DE102007008542B4 - Method and device for controlling the light output of a vehicle - Google Patents

Abstract

Method for controlling the light output of a vehicle, in which
several images are captured sequentially as a sequence of images,
tracking objects over multiple images to obtain an image sequence with tracking data;
at least a part of the images of the obtained image sequence is analyzed by means of a first analysis method (62);
at least a part of the images of the obtained image sequence is analyzed using at least one second analysis method (62 to 68) different from the first analysis method (62), wherein at least one analysis result is generated by each analysis method (62 to 68);
the analysis results are evaluated by means of an evaluation procedure, the evaluation process issuing at least one evaluation result;
and in which, depending on the evaluation result, the light output is controlled by at least one headlight (20, 22) of the vehicle (18).

Description

The invention relates to a method and a device for controlling the light output of a vehicle, in which a plurality of images of the surroundings of the vehicle are detected and analyzed.

From the document DE 10 2004 063 836 A1 It is known to arrange for imaging a camera in the region of the interior mirror of the motor vehicle, which detects the light of oncoming traffic and the light from other external light sources. The traffic situation in front of the vehicle is analyzed and depending on the analysis result, the range of the low beam headlight of the motor vehicle is adjusted. A variable adjustment of the range of the low beam can be done depending on the design of the headlamp by a headlamp leveling, by rotating a diaphragm shaft, by a sliding lowering of a shutter by a change in performance of one or more light sources, and by connecting one or more light sources. With the help of an already mentioned diaphragm shaft, which is also referred to as VarioX roller and which is arranged in the emission in front of the light source of the headlamp, the cut-off of the headlamp of the motor vehicle can be moved.

From the document EP 0 935 728 B1 a headlight for vehicles is known in which a diaphragm shaft is rotatable in several rotational positions to produce different light patterns. With the help of these light figures, a light-dark border can be varied.

From the document EP 0 971 829 B1 For example, a control system for automatically dimming vehicle headlights is known which comprises an optical system and an image processing system. The optical system is able to distinguish between headlamps and taillights, with the light beams from the headlamps and from the taillights focused on different areas of a sensor pixel array. Depending on the vertical and horizontal angular position as well as the distance of other vehicles, the high beam of the vehicle equipped with the control system is controlled.

The object of the invention is to provide a method and a device for controlling the light output of a vehicle, in which the light output can be controlled by at least one headlight of the vehicle in a simple manner to optimally illuminate the traffic space in front of the vehicle.

This object is achieved by a method having the features of patent claim 1 and a device having the features of patent claim 11. Advantageous developments of the invention are specified in the dependent claims.

In the method for controlling the light emission of a vehicle having the features of claim 1, a plurality of images are sequentially detected as an image sequence. At least a part of the images of the image sequence is analyzed by means of a first analysis process. Furthermore, at least a part of the images of the image sequence is analyzed by means of at least one second analysis method different from the first analysis method, wherein each analysis method generates at least one analysis result.

The analysis results are evaluated by means of a rating procedure and an evaluation result is output. Depending on the evaluation result, the light output is controlled by at least one headlight of the vehicle.

In this method according to the invention, a plurality of analysis methods are preferably carried out in parallel, in the multitasking mode of a control unit or in succession. The analysis results of the analytical methods are evaluated using the evaluation method. In particular, the evaluation method can take account of preset evaluation criteria, by means of which, in particular, legal or other regulations, which are specified in specifications, for example, can be implemented. By means of the method having the features of patent claim 1, it is possible to set the light output by at least one headlight of the vehicle automatically as a function of the surroundings of the vehicle analyzed with the aid of the captured images and the analysis methods. In particular, the area in front of the vehicle is detected and evaluated with the aid of the images in that the light beams emitted by the at least one headlight of the vehicle can hinder other road users. By the method, the light emission of the headlights of a vehicle and thus the illumination of the area in front of the vehicle can be optimally adjusted, obstructing other road users is avoided.

