination range is formed, and on both sides of this central region, side regions with a longer illumination range are formed, i.e., the second overall light distribution is provided. In this case, the illumination range in a side region is controlled depending on the horizontal angle between the direction of travel of the vehicle and the line connecting the vehicle to the detected road user or the other detected road user.

[0055] When controlling the beam range depending on the horizontal angle, a hysteresis can also be run through.

[0056] According to the fifth aspect of the invention, a headlight assembly for a vehicle is further provided, comprising at least two spaced-apart headlights for generating an overall light distribution and a device for detecting a road user in the direction of travel in front of the vehicle. The headlight assembly further comprises a control device coupled to the road user detection device and with which an overall light distribution can be generated that is controllable such that it has a beam range in the direction of at least one detected road user that is less than the distance to the detected road user, and that has a beam range in another direction that is greater than the distance to the detected road user.The headlight arrangement is characterized in that the headlights can be controlled by the control device in such a way that the beam range in the other direction can be regulated as a function of the horizontal angle between the direction of travel of the vehicle and the connecting line from the vehicle to the detected road user or another detected road user.

[0057] The method and headlight arrangement of the fifth aspect of the invention allow the light distribution to be controlled, in particular, when a vehicle is traveling in front of the driver's vehicle and another vehicle is approaching, or when the driver's vehicle is overtaking another vehicle traveling in front. This ensures that the overall light distribution generated does not change too abruptly.

[0058] According to a sixth aspect of the present invention, a method for controlling a headlight assembly of a vehicle is provided, with which at least two overall light distributions can be generated, wherein it is possible to switch back and forth between the two overall light distributions. In the method, road users in the direction of travel in front of the vehicle are detected. According to the invention, one of the two overall light distributions is selected depending on the detection rate of other road users.

[0059] The method according to the sixth aspect of the invention achieves the generation of a specific overall light distribution when traffic density is too high, i.e., when a large number of road users are detected within a time interval. The two overall light distributions are, in particular, the aforementioned second overall light distribution, i.e., the masked permanent high beam, and the third overall light distribution, i.e., the variable beam range. If the detection rate exceeds a specific limit, the variable beam range is generated as the overall light distribution. This prevents the areas of the masked permanent high beam, which have a very large beam range, from having to be switched on and off too frequently.

[0060] According to one embodiment of the method according to the invention, the selection of the overall light distribution is further dependent on the position of other detected road users and / or the vehicle speed. For example, a first step size is generated for each detected road user. The first step sizes generated for each detected road user are integrated, and a second step size, which depends on the vehicle speed, is subtracted from the integrated step size. This generates a first output signal that characterizes one of the overall light distributions. The first step size can also depend on the vehicle speed.

[0061] According to a further embodiment of the method according to the invention, the overall light distribution is selected alternatively or additionally depending on temporal changes in the vehicle's steering angle. In particular, a first steering angle value is generated depending on steering angle changes. Furthermore, a second steering angle value is generated depending on the vehicle speed and the steering angle. A second output signal is then generated depending on the first and second steering angle values, which characterizes one of the overall light distributions. By taking the temporal changes in the steering angle into account, a winding road can be detected. The use of certain overall light distributions, such as the use of masked permanent high beam, is disadvantageous on winding roads. The method according to the invention can therefore also be used to generate an overall light distribution suitable for winding roads on such roads.For example, in this case a sliding beam range can be created.

[0062] According to a further embodiment of the method according to the invention, the third overall light distribution is generated when the first or second output signal characterizes this overall light distribution. Furthermore, a hysteresis can be applied when switching back and forth between the two overall light distributions to avoid excessive switching.

[0063] According to the sixth aspect of the invention, a headlight assembly for a vehicle is further provided, comprising at least two spaced-apart headlights for generating two overall light distributions, a control device with which the two overall light distributions can be switched back and forth, and a device for detecting road users in the direction of travel in front of the vehicle. The headlight assembly according to the invention is characterized in that the control device is designed such that one of the two overall light distributions can be selected depending on the detection rate of other road users.

[0064] The headlight arrangement according to the invention is particularly designed such that it can carry out the steps of the method of the sixth aspect of the invention completely or partially.

[0065] The following describes possible embodiments and developments of the invention that can be combined with all of the aforementioned aspects of the invention. Furthermore, the embodiments and developments described below can be combined with one another as desired.

[0066] In one embodiment of the method according to the invention, switching from one overall light distribution to another overall light distribution occurs automatically, in particular depending on the detection of a road user. Furthermore, it is also possible for switching from one overall light distribution to another overall light distribution to be triggered by a user's actuation.

[0067] According to a further embodiment of the method according to the invention, the energy consumption of at least one overall light distribution can be adjusted. In particular, the second overall light distribution can be operated in an energy-saving mode as a so-called high-beam assistant, which only switches between a low beam and a conventional high beam. In this energy-saving mode, energy is saved by the reduced frequency of movement, in particular by less frequent actuation of actuators.

[0068] The method according to the invention can also detect whether a multi-lane road is being traveled. A multi-lane road is understood here to mean several adjacent lanes assigned to the same direction of travel. If a multi-lane road is detected, the system switches to a lighting state for the adjacent lane, in which this adjacent lane is illuminated with a shorter beam range. For example, the system can switch to the third overall light distribution for the variable beam range.

[0069] According to one embodiment of the method according to the invention, a hysteresis is passed through when switching back and forth between two overall light distributions. In particular, in the fourth aspect of the method according to the invention, a hysteresis is passed through when switching back and forth between the two lighting states for the light range in the direction of the approaching roadway. This embodiment ensures that excessive switching back and forth between two overall light distributions is avoided. This leads to a stabilization of the light pattern.

[0070] With the method according to the invention and with the headlight arrangement according to the invention, other road users can be detected by a camera, in particular a CCD camera with downstream image processing software and / or a laser, infrared, and / or radar sensor. These sensors detect whether another road user is within the detection range. If so, the position of the other road user relative to the vehicle is also recorded. In this way, not only illuminated road users can be detected, but also road users who do not have their own light sources, such as pedestrians.

[0071] In all the aforementioned aspects of the method according to the invention, the road user, if it is an illuminated vehicle, can be detected in the direction of travel in front of the vehicle in particular by taking an image of a traffic area in the visible spectral range, extracting from the image contiguous areas with a brightness that exceeds a threshold value, classifying the areas at least as a function of their size, forming for each area a confidence value that represents a measure of the similarity of the area of ​​the image to a vehicle light from the classification of the area and a physical quantity that is assigned to the area, and finally determining whether an area is assigned to a vehicle light as a function of the confidence value.

[0072] The method according to the invention takes into account that bright areas of the image originating from distant light sources differ from bright areas originating from nearby light sources. Based on this distinction, the bright areas of the image are classified. However, since classification often does not allow for a clear assignment of an area to a vehicle light, a confidence value is subsequently determined at least for those areas that cannot be clearly assigned to a vehicle light. Based on this confidence value, it can be determined very reliably whether an area can be assigned to a vehicle light.

[0073] According to one embodiment of the method according to the invention, a classification value is obtained from the properties of the area during the classification of the areas. This classification value is assigned to each area, and the confidence value is calculated from the classification value of the area and the physical quantity. The classification value describes how well or clearly the area can be assigned to the respective class during the classification. The classification value thus represents a distinction within a class.

[0074] For the classification of each area, specific properties are determined. These properties may include, for example, the brightness of the area, the shape or border of the area, and / or the color within the area. Furthermore, the properties may include values ​​for the center of gravity, the extent, and / or the principal axes of the area, and additionally or alternatively, the intensity of monochrome pixels in the area. The intensity of monochrome pixels in the area may include the maximum intensity, a mean value, the standard deviation, the position of the maximum within the area, the distribution of the histogram, and / or the value of the mean gradient. Furthermore, these values ​​may alternatively or additionally be determined only for pixels of a color that corresponds to the color of a vehicle's taillight, i.e., typically red.

[0075] Furthermore, properties that indirectly obtain color information can be taken into account by determining various comparisons between monochrome and red pixels, e.g., the ratio between the mean value of the monochrome level and the mean value of the red level. "Monochrome" in this context also refers to the gray value or brightness in this range.

[0076] Finally, the method according to the invention makes it possible to calculate an average of the properties taking into account several consecutive images.

[0077] According to one embodiment of the method according to the invention, the classification values ​​of the regions are obtained using a learning algorithm, and the classification values ​​are subsequently assigned to discrete weighted classification values. The confidence value is then calculated from the weighted classification value of the region and the physical quantity.

[0078] According to one embodiment of the method according to the invention, the physical quantity used to calculate the confidence value is the maximum gray value within the range, in particular the maximum gray value within the range normalized over the maximum possible gray value. When determining the confidence value in this way, it is taken into account that vehicles that are closer to the host vehicle produce brighter areas in the image than vehicles that are further away. Accordingly, areas assigned to vehicles that are close to the host vehicle have a higher confidence value than areas assigned to vehicles that are further away. Furthermore, a bright area resulting from a reflection of the vehicle's own light receives a low confidence value, whereby this confidence value decreases further if the reflection originates from an infrastructure element that is very far away.

[0079] According to a further embodiment of the method according to the invention, the classification is used to determine whether an area can be assigned to a vehicle light or the lights of a vehicle. Subsequently, the confidence value is calculated only for those areas to which a vehicle light cannot be clearly assigned based on the classification.

