DE102010045511A1 - Method and device for detecting lane markings in an image - Google Patents

Abstract

The invention relates to a method for the detection of lane markings (FM1, FM2) in an image (B), the surroundings of a vehicle (1) being recorded in at least one image (B), lane markings (FM1, FM2) in the image ( B) are detected and a lane course is determined from the lane markings (FM1, FM2), shadow areas (S1, S2) present in the image (B) being taken into account when the lane markings (FM1, FM2) are detected. According to the invention, a direction of incidence of light (R) is determined and, on the basis of the direction of incidence of light (R), a probability is determined whether in the image (B) a shadow area (S1, S2) similar to a lane marking (FM1, FM2) and / or a lane marking (FM1, FM2) similar shadow gap (SL) results from a shadow cast by one or more objects (O1, O2) in the vicinity of the vehicle (1) due to a light (L) from a light source with the determined direction of incidence of light (R), with a plausibility check of as potential lane markings in the image (B) recognized structures is carried out and depending on the size of the probability as a shadow area (S1, S2) and / or shadow gap (SL) identified potential lane markings are discarded when determining the course of the lane. The invention further relates to a device (2) for the detection of road markings (FM1, FM2) in an image (B).

Description

The invention relates to a method for detecting lane markings in an image, wherein an environment of a vehicle is detected in at least one image, wherein lane markings are detected in the image and from the lane markings a lane course is determined, wherein upon detection of the lane markings in the image existing Shadow areas are taken into account.

The invention further relates to a device for detecting lane markings in an image, comprising at least one image capture unit for detecting an environment of a vehicle in at least one image, comprising an evaluation unit for detecting the lane markings in the image, for determining a lane course from the lane markings and for consideration of existing in the image shadow areas in the detection of the lane markings.

From the US 2010/0041714 A1 For example, a method and an apparatus for detecting lane markings on a roadway on which a vehicle is moving are known. By means of an image acquisition unit arranged on the vehicle, images of the surroundings of the vehicle are acquired. The captured images are normalized for brightness before the lane markers within the images are detected in main directions. In each main direction, possible candidates for lane markings are subsequently identified in the images by means of filters using edge detection and line detection, and the lane markings are identified from the candidates by means of a transformation process. Road markings in sharp turns are distinguished from various shadows of buildings and trees as distracting factors.

Furthermore, from the US 2006/0274917 A1 a method for reducing the optical impact of shadows in the processing of images in a video-based traffic detection device known. In the method, first an image area with a relatively constant brightness is determined. Subsequently, this area is assigned to a shadow area and the optical influence of the image area is removed from the image. A delineation of the shadow area from an environment takes place in particular by detecting edges between the shadow area and the surroundings.

Furthermore, the describes US Pat. No. 6,819,779 B1 a method for detecting lane markings in an electronically stored image. The image is captured by an image capture unit, displays at least one lane marker and its surroundings, and is stored as an array of pixels. Edge detection is used to generate a list of characteristic edges that shape the image. From the list, pairs of edges belonging to lane markings are filtered according to predetermined criteria. These pairs of edges are sorted into lists that belong to individual lane markings. These individual lane markings, which respectively define a left and a right border of a lane, are identified geometrically. Three-dimensional curves are placed along the right and left lane markings to form a lane course. Additionally, shadows are removed from the images to detect the lane markers.

The DE 103 38 764 B4 discloses a method of activating a night vision system in a vehicle. In a first step, a current position of the vehicle is determined before, in a second step, a position of the sun relative to the vehicle is determined using the determined position data of the vehicle. In a third step, a need is identified to assist a driver of the vehicle by a method of improving visibility in the vehicle, determining the need for assistance using the detected position of the sun relative to the vehicle. The need to improve visibility is derived from whether the sun is above or below the horizon in relation to the vehicle.