By a device for controlling the light emission of a vehicle having the features of claim 11, the same advantages can be achieved as by the method according to claim 1. Furthermore, this device can be developed in the same way as this method.

Further features and advantages of the invention will become apparent from the following description which, taken in conjunction with the accompanying drawings Invention explained with reference to an embodiment.

• 1 eine schematische Darstellung einer Verkehrssituation mit mehreren Kraftfahrzeugen;
• 2 ein Blockdiagramm zur verkehrsabhängigen Steuerung der Lichtabgabe eines Kraftfahrzeugs;
• 3 das mit Hilfe einer Kamera aufgenommene Bild eines vorausfahrenden Fahrzeugs bei Dunkelheit, wobei die Rücklichter des vorausfahrenden Fahrzeugs sichtbar sind;
• 4 das mit Hilfe einer Kamera aufgenommene Bild eines entgegenkommenden Fahrzeugs bei Dunkelheit, wobei ein Frontscheinwerfer des Fahrzeugs sichtbar ist;
• 5 das mit Hilfe einer Kamera aufgenommene Bild einer vom Fahrzeug weit entfernten Straßenlampe bei Dunkelheit;
• 6 das mit Hilfe einer Kamera aufgenommene Bild einer nahe dem Fahrzeug angeordneten Straßenlampe;
• 7 das mit Hilfe einer Kamera aufgenommene Bild von insgesamt zwei Leitpfosten, wobei ein erster Leitpfosten nahe dem Fahrzeug und ein zweiter Leitpfosten weiter entfernt vom Fahrzeug angeordnet sind;
• 8 das mit Hilfe einer Kamera aufgenommene Bild eines entgegenkommenden Fahrzeugs bei Dunkelheit, wobei die Frontscheinwerfer des Fahrzeugs sichtbar sind;
• 9 das Tracking eines Objekts über mehrere Bilder zum Ermitteln der Zusammengehörigkeit von in den einzelnen Bildern ermittelten Objekten; und
• 10 eine grafische Darstellung der Positionsbestimmung eines Objekts mit Hilfe einer bewegten Kamera, um die Hypothese zu bestätigen oder zu widerlegen, dass das Objekt ein ortsfestes Objekt ist.

Show it:

• 1 a schematic representation of a traffic situation with multiple vehicles;
• 2 a block diagram for traffic-dependent control of the light output of a motor vehicle;
• 3 the image taken by a camera of a vehicle ahead in the dark, the taillights of the vehicle in front are visible;
• 4 the image taken by a camera of an oncoming vehicle in the dark, wherein a headlight of the vehicle is visible;
• 5 the image taken with the aid of a camera of a street lamp far away from the vehicle in the dark;
• 6 the image taken with the aid of a camera of a street lamp arranged near the vehicle;
• 7 the image taken with the aid of a camera of a total of two guide posts, wherein a first guide post near the vehicle and a second guide post are arranged further away from the vehicle;
• 8th the image taken by a camera of an oncoming vehicle in the dark, the headlights of the vehicle being visible;
• 9 tracking an object across multiple images to determine the identity of objects found in each image; and
• 10 a graphical representation of the position of an object using a moving camera to confirm the hypothesis or disprove that the object is a stationary object.

In 1 is a schematic representation of a traffic situation 10 with several vehicles 18 . 24 . 26 . 28 on the lanes 14 . 16 the roadway 12 shown. On the lane 14 is the vehicle 18 , which is a device for controlling the light output of the headlights 20 . 22 of the vehicle 18 having. In front of the vehicle 18 is on the lane 14 the vehicle 24 , The vehicle 24 forms the vehicle 18 nearest traffic ahead. On the lane 16 come to the vehicle 18 the vehicles 26 . 28 contrary to oncoming traffic. There are also two street lamps 30 . 32 on the right side of the track and two streetlights 34 . 36 on the left lane in the area in front of the vehicle 18 , From the vehicle 18 Seen from are still on the right lane two Leitposts 38 . 40 and on the left side of the road two guide posts 42 . 44 available. The street lamps 30 to 36 serve to illuminate the lanes 14 . 16 within a closed village. The guide posts 38 to 44 serve to mark the edges of the road outside closed villages.