[0080] According to a further embodiment of the method according to the invention, to improve the reliability of the method, the surroundings of the areas can be examined, determining whether another area is located near an area, so that the two areas can be assigned to two headlights or taillights of a vehicle. Thus, areas are paired. From this, indications of duplicate areas can be obtained, which can be assigned, in particular, to motor vehicle lights. Based on this examination, the classification value of an area can be adjusted.

[0081] According to a further embodiment of the method according to the invention, the temporal development of a region can be tracked using an image sequence. However, since tracking the regions is often difficult, the method according to the invention can also be carried out independently of such tracking of the regions by determining the temporal coherence of the confidence values. In this case, a confidence value is changed depending on its temporal coherence. For this purpose, in particular, an accumulation field is formed in which the confidence value is accumulated for the pixels of the image. For example, in the accumulation field, during the transition from one image to a subsequent image, the confidence value for a pixel can decrease by a fixed value and increase by the confidence value of the corresponding pixel in the subsequent image.Furthermore, the accumulation field can expand during the transition from one image to the next, depending on the expected movement of an object assigned to the area. Finally, the temporal evolution of the accumulation field can be used to determine whether an area is assigned to a vehicle light, with the confidence values ​​of the accumulation field being subject to temporal hysteresis.

[0082] The advantage of the inventive method for determining the temporal coherence of confidence values ​​is that the difficult task of tracking a region within an image sequence is eliminated. It is sufficient to consider the accumulation field for only a very few frames to reliably assign the regions to vehicle lights, allowing the assignment to be performed very quickly.

[0083] According to a further embodiment of the method according to the invention, the traffic area is recorded at a horizontal aperture angle of more than 40°. The advantage of using such an aperture angle is that the image can be used not only to detect vehicles in the dark, but can also be utilized by other vehicle applications, in particular driver assistance systems. However, with such an aperture angle, it is difficult to detect vehicle lights that are far away. According to a development of the method according to the invention, the image is therefore recorded with a sensor that has one region that is only sensitive in the wavelength range that corresponds to the rear light color of a vehicle, i.e. usually red, and another region that detects the brightness of the incident light, in particular in the visible spectrum.The area that detects the brightness of the incident light does not detect any light from the near infrared range.

[0084] The area that is only in the wavelength range corresponding to the taillight color of a vehicle takes up, for example, 25%.

[0085] The image is taken in particular by a monochrome camera.

[0086] The invention will now be explained using exemplary embodiments with reference to the accompanying drawings. Fig. 1 shows schematically an embodiment of a headlight of the headlight arrangement according to the invention, Fig. 2 shows schematically an embodiment of the headlight arrangement according to the invention, Fig. 3 shows the radiation characteristic of the first total light distribution on a measuring screen, which is generated by an embodiment of the method according to the invention, Fig. 4 shows the radiation characteristic of the first total light distribution on the road, which is generated by the embodiment of the method according to the invention, Fig. 5 shows the radiation characteristic of the second total light distribution on a measuring screen, which is generated by the embodiment of the method according to the invention, Fig. 6 shows the radiation characteristic of the second total light distribution on the road, which is generated by the embodiment of the method according to the invention, Fig. 7 shows the third overall light distribution on the road produced by the embodiment of the method according to the invention, Fig. 8 shows schematically the structure of the device for detecting other road users, Fig. 9 shows the method steps for detecting other road users, which are carried out in the embodiment of the method according to the invention, Fig. 10 shows a hysteresis process which is carried out when detecting other road users, Fig. 11 shows further possible method steps that are carried out when detecting another road user in an embodiment of a method according to the invention, Fig. 12 shows a light distribution in a known method when switching from the first total light distribution to the second total light distribution, Fig. 13 shows the light distribution when pivoting from the first total light distribution to the second total light distribution according to the embodiment of the method according to the invention, Fig. 14 shows an example of the second total light distribution, Fig. 15 shows the change in the luminous range in the central region of the second overall light distribution according to the embodiment of the method according to the invention, the Fig. 16 and Fig. 17 show the change in the second overall light distribution in the left side area of ​​an oncoming vehicle according to the embodiment of the method according to the invention, the Fig. 18 and Fig. 19 show the change in the right side area of ​​the second overall light distribution during an overtaking maneuver according to the embodiment of the method according to the invention and Fig. 20 shows a diagram from which the luminous range in a side area in the second total light distribution results as a function of the horizontal angle to a detected road user according to the embodiment of the method according to the invention. Fig. 21 shows a diagram illustrating the fade time when switching from the first total light distribution to the second total light distribution according to the embodiment of the method according to the invention, Fig. 22 shows a flowchart for calculating a delay time, Fig. 23 shows a flow chart for generating a switching process between the second and the third total light distribution according to the embodiment of the method according to the invention,

[0087] The headlight arrangement, which is generally Fig. 2, comprises two spaced-apart projection headlights 1 and 2, which are arranged at the front on the right and left sides of the vehicle in a manner known per se. One of these projection headlights 1, 2 is shown in Fig. 1. The projection headlight 2 arranged on the other side is essentially identical in design.

[0088] In Fig. 1 shows a section through projection headlight 1 in a plane parallel to the plane spanned by the vehicle's longitudinal axis and the vertical V. Projection headlight 1 comprises, in a manner known per se, a light source 3 surrounded by a reflector 6 designed as an ellipsoid of revolution. Reflector 6 thus has two focal points. Light source 3 is located at one of the focal points of reflector 6. The light emitted by light source 3 is reflected by reflector 6 in the light emission direction L of projection headlight 1 toward a projection lens 7. A diaphragm arrangement with flat diaphragms 8 and 9 is arranged at the focal point of projection lens 7 and near the second focal point of reflector 6. The normals to flat diaphragms 8 and 9 are aligned essentially parallel to the light emission direction L.The light source 3, the reflector 6, the lens 7, and the apertures 8, 9 are arranged within a housing 4, which is enclosed by a lens 5. The shape of the cut-off line of the overall light distribution of the projection headlight 1 can be changed by moving the apertures 8 and 9 in the vertical and / or horizontal direction.

[0089] The manner in which different overall light distributions can be generated by the movement of the diaphragms 8 and 9 is described, for example, in DE 10 2007 045 150 A1, the relevant content of which is incorporated into the present application by reference.

[0090] With reference to Fig. 2, an embodiment of the headlight arrangement is described below, which comprises a headlight 1, 2 on the right and left side, as shown in Fig. 1 is shown.

[0091] The right headlight 1 of the headlight arrangement is connected to a control unit 13, the left headlight 2 to a control unit 14. The control units 13 and 14 control the partial light distributions of the headlights 1 and 2, which, when superimposed, result in an overall light distribution.

[0092] The control units 13 and 14 control a headlight range adjustment for the headlights 1 and 2, in which the headlights 1 and 2 can be pivoted about a horizontal axis 37 by means of the actuator 19 and 22, respectively. The light emission direction L of the headlights can thus be adjusted in the direction of the arrow A ( Fig. 1). Furthermore, control units 13 and 14 control actuators 20 and 23, respectively, with which headlights 1 and 2 can be pivoted about a vertical axis 38. By means of actuators 20 and 23, the light emission direction L of headlight 1 or headlight 2 can be pivoted in the direction of arrow B. Actuators 20 and 23 are, for example, part of an existing cornering light. Finally, control units 13 and 14 control the vertical and / or horizontal position of the apertures 8 and 9 of the aperture assemblies for the right and left headlights 1, 2, by means of actuators 21 and 24.

[0093] In the following, with reference to the Fig. 3 - 7 different overall light distributions are described, which can be generated by the headlights 1 and 2 of the headlight arrangement according to the invention: The first total light distribution 39 produced by the headlight arrangement is in Fig. 3 as an isolux diagram on a measuring screen and in Fig. 4 using a cut-off line on a road. The overall light distribution 39 is asymmetrical with respect to a central axis 36, which, in the case of a straight road, separates the roadway 26 of the vehicle 10, which includes the headlight arrangement, from the oncoming lane 27. The luminous range in the area of ​​the oncoming lane 27 is much smaller than the luminous range in the area of ​​the roadway 26 and, in the case of right-hand traffic, also smaller than on the right side next to the roadway 26. This asymmetry is also present in the Fig. 3 is visible on the measuring screen. On the right side of the first overall light distribution 39, there is a rise 42 that forms an angle of 15° with the horizontal 35. In the present embodiment, the course of the cut-off line or the light distribution on the measuring screen corresponds to the specifications of the current European standard ECE-R 112 for a low beam of a headlight.

[0094] The second total light distribution 40 generated by the headlight arrangement according to the invention or the method according to the invention is shown as an Isolux diagram on a measuring screen in Fig. 5 and as a light-dark boundary on the road in Fig. 6. The second overall light distribution 40 is a masked permanent high beam. This overall light distribution 40 is characterized in that it has a luminous range in the direction of a detected road user 12, i.e. in the area M of the overall light distribution 40, which is controlled such that it is at least less than the distance to the detected road user 12 and in particular extends to the other road user 12. If the other road user 12 is a vehicle traveling ahead, the luminous range in the area M of the overall light distribution 40 can, for example, extend to the rear bumper of the vehicle 12 traveling ahead.