In addition is from the DE 103 38 766 B4 a vehicle data bus system with a data bus, wherein over the data bus several connected bus subscribers are in data transmission connection. Furthermore, a locating unit embodied as a bus subscriber is provided, which is set up for determining at least vehicle position data and for outputting the determined vehicle position data on the data bus, the locating unit having a locating sensor system comprising at least one GPS receiver. The locating unit is set up in addition to the determination of sun position data using the determined vehicle position data and outputting the ascertained sun position data on the data bus to derive from the direction of solar radiation into the vehicle, which type of visual obstruction for the driver by the sun is possible. In addition, as a further bus subscriber, means for receiving sun position data from the data bus for determining the necessity of assisting a driver of the vehicle by activating an arrangement for improving the visibility in a vehicle is provided. The arrangement for improving the visibility in a vehicle includes a camera for observing a solid angle range and a display device in the interior of the vehicle for displaying the images of the camera.

The invention has for its object to provide a comparison with the prior art improved method for detecting lane markings in a picture. Furthermore, the invention has for its object to provide an improved device for the detection of lane markings in a picture.

With regard to the method, the object is achieved by the features specified in claim 1 and in terms of the device by the features specified in claim 8.

Advantageous embodiments of the invention are the subject of the dependent claims.

In a method for detecting lane markings in an image, an environment of a vehicle is detected in at least one image, wherein lane markings are detected in the image and a lane course is determined from the lane markings, wherein shadow regions present in the detection of the lane markings in the image are taken into account.

According to the invention, a light incidence direction and, based on the direction of light incidence, a probability are determined as to whether a shadow area similar to a lane marking in the image and / or a shadow gap similar to a lane marking result from a shadow cast of one or more objects in the surroundings of the vehicle due to a light beam of a light source with the detected light incidence direction result in a plausibility of structures recognized as potential lane markers in the image is performed and depending on the size of the probability identified as a shadow area and / or shadow gap potential lane markings are discarded in the determination of the lane course.

This results in the advantage that the detection and detection of lane markings in the image, d. H. in the vicinity of the vehicle, and reliably distinguishes potential shadow markings from shadow areas and / or shadow gaps from real road markings. This results in a particularly advantageous manner that false warnings and / or incorrect interventions are avoided when using the method to assist a driver in the tracking of his vehicle.

The direction of light incidence is determined in particular relative to the vehicle and / or relative to a cardinal direction, for example the sky direction "north". The sense of direction of the light incidence direction is made up of a horizontal angle, i. H. the so-called azimuth, and a vertical angle, d. H. the so-called elevation, together.

In a preferred embodiment of the method according to the invention, a sun position is determined on the basis of a position of the vehicle, an orientation of the vehicle relative to a cardinal direction, a calendar date and / or a current time when determining the light incident direction. This determination is characterized on the one hand by the fact that it is particularly simple and feasible without much additional effort. On the other hand, the position of the sun can be determined very precisely.

Furthermore, it is checked according to an embodiment of the method according to the invention, whether there are one or more lane markings in the shadow area and / or the shadow gap. In shadow areas, lane markings are more difficult to detect due to the reduced brightness and consequent reduction in contrast between the lane markings and a road surface. For this complex procedures are required. Due to the detection of the shadow area and / or the shadow gaps, it is possible to carry out these complex procedures only in limited areas of the image, ie the shadow areas, and thus to reduce the effort in determining the lane markings with increasing reliability.

In order to determine whether the light source, in particular the sun, is obscured by clouds or objects located in the surroundings of the vehicle, it is preferable to determine a luminous intensity that is taken into account in the plausibility check of the potential lane markings. Thus, the detection of the lane markings is further improved. In particular, it is also taken into account that with less intensity of the light source of the light source, fewer shadow areas and / or shadow gaps are formed or formed less pronounced, so that their contrast to unimpeded illuminated areas or to the respective shadow area is reduced.

In order to further determine quickly and reliably whether the determined potential lane markings result from a shadow and / or a shadow gap of one or more objects, in determining the potential lane markings and / or plausibility is preferably a size and / or shape of the objects and / or a distance of the at least one image acquisition unit to the objects taken into account.