In front of the vehicle 18 are those with the help of the headlights 20 . 22 of the vehicle 18 illuminable areas for conventional dipped beam (area 46 ) for headlamps equipped with headlamp leveling and a VarioX roller shutter 20 . 22 of the vehicle 18 illuminable area 48 and the area which can be illuminated by means of the high beam 50 as a projection onto the roadway 12 shown schematically. The vehicle 18 has an image capture unit 52 on, preferably near the rear-view mirror behind the windshield of the vehicle 18 is arranged. The image capture unit 52 preferably generates charge images with the aid of a CMOS pixel array, which are converted into digital image data with the aid of an analog-to-digital converter. For each pixel of the CMOS pixel array, a brightness value is determined, preferably using the image acquisition unit 52 Grayscale images and no color images captured. The image capture unit 52 has a suitable optical system for imaging the entire relevant area in front of the vehicle 18 on, so that on with the help of the image capture unit 52 captured images the preceding vehicle 24 , the oncoming vehicles 26 . 28 , the street lamps 30 to 36 as well as the guide posts 38 to 44 are shown. In the dark are from the image capture unit 52 only the currently light-emitting light sources and the currently light-reflecting elements of these objects are detected as bright areas in the recorded images.

The headlights 20 . 22 have actuators for headlamp leveling and / or for setting a desired position of the respective diaphragm roller or at least allow the switching on and off of the high beam via a data bus. Preferably, the headlights 20 . 22 of the vehicle 18 Control units connected via a CAN bus system with other control units of the vehicle 18 are connected. The image capture unit 52 is also via the CAN bus system with other control units of the vehicle 18 , in particular the control units of the headlamp systems 20 . 22 , connected.

The light control of the vehicle 18 has to be done in such a way that neither the preceding traffic (vehicle 24 ) still the oncoming one Traffic (vehicles 26 . 28 ) and thus obstructed.

The image capture unit 52 continuously takes pictures of the relevant area in front of the vehicle 18 on, which are evaluated by an image processing unit, as will be explained in more detail below in connection with the other figures. The analysis results of several analytical methods are evaluated using a suitable evaluation method. An evaluation result of the evaluation process then indicates which headlamp setting of the headlamp 20 . 22 is possible to achieve the optimal road illumination. The optimal road illumination is the maximum illumination of the traffic area that can be generated by the headlights, in which the other road users are not dazzled and obstructed. The optimum illumination can be dispensed with, for example, during day trips and / or in villages in the dark. At the in 1 illustrated traffic situation 10 is the optimal possible illumination through the area 51 given, since in this illumination no road users of the vehicles 24 . 26 . 28 through the headlights 20 . 22 of the vehicle 18 emitted light is impaired.

In 2 is a block diagram with functions for image acquisition, image processing and control of the headlights 20 . 22 shown. With the help of the image acquisition unit 52 will, as already related to 1 described, an image sequence with images of the area in front of the motor vehicle 18 added. Preferably, when driving the motor vehicle 18 with the light system switched on or pictures taken in the dark. The picture is taken with the help of the function block 54 controlled by the function block 54 both the timing of the image acquisition and the conversion of the recorded images is controlled in digital image data. The images of the image sequence can be at the same time intervals or in of the vehicle speed of the motor vehicle 18 dependent time intervals are recorded. With the help of the image acquisition unit 52 and the image pickup control 54 there is a data collection. In the captured images, areas with higher brightness than the rest of the image areas are rendered using the Blobextraktionsfunktion 56 extracted. These areas of greater brightness are radiation sources, such as light sources or light-reflecting objects, ie, for example, traffic signs or guide posts, as well as warning signs and reflectors on motor vehicles.