[0095] On at least one side adjacent to the central region M in the direction of the detected road user 12, a side region S1 is formed, in which the illuminating range is greater than the illuminating range in the region M of the overall light distribution 40. The light thus shines past the road user 12 in order to provide the driver of the vehicle 10 with better illumination of the traffic area in front of the vehicle 10. The illuminating range in the side region S1 can, for example, correspond to the illuminating range with a conventional high beam function. Preferably, a side region S2 is also formed on the other side of the second overall light distribution 40, which also has a greater illuminating range than the illuminating range in the central region M.The illumination range of the side area S2 can also correspond to the illumination range of a conventional high beam function, so that the overall light distribution 40 can correspond to a conventional high beam, in which an area near the detected road user 12 and in the direction of travel in front of the road user 12 is cut out in the light distribution. In this way, the driver of the vehicle 10 can be provided with optimal illumination of the traffic area without dazzling the other road user 12. If another road user, such as an oncoming vehicle 11, is detected in the illumination area of ​​the headlight arrangement, the illumination range can also be regulated in the direction of this road user 11 so that it only reaches as far as this road user 11.

[0096] In the Fig. 6, the beam range in the side area S2 of the second overall light distribution 40 is continuously adjusted to the distance of the host vehicle 10 to the oncoming vehicle 11. Furthermore, according to another embodiment, it is possible that the width of the central area M is selected such that all detected road users are located in the corridor between the side areas S1 and S2, ie for example the Fig. 6 shown vehicles 11 and 12. However, in this case too, the headlight range for the second side area S2 can be regulated depending in particular on the position of an oncoming vehicle 11, as will be explained later.

[0097] As will be described in detail later, for the control of the headlight range in the central area M and, if applicable, in the side areas S1 and S2, data on other road users in the direction of travel in front of the vehicle 10 are continuously transmitted to the control unit 16 from an image processing device 15. Depending on the position of other detected road users 12 or 11, the control unit 16 transmits control signals to the control units 13 and 14 for the headlights 1 and 2. The control units 13 and 14 then control the actuators 19 to 21 for the right headlight 1 and the actuators 22 to 24 for the left headlight 2 such that the desired second overall light distribution 40 is generated. The vertical cut-off line is generated on the one hand by pivoting the headlights 1 and 2 about the vertical axes 38 and on the other hand by actuating the diaphragms 8 and 9 by means of the actuators 21 and 24.The horizontal light-dark boundary in the central area M, i.e. the headlight range in the central area M, is not generated by a displacement of the diaphragms 8 and 9, but preferably exclusively by the headlight range adjustment, i.e. by pivoting the headlights 1 and 2 about the horizontal axis 37 by means of the actuators 19 and 22.

[0098] The terms horizontal and vertical used in relation to the cut-off line refer to the light distribution on a measuring screen arranged perpendicular to the light emission direction L. In this case, a horizontal cut-off line is parallel to the horizontal axis 35, which, with a screen 10 meters away, is located 10 centimeters below the mounting height of headlights 1 and 2. A vertical cut-off line runs perpendicular to the horizontal axis 35.

[0099] In Fig. Figure 7 shows a third overall light distribution 41 generated by the headlight arrangement according to the invention and the method according to the invention as a cut-off line on a road. This third overall light distribution 41 is also referred to as a variable light range. It is characterized in that the maximum light range of the light distribution 41 is regulated such that it extends to a detected road user 12 in the direction of travel in front of the vehicle 10. With this third overall light distribution 41, optimal illumination is thus guaranteed up to the other detected road user 12 without dazzling this road user 12. However, the side areas S1 and S2 of the masked permanent high beam are not provided. The third overall light distribution 41 is essentially symmetrical to a vertical plane running between the headlights 1 and 2 in the direction of light emission direction L.The maximum luminous range is generated in the third overall light distribution 41 essentially over the entire opening angle, as shown in . Fig. 7. However, the aperture angle is selected such that oncoming vehicles 11 in the opposite lane are not dazzled when they pass vehicle 10 with the third overall light distribution 41. The maximum beam range of the third overall light distribution 41 is controlled by means of the diaphragms 8 and 9, i.e., differently than the beam range control in the central area M of the second overall light distribution 40.

[0100] Furthermore, a device is provided for detecting a road user in the direction of travel of the vehicle, ie usually in the direction of the light emission L of the headlights 1, 2. This detection device can be - as in Fig. 2 - a camera 18 with a connected image processing unit 15, which detects the lights of leading and oncoming vehicles 11, 12. Using the image processing device 15, the direction of these lights can be detected in both a horizontal and a vertical direction. The image processing unit 15 analyzes the scene recorded by the forward-facing camera. In this scene, the position of the lights of leading and oncoming vehicles 11, 12 is detected. Based on the horizontal distance between two headlights or taillights of another vehicle 11, 12, the image processing can also deduce the width of the vehicle. Furthermore, light sources of other road users can be detected. Finally, light sources can be detected that indicate street lighting or a town.Streetlights can usually be distinguished from vehicle lights by their position in the camera image or by their intensity, which is frequency-modulated by the network. The camera's aperture angle preferably corresponds to the aperture angle of the headlight array.

[0101] According to another embodiment of the system, the detection device 15 is designed as a laser or radar sensor, with which the distance of objects in the direction of light emission L can be measured. In particular, unlit or insufficiently lit road users, such as pedestrians and possibly also cyclists, can be detected here. Furthermore, the distance measurement can be used to specifically detect road users who are within the glare limits of a light distribution. Finally, by measuring the distance, speed and direction of movement of the road users, a good classification of vehicles or road users can be achieved, thereby avoiding incorrect control of the headlight arrangement. The vertical angle, i.e. the distance from the host vehicle, can be calculated from the distance of the object and can be used as a control variable for the headlight arrangement.

[0102] With distance measurement using a laser or radar sensor, it is also possible to distinguish a moving vehicle from stationary objects by detecting their speed. Furthermore, a scanning laser rangefinder can measure the width of detected objects, allowing for a more accurate determination of the type of object—i.e., whether it is a road user, a motor vehicle, a cyclist, or a control post.

[0103] The laser or radar sensor can also be combined with a camera to increase detection reliability, particularly with regard to detecting the road user's vertical position. Since cameras, laser, or radar sensors are increasingly being used in vehicles with driver assistance systems, these sensors can also be used to control the headlight arrangement at no additional cost.

[0104] In the following, with reference to the Fig. 8 to 11 describe in detail a further possible embodiment of the device for detecting a road user according to an embodiment of the invention: In this case, the detection device comprises a sensor that captures an image of a traffic area in the visible spectral range. The sensor can, for example, be positioned behind the windshield, facing the road in front of the vehicle in the direction of travel. The sensor is a monocular image recording system capable of capturing light sources located more than 600 meters in front of the vehicle in real time. In particular, it can detect headlights of another vehicle located more than 600 meters away and taillights of another vehicle located more than 400 meters away.

[0105] The sensor can be integrated into the camera 18. It can be used not only for image capture, which, after image processing, is used to control the vehicle's headlights 1, 2. Rather, the sensor's image capture can also be used for other driver assistance systems, such as lane departure warning and traffic sign recognition. This multiple use of the sensor reduces vehicle manufacturing costs.

[0106] Vehicles are typically detected in the dark based on the light emitted by their headlights or taillights. It has been shown that simply thresholding the light intensity in the captured image does not lead to reliable vehicle detection. The light emitted by the vehicle itself, which is reflected back to the vehicle by various objects located in front of the vehicle in the traffic area, makes it difficult to distinguish between such self-reflected light and other vehicles at medium and long distances. For this reason, it is necessary to adapt both the sensor and the subsequent image processing of the image captured by the sensor to this problem.

[0107] In order to effectively distinguish between a headlight and a taillight, the sensor has an area that is only sensitive in the wavelength range corresponding to the color of a vehicle's taillight, i.e. this sensor area is only sensitive to red light. However, since color sensors are more sensitive than monochrome sensors, i.e. sensors that measure brightness or gray value, the sensor also includes an area that detects brightness in the visible spectral range. In this sensor, the sensor area for brightness takes up 75% of the pixels and the area for red light takes up 25% of the pixels. With this type of sensor, images can be captured that can also be used for other applications.

[0108] The horizontal aperture angle of the sensor, for example, is greater than 40°. With such an aperture angle, it is difficult to detect distant taillights. For example, a taillight measuring 10 cm x 10 cm will be imaged on less than one pixel of the sensor at a distance of more than 100 m. However, the light emitted by the taillight forms a larger cone, so that a taillight at a distance of around 400 m still hits an area of ​​approximately 4 to 10 pixels. However, to extract color information from the image acquired by the sensor, the filter arrangement presented here with one red and three pixels without a color filter, or a so-called Bayer matrix, is used. The sensor contains 2 x 2 blocks with a single sector sensitive in the red spectral range and three conventional monochrome detectors, which have a higher sensitivity than color-sensitive detectors but also detect light in the red spectral range.With such a sensor, it is possible to distinguish red light sources from white light sources, while simultaneously providing high sensitivity for the pixels without a color filter. It is noted that the sensor is insensitive in the near-infrared range and operates at 10 bits per pixel using a logarithmic curve, which essentially prevents fully saturated light spots where information is lost.

