DE102019132973A1 - Method for reorganizing sampling points in a point cloud generated by a detection system of a vehicle - Google Patents

Abstract

The present invention relates to a method for reorganizing scanning points (16) in a point cloud generated by a detection system (1) of a vehicle, a scene being measured by means of electromagnetic waves by a detection system (1) of a vehicle and at least one scanning of the scene is generated with a plurality of sampling points (16). The sampling points (16) are grouped to form at least one chain (20), the grouping being carried out by a sampling point-to-chain assignment algorithm. If no chain (20) exists, a new chain (20) is created. If a chain (20) is present, a sampling point (16) from a volley (15) is assigned to the chain (20), the assignment being based on a distance between the given sampling point (16) from a step k and a last sampling point ( 16) is based on a step k-1 in the chain (20), the chain (20) being assigned a sampling point (16) if the distance between them is the lowest of the sampling points (20) in the volley (15), and is below a predefined allocation threshold. The chain (20) is skipped for the scanning point (16) if the distance exceeds the predefined allocation threshold value, and the last scanning point (20) is marked as the end point (28). The chain (20) is then saved for further processing. The present invention also relates to a driver assistance system and a computer program product.

Description

The present invention relates to a method for reorganizing sample points in a point cloud generated by a detection system of a vehicle. The present invention also relates to a driver assistance system. The present invention also relates to a computer program product.

In automotive applications such as obstacle detection and avoidance during autonomous journeys or adaptive front lighting, the three-dimensional environment around a vehicle is recorded. To measure the surroundings, the vehicle is typically equipped with suitable sensors in the form of 3D scanners such as so-called lidar sensors (light detection and ranging) or radar sensors. With light detection and distance measurement, the distance to objects is determined by illuminating the environment and thus the objects in it with pulsed laser light and capturing the reflected laser light. The return time of the laser light is a measure of the distance to the surface of an object in the vicinity. An intensity of the reflection can be processed to obtain more information about a surface that is reflecting the laser light.

A 3D scanner creates a set of data points in three-dimensional space called a point cloud. A point cloud is a geometric data structure. Each (data) point of the point cloud corresponds to a physical point on the outer surface of an object in the vicinity of a vehicle and typically includes the coordinates X, Y and Z of the physical point in a three-dimensional Cartesian coordinate system as well as optional additional features. A 3D scanner typically outputs the measured point cloud as a data structure or data file. 5 shows an example of a point cloud obtained when scanning the three-dimensional surroundings of a vehicle with a lidar sensor. In general, point clouds are not limited to a three-dimensional coordinate system, but can be of higher or lower dimension.

In order to understand the environment, it is important to recognize the objects by semantically segmenting each point of an object and classifying the objects. Object recognition, semantic segmentation and classification are identified as three basic problems / tasks for scene understanding in the field of computer vision.

One task is the detection and tracking of road markings on the basis of lidar data. The detection of lane markings is a prerequisite for many driver assistance systems as well as for autonomous vehicles.

The problem with the LIDAR algorithms for single-scan-based lane marker recognition is, for example, that these algorithms require a linear organization of the scanning points within the point cloud. If the LIDAR delivers more than one echo per recording, the point cloud must be transformed in order to be compatible with an algorithm for recognizing road features.

It is an object of the present invention to provide a method for reorganizing sample points in a point cloud.