The Blobextraktion 56 is performed for each image used for further analysis. With the help of a subsequent blobtracking function 58 the same radiation sources are determined in several consecutively recorded images and their movement is tracked over several images. The blobtracking thus includes processing steps that serve the tracking of objects across multiple images and thereby determine the information about the course of the object movement and the position of the object in space. With the help of blobtracking, a prediction of the position and movement information can be made on the basis of the motion information and position information of an object already determined over several images in conjunction with physical and mathematical laws. This process is also called prediction.

The association preferably encompassed by blobtracking identifies individual objects over several measurement cycles or over a plurality of images, as a result of which an object determined in an image currently to be evaluated can be assigned to an object that has already been detected and tracked over several images. On the basis of the motion profiles determined for individual objects, the object can be assigned to a specific object model, as a result of which subsequently at least one object property of the object can be determined. The blob tracking function 58 will provide additional calibration information 60 provided to prevent blobtracking errors.

The of the blobtracking function 58 determined tracking results are subsequently several analysis functions 62 to 68 fed, each performing an analysis method. Every analysis function 62 to 68 become additional information 62a to 68a provided in the form of model knowledge in order to be able to process and evaluate the supplied information accordingly. In the analysis function 62 The 2D properties of objects are recorded, in particular the shape, the structure, the size etc. For this purpose, the Blobstruktur evaluated, in particular the gray value and the gray value distribution of each blob, ie each brightness blob detected in the captured image or in the captured images and is evaluated. However, the gray values and the gray value distribution of the brightness blots are strong from the distance and the viewing angle from the image capture unit 52 depending on the radiation source of the blob and the design of the brightness point. Examples of possible blobs whose two-dimensional properties ( 2D Properties), ie their Blobstruktur, analyzed are below in the 3 to 8th shown by way of example. However, the analysis of the 2D properties gives a weakly significant feature, which only leads to a meaningful evaluation result in combination with the results of further analyzes.

The analysis function 64 determines bright areas belonging together from the blob tracking data, preferably over a plurality of successively recorded images of a sequence and analyzes a three-dimensional structure (3D structure) of the individual light-emitting elements in the room and can therefore in particular associate associated elements with an object or combine them into an object. In this analysis, a modeling of expected structures takes place. In particular, pairs of lights, such as headlight pairs and pairs of taillights, are determined. For example, with the 3D structural analysis, vehicles with numerous auxiliary lights can be identified as trucks and a series of illuminated spots as street lights or guide posts. For individual light sources, such. For example, the tail light or front headlamp of motorcycles or cars or trucks with only one working taillight or only one functioning headlamp can not be analyzed more accurately using 3D structural analysis so that 3D structural analysis can not determine any further information for these objects. The analysis function 64 to determine the 3D structure can also be used as part of a so-called Lane Departure Warning (LDW function).

A tracking can then preferably take place for all points of light associated with an object. Using the power indicator function 56 For example, mains operated light sources such as traffic lights, street lighting, illuminated billboards, and building leaked light may be distinguished from non-mains powered light sources. The mains-operated light sources are fixed light sources, which are irrelevant for a light control according to the invention, since these lights are not attributable to road users, in the light distribution of the motor vehicle 18 radiated light would have to be considered. The sampling frequency, ie the image acquisition frequency, which is determined by means of the power indicator function 66 Evaluated images can be well below the mains frequency of 50 or 60 Hz, with which the mains-powered light sources are supplied with energy. A concrete embodiment of the power indicator is detailed in the document WO 2006/069978 A2 described.