According to an expedient embodiment of the method according to the invention, when determining the potential lane markings and / or plausibility check, a street category of a road on which the vehicle is moving is taken into account. Thus, it is possible that a likelihood of the presence of certain objects, such as crash barriers and / or guide posts, which generate certain shadow areas and / or shadow gaps, is included in the determination and / or plausibility check. For example, the probability of guardrails on highways is higher than on local roads. Thus, the likelihood on highways is higher than on local roads, that due to a shadow of the guardrail shadow areas and / or shadow gaps are determined as potential lane markings.

Alternatively or additionally, it is considered whether the vehicle is located on, below, above and / or within one or more structures for guiding traffic routes. This embodiment in turn leads to the advantage that a probability of the presence of certain objects, such as crash barriers and / or guide posts, can be included in the determination and / or plausibility check. In this case, a probability for the presence of crash barriers on bridges is particularly high. By contrast, when driving through tunnels, it is unlikely or impossible for shadow areas and / or shadow gaps created by sunlight to be recognized as potential lane markings.

In a development of the method, it is determined on the basis of a terrain model of a digital map of a navigation device and the determined light incidence direction and / or a city model of a digital map of a navigation device and the determined direction of light incidence whether terrain structures, plants and / or in a line of sight between a vehicle position and the light source or structures, taking into account the identified off-road structures, plants and / or structures in determining the potential lane markings and / or the plausibility check. This makes it possible to determine in a simple and rapid manner whether terrain structures, plants and / or structures are present in the surroundings of the vehicle, which generate interpretable shadow areas and / or shadow gaps as road markings depending on the light incident direction or an area in which the vehicle is completely shaded, so that the probability of such misinterpretable shadow areas and / or shadow gaps is very low.

The device for detecting lane markings in an image comprises at least one image capture unit for detecting a surroundings of a vehicle in at least one image and an evaluation unit for detecting the lane markings in the image, for determining a lane course from the lane markings and for taking account of shadow regions present in the image the detection of lane markings.

According to the invention, the evaluation unit is coupled to a detection unit for determining a direction of light incidence, whereby a probability can be determined by means of the evaluation unit based on the light incidence direction, whether a shadow area similar to a lane marking and / or a shadow gap similar to a lane marking from a light source of a light source By means of the evaluation unit, a plausibility check of structures identified as potential lane markings in the image and, depending on the size of the likelihood, a rejection of the potential lane markings identified as the shadow area and / or shadow gap contribute to the detected light incidence direction the determination of the lane course are feasible.

The device is characterized by a simple structure and it is in a particularly advantageous manner only a small effort to realize the function of the device required. Furthermore, in particular, a driver assistance system for assisting a driver of a vehicle in the tracking of the vehicle can be improved in its function, and false warnings and / or incorrect interventions are avoided or at least reduced.

The determination unit preferably comprises at least one light sensor for detecting a strength of a light beam or is coupled thereto. By means of the light sensor, which is designed for example as a photoresistor, photodiode, phototransistor, pyroelectric sensor, photocell and / or solar cell, the intensity of the light source of the light source can be determined very accurately and at the same time simply and with little effort. Thus, it can be determined whether, for example, the sun is hidden as a light source of clouds or objects or the light is unhindered.

In a particularly preferred embodiment of the device according to the invention, the evaluation unit is provided with a locating unit, a navigation device and / or a receiving unit for receiving position data of the vehicle Calendar date, a time and / or an orientation of the vehicle coupled to a cardinal direction. Because of this design of the device, a position of the sun is very simple and can be determined very accurately. In particular, when using the device in modern vehicles with navigation devices, no additional arrangements are required for determining this data. Alternatively or additionally, the additional installation of the locating unit and / or the receiving unit for receiving position data of the vehicle is characterized by a low material and assembly costs.