• - Messen einer Szene mittels elektromagnetischer Wellen aus einem Detektionssystem eines Fahrzeugs und Erzeugen mindestens einer Abtastung der Szene mit einer Vielzahl von Abtastpunkten, wobei die Gesamtheit aller Abtastpunkte zusammen mit den zugehörigen Punktdaten der erzeugten Abtastung eine Punktwolke bildet, wobei elektromagnetische Wellen, die von einem Objekt innerhalb eines vorbestimmten Polarwinkels eines kugelförmigen Koordinatensystems des Detektionssystems reflektiert werden, eine Schicht von Abtastpunkten bilden, wobei elektromagnetische Wellen, die von einem Objekt unter demselben Azimutwinkel für den vorbestimmten Polarwinkel des kugelförmigen Koordinatensystems reflektiert werden, eine Salve bilden,
• - Gruppieren der Abtastpunkte zu mindestens einer Kette, wobei die Gruppierung durch einen Abtastpunkt-zu-Kette-Zuordnungsalgorithmus erfolgt, wobei jede Kette einen Startpunkt und einen Endpunkt aufweist, wobei die Anpassung durch den Algorithmus Salve für Salve ausgehend von der Salve mit Index 0 erfolgt, wobei jede Kette bis zu einem Abtastpunkt pro Salve aufnehmen kann,
• - wenn keine Kette vorhanden ist, Erstellen einer neuen Kette und Zuordnen eines Abtastpunktes von der Salve mit Index 0 zur neuen Kette und Markieren des Abtastpunktes als Startpunkt,
• - wenn eine Kette vorhanden ist, Zuordnen eines Abtastpunkts von der Salve zu der Kette, wobei die Zuordnung auf einem Abstand zwischen dem gegebenen Abtastpunkt von einem Schritt k und einem letzten Abtastpunkt von einem Schritt k-1 in der Kette basiert, wobei der Kette ein Abtastpunkt zugeordnet wird, wenn der Abstand zwischen ihnen der niedrigste der Abtastpunkte in der Salve ist, und unterhalb eines vordefinierten Zuordnungsschwellenwerts ist,
• - Überspringen der Kette n für den Abtastpunkt m, wenn der Abstand den vorgegebenen Zuordnungsschwellenwert überschreitet, und Markieren des letzten Abtastpunktes als Endpunkt,
• - Speichern der Kette zur weiteren Verarbeitung.

This goal is achieved by the independent claims. Advantageous embodiments are specified in the dependent claims. In particular, the present invention provides a method for reorganizing sample points in a point cloud generated by a detection system of a vehicle, comprising the following steps:

• - Measuring a scene by means of electromagnetic waves from a detection system of a vehicle and generating at least one scan of the scene with a plurality of scan points, the entirety of all scan points together with the associated point data of the generated scan forming a point cloud, with electromagnetic waves generated by an object are reflected within a predetermined polar angle of a spherical coordinate system of the detection system, form a layer of scanning points, with electromagnetic waves reflected by an object at the same azimuth angle for the predetermined polar angle of the spherical coordinate system forming a volley,
• - Grouping of the sampling points into at least one chain, the grouping being carried out by a sampling point-to-chain assignment algorithm, each chain having a starting point and an end point, the adaptation being carried out by the algorithm salvo by salvo starting from the salvo with index 0 each chain can accommodate up to one sampling point per salvo,
• - if there is no chain, create a new chain and assign a sampling point from the salvo with index 0 to the new chain and mark the sampling point as the starting point,
• - if a chain is present, assigning a sample point from the volley to the chain, the assignment being based on a distance between the given sample point from step k and a last sample point from step k-1 in the chain, the chain being a sample point is assigned if the distance between them is the lowest of the sampling points in the salvo and is below a predefined assignment threshold,
• - Skipping the chain n for the sampling point m, if the distance exceeds the predefined allocation threshold value, and marking the last sampling point as the end point,
• - Saving the chain for further processing.

In view of the fact that each photomultiplier can provide several sampling points from separate reflectors and that the complexity of the feature extraction algorithms increases in the presence of ambiguities in the assignment of sampling points, so-called sampling point chains were introduced as an elementary block of the sampling point layers. To reduce the complexity of the pattern recognition algorithms, linear sequences of sample points that repeat the shape of the features can be provided as input. The basic idea of the invention is therefore to group sampling points which possibly belong to the source or to the feature and to assign them to a chain. The algorithm provided can be interpreted as an “online” temporal data classifier in which the previous number of classes or chains is not known, the maximum number of classes or chains being derived from the configuration parameters. The number of classes and their localization can vary with each update. The algorithm is a recursive online algorithm in which only one measurement per class is stored in a training vector. The training vector is updated with the assigned points and predictions from a previous step. If the distance to the next point exceeds a threshold value, a new class is created; if the class was not assigned the new point in several predefined steps, it is removed from the training vector.

According to a preferred embodiment of the invention, the assignment of a sampling point m to a chain n is based on an MxN confusion matrix, where N is the number of active chains in step k and M is the number of new sampling points to be assigned to the chain.

According to a further preferred embodiment of the invention, the distance is a radial distance between the scanning points.

According to a preferred embodiment of the invention, the point cloud is a filtered point cloud, a sampling point with a certain confidence level being regarded as a sampling point which is reflected from a floor surface in front of the detection system.