With the help of the analysis function 68 a distinction can be made from stationary, ie stationary, and moving objects, by the position of a light-emitting object in space with the aid of the image acquisition unit 52 taken pictures in a known relative movement of the image acquisition unit 52 is determined to the object. Such a position determination is with the aid of an image acquisition unit 52 , which has only a monocamera system and no stereo camera system, only possible for fixed objects. Can for such an object on at least three successive recorded images, between which a relative movement of the image acquisition unit 52 If a single position has been plausibly determined, the object is a stationary object. In this analysis function 68 is preferably hypothesized that the object to be analyzed is a stationary stationary object. The position of the object is thereby when using a mono camera as an image capture unit 52 is determined by means of two successive recorded images and checked with at least one third image by the position of the object is further determined using the first and third and / or second and third image. If the determined positions match, the object is stationary and stationary in the period considered. Thus, the hypothesis that the object is stationary is checked and verified with the aid of the third and / or further images. If there are other positions for the object, so that the hypothesis can not be verified, the object is a moving object. The analysis function 68 In addition to model knowledge, information about a vehicle dynamics model is also provided for the determination of stationary objects 68a , via the CAN bus system 69 of the motor vehicle 18 current vehicle data is made available, in particular the speed and / or position of the vehicle 18 affect. From this data, the position change of the image acquisition unit 52 between the recording of two used for determining position images.

The arialysis results of the analysis function 62 to 68 become the classification function 70 fed by a summary of the through the individual analysis functions 62 to 68 carried out individual evaluations. In this case, an assignment and a differentiated evaluation are carried out in predefined object classes. Such object classes may be, for example, taillight, headlights, street lamp, reflector, road sign, etc. Based on preset rules by which functional rules of the overall system for controlling the light output are implemented by the function categorization, an overall evaluation of the analysis results of the analysis functions. Specifically, specifications and legal regulations can be taken into account and actions for lighting control can be defined. Subsequently, due to the adjustment possibilities of the lighting system or the headlights 20 . 22 of the motor vehicle 18 the optimal illumination 51 the relevant traffic space in front of the vehicle 18 taking into account the other road users 24 . 26 . 28 certainly. This information is used as a parameter or actuating signal via the CAN bus system of the motor vehicle 18 transmitted to the light system, through which the light output of the headlights 20 . 22 is adjusted accordingly. About a human-machine interface information about the current setting of the lighting system of the motor vehicle 18 to the driver of the motor vehicle 18 be issued.

In the 3 to 8th is one each with the help of the image acquisition unit 52 of the motor vehicle 18 recorded image in different traffic situations, where in each case the analysis result of the analysis function 62 is specified by which the determined object is more precisely determined and assigned to an object class. At the in 3 The light-emitting elements shown are the taillights and license plate illumination of a vehicle ahead 24 , At the in 4 shown light emitting area is a headlight of an oncoming vehicle. At the in 5 shown light emitting area is a remote from the motor vehicle 18 or removed from the image capture unit 52 arranged street lamp. At the in 6 shown light emitting area is a close to the motor vehicle 18 or near the image capture unit 52 arranged street lamp. At the in 7 shown light emitting areas is one near the motor vehicle 18 arranged first guide post and one away from the motor vehicle 18 arranged second guide post. At the in 8th shown light emitting areas are remote from the motor vehicle 18 arranged headlights oncoming motor vehicles.

In 9 For example, the analysis result of the 3D structure analysis function 64 is shown for a total of 12 street lamps, with the street lamps in the camera coordinate of the image acquisition unit 52 provided right-hand diagram as points of light and in the left diagram as points in the coordinate system of a plane are shown. The positions of the light points or the street lamps in the plane are in the left diagram of 9 starting from the image acquisition unit 52 or to the motor vehicle 18 indicated, wherein the image acquisition unit 52 is arranged in the coordinate origin.

In 10 is the position determination of a stationary object 80 with the help of the analysis function 68 shown. The image capture unit 52 is with the vehicle 18 emotional. The movement of the image acquisition unit 52 in the plane is represented by several points, which are arranged on a straight line parallel to the Y-axis. The relative movement of the image acquisition unit 52 to the object 80 is due to the movement of the vehicle 18 set because the object 80 is assumed to be stationary. Will for the object 80 With the help of several position determinations with different recorded pictures the same position is determined, the assumption is correct, that it is the object 80 is a fixed object. If different positions or non-plausible positions are determined, this is the object 80 not a stationary object.