Embodiments of the method are explained in more detail below with reference to drawings:

Showing:

1 schematically a vehicle with a device according to the invention for the detection of lane markings in an image, and

2 schematically an image of an environment of the vehicle according to 1 ,

Corresponding parts are provided in all figures with the same reference numerals.

In 1 is a vehicle 1 with a device according to the invention 2 for the detection of road markings FM1, FM2 in an in 2 shown image B. The road markings FM1, FM2 may be structural limitations of a 2 roadway FB described in detail and act on the road FB applied lines.

The device 2 includes an image capture unit 2.1 for detecting an environment of a vehicle 1 in at least one picture B. The device 2 further includes one with the image capture unit 2.1 coupled evaluation unit 2.2 to evaluate the image B.

In addition, the device 2 a determination unit 2.3 to determine an in 2 Light incident direction R of a light beam L closer to a light source not shown in more detail, wherein the detection unit 2.3 with the evaluation unit 2.2 is coupled. The determination unit 2.3 includes a light sensor 2.3.1 for detecting a strength of the light sheet L, wherein the light sensor 2.3.1 as a photoresistor, photodiode, phototransistor, pyroelectric sensor, photocell and / or solar cell is formed.

The evaluation unit 2.2 is further with a navigation device 3 coupled, wherein by means of the navigation device 3 a position POS of the vehicle 1 , a calendar date D, a time T and an orientation A of the vehicle 1 to a cardinal direction, for example a north orientation, to the evaluation unit 2.2 are transferable.

The device 2 for the detection of the lane markings FM1, FM2 is with a driver assistance device 4 to assist a driver of the vehicle 1 in a tracking of the vehicle 1 coupled. The driver assistance device 4 is designed such that an automatic and active tracking within lanes FS1, FS2 limiting road markings FM1, FM2 is done by means of steering interventions and / or that the driver warned before leaving the lane FS1, FS2 or lane FB by optical, acoustic and / or haptic signals becomes. For this purpose, from the road markings FM1, FM2 by means of the evaluation unit 2.2 determines a lane course.

2 shows by means of the image capture unit 2.1 taken picture B of the environment of the vehicle 1 , where the vehicle 1 on a carriageway FB with two lanes FS1, FS2 moved.

A right lane FS1 is bounded on the right side by a first lane marking FM1, which is formed as a solid line. On the left side, the right lane FS1 is bounded by a second lane marking FM2, the second lane marking FM2 simultaneously separating a left lane FS2 from the right lane FS1. The second lane marking FM2 is formed as a broken line. On the left side, the left lane FS2 is bounded by an object O1 designed as a guardrail, in particular by an obstacle designed as two guardrails arranged one above the other.

The detection of the lane markers FM1, FM2 in the image B of the surroundings of the vehicle 1 takes place by means of the evaluation unit 2.2 wherein the lane markers FM1, FM2 are detected in the image by well-known methods. These methods comprise, in particular, an edge detection and / or a determination of the road markings FM1, FM2 on the basis of a contrast ratio between the road markings FM1, FM2 and the roadway.

Due to the light beam L of the sun with the light incidence direction R, a shadow cast of the guardrail is produced in the left-hand area of the left lane FS2, so that a shadow area S1 is formed. Due to the non-entire surface of the guardrail this is illuminated in some areas of the light L, so that the shadow area S1 also not full surface is trained. Shadow gaps SL are formed in the shadow area S1.

The illustrated shadow gaps SL are similar in their arrangement and shape of the formed as a broken line pavement marking FM2. So that the shadow gaps SL from the evaluation unit 2.2 are not detected as lane markers FM1, FM2 and incorrect interventions and / or false warnings of the driver assistance device 4 lead, it is determined whether a light L with a light incidence direction R is present, which could cause the shadow area shown S1 with the shadow gaps SL.