According to a further preferred embodiment of the invention, the step of storing the chain for further processing comprises the step of linearly storing the chain in a memory of a computing unit of the vehicle.

In a preferred embodiment of the invention, the method comprises the step of providing the stored chain to an extraction algorithm for road features.

According to a further preferred embodiment of the invention, the detection system is a LIDAR system.

According to a preferred embodiment of the invention, the azimuthal angle is defined by an angle of a mirror of the LIDAR system and the polar angle is defined by a laser diode or a photodiode position of the LIDAR system in relation to the mirror.

The present invention also provides a driver assistance system for carrying out the method described above.

In a preferred embodiment of the invention, the driver assistance system comprises a detection system.

According to a further preferred embodiment of the invention, the detection system comprises a LIDAR system.

In addition, the present invention provides a computer program product containing instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the method described above.

These and other aspects of the invention will be apparent and explained with reference to the embodiments described below. Individual features that are disclosed in the embodiments can alone or in combination constitute an aspect of the present invention. Features of the various embodiments can be transferred from one embodiment to another embodiment.

• 1 zeigt verschiedene schematische Ansichten eines Detektionssystems gemäß einer Ausführungsform der Erfindung,
• 2 zeigt verschiedene schematische Ansichten des Sichtfeldes eines Detektionssystems gemäß einer Ausführungsform der Erfindung,
• 3 zeigt einen r-z Querschnitt des Detektionssystems mit verschiedenen Detektionssch ichten,
• 4 zeigt ein schematisches Schema der Indexierung einer Punktwolke,
• 5 zeigt eine schematische Ansicht des Sichtfeldes eines Detektionssystems mit Abtastpunkten, die von einer Bodenoberfläche gemäß einer Ausführungsform der Erfindung reflektiert werden,
• 6 zeigt ein Beispiel für eine LIDAR-Punktwolke,
• 7 zeigt ein Beispiel für eine Punktwolke, die von einer Bodenoberfläche erzeugt wird,
• 8 zeigt verschiedene Szenarien, die zu einer Mehrfachechoerzeugung führen,
• 9 zeigt ein Flussdiagramm, das ein Verfahren zum Umorganisieren von Abtastpunkten in einer Punktwolke veranschaulicht, die von einem Detektionssystem eines Fahrzeugs gemäß einer Ausführungsform der Erfindung erzeugt wird,
• 10 zeigt die Ausgabe des Verfahrens zur Umorganisation von Abtastpunkten in einer Punktwolke gemäß einer Ausführungsform der Erfindung,
• 11 zeigt verschiedene Beispiele für übereinstimmende Ketten, wenn das Verfahren zur Umorganisation von Abtastpunkten in einer Punktwolke gemäß einer Ausführungsform der Erfindung ausgeführt wird,
• 12 zeigt ein Speicherlayout der Ketten innerhalb einer einzelnen Schicht gemäß einer Ausführungsform der Erfindung.

In the drawings:

• 1 shows various schematic views of a detection system according to an embodiment of the invention,
• 2 shows various schematic views of the field of view of a detection system according to an embodiment of the invention,
• 3rd shows a rz cross section of the detection system with different detection layers,
• 4th shows a schematic diagram of the indexing of a point cloud,
• 5 shows a schematic view of the field of view of a detection system with scanning points reflected from a ground surface according to an embodiment of the invention;
• 6th shows an example of a LIDAR point cloud,
• 7th shows an example of a point cloud generated by a ground surface,
• 8th shows different scenarios that lead to multiple echo generation,
• 9 FIG. 13 shows a flow diagram illustrating a method for reorganizing sample points in a point cloud generated by a detection system of a vehicle according to an embodiment of the invention.
• 10 shows the output of the method for reorganizing sample points in a point cloud according to an embodiment of the invention,
• 11 shows various examples of matching chains when performing the method for reorganizing sample points in a point cloud according to an embodiment of the invention,
• 12th Figure 12 shows a memory layout of the chains within a single layer according to an embodiment of the invention.

1 shows various schematic views of a detection system 1 according to one embodiment of the invention. In one embodiment of the invention, for example, the use of a LIDAR system as a detection system 1 are provided. LIDAR (range detection and ranging) is a device that measures the distance to the target using the Time of Flight (ToF) method. A lidar system 1 consists of four main detection components, a laser diode 2 , which emits light pulses, the initially collimated laser beam being defocused to produce a beam with a non-zero angle of divergence, a rotating mirror 4th , which defines the direction in which the detection is to take place, a focusing lens 7th that the reflected light hit the photodetector 8th projected and a group of photodetectors that collects the reflected light and converts it into the electrical signal.