In this analysis, then the distance between the motor vehicle 18 and the object 80 be determined with the help of a mono camera, if it is the object 80 is a fixed object. The classification of objects is preferably carried out by means of an evaluation method which carries out the evaluation of the objects to be detected on the basis of the analysis results output by the analysis functions. In particular, a standard classification method can be used as the evaluation method, in particular using the Bayes classification, a fuzzy logic, a support vector machine and / or another suitable evaluation method.

Claims (11)

Method for controlling the light output of a vehicle, in which several images are captured sequentially as a sequence of images, tracking objects over multiple images to obtain an image sequence with tracking data; at least a part of the images of the obtained image sequence is analyzed by means of a first analysis method (62); at least a part of the images of the obtained image sequence is analyzed using at least one second analysis method (62 to 68) different from the first analysis method (62), wherein at least one analysis result is generated by each analysis method (62 to 68); the analysis results are evaluated by means of an evaluation procedure, the evaluation process issuing at least one evaluation result; and in which, depending on the evaluation result, the light output is controlled by at least one headlight (20, 22) of the vehicle (18). Method according to Claim 1 , characterized in that the analysis methods independently analyze the respective part of the images, wherein at least a part of the images used by the first analysis method (62) is also used for the analysis by means of the second analysis method (64 to 68). Method according to one of the preceding claims, characterized in that the images of the image sequence with the aid of a single camera (52) or a stereo camera are detected, wherein at least the images analyzed using the first analysis method (62) are preferably grayscale images. Method according to one of the preceding claims, characterized in that the light distribution generated by the headlamps (20, 22) is controlled, the light beams emitted by at least one headlamp (20, 22) of the vehicle (18) being adjusted by changing the angle of inclination , the yaw angle and / or an optical system for shaping the emitted light beams of at least one headlight (20, 22) and / or at least one light source is controlled. Method according to one of the preceding claims, characterized in that the evaluation method with the aid of the analysis results performs a classification with the aid of features contained in the analysis results of detected objects (80), preferably by means of a standard classification method, in particular using the Bayes classification Fuzzy logic or a support vector machine. Method according to Claim 5 , characterized in that further information provided by the vehicle (18) is taken into account in the classification (68a). Method according to one of the preceding claims, characterized in that a categorization of objects (80) determined with the aid of the images is carried out with the aid of the first analysis method (62), the second analysis method (64 to 68) and / or the evaluation method. Method according to one of the preceding claims, characterized in that the analysis methods (62 to 68) comprise an analysis method for power-lamp recognition, an analysis method for three-dimensional structure analysis for detecting associated objects, an analysis method for determining two-dimensional properties of an object, such as shape, structure and / or size, an analysis method for determining three-dimensional properties of an object, preferably for determining the object status moved or unmoved, and / or at least one further analysis method. Method according to Claim 8 , characterized in that the determined properties of an object (80) are compared by at least one of the analysis methods (62, 60) with properties of an object model (62a to 66a), whereby due to the matching and / or deviating properties between the actual properties of the object (80) and an object model (62a to 66a) an evaluation criterion as an analysis result of the evaluation method is provided. Method according to one of the preceding claims, characterized in that the evaluation method in the processing of the analysis results performs a weighting of the analysis results and / or includes error tolerances in the evaluation of the analysis results. Device for controlling the light emission of a vehicle, with an image acquisition unit (52) for capturing images of a sequence of images, wherein a tracking of objects over a plurality of images takes place in order to obtain an image sequence with tracking data; with an evaluation unit which analyzes a part of the images of the obtained image sequence with the aid of a first analysis method (62), analyzing at least part of the images of the obtained image sequence using at least one second analysis method (64 to 68) different from the first analysis method (62), evaluating, by means of a rating procedure, the results of the analysis produced by each analysis method and outputting an evaluation result, and which, depending on the evaluation result, generates an actuating signal by which the emission of light beams of at least one headlight (20, 22) of a vehicle (18) is controlled.

Images