[0109] The sensor is connected to an image processing device 15. The image processing device 15 analyzes the image captured by the sensor such that bright image areas of the captured image can be assigned to vehicles in the traffic area of ​​the host vehicle. During image processing, it is taken into account that bright areas captured by the sensor differ significantly for vehicles close to the host vehicle from such areas for vehicles far away. Corresponding differences also arise for bright areas resulting from headlights and those resulting from taillights. The intensity of bright areas resulting from headlights is higher and whiter, whereas bright areas resulting from taillights are of lower intensity and redder.Furthermore, the image processing takes into account that the device must react more quickly to vehicles that are closer to the own vehicle than to vehicles that are further away from the own vehicle.

[0110] In the following, the individual units of the image processing device 15 and the steps carried out in a first embodiment of the method for detecting another road user with reference to the Fig. 8 to 10 explains: First, in step 100, as described above, an image of the traffic area in the visible spectrum in front of the own vehicle 10 is recorded using the sensor.

[0111] In step 110, the extraction unit 31 determines which contiguous regions of the image have a brightness that exceeds a specific threshold. These image regions are extracted. It must then be determined whether they are associated with another vehicle or another object that is not a vehicle. Since the extraction is based on simple thresholding, it can be performed very quickly by the image processing device 15, i.e., preferably in real time. The contiguous regions of the image whose brightness exceeds the threshold are also referred to as blobs.

[0112] For each area, the following steps 120 to 140 are then carried out: In step 120, the regions are classified by a classifier 32. Furthermore, they are assigned a classification value. For this purpose, various properties of the region are first determined. Of these properties, the maximum gray value of the region and the red component of the region are particularly important. In the method, the following further potentially useful properties of the region are also determined and used in the classification: (1) the binary values ​​of the region, in particular the area, the center of gravity, the extent, the border and / or the principal axes; (2) the intensity orthe gray value resulting only from monochrome pixels, in particular the maximum, the mean, the standard deviation, the position of the maximum within the range, the distribution of the histogram, and / or the magnitude of the mean gradient; (3) the same properties only for red pixels; and (4) properties from which color information can be obtained through various comparisons between monochrome and red pixels, e.g., the ratio between the mean of the monochrome level and the mean of the red level.

[0113] Since it is usually not possible to reliably determine that a vehicle is the light source based on a single property or the properties described above, the properties of the regions are subjected to a learning algorithm. The learning algorithm ultimately yields a classification with classification values ​​for the regions and, finally, with discrete weighted classification values ​​for the regions. The learning algorithm used is the Real-AdaBoost algorithm, as described in R. Schapire and Y. Singer, "Improved boosting using confidence-rated predictions" in Machine Learning, Vol. 37, No. 3, pages 297-336, 1999.This learning algorithm takes into account obvious differences in the properties of vehicle lights to classify the different areas a priori: Small and non-small areas look different, and areas resulting from headlights differ from those resulting from taillights. Following this approach, the areas were divided into four basic classes: • C h,s : for the detection of oncoming vehicles 11 that are at a medium to long distance from the own vehicle; • C h,ns : for the detection of oncoming vehicles 11 located at distances close to the own vehicle up to a medium distance; • C t,s : for the detection of preceding vehicles 12 at medium to long distances; • C t,ns: for the detection of preceding vehicles 12 that are close to the own vehicle up to a medium distance.

[0114] According to the principle that it is better to create an incorrect assignment to a vehicle than to miss a correct assignment, after dividing the areas into classes, the maximum value of the four output signals of the learning algorithm is taken.

[0115] In step 130, based on this basic classification, an initial decision can be made as to whether or not an area should be assigned to a vehicle. If such a rough assignment is possible, the method continues with step 140. If such a rough assignment is not possible, the method continues with step 150, which will be explained later.

[0116] Even if a rough assignment to a vehicle is possible in step 130, it has been found that this rough assignment does not lead to 100% vehicle detection. Furthermore, it cannot be ruled out that a reflection of the light itself or the light emission of another object is mistaken for a vehicle light.

[0117] In the method according to the invention, a confidence value is determined for each region in step 140 using a confidence unit 33. This confidence value represents a measure of the similarity of the region of the image to a vehicle light. This confidence value can be calculated from a previously determined classification value of the region and a physical quantity associated with the region. Preferably, however, discrete, weighted classification values ​​for the regions are obtained from the classification values ​​using the learning algorithm.

[0118] The classification value is derived from the Real-AdaBoost algorithm described above using examples of areas originating from vehicle light sources and corresponding counterexamples. This results in a classification value c, where positive classification values ​​indicate that an area is similar to a vehicle light, and negative classification values ​​indicate the opposite. The classification values ​​c are then discretized into corresponding weighted classification values, with the weights assigned as follows: ω={ω+if c≥t+appliesω0ift0≤c <t+giltω−wennt−≤c<t0gilt0wenn c<tgilt, where t + , t0 and t - are thresholds set for each class, and ω + , ω0 and ω - corresponding weights are defined for each class. Above t + one is sure that an area can be assigned to a vehicle, and below t -one is sure that this area cannot be assigned to any vehicle. The area of ​​t - are + is considered as an uncertain outcome of the classification, where t - to t0, the range is assumed to be more similar to a light source that does not originate from another vehicle, whereas in the range from t0 to t - It is assumed that the areas are more similar to vehicle headlights. The following table shows an example of the thresholds used:

[0119] It turns out that at these thresholds, no regions are misclassified for the above equation. Furthermore, the correct classification for non-small regions is above 90%.

[0120] From the weighted classification values, a confidence value v is calculated for each region. Assuming that g is the maximum gray value of a given region normalized over the maximum possible gray value for a region, the confidence value is defined as follows: v=ω×g, where ω is the weighted classification value assigned to the region. The confidence value is thus the classification certainty multiplied by the certainty of a physical quantity, which in this case is the maximum gray value of the region, specifically the maximum gray value of a given region normalized over the maximum possible gray value. The maximum gray value thus influences both the weighted classification value and the physical quantity. However, when obtaining the weighted classification values, thresholds learned using an algorithm are used based on various properties of the gray value, whereas the gray value is used directly when calculating the physical quantity.

[0121] The confidence value calculation given above takes into account that vehicles closer to the host vehicle will produce brighter areas in the image captured by the sensor than vehicles farther away. Consequently, areas classified as belonging to a vehicle near the host vehicle will have a higher confidence value. If a vehicle is farther away, the associated area of ​​the sensor image will have a high confidence value, but a lower one than the area associated with a closer vehicle. A bright area resulting from a reflection will not have as high a confidence value if correctly classified, and this confidence value will be even lower for areas resulting from reflections from distant objects.

[0122] If it has been determined in step 130 that the basic classification for a vehicle cannot be established, additional information is used for the classification in step 150. In particular, the surroundings of the area can be examined. During this examination, it can be determined whether there is another area near an area, so that the two areas can be assigned to two headlights or taillights of another vehicle. For example, for the areas for which the basic classification in step 120 showed that t - ≤ c ≤ t +In particular, such a surrounding area inspection must be conducted. If the surrounding area inspection reveals a pair of lights, this indicates a twin area originating from the right and left headlights or the right and left taillights of a vehicle. Of course, this inspection only produces a positive result for vehicles with two headlights or taillights, but not for motorcycles or similar vehicles.

[0123] When examining the environment, for each region, a window of a size proportional to the region's border is placed on the left and right sides of the region. Within each window, on the left and right sides, other regions are searched for whose centroid lies within one of the windows. Once such a possible twin region is found, some of the properties described above are compared. Each comparison involves calculating a ratio, e.g., the ratio of the maximum gray values ​​of the two regions. These ratios form a set of properties that can be combined with other properties such as the distance between the centroids, the absolute value of the maximum gray value, etc. These values ​​form a set of properties that are used for the Real-AdaBoost classification algorithm described above.The new classification result of the learning algorithm is used to modify the weighted classification value ω obtained during the classification described above. The following rule is used: If the pairing classification results in a negative value (no twin regions), the weighted classification value ω is not modified; otherwise, this weighted classification value is incremented, i.e., ω. - becomes ω0 or ω0 becomes ω + These weighted classification values ​​are then used as a basis for the subsequent procedural steps.

[0124] After the confidence values ​​for the regions have been determined in step 140, the temporal coherence of the confidence values ​​is determined in step 160. For this purpose, a temporal coherence analysis is performed. It should be noted that the confidence values ​​for the subsequent temporal coherence analysis can also be obtained in a manner other than that described above. On the other hand, the temporal coherence analysis can also be omitted if the weighted confidence values ​​have been determined as described above.

[0125] One possibility for coherence analysis is based on target tracking, i.e., the temporal change of the regions in the image captured by the sensor is tracked. Such target tracking is optionally performed in the second embodiment of the method according to the invention. However, it is difficult to implement and not error-free. Therefore, in the present first embodiment, a coherence analysis is performed that does not require target tracking.

[0126] In the temporal coherence analysis of this design, an accumulation field A is formed to determine whether the assigned confidence values ​​are temporally coherent. The accumulation field A has the same dimensions as the original image. The following steps are performed when renewing the accumulation field: 1. The values ​​of the accumulation field A range from 0 to a given value M A , starting at A = 0. 2. When a new image k has been obtained: (a) Decreasing A. A decay of the accumulation is performed by the calculation step A = max(0,A - d), where d is a fixed number that determines the decay rate: Starting at M A , you need M A / d steps to obtain 0. Here, d can take different values ​​from two possibilities for different cells of the accumulation field A and through a hysteresis process. (b) Propagating A. The values ​​of each cell are propagated according to the expected motion of the targets to combine confidence values ​​resulting from the same target from frame to frame. (c) Increasing A. Assuming that A i (k) is a set of coordinates forming the area i detected in the current image k, and v i (k) is the corresponding confidence value, the following renewal formula is used: A Ai (k)= min(A (k-1) Ai + v i (k) , M A ), where A Ai stands for the cell of A whose coordinates A i are equivalent to.