This is the evaluation unit 2.2 with the investigative unit 2.3 coupled, by means of which the light incident direction R of the light L is determined. In the illustrated embodiment, a sun position is determined in determining the light incidence direction R, since the generated shadow area S1 and the shadow gaps SL arise due to the light L with the light incident direction R of the sun on the road FB. In addition, it is possible to detect light beams L and their light incidence directions R of other light sources, for example street lights and / or headlights.

The determination of the position of the sun is made by means of the evaluation unit 2.2 carried out, for which purpose one or more algorithms are provided, by means of which the sun's position, in particular in coordinates relative to the vehicle 1 or in coordinates relative to the earth is determined. The determination of the position of the sun is based on the position POS of the vehicle 1 , the calendar date D, the time T and the orientation A of the vehicle 1 to the direction "North", this data from the navigation device 3 to the evaluation unit 2.2 be transmitted. The transmission preferably takes place via a data bus (not shown) of the vehicle 1 , for example, a CAN bus or LIN bus. The determination of the position of the sun is in particular by means of one of DE 103 38 764 B4 known method and / or by means of one of the DE 103 38 766 B4 performed known method.

Alternatively to the determination of the position of the sun using the evaluation unit 2.2 this is done by means of the navigation device 3 , Where the sun is then in a manner not shown in coordinate form to the evaluation 2.2 is transmitted.

Based on the direction of light incidence R of the light L of the sun, ie the position of the sun, is determined by means of the evaluation unit 2.2 determines a likelihood, whether in the image of a road marking FM1, FM2 similar shadow area S1 and / or a road marking FM1, FM2 similar shadow gap SL are present, resulting from the generated due to the light sheet L with the detected light incidence direction R shadow of the guardrail. In other words, it is determined whether the sun is at an appropriate angle, ie, elevation, and a suitable direction, ie, azimuth, to produce shadow areas S1 and shadow gaps SL formed similarly to lane markers FM1, FM2.

In order to use the shadow area S1 and the shadow gaps SL as incorrectly identified road markings FM1, FM2 for operating the driver assistance device 4 to be avoided, first by means of the evaluation unit 2.2 all structures on the roadway FB, which could represent road markings FM1, FM2, deposited as potential lane markings.

These potential lane markings become one by means of the evaluation unit 2.2 plausibility check carried out, it being determined based on the plausibility check whether the structures recognized as potential lane markings in the image B lanes road markings FM1, FM2 or shadow areas S1 and / or shadow gaps SL. Depending on the determined size of the probability identified as shadow areas S1 and / or shadow gaps SL potential lane markings are discarded and not for the operation of the driver assistance device 4 used.

To the reliability of the device 2 in the determination of the lane markings FM1, FM2 and the detection of shadow areas S1 and shadow gaps SL is further improved when determining the position of the sun, first by means of the light sensor 2.3.1 a strength of the light L determined. The light intensity is used to determine whether the sun is obscured by clouds or other objects, or whether the surroundings are illuminated unhindered.

The objects are terrain structures, plants and / or structures. Based on a terrain model of a digital map of the navigation device 3 , a city model of the digital map of a navigation device 3 and the determined light incidence direction R determines the evaluation unit 2.2 whether terrain structures, plants and / or structures are located in a line of sight between a current vehicle position and the light source, which surrounds the vehicle 1 shadow and thus lead to low levels of light L. In particular, it is determined whether on a road side, from whose direction the Sun shines in, forest is located, which can lead to Schattenwürfen.

If an object is in the line of sight between the vehicle position and the sun and / or the sun is obscured by clouds, the intensity of the light L is low. At low magnitudes of the light L, it is unlikely that a contrast between the shadow gaps SL and the shadow area S1 is sufficiently large that the evaluation unit 2.2 the shadow gaps SL identified as lane markers FM1, FM2. This is taken into account in the plausibility check of the potential lane markings, whereby if a lumen intensity limit L is exceeded, potential lane markings are recognized as real lane markings FM1, FM2.