1 a) shows the projection of laser light onto an object. A laser diode 2 emits laser beams by a rotating mirror 4th be aimed at the object. In 1 a) is the projected laser light 5 shown.

1 b) shows the propagation of the reflected light. The laser light is reflected from the object and by means of a focusing lens 7th on a photodetector 8th directed. In 1 b) the photodetector components of the LIDAR system are made by an array of photodetectors 8th shown, which are aligned parallel to the z-axis. Any photodetector 8th within the arrangement of the photodetectors collects the light reflected from a predefined polar angle of the sensor coordinate system and forms the so-called layer 14th the sampling points 16 . The photodetector 8th can be an avalanche photodiode or an APD. In 1 b) is also the area 6th represented by one of the photodetectors 8th from the range of photodetectors is covered.

2 shows various schematic views of the field of view 9 a detection system 1 according to one embodiment of the invention. 2 a) shows the xy cross-section of a field of view 9 of a LIDAR system. The xy cover is made through the use of a rotating mirror 4th reached. The rotating mirror changes with each laser burst 4th its orientation, which is predefined according to the azimuthal resolution of the LIDAR. In 2 B) is the direction of mirror rotation 10 and the resulting direction of the emitted light depending on the orientation of the mirror 4th shown.

3rd . shows a rz cross section of a LIDAR system. The LIDAR system scans a floor surface 11 , e.g. after road markings. The light reflected from a predefined polar angle of the sensor coordinate system defines different layers 14th . In 3rd are the layers 12th that have no sampling point 16 from the soil surface 11 generate, as well as the layers 13th having a sampling point from the ground surface 11 generate, shown.

4th shows a schematic diagram of the indexing of a point cloud. A 3D scanner creates a set of data points in three-dimensional space called a point cloud. A point cloud is a geometric data structure. Each (data) point of the point cloud corresponds to a physical point on the Outer surface of an object in the vicinity of a vehicle and typically includes the coordinates X, Y and Z of the physical point in a three-dimensional Cartesian coordinate system. Electromagnetic waves reflected from the physical point on the outer surface are called echoes, where electromagnetic waves reflected from an object at the same azimuthal angle for a predefined polar angle of the spherical coordinate system are called a volley 15th form. Any echo or sample point 16 within a point cloud carries the following indices, an index of the volley 15th , which corresponds to the mirror angle, a layer index 14th which corresponds to the polar angle of the photomultiplier field of view axis, and an echo index. Each block in the in 4th The scheme shown corresponds to the arrangement of the echoes generated by the same photomultiplier.

5 shows a schematic representation of the field of view of a detection system 1 with sampling points 16 reflected from a floor surface according to an embodiment of the invention. In 5 the xy cross-section of a field of view of a LIDAR system is shown. In addition, there are different layers 14th of sampling points 16 reflected from the ground surface.

6th shows an example of a point cloud obtained when scanning the three-dimensional surroundings of a vehicle with a LIDAR system. In general, point clouds are not limited to a three-dimensional coordinate system, but can be of higher or lower dimension. The example of a LIDAR generated point cloud in 6th was within one mirror rotation cycle 4th receive. In 6th is the echo of different layers 14th , as in 5 described to see.

7th shows an example of a point cloud generated from a ground surface. The vertical lines connect echoes that come to the same layer 14th and volley 15th and were generated by the same photomultiplier for the same laser beam.

It is noteworthy that a separation lane marking can generate up to three echoes, for example at x = 200cm and y = 3000-3500cm, during scan points 16 can be recognized by continuous lane markings as the furthest or second echoes.

8th shows different scenarios that lead to the generation of multiple echoes by a single photomultiplier. Laser beams are emitted by a LIDAR system and reflected on the surface of the ground, with in 8th Echoes are shown at the same azimuth angle. In the 8 a) to c ) represent echoes of different orders, first, second and third. In 8 a) is a vertical continuous road marking 17th shown, with the continuous lane marking 17th can be recognized as the most distant or second echo. In 8 a) make the circles 18th First order echoes and the solid points 19th Second order echoes. In 8 b) are two vertical road markings 17th and the echoes 18th , 19th first and second order shown. In 8 c) are three horizontal road markings 17th shown, the echoes 18th , 19th generate first, second and third order.