[0127] During the propagation step, a type of dilation is performed, similar to the mathematical morphology of gray values, but whose structuring element differs for different accumulation cells. The difference stems from the expected movement for the different targets in image space: targets near the horizon remain relatively static from frame to frame, whereas the position of closer targets varies more significantly. Furthermore, oncoming vehicles move faster towards the bottom of the image, whereas leading vehicles do not exhibit such behavior. Therefore, the accumulation field is used specifically for areas that are more similar to those resulting from headlights than to those resulting from taillights, and a different field is used for the remaining areas. This different field is chosen according to the classification if it had a higher output during the base classification.

[0128] The regions for image k are finally classified as vehicle lights or no vehicle lights according to a hysteresis criterion related to the corresponding positions of the accumulation field. This means that each accumulation field has an associated state field S of the same dimension. The hysteresis is then calculated as in Fig. 10 shown, carried out as follows: • If cell A i,j is zero, the corresponding value of the state field S i,j set to FALSE. The next expiration share for A i,j is d = d f , where d f is a fixed number; • if A i,j ≥ M A / 2, S i,j set to TRUE and d = d t , where d t is a fixed value; • if 0 < A i,j < M A / 2 applies, S i,j not changed and there is no associated expiry of Ai,j .

[0129] Consequently, for a given area i with A i (k) at image k the logical operation OR on S Ai (k) applied to definitively assign the areas to a vehicle or no vehicle.

[0130] During the temporal coherence analysis, the maximum value for the hysteresis criterion is set to M A = 2, so that the hysteresis assigned to a cell of the accumulation field is TRUE above M A / 2 = 1 and does not reach the state FALSE before reaching the value zero again, as in Fig. 10. The following values ​​are chosen to control the decay: (dt, d f) = (45, 15) images, which means that when a vehicle disappears, the system continues to illuminate the newly clear area for approximately two seconds. The same values ​​are chosen for the accumulation field of the headlight-like areas and for the taillight-like areas.

[0131] As explained, to determine temporal coherence, an accumulation field is formed in which the confidence value is accumulated for the pixels of the image. During a transition from one image to a subsequent image, the confidence value for a pixel is reduced by a fixed value and increased by the confidence value of the corresponding pixel of the subsequent image. During the transition from one image to a subsequent image, the area will expand depending on the expected movement of an object assigned to the area. The final decision as to whether an area is assigned to a vehicle light is determined based on the temporal development of the accumulation field, whereby the confidence values ​​of the accumulation field are subject to temporal hysteresis. This final assignment is made by the assignment unit 34 in step 170.

[0132] In the following, Fig. 11 describes a second embodiment of the process steps for detecting other road users: As in the first embodiment, a sensor for capturing an image of a traffic area in the visible spectrum and an image processing device 15 are used. The detection of road users, e.g., vehicles, is carried out based on the vehicle lights, ie, for oncoming vehicles 11 based on the light emission of the front headlights, and for preceding vehicles 12 based on the light emission of the rear lights.

[0133] In step 180, as in step 100 of the first embodiment, an image of the traffic area in front of the host vehicle 10 is captured. In step 190, as in step 110 of the method of the first embodiment, contiguous areas with a brightness that exceeds a threshold are extracted. Thus, in the image generated by the sensor, contiguous bright pixels are searched for that exceed a predetermined intensity and, if necessary, also a predetermined size. Such image areas can belong to light sources of another vehicle, but also to other light sources or reflections of the light emission of the host vehicle. The result of step 190 is a binary image in which contiguous areas with bright pixels are defined by a border. Furthermore, as in the first embodiment, properties of the areas, such as size, brightness, color, etc., are determined.If the analysis of an individual image reveals that an area can be associated with a vehicle light with a high probability, the method for these areas can proceed directly to step 250. Furthermore, the areas can be classified based on the properties of the areas in step 240. Finally, further information about the areas can be obtained, as described below. In step 200, an environmental analysis of the area is performed for each area. This environmental analysis can, for example, include the pairing based on twin areas described in the first embodiment, in which two bright areas are assigned to a pair of headlights or a pair of taillights of a vehicle. The result of the environmental analysis in step 200 is further processed in step 240.

[0134] A further optional processing step involves tracking one or more regions in an image sequence and determining the movement of the region(s). To facilitate relocating a region within the image sequence, a global motion estimation is performed in step 230. The vehicle's 10 own motion is calculated and taken into account accordingly when tracking the regions. The global motion can be determined in step 230 based on the correlation of the movement of several objects in the image. It can also be determined using values ​​provided by the odometry in the vehicle or other sensors in the vehicle, such as acceleration sensors. A combination of both approaches is also possible. For this purpose, the image processing device 15 can be coupled to the vehicle bus 17, for example, via the control unit 16.In step 220, the movement of the individual regions in the image sequence is determined taking into account the global movement of the vehicle determined in step 230.

[0135] If the regions are tracked across multiple images using the image processing device 15, it is also possible in step 210 to stabilize the internal properties of the regions, which may fluctuate slightly from image to image, across multiple images. For this purpose, the properties of the regions are averaged over time across multiple images in step 210. This method step is particularly advantageous for determining the color of a very small region: In a small image region, the light only hits very few pixels on the sensor. In a color image sensor, the individual pixels are each sensitive to only one color, usually red, green, or blue. Color determination is only possible when the region hits enough pixels of all color components. If the size of the region is not sufficient for this, the pixels of the region that were determined for several consecutive images can be used in step 210.

[0136] The data obtained in steps 190, 200, 210, and / or 220 are further processed in step 240. As described in the first embodiment, the regions are classified depending on their size and the other properties obtained in the preceding steps, and a confidence value, in particular the weighted confidence value, is determined for the regions as described above. The higher the confidence value, the higher the probability that the region can be assigned to a vehicle light, i.e., the headlight of an oncoming vehicle 11 or the taillight of a preceding vehicle 12.

[0137] If the observation of an area in one image is not sufficient to make a clear classification, the confidence value for each area is accumulated across multiple images in step 260. This takes into account the movement of each area determined in step 220. If an area can be tracked across multiple images, a new confidence value is determined by accumulating or reducing the confidence values ​​for the area in several consecutive images.

[0138] In parallel or alternatively, the confidence value for each pixel can be accumulated or reduced. Tracking of the regions is not required for this. In step 250, the image regions, i.e., the pixels within the border of a specific region determined by the preceding steps, are assigned the confidence value of that region. For a subsequent image, a specific value, as explained in the first embodiment, is automatically subtracted from the confidence value of the region, so that the confidence value in regions in which there are no longer any bright regions in subsequent images decreases to zero over time. At the same time, for each subsequent image, the accumulated confidence of the region below is added to the confidence value of a region.This results in a new confidence value that not only contains the current confidence of the area, but also, via a two-dimensional accumulation, the confidences of recently preceding areas in this image area. Furthermore, in step 250, this two-dimensional confidence field of an area can be enlarged as it decays from image to image. The direction of propagation during the enlargement of this confidence field can be adapted to the expected object movement or the expected movement of the host vehicle 10. This ensures that, despite the movement of an area, the area still lies at confidence values ​​that were generated by the same area in the previous image.

[0139] If the classification generated in step 240 ultimately results in a confidence value that allows a clear decision for an area as to whether it can be assigned to a vehicle light or not, the result of step 240 can also be processed directly. The confidence values ​​determined in this way in steps 240, 250, and 260 are combined in step 270, and a final decision is made as to whether an area is a relevant object, i.e., whether the brightness in this area originates from a light source of another vehicle or not.

[0140] Finally, in step 280, areas that can be assigned to other vehicles 11, 12 are output with the corresponding coordinates relative to the host vehicle 10. These coordinates yield, in particular, the angular positions of the detected preceding or oncoming vehicles 11, 12 relative to the host vehicle 10.

[0141] It should be noted that, depending on the result of step 190, steps 200, 210, and 220 may be omitted entirely, or only one or more of these steps may be performed. Furthermore, steps 250 and 260 may be executed in parallel or alternatively. Finally, they may also be omitted entirely, depending on the result of step 240.

[0142] Finally, the hysteresis explained with reference to the first embodiment can be used. If a confidence value is exceeded, the area is classified as a relevant object. If a lower value is undershot, the area is no longer classified as a relevant object. The accumulation of confidence values ​​in processing steps 250 and 260, respectively, or the accumulation in processing step 210, is truncated at a set maximum value. This ensures that the lower hysteresis threshold is undershot again within a sufficient time.

[0143] Using the camera 18, which includes the sensor described above, and the image processing device 15, it is thus detected whether another road user, in particular one with an illuminated light, is located in front of the driver's own vehicle 10 in the direction of travel. Furthermore, the position of such a road user relative to the driver's own vehicle is determined. The data relating to a detected other road user is transmitted from the image processing device 15 to the control unit 16.

[0144] The control unit 16 is further connected to a vehicle bus 17, via which further data acquired in the vehicle can be transmitted to the control unit 16. For example, geographical data from a navigation system 28 can be transmitted to the control unit 16 via the vehicle bus 17. The navigation system 28 can determine the current position of the vehicle 10 using a receiving sensor, which can be embodied, for example, as a GPS (Global Positioning System) receiver 29.