Alternatively or additionally, in the case of the plausibility check, a size and shape of the object O1 designed as a guardrail and a distance of the image capture unit are further determined 7 taken into account for the object O1. For this purpose, the guardrail by means of the evaluation unit 2.2 measure and the distance to the image acquisition unit 7 is determined. From the size, shape, the distance and the determined light incidence direction R, it is determined whether the object O1 can produce a shadow, from which a shadow area S1 with shadow gaps SL arises, which can be interpreted as lane markings FM1, FM2. Depending on whether the object O1 can produce such a shadow, the size of the probability is adjusted.

Also alternatively or additionally, the plausibility check is a street category of a road on which the vehicle 1 moved, taken into account. Depending on the road category, it is estimated how likely it is that an object O1 generating a shadow area S1 and / or at least one shadow gap SL, such as the guardrail, is on the busy road. Because guardrails are very common on motorways, but not on local roads, on motorways it is more likely that a shadow area S1 caused by crash barriers and / or a shadow gap SL will be identified as lane markers FM1, FM2 than on municipal roads. The data of the road category is preferably from the navigation device 3 to the evaluation unit 2.2 transfer. Depending on the results derived from the road category, the probability of a plausibility check is adjusted again.

Furthermore, it is preferably considered whether the vehicle 1 located on, below, above and / or within one or more structures for traffic routes. In particular, it is detected whether the vehicle 1 is located on a bridge, since the probability of crash barriers, which cause the shadow areas S1 and shadow gaps SL shown, is particularly high on bridges. Furthermore, it is detected in particular whether the vehicle 1 moved in a tunnel. In a tunnel, shadow areas S1 and shadow gaps SL are unlikely to be generated by a sunshine L generated by the sun. Rather, a light L is generated in a tunnel of artificial light sources. The data, whether the vehicle 1 on, below, above and / or within one or more of such structures are again preferably by the navigation device 3 provided and to the evaluation unit 2.2 transfer.

It is further preferably checked whether shadow areas S2 are generated by an object O2 moving with the vehicle. In the illustrated embodiment, object O2 is a truck. Since detection of the lane markers FM1, FM2 is difficult in this shadow area S2 and the shadow area S1 generated by the guardrail due to reduced contrasts between lane markers FM1, FM2 and the lane FB, adjustment of optical parameters of the image acquisition unit becomes 2.1 , in particular contrast settings, adapted to the properties environment in the shadow areas S1, S2. Furthermore, the detection of the lane markings FM1, FM2 in the shadow areas S1, S2 is preferably carried out alternatively or additionally in a more detailed and complex manner in order to enable a reliable detection of the lane markings FM1, FM2.

LIST OF REFERENCE NUMBERS

1

vehicle

two

contraption

2.1

Image capture unit

2.2

evaluation

2.3

determining unit

2.3.1

light sensor

3

navigation device

4

Driver assistance device

A

alignment

B

image

D

calendar date

FB

roadway

FM1

road marking

FM2

road marking

FS1

lane

FS2

lane

L

glow

O1

object

O2

object

POS

position

R

Light direction

S1

shadow area

S2

shadow area

SL

shadow gap

T

time of day

QUOTES INCLUDE IN THE DESCRIPTION

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

Cited patent literature

• US 2010/0041714 A1 [0003]
• US 2006/0274917 A1 [0004]
• US 6819779 B1 [0005]
• DE 10338764 B4 [0006, 0043]
• DE 10338766 B4 [0007, 0043]

Claims (10)