9 illustrates the entire logical sequence for performing the method according to one or more embodiments of the invention. The method of reorganizing sample points 16 in one of a detection system 1 point cloud generated by a vehicle starts with step 100 . In step 100 becomes a scene surrounding a vehicle by means of electromagnetic waves from a detection system 1 of the vehicle is measured and generates at least one scan of the scene with a plurality of sampling points 16 , where the total of all sample points 16 forms a point cloud together with the associated point data of the generated scan. Electromagnetic waves emitted by an object within a predefined polar angle of a spherical coordinate system of the detection system 1 are reflected, form a layer 14th of sampling points 16 , wherein electromagnetic waves reflected from an object at the same azimuthal angle for the predefined polar angle of the spherical coordinate system, a salvo 15th form.

In step 110 become the sampling points 16 to at least one chain 20th grouped, the grouping being done by a sample-point-to-chain mapping algorithm, each chain 20th a starting point 26th and an endpoint 28 having, the adaptation by the algorithm volley by volley starting from the volley 15th with index 0, with each chain 20th up to a sampling point 16 per volley 15th can accommodate.

If no chain 20th is present in step 140 a new chain 20th created and the new chain in step 150 a sampling point 16 assigned from the salvo with index 0 and the sampling point 16 as a starting point 26th in step 160 marked.

If a chain 20th is present, the chain becomes 20th in step 170 a sampling point 16 from the volley 15th assigned, the assignment being based on a distance between the given sampling point 16 from a step k and a last sampling point 16 from a step k-1 in the chain 20th based, being the chain 20th in step 200 a sampling point 16 is assigned when the distance between them is the lowest of the sample points 16 in the volley 20th and below a predefined allocation threshold 190 and the method with step 170 starts all over again.

If the distance exceeds the predefined allocation threshold value 180 , will be in step 210 the chain 20th for the sampling point 16 skipped and the last sampling point 16 as an end point 28 marked.

In step 220 becomes the chain 20th saved for further processing. In one embodiment of the invention, the chain 20th linearly stored in a memory. The memory can be, for example, the memory of a computing unit of the vehicle, wherein the computing unit can be, for example, a driver assistance system that receives the chains and stores them in a memory for further processing.

10 Fig. 10 illustrates an output of the algorithm according to an embodiment of the invention. The volley indices are on the X-axis in 10 shown. The iteration steps k and k-1 of the method and four detected chains 20th are shown. At iteration step k, the algorithm receives two sampling points as input and has two active chains 20th from iteration step k-1 to which these two new sampling points 16 can be assigned if the matching criteria are met, or a new chain is created 20th if the matching criteria are not met. The general idea is for each iteration step k, of each class, is used when assigning the new sampling points 16 only one sampling point 16 (last sampling point in the chain 20th ) is taken into account, the assignment being based on the MxN confusion matrix, where N is the number of active chains in step k and M is the number of new sampling points to be assigned 16 is. One sampling point 16 becomes a chain 20th allocated only if the distance between them is the lowest in the series and is below the allocation threshold. Chains can therefore be used to maintain the shapes of the road features 20th of sampling points 16 algorithms used to extract road features instead of the raw point cloud.

11 shows different examples of matching chains 20th when the procedure for reorganizing sample points 16 is carried out in a point cloud according to an embodiment of the invention. The solid lines in 11 a) to d ) represent the generated chains 20th represent, dashed lines 21 show volleys 15th with multiple sampling points 16 , with more than one echo recorded per photomultiplier. 11 a) and b) shows the initialization of a new chain 20th due to the presence of second echoes 19th . 11 c) shows a continuous chain with noise sample points assigned to the "stand-alone" chain. This can be the case with shifts 14th occur with a high level of electronic noise. In 11 d) and e) is the end of a chain 20th shown.

12th Figure 3 shows a chain store layout within a single layer. After a whole shift 14th was processed with the algorithm according to the invention and the sampling points 16 the corresponding chains 20th are assigned to the chains 20th stored linearly in a memory. The memory can be, for example, the memory of a computing unit of the vehicle, wherein the computing unit can be, for example, a driver assistance system that includes the chains 20th receives and stores for further processing in a memory. With this layout, an ambiguity between multiple echoes and possible features is resolved and feature extraction algorithms can be used in the further processing of the sampling points 16 all available sample points 16 consider.