[0145] The headlight arrangement can further comprise a device 25 for detecting the driving behavior of the vehicle 10. The device 25 can be coupled, for example, via the vehicle bus 17 to an internal clock and a speedometer of the vehicle and, based on the transmitted speed using the time signal of the internal clock, can determine the acceleration of the vehicle 10. Based on the temporal development of the magnitude of the acceleration or the magnitude of the absolute value of the acceleration, the device 25 can determine driving dynamics and assign them to a specific class for the driving behavior. Furthermore, an input signal, for example from the driver, can be transmitted to the device 25, which specifies the driving behavior. Finally, the device 25 can also determine the driving behavior depending on a driver type. For this purpose, the device 25 can be coupled via the vehicle bus 17 to a device for determining the identity of the driver.Historical data on this driver can be stored in the device 25, allowing the device 25 to assign a specific driver type to the current driver, which determines their driving behavior. The driving behavior determined by the device 25 is transmitted to a timer 27. Depending on the determined driving behavior, the timer 27 determines a time interval ΔT for the transition from one overall light distribution to another. The determined time interval ΔT is transmitted from the timer 27 to the control unit 16, which then processes it further.

[0146] The control unit 16 is further connected to a delay unit 26. The delay unit 26 determines a delay time by which switching between two overall light distributions or between two lighting states for the headlight range is delayed. The determination of the delay time by means of the delay unit 26 will be described in detail later.

[0147] The following describes how the headlight arrangement described above is controlled or regulated in an embodiment of the method according to the invention: With reference to the Fig. 12 and Fig. Figure 13 explains how to switch from the first overall light distribution 39 to the second overall light distribution 40. For example, the vehicle 10 is traveling on a single-lane roadway 26. The headlight arrangement of the vehicle 10 generates the first overall light distribution 39. Based on other detected road users 11 and 12, a control signal is transmitted to the control unit 16 by means of the camera 18 and the image processing device 15, indicating that the first overall light distribution 39 should be switched to the masked permanent high beam of the second overall light distribution 40.

[0148] During the transition from the first overall light distribution 39 to the second overall light distribution 40, the actuator 23 is controlled by the control unit 14 of the left headlight 2 such that the light emission direction L of the second headlight 2 is pivoted horizontally outwards in the direction of the oncoming lane away from the light emission direction L of the first headlight 1. Since, in the first overall light distribution 39, the greater light range in the right area of ​​the overall light distribution 39 is generated, among other things, by the light emission of the second, left headlight 2, the problem arises that if the left headlight 2 pivots about the vertical axis 38, other road users 11 and 12 would be dazzled. This situation is shown schematically in Fig. 12 shown.

[0149] In order to avoid this dazzling of other road users 11 and 12 during switching from the first overall light distribution 39 to the second overall light distribution 40, in the method according to the invention, the beam range of the left headlight 2 (in right-hand traffic) is initially reduced on the right side of the center axis 36 between the lanes 26 and 27 at least to such an extent that it is smaller than the distance to a detected road user 11 or 12. Only then is the second overall light distribution 40 generated by increasing the beam range in the left side area next to detected road users 11 and 12 in order to create a corridor with a shorter beam range for one of the road users 11 or 12 or for both road users 11 and 12. Fig. Figure 13 illustrates the switching from the first overall light distribution 39 to the second overall light distribution 40 according to the method of the invention. It can be seen that other road users are not dazzled even during the switching process.

[0150] Conversely, if the system switches back from the second overall light distribution 40 to the first overall light distribution 39, any left side area S2 of the second overall light distribution 40 is initially masked out, so that the light range in this area corresponds to the light range in the central area M, then the left headlight 2 is pivoted about the vertical axis 38 back into the position for the first overall light distribution 39 and only then is the asymmetrical light distribution characteristic of the first overall light distribution 39 also generated by the left headlight 2.

[0151] If a so-called city light function can be generated with the headlight arrangement, in which the asymmetry of the first light distribution 39 is canceled out and such a small maximum light range is provided that other road users are not dazzled, when switching from the first overall light distribution 39 to the second overall light distribution 40, the city light function can also be switched on first, then the left headlight 2, and possibly also the right headlight 1, can be pivoted outwards about the vertical axis 38 in order to then finally generate the masked permanent high beam of the second overall light distribution 40.

[0152] With reference to the Fig. 14 and Fig. 15 explains the control of the light range in the central area M in the second overall light distribution 40. In this case, the second overall light distribution 40 is different from the Fig. In the example shown in Figure 10, a corridor with reduced beam range 43 is generated in a central region M, in which both a preceding vehicle 12 and oncoming vehicles 11a, 11b are located. A second overall light distribution 40 adapted in this way can be generated by a control signal from the image processing device 15 when there is increased traffic density, at which continuous control of the beam range in the left side region S2 would be disadvantageous.

[0153] The side areas S1 and S2 of the Fig. 14 and Fig. 15 are generated, for example, by pivoting the two headlights 1 and 2 further apart about the vertical axis 38 with respect to their light emission direction L. The inner vertical light-dark boundaries in the corridor for the vehicles 11 and 12 are generated by displacing the diaphragms 8 and 9 by means of the actuators 21 and 24. The horizontal light-dark boundary in the central area M, i.e. the headlight range 43 in this central area M, is, however, generated exclusively by the headlight range control by pivoting the headlights 1 and 2 about the horizontal axis 37 by means of the actuators 19 and 22. The control is carried out in such a way that the light-dark boundary 43 is adapted to the distance to the detected road users 11 and 12. For example, based on the traffic situation, which in Fig. 14, the oncoming vehicle 11a has passed, the headlight range is set to the value shown in Fig. 15 is brought closer to the value 44 shown in FIG. 15 to the vehicle 12 in front. This type of control of the headlight range 43 in the central area M of the second overall light distribution is particularly advantageous when driving uphill and downhill.

[0154] In the Fig. 16 to 19 shows a further development of the control of the second total light distribution 40. As in the case of the Fig. 12 and Fig. In the example explained in Figure 13, the beam range in the central area M is regulated so that it extends to the next detected road user 11, 12. The aperture angle for the central area M is thus selected so that it covers the direction of all detected road users 11 and 12 in the direction of travel in front of the vehicle 10. On the right side, next to the central area M, a side area S1 with a greater beam range is formed. Similarly, on the left side, next to the central area M, a second side area S2 with an increased beam range is formed.

[0155] At the Fig. In the situation shown in Figure 16, in which a preceding vehicle 12 was detected and an oncoming vehicle 11a was detected, the luminous range in the left side area S2 is maximum, ie L max. If the oncoming vehicle 11a continues to approach the vehicle 10 with the headlight arrangement, the problem arises that the driver of the vehicle 11a is dazzled by the left side area S2 if the central area M is formed with as small an opening angle as possible. As the oncoming vehicle 11a approaches the host vehicle 10, the opening angle for the central area M is increased. However, if an oncoming vehicle 11a reaches the left vertical cut-off point to the side area S2, the left side area S2 must be switched off, i.e. the beam range of the left side area S2 must be reduced to the beam range of the central area M in order to avoid dazzling the passing vehicle 11a.The closer the vehicle 11a is to the own vehicle 10, the higher the speed at which the headlight range of the left side area S2 must be lowered in order to avoid glare, since the relative angular speed of the passing vehicle 11a becomes very high and thus very little time is available for lowering the headlight range in the left side area S2.

[0156] To avoid the risk of dazzling a passing vehicle 11a, in the method according to the invention, the headlight range LW2 in the direction of the left side region S2 of the second overall light distribution is regulated as a function of the horizontal angle ϕ between the direction of travel FR of the vehicle 10 and the connecting line from the vehicle 10 to the oncoming vehicle 11a. The headlight range LW2 in the left side region S2 of the second overall light distribution 40 is reduced in time so that there is no abrupt change in the second overall light distribution 40 when an oncoming vehicle 11a passes the host vehicle 10. The headlight range in the right side region S1 can remain unchanged. In this case, the inner vertical cut-off line of the second side region S1 is regulated as a function of the position of the preceding vehicle 12.

[0157] With reference to the Fig. 18 and Fig. Figure 19 shows the control of the second overall light distribution during an overtaking maneuver. In front of the host vehicle 10 there is a vehicle 12 traveling ahead, in the direction of which the headlight range is shorter than the distance to the vehicle 12 traveling ahead. During an overtaking maneuver, the vehicle 10 drives into the oncoming lane and approaches the vehicle 12 traveling ahead. In this case, the headlight range LW1 in the right-hand side region S1 of the overall light distribution 40 is controlled as a function of the horizontal angle ϕ between the direction of travel of the vehicle 10 and the connecting line from the vehicle 10 to the vehicle 12 traveling ahead. The closer the vehicle 10 comes to the vehicle 12 traveling ahead, the further the headlight range LW1 in the right-hand side region S1 is lowered. The control for lowering the right-hand side region S1 can be carried out automatically based on the data transmitted by the image processing device 15.However, the control can also be induced by setting the direction indicator in the direction of the oncoming lane, if necessary in conjunction with data on the acceleration of the vehicle 10.

