Method and device for determining and maintaining a lateral distance of a first vehicle to a second vehicle during an overtaking manoeuvre

Abstract

The invention relates to a method and a device for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking manoeuvre with a distance measuring unit (14) directed in a direction of travel of the first vehicle (10), a camera (16) directed in the direction of travel of the first vehicle (10), an image processing unit (18) and a computing unit (15), in which method by way of the distance measuring unit (14) an actual distance E(t) of the first vehicle (10) to the second vehicle (12) is continuously or cyclically measured, and by way of the camera (16) an image of the second vehicle (12) is continuously recorded, and in the image processing unit (18) a current actual offset between a lateral face (Z) of the second vehicle (12) closest to the middle of the road and the camera axis (X) is determined cyclically from the image data and from the determined actual distance E(t), and in the computing unit (15) an actual lateral distance S(t) is calculated from this actual offset. Furthermore, the distances determined by way of the camera (16) and the distance measuring unit (14) can be verified in order to improve the accuracy of the calculated lateral distance S(t).

Description

[0001] Method and device for determining and maintaining a lateral distance of a first vehicle to a second vehicle during overtaking

[0002] The invention relates to methods and a device for determining and maintaining a lateral distance of a first vehicle from a second vehicle during an overtaking maneuver, comprising a distance measuring unit directed in the direction of travel of the first vehicle, a camera directed in the direction of travel of the first vehicle, an image processing unit, and a computing unit.

[0003] Modern vehicles have a variety of assistance systems, such as distance control systems, lane change assistance systems,

[0004] Collision warning systems and / or lane departure warning systems are also available. These systems typically use both radar measurements and camera images, which can be evaluated by an image processing unit.

[0005] Various methods have become known which, in addition to distance measurement using a radar unit, which usually emits electromagnetic waves and receives the electromagnetic waves reflected from the vehicles ahead and then uses a computing unit to determine a distance using the detected time delay or frequency shift, also or alternatively carry out a distance measurement using a camera. The simplest way to measure a distance is by evaluating the images from a stereo camera, but methods have also become known with which a distance can at least be estimated using a mono camera by taking the camera's focal length into account. Width measurements can also be carried out using mono cameras by evaluating successive image sequences, or width measurements can be carried out by identifying known quantities in the image so that a conversion can be carried out.Various methods for measuring width using stereo cameras are also known, with which a three-dimensional image is created that contains the calculated dimensions of the bodies or surfaces present therein.

[0006] To protect other road users or due to legal changes, it may become necessary in the future to not only measure the distance to vehicles ahead or following, or to verify that a lane marking has been crossed, but also to specifically determine the lateral distance to vehicles being overtaken. In particular, it will be necessary to differentiate between vehicle types, i.e., to be able to distinguish between a single-track vehicle, such as a bicycle, and a two-track vehicle, such as a car, since different minimum lateral distances are required depending on the vehicle type, which also depends on whether the vehicle is located within or outside of a built-up area.

[0007] A corresponding determination of existing lateral distances using only forward-facing measuring units is currently unknown. In particular, there is no known way to verify determined lateral distances to ensure a high degree of certainty in maintaining a corresponding distance. This can be helpful for vehicle drivers, but is essential for autonomous vehicles to reliably comply with traffic regulations.

[0008] The task therefore arises to provide a method for determining and maintaining a lateral distance of a first vehicle to a second vehicle during an overtaking maneuver, with which on the one hand a lateral distance can be reliably determined with a high degree of accuracy and, if possible, predicted and on the other hand, in particular for autonomous vehicles, the determined values ​​of forward-facing measuring units can be verified in order to exclude errors in the measurement and consequently in the control of an autonomous vehicle.

[0009] This object is achieved by a method for determining and maintaining a lateral distance of a first vehicle from a second vehicle during an overtaking maneuver with the features of the main claim 1.

[0010] The first vehicle is typically a two-track vehicle, particularly a car, truck, bus, or other special-purpose vehicle, while the second vehicle can be a single- or two-track vehicle. This includes, in particular, bicycles, motorcycles, scooters, but also construction machinery or buses. The overtaking maneuver is considered a potentially possible overtaking maneuver and thus begins as soon as a vehicle ahead is detected. Alternatively, the process can also be initiated by activating the vehicle's turn signal or run continuously in the background.

[0011] To implement the method, the first vehicle has a distance measuring unit directed at least in the direction of travel of the first vehicle, via which vehicles traveling ahead can be detected. In addition, the first vehicle has a camera directed in the direction of travel of the first vehicle, via which image data from the space ahead can be continuously recorded and processed. These can be various types of cameras, such as mono cameras, stereo cameras, light field cameras, etc., which enable the digital data of each recorded image to be processed and evaluated in an image processing unit. Furthermore, the vehicle has a computing unit for calculating distances, lengths, widths, and the like from the acquired data.Of course, the image processing unit and the computing unit do not have to be two separate units, but can also be combined into a single unit. The computing unit is connected to the image processing unit and the distance measuring unit for data exchange. According to the invention, the distance measuring unit continuously or cyclically measures the shortest distance between the first vehicle and the second vehicle driving ahead. During an overtaking maneuver, this actual distance and thus the shortest connection between the distance measuring unit and the vehicle driving ahead is usually the left edge of the rear of the vehicle driving ahead or, before the overtaking maneuver, the rear of the vehicle. An image of the second vehicle is continuously recorded via the camera, and image data is determined from the images in the image processing unit cyclically, for example 10 times per second.This image data is transmitted to the computing unit, which uses the image data and the determined actual distance to determine the current actual offset of a side surface of the second vehicle facing the road center from the camera axis. From this determined actual offset, the computing unit calculates an actual lateral distance. Various image evaluation methods are known for this purpose. For example, the focal length of the camera can be used for the calculation to convert the offset of the second vehicle driving ahead from the offset shown in the image to the actual offset. Furthermore, consecutive image sequences can be used for this purpose, or known methods such as the computer stereo vision method for evaluating images from a stereo camera with a CCD sensor.