List of reference symbols

1

Detection system

2

Laser diode

3

Laser beam deviation

4th

rotating mirror

5

projected laser light

6th

Areas covered by the photodetector

7th

Focusing lens

8th

Photodetector

9

Field of view

10

Direction of rotation of the mirror

11

Soil surface

12th

Layers that do not generate scan points from the road

13th

Layers that generate scan points from the road

14th

layer

15th

volley

16

Sampling point

17th

Road markings

18th

Sampling point with the furthest echo

19th

Sampling point with second echo

20th

Chain

21

Salvo with multiple sampling points

26th

Start sampling point

27

connecting sampling points

28

End sampling point

100

Measure a scene

110

Grouping of the sampling points

120

when there is no chain

130

when a chain exists

140

Create a chain

150

Assigning a sampling point to the chain

160

Mark the sampling point as the start

170

Assignment of the scanning point to the chain

180

Distance exceeds the predefined allocation threshold

190

Distance is below a predefined allocation threshold

200

Assignment of sampling point to chain if the distance is below a predefined assignment threshold

210

skip the chain and mark the last point as the end point

220

Save the chain

Claims (12)

Method for reorganizing scanning points (16) in a point cloud generated by a detection system (1) of a vehicle, comprising the following steps: - Measuring a scene by means of electromagnetic waves from a detection system (1) of a vehicle and generating at least one scan of the scene with a plurality of scan points (16), the entirety of all scan points (16) forming a point cloud together with the associated point data of the scan generated , wherein electromagnetic waves that are reflected by an object within a predetermined polar angle of a spherical coordinate system of the detection system (1) form a layer (14) of scanning points (16), wherein electromagnetic waves that are reflected by an object at the same azimuthal angle for the predetermined polar angle of the spherical coordinate system are reflected, form a volley (15), - Grouping of the sampling points (16) to form at least one chain (20), the grouping being carried out by a sampling point-to-chain assignment algorithm, each chain (20) having a starting point (26) and an end point (28), the Adaptation by the algorithm salvo (15) for salvo (15) is carried out starting from the salvo (15) with index 0, whereby each chain (20) can record up to one sampling point (16) per salvo (15), - If there is no chain (20), create a new chain (20) and assign a scanning point (16) from the volley (15) with index 0 to the new chain (20) and mark the scanning point (16) as the starting point ( 26), - If a chain (20) is present, assigning a scanning point (16) from the volley (15) to the chain (20), the assignment being based on a distance between the given scanning point (16) of a step k and a last scanning point (16) is based on a step k-1 in the chain (20), the chain (20) being assigned a sampling point (16) if the distance between them is the lowest of the sampling points (20) in the volley (15) , and is below a predefined allocation threshold, - Skipping the chain (20) for the scanning point (16) if the distance exceeds the predefined allocation threshold value, and marking the last scanning point (20) as the end point (28), - Saving the chain (20) for further processing. Procedure according to Claim 1 wherein the assignment of a sample point (16) to a chain (20) is based on an MxN confusion matrix, where N is the number of active chains (20) at step k and M is the number of new sample points (16) that are to be assigned to the chain. Method according to one of the preceding claims, characterized in that the distance is a radial distance between the scanning points (16). Method according to one of the preceding claims, characterized in that the point cloud is a filtered point cloud, a sampling point (16) being considered with a certain confidence level as a sampling point (16) which is reflected from a ground surface (11) in front of the detection system (1) . Method according to one of the preceding claims, characterized in that the step of storing the chain (20) for further processing comprises the step of linearly storing the chain (20) in a memory of a computer unit of the vehicle. A method according to any preceding claim, wherein the method comprises the step of providing the stored chain (20) to a road feature extraction algorithm. Method according to one of the preceding claims, characterized in that the detection system (1) is a LIDAR system. Procedure according to Claim 7 , characterized in that the azimuthal angle is defined by an angle of a mirror of the LIDAR system and the polar angle by a laser diode (2) or a photodiode position (8) of the LIDAR system in relation to the mirror (4). Driver assistance system for performing the method according to one of the Claims 1 to 8th . Driver assistance system according to Claim 9 , characterized in that the driver assistance system comprises a detection system. Driver assistance system according to Claim 10 , characterized in that the detection system comprises a LIDAR system. Computer program product, comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the method according to the Claims 1 to 8th executes.

Images