[0158] During an overtaking maneuver, vehicle 10 enters the oncoming lane and approaches the vehicle 12 in front. In this case, the headlight range LW1 in the right-hand side region S1 of the second overall light distribution 40 is controlled as a function of the horizontal angle ϕ between the direction of travel of vehicle 10 and the connecting line from vehicle 10 to the vehicle 12 in front. The closer vehicle 10 comes to the vehicle 12 in front, the further the headlight range LW1 in the right-hand side region S1 is lowered. The control for lowering the right-hand side region S1 can be carried out automatically based on the data transmitted by image processing device 15. However, the control can also be induced by setting the direction indicator in the direction of the oncoming lane, possibly in conjunction with data on the acceleration of vehicle 10.

[0159] In Fig. Figure 20 shows an example of a functional relationship between the headlight range LW, i.e., LW1 or LW2, and the horizontal angle ϕ to another road user 11 or 12. In this example, the horizontal angle ϕ is positive both in the direction of an oncoming vehicle 11 and in the direction of a preceding vehicle 12 that is being overtaken. Up to an angle of 0.5 degrees, the headlight range LW is maximum in the side area S1 or S2, i.e., it corresponds to the headlight range L max At larger horizontal angles ϕ, the headlight range LW is reduced linearly until it reaches a minimum value at 4.5 degrees, at which other road users are no longer dazzled. The headlight range is determined as follows: LW=Lmax for Φ<0.5 degrees LW=Lmax(−0.25⋅Φ+1.125) for 0.5 degrees≤Φ≤4.5 degrees, LW=0 for Φ>4.5 degrees

[0160] The increase in Fig. The curve shown in Figure 20 can also be varied. For example, it can be adjusted to 0.4 degrees -1 or up to 0.167 degrees -1 The critical angles for the loss of the beam range and the attainment of the minimum beam range can also be varied in this case.

[0161] As explained above, the control unit 16 can also specify the time interval for the transition from one overall light distribution to another. This time interval is transmitted to the control unit 16 by the timer 27. The time interval determines, in particular, a switch from the first overall light distribution 39 or from the third overall light distribution 41 to the second overall light distribution 40.

[0162] In Fig. Figure 21 shows a diagram that plots the change in headlight range ΔLW per unit of time as a function of vehicle acceleration, expressed in % / s. In this case, 0% means that the headlight range LW corresponds to the headlight range of a low beam function, and 100% corresponds to the headlight range of a high beam function. A headlight range change of 100% per second thus results in a time interval ΔT of one second for the change from a headlight range for a low beam function to a headlight range for a masked permanent high beam. Accordingly, a headlight range change of 20% per second means a time interval ΔT of 5 seconds until the masked permanent high beam is fully activated.

[0163] The change in the luminous range ΔLW and thus the time interval ΔT for the change from one total light distribution to another total light distribution can be calculated from the acceleration B as follows: ΔLW=B⋅k3 where for k3 applies 10%sm≤k3≤200%sm in particular 40%sm≤k3≤80%sm and preferably: k3=50%sm and where: 20%m≤ΔLW≤150%s

[0164] In the following, with reference to Fig. 22 describes how the delay unit 26 determines a time interval Tv by which the switching between two overall light distributions or two lighting states of an overall light distribution is determined depending on the detection rate of other road users 11, 12. The data on the occurrence and positions of other road users 11, 12 are transmitted to the delay unit 26 from the image processing device 15 via the control unit 16 and the vehicle bus 17. The control unit 16 processes the delay time transmitted by the delay unit 26 such that a switching process is delayed by this time interval Tv.

[0165] The following describes an example in which the delay unit 26 determines a time interval Tv for the delay time for switching from the first overall light distribution 39, i.e., for example, a low beam, to the second overall light distribution 40, i.e., for example, a masked permanent high beam. For this purpose, the delay unit 26 comprises a signal transmitter 45 which detects a change from the second overall light distribution 40 to the first overall light distribution 39 and sends a signal to a switch 46 with each change. With each change from the second overall light distribution 40 to the first overall light distribution 39, the signal from the signal transmitter 45 operates the switch 46 such that a positive time value, which is generated by a time incrementer 47, is switched to its first input.For example, during such a change, a time value of 7 seconds can be applied to the first input of switch 46, which passes this value to its output.

[0166] Following a signal from signal generator 45 and the one-time passing of the time value from time incrementer 47, switch 46 switches to a second input, which is connected to the output of another switch 48. Switch 48 is connected to another signal generator 49. Signal generator 49 detects whether the first total light distribution 39 is active. If this is the case, it switches the first input of switch 48 to the output of a first timer 50 for a decay rate. This first timer 50 continuously outputs a specific negative time value per time unit to the first input of switch 48. For example, a decay rate of -0.3 s is detected by the first timer 50. -1to the first input of switch 48, which feeds this value via its output to the second input of switch 46.

[0167] If the signal generator 49 detects that the first total light distribution 39 is not active, it switches the switch 48 to the second input, which is connected to a second timer 51 for the decay rate. The second timer 51 causes a faster reduction in the delay time. For example, the second timer continuously transmits a decay rate of -0.8 s to the second input of the switch 48. -1, which is forwarded to the second input of the switch 46 when the first overall light distribution 39 is not active.

[0168] The time values ​​generated at the output of switch 46 are transmitted to an integrator 52, which sums the values. However, integrator 52 contains a limit for the delay time Tv. The delay time Tv can only be within a range between 2 seconds and 200 seconds. The delay time Tv is output by integrator 52. This delay time Tv is transmitted by delay unit 26 to control unit 16.

[0169] It should be noted that the signal generator 45 detects a change from the second overall light distribution 40 to the first overall light distribution 39, in particular when another road user, for example an oncoming vehicle 11 or a vehicle 12 traveling ahead on the adjacent lane, has been detected. Thus, the delay time Tv depends on the detection rate of other road users. Alternatively, however, it is also possible for the signal generator 45 to output a signal to the switch 46, regardless of an actual change in the overall light distribution, when another road user has been newly detected. In this case, the other newly detected road user is, in particular, a vehicle 12 traveling ahead on the adjacent lane that is overtaking the driver's own vehicle 10, or an oncoming vehicle 11 that is leaving the camera's detection range and driving past the driver's own vehicle.

[0170] The delay time Tv can be used not only to delay switching between two overall light distributions, but also to switch between two lighting states of an overall light distribution. As explained above, for example, the second overall light distribution 40, i.e., the masked permanent high beam, can have two lighting states. In a first standard lighting state, a corridor is formed for a detected road user or several detected road users 11, 12, in front of which a central area M with a shorter lighting range is formed, and to the side of which side areas S1 and S2 with a longer lighting range are formed (see Fig. 14 and Fig. 15). In a different lighting condition, for example, the left side area S2 (in right-hand traffic) can be omitted. In this area, the same light range is then provided as in the central area M. In this case, it is not necessary, as in relation to the Fig. 16 and Fig. 17 explains how to regulate the headlight range in the left area depending on the horizontal angle ϕ to a detected road user. If, in this case, the signal generator 45 detects, for example, an oncoming vehicle 11 too frequently, the headlight range in the side area S2 is no longer regulated depending on the horizontal angle Φ to an oncoming vehicle 11, but the increase in the headlight range in the area S2 is delayed by the time interval Tv after an oncoming vehicle 11 has passed the own vehicle 10. If oncoming vehicles 12 are detected again within this time interval Tv, and therefore - as with reference to Fig. 22 explains - the signal generator 45 feeds the time value of the time incrementer 47 to the integrator 52, the light range in the side area S2 will remain minimal for a longer period of time and will not be increased in the meantime.

[0171] Furthermore, the control unit 16 can determine under which conditions switching occurs between two overall light distributions or two lighting states of an overall light distribution. When deciding whether switching occurs at all, the control unit 16 can take into account the statistics of the detected road users and / or the temporal development of the steering angle. This will be explained below using the switching between the second overall light distribution 40 and the third overall light distribution 41 with reference to Fig. 23 explains: A signal is transmitted from the image processing device 15 to the control unit 16 when a new road user, for example a new vehicle, has been detected. This signal is fed to the switch 57 as a control signal. The switch 57 then switches the Fig. 23, which is connected to the output of the logic unit 56. The logic unit 56 has two inputs via which step sizes are supplied. A position-dependent step size is supplied to the logic unit 56 via the first input. For this purpose, the position of the object transmitted by the image processing device 15 is supplied to a position-dependent first step size incrementer 53, which assigns a step size to the position of the new object using a characteristic curve and outputs it to the logic unit 56. A speed-dependent step size is supplied to the second input of the logic unit 56. For this purpose, the current vehicle speed is supplied to a speed-dependent second step size incrementer 55 by means of the speedometer 54.Using a characteristic curve, the second step size incrementer 55 determines a step size as a function of the vehicle speed and transmits it to the logic unit 56. The logic unit 56 determines the maximum step size transmitted via the two inputs and transmits this maximum step size to the first input of the switch 57, which forwards it to an integrator 59.

[0172] Then switch 57 switches to the second - in Fig.23 below, which is connected to a speed-dependent step size decrementer 58. The current vehicle speed is also fed to the step size decrementer 58 via the speedometer 54. Based on a characteristic curve, the step size decrementer 58 determines a step size value by which the step size of the integrator 59 is to be decremented per unit of time. The step size decrementer 58 thus specifies a speed-dependent reduction rate for the step size.