Method for detecting lane markings (FM1, FM2) in an image (B), wherein an environment of a vehicle ( 1 ) is detected in at least one image (B), wherein lane markings (FM1, FM2) in the image (B) are detected and from the lane markings (FM1, FM2) a lane course is determined, wherein in the detection of the lane markings (FM1, FM2 ) in the image (B) existing shadow areas (S1, S2) are taken into account, characterized in that a light incidence direction (R) is determined and that on the basis of the direction of light incidence (R) a probability is determined whether in the image (B) of a lane marking ( FM1, FM2) similar shadow area (S1, S2) and / or a road marking (FM1, FM2) similar shadow gap (SL) from a due to a light (L) of a light source with the detected light incidence direction (R) generated shadow cast one or more objects (O1, O2) in the vicinity of the vehicle ( 1 ) and that a plausibility check of structures recognized as potential lane markings in the image (B) is carried out, wherein potential lane markings identified as shadow area (S1, S2) and / or shadow gap (SL) depend on the size of the likelihood Lane course be discarded. A method according to claim 1, characterized in that when determining the direction of light incidence (R) a position of the sun based on a position (POS) of the vehicle ( 1 ), an orientation (A) of the vehicle ( 1 ) relative to a direction, a calendar date (D) and / or a current time (T) is determined. A method according to claim 1 or 2, characterized in that it is checked whether one or more lane markings (FM1, FM2) are in the shadow area (S1, S2) and / or the shadow gap (SL). Method according to one of the preceding claims, characterized in that a strength of the light (L) is determined. Method according to one of the preceding claims, characterized in that when determining the potential lane markings and / or plausibility a size and / or shape of the objects (O1, O2) and / or a distance of the at least one image capture unit ( 2.1 ) is taken into account for the objects (O1, O2). Method according to one of the preceding claims, characterized in that in the determination of the potential lane markings and / or plausibility check a street category of a road on which the vehicle ( 1 ), takes into account and / or takes into account whether the vehicle ( 1 ) located on, below, above and / or within one or more structures for traffic routes. Method according to one of the preceding claims, characterized in that based on a terrain model of a digital map of a navigation device ( 3 ) and the determined light incidence direction (R) and / or a city model of a digital map of a navigation device ( 3 ) and the determined direction of light incidence (R) is determined whether in a line of sight between a position (POS) of the vehicle ( 1 ) and the light source terrain structures, plants and / or structures, wherein the determined terrain structures, plants and / or structures are taken into account in the determination of the potential lane markings and / or plausibility. Contraption ( 2 for detecting lane markings (FM1, FM2) in an image (B), comprising at least one image capture unit ( 2.1 ) for detecting an environment of a vehicle ( 1 ) in at least one image (B), comprising an evaluation unit ( 2.2 for detecting the lane markings (FM1, FM2) in the image (B), for determining a lane course from the lane markings (FM1, FM2) and for taking into account the shadow areas (S1, S2) present in the image (B) in the detection of the lane markings (FM1, FM2), characterized in that the evaluation unit ( 2.2 ) with an identification unit ( 2.3 ) is coupled to determine a direction of light incidence (R), wherein by means of the evaluation unit ( 2.2 ) and based on the direction of light incidence (R) a probability can be determined whether in the image (B) a shadow area (S1, S2) similar to a lane marking (FM1, FM2) and / or a shadow lane (SL) similar to a lane marking (FM1, FM2) from a shadow cast of one or more objects (O1, O2) in the surroundings of the vehicle generated by a light beam (L) of a light source with the determined light incident direction (R) ( 1 ) and that by means of the evaluation unit ( 2.2 ) a plausibility of recognized as potential lane markers in the image (B) structures and depending on the size of the probability of a rejection of the shadow area (S1, S2) and / or shadow gap (SL) identified potential lane markers in determining the lane course are feasible. Contraption ( 2 ) according to claim 8, characterized in that the determination unit ( 2.3 ) at least one light sensor ( 2.3.1 ) for detecting a strength of the light (L) or is coupled thereto. Contraption ( 2 ) according to claim 8 or 9, characterized in that the evaluation unit ( 2.2 ) with a locating unit, a navigation device ( 3 ) and / or a receiving unit for receiving a position (POS) of the vehicle ( 1 ), a calendar date (D), a time (T) and / or an orientation (A) of the vehicle ( 1 ) is coupled to a cardinal direction.

Images