[0173] The step size determined by the integrator 59 can, for example, be limited to a range of 30 to 300. The degradation rate specified by the step size decrementer 58 can, in this case, be in a range of -40 to -30 per second.

[0174] Furthermore, a speed-dependent shift of the step size value occurs in order to directly influence the on and off thresholds for the third overall light distribution 41 via the vehicle speed. For this purpose, the current speed from the speedometer 54 is further fed to the shift unit 61, which determines a shift for the step size using a characteristic curve. The step size output by the integrator 59 is fed to a positive input of a computing element 60, and the step size output by the shift unit 61 is fed to a negative input of the computing element 60. The computing element 60 subtracts the step size transmitted by the shift unit 61 from the step size output by the integrator 59 and outputs the result to a hysteresis unit 62. The hysteresis unit 62 specifies switching thresholds so that excessive switching between the two overall light distributions is prevented.

[0175] Finally, depending on the step size output by the computing element 60, a characteristic signal for either the second or the third overall light distribution 40, 41 is output via the first output unit 63. The control element 16 can control the headlight arrangement depending on this characteristic signal. However, it is also possible for it to also take into account changes in the steering angle of the vehicle 10, as explained below: In this case, the vehicle's steering angle is continuously detected by a steering angle detection unit 64. The detected steering angle is fed to a generation unit 65 for generating a weighting factor. The generation unit 65 outputs a steering angle-dependent weighting factor to the first input of a logic unit 66.

[0176] Furthermore, the speed determined by the speedometer 54 is fed to a second generating unit 67 for generating a further weighting factor. Using a characteristic curve, this unit 67 generates a weighting factor, which is fed to the second input of the logic unit 66. The logic unit 66 determines which of the two supplied weighting factors is the lower and feeds it to the first input of a weighting unit 68.

[0177] The steering angle determined by the steering angle detection unit 64 is fed to a differentiator 69, which determines the temporal change in the steering angle. In the computing element 70, the absolute value of the angle change is determined and fed to the second input of the weighting unit 68. In the weighting unit 68, a characteristic variable is thus determined as a function of characteristic curves for the vehicle speed and as a function of the steering angle or the change in the steering angle, which is fed to a positive input of a further computing element 71. The current vehicle speed is further fed from the speedometer 54 to a threshold value generator 72, which generates a speed-dependent threshold value using a characteristic curve and feeds it to a negative input of the computing element 71. In the computing element 71, the threshold value is subtracted from the characteristic variable, and the result is fed to an integrator 73.The threshold value generated by the threshold value generator 72 is selected such that, in the case of no or only a slight change in the steering angle, an input value for the weighting unit 71 results, which causes a reduction of the characteristic value determined by the integrator 73. The threshold for reducing the characteristic value for the steering angle change determined by the integrator 73 is in a range between 2° per second and 3° per second.

[0178] The characteristic variable integrated by the integrator 73 represents a change in the steering angle. The integrator 73 transmits the determined change in the steering angle to a hysteresis unit 74, which implements on and off thresholds for the switching process to prevent excessive switching back and forth. The hysteresis unit 74 is set, for example, such that above integrated steering angle changes of approximately 200 degrees, a switchover from the second overall light distribution 40 to the third overall light distribution 41 occurs. If the integrated steering angle change falls below approximately 100 degrees, the second overall light distribution becomes active again. The hysteresis unit 74 outputs an identification signal via the second output unit 75, which identifies either the second or the third overall light distribution 40, 41.

[0179] The control unit 16 can determine which overall light distribution is to be controlled using the first output unit 63 or the second output unit 75. A further logic unit 76 can also be provided in the control unit 16, to which the identification signals from the first output unit 63 and the second output unit 75 are fed. From these two input signals, the logic unit 76 determines an identification signal, which is output via the third output unit 77. In this case, an identification signal for the third overall light distribution 41 is generated if such an identification signal is output by the first output unit 63 or the second output unit 75. The control unit 16 can then use this identification signal to generate the corresponding overall light distribution using the headlight arrangement. List of reference symbols 1 right headlight 2 left headlights 3 Light source 4 housings 5 Lens 6 Reflector 7 Projection lens 8 first aperture 8a Shading edge of the first aperture 9 second aperture 9a Shading edge of the second aperture 10 Vehicle with headlight arrangement 11 oncoming vehicle 12 vehicle ahead 13 Control unit for the right headlight 14 Control unit for the left headlight 15 Image processing device (device for detecting other road users) 16 Control unit 17 vehicle bus 18 Camera 19 first actuator for the right headlight 20 second actuator for the right headlight 21 third actuator for the right headlight 22 first actuator for the left headlight 23 second actuator for the left headlight 24 third actuator for the left headlight 25 Device for recording driving behavior 26 Delay unit 27 timers 28 navigation device 29 GPS receivers 30 Central axis 31 Extraction unit 32 Classifier 33 Confidence unit 34 Allocation unit 35 Horizontal plane 36 Central axis 37 horizontal axis 38 vertical axis 39 first overall light distribution / low beam 40 second total light distribution / masked permanent high beam 41 third total light distribution / sliding beam range 42 15° increase in the first total light distribution 43 Beam range, cut-off line 45 signal generators 46 switches 47 time incrementers 48 switches 49 signal generators 50 first timer for degradation rate 51 second timer for degradation rate 52 Integrators 53 position-dependent first step size incrementer 54 speedometers 55 position-dependent second step size incrementer 56 Logic unit 57 switches 58 position-dependent step size incrementer 59 Integrator 60 arithmetic terms 61 Shifting unit 62 Hysteresis unit 63 first output unit 64 Steering angle detection unit 65 first weighting factor generation unit 66 Logic unit 67 second weighting factor generation unit 68 weighting unit 69 Differentiator 70 arithmetic terms 73 threshold generators 100-170 process steps 180-280 process steps

Claims

[1] Method for controlling a headlight arrangement for a vehicle (10), the headlight arrangement comprising two spaced headlights (1, 2), in which - road users (11, 12) are detected in the direction of travel in front of the vehicle (10) and - a first overall light distribution (39) can be generated, in which the luminous range on a first side of a central axis (36) is greater than on the other, second side of this central axis (36), the greater luminous range on the first side of the central axis (36) being generated by the light emission of both headlights (1, 2), and - a second overall light distribution (40) can be generated, in which the overall light distribution (40) can be controlled such that it has a central region (M) with a luminous range in the direction of at least the detected road user (12) that is less than the distance to the detected road user (12), and has side regions (S1, S2) on both sides of this central region (M) with luminous ranges that are greater than the distance to the detected road user (12), characterized by , that - when switching from the first overall light distribution (39) to the second overall light distribution (40), the luminous range of at least the second headlight (2) on the first side of the central axis (36) is first reduced at least to such an extent that it is smaller than the distance to the detected road user (12), then the light emission direction (L) of the second headlight (2) is pivoted about a vertical pivot axis (38) and then the second overall light distribution (40) is generated. [2] Method according to claim 1, characterized by that the total light distributions (39, 40) are generated by the superposition of a first partial light distribution of the first headlight (1) and a second partial light distribution of the second headlight (2). [3] Method according to claim 1 or 2, characterized by that the first overall light distribution (39) provides a low beam function. [4] Method according to one of the preceding claims, characterized bythat in the second overall light distribution (40) the luminous range in the direction of the detected road user (12) is regulated so that it reaches the detected road user (12). [5] Method according to one of claims 1 to 4, characterized by that in the second overall light distribution (40) the side regions (S1, S2) each form a vertical light-dark boundary to the central region (M). [6] Method according to one of the preceding claims, characterized by that the switching from one overall light distribution (39) to the other overall light distribution (40) takes place automatically depending on the detection of the road user (12). [7] Method according to one of the preceding claims, characterized by that the road user (11, 12) in the direction of travel in front of the vehicle (10) is detected by - an image of a traffic area is taken in the visible spectral range, - extracting contiguous areas from the image with a brightness that exceeds a threshold, - the areas are classified at least according to their size, - for each area, a confidence value representing a measure of the similarity of the area of ​​the image to a vehicle light is formed from the classification of the area and a physical quantity associated with the area, and - depending on the confidence value, it is determined whether an area is assigned to a vehicle light. [8] Headlight arrangement for a vehicle with - at least two spaced headlights (1, 2) for generating an overall light distribution (39, 40), - a device (15, 18) for detecting road users (11, 12) in the direction of travel in front of the vehicle (10) and - a control device which is coupled to the road user detection device (18, 20) and with which a first overall light distribution (39) can be generated, in which the luminous range on a first side of a central axis (36) is greater than on the other, second side of this central axis (36), wherein the greater luminous range on the first side of the central axis (36) is generated by the light emission of both headlights (1, 2), and a second overall light distribution (40) can be generated, in which the overall light distribution (40) can be controlled such that it has a central region (M) in the direction of the detected road user (12) with a luminous range that is less than the distance to the detected road user (12), and on both sides next to this central region (M) has side regions (S1, S2) with a luminous range that is greater than the distance to the detected road user (12), characterized by , that - the headlights (1, 2) can be controlled by the control device (16) when switching from the first overall light distribution (39) to the second overall light distribution (40) in such a way that firstly the luminous range of at least the second headlight (2) on the first side of the central axis (36) is reduced at least to such an extent that it is smaller than the distance to the detected road user (12), then the light emission direction (L) of the second headlight (2) is pivoted about a vertical pivot axis (38) and then the second overall light distribution (40) is generated.

Images