[0012] Accordingly, in the device according to the invention, the image processing unit serves to determine image data, and the computing unit is connected to the image processing unit to receive the image data and serves to calculate an actual

[0013] Offset of a side surface of the second vehicle furthest towards the road centre to the camera axis from the image data and the actual distance determined by the distance measuring unit and an actual

[0014] Side distance from the determined actual offset.

[0015] Preferably, the computing unit compares the determined actual lateral distances with the stored target lateral distances. Accordingly, the driver can either be advised to maintain an additional distance or, in the case of an autonomous vehicle, this difference can be used to correct the position of the first vehicle.

[0016] Preferably, an angle between the camera axis and a side surface of the second vehicle facing the road center is calculated in the computing unit from the continuously recorded image data of the image processing unit. From this, the actual lateral distance is continuously calculated in the computing unit by multiplying the actual distance determined by the distance measuring unit by the sine of the calculated angle and subtracting a lateral distance of the camera from a side of the first vehicle facing the roadside. It can be assumed that the accuracy of the angle calculated from the image data is relatively high, and thus the determined actual lateral distance is also highly accurate using the actual distance determined by the distance measuring unit, which also has a high degree of accuracy.

[0017] In a further embodiment, a stereo camera or a light field camera is used to determine the angle. Using these two cameras, the angle of the camera axis to the vehicle in front can be determined with good accuracy in a known manner. Of course, the angle to the central axis between the two cameras is calculated, and thus to the vehicle center if both cameras are positioned with the same offset from the vehicle center. A similar method for determining the angle can be achieved using a mono camera and optical deflection devices. Using the optical deflection devices, the mono camera can also capture and compare two images from different perspectives, and then calculate distances and angles in the same way as with a stereo camera.

[0018] A mono-camera and its focal length can also be used to determine the angle. Using the focal length of the mono-camera, it is then possible to convert the lengths of the image into real-world lengths, assuming a known number of pixels and resolution. In particular, the angle to the camera axis can be determined with high accuracy in this way.

[0019] It is particularly preferred if the distance measuring unit is a radar unit. A radar unit can determine the distance to the vehicle ahead with high accuracy.

[0020] In an alternative embodiment, the distance measuring unit is the camera, which is also used to determine the angle. In particular, a stereo camera can also be used to determine the distance to the vehicle ahead in a known manner. This is achieved by using the different perspectives created by the stereo camera to the second vehicle ahead in the image processing unit, for example, using computer stereo vision, to generate a three-dimensional image that allows distances to be calculated.

[0021] When using a mono camera, it is also possible to calculate the distance to the second vehicle from two consecutive images based on the difference in size of an object in the image whose size is known. For example, the size of a license plate or similar object can be used as a reference. In this case, both the distance and the angle can be determined using the camera alone.

[0022] It is particularly preferred if a second actual distance to the second vehicle is calculated from the data of the image processing unit in the computing unit and compared and verified with the first actual distance determined by the radar unit at an identical time. In this way, errors in calculating the distance and, consequently, also the angle and thus the actual lateral distance can be avoided.

[0023] For this purpose, especially when the actual distances determined by the image processing unit and the radar unit are identical, a correct measurement can be concluded and used to calculate the actual lateral distance. Such verification is particularly essential for autonomous vehicles to prevent errors in automatic steering that could lead to accidents or hazards, especially for overtaking cyclists.

[0024] In a preferred embodiment of the invention, the calculated actual lateral distance to the second vehicle is displayed in the first vehicle. Accordingly, if a driver detects that the displayed actual lateral distance is insufficient, they can intervene to correct the situation.

[0025] In an alternative preferred embodiment of the method, images from a monocamera are used, and in the image processing unit, an offset of the side surface of the second vehicle facing the road center relative to the camera axis is measured from image data. The calculation unit compares this offset with reference images from a map in which the actual offsets to the determined lateral offsets are stored for different distances. The actual offset A(t) is determined from this comparison. In this way, the necessary computing power is minimized, yet it is still possible to determine whether a sufficient lateral distance exists. Neural networks, for example, can also be used to learn these reference images.

[0026] The determined difference can be displayed or processed as a scaled deviation in the first vehicle. A positive deviation can confirm sufficient distance, while a negative indication indicates that additional distance is necessary. A driver of an overtaking vehicle can react accordingly to such a display, while an autonomous vehicle can be controlled accordingly until a zero deviation is reached, without having to learn specific measurements.

[0027] Furthermore, it is advantageous if the computing unit identifies the second vehicle as a single- or dual-track vehicle based on the data from the image processing unit. This is particularly possible through a simple width measurement or through object recognition, for example, via a neural network. The computing unit can store relevant data specified by the road traffic regulations, which may differ for two-wheeled vehicles or four-wheeled vehicles being overtaken. This ensures not only that an absolute target lateral distance is reliably maintained, but that it can also be adjusted depending on the vehicle being overtaken.

[0028] Furthermore, it is determined, preferably via a navigation device or traffic sign recognition, whether the first and second vehicles are located within or outside of a built-up area. This information is also used to set the target lateral distance, since a different lateral distance is legally required outside built-up areas and therefore at higher expected speeds. Of course, this minimum lateral distance can also be set depending on the speed of the overtaking vehicle. The various target lateral distances are stored in the processing unit accordingly for the various vehicle types and depending on the GPS or navigation data and are taken into account when comparing them with the calculated actual lateral distance.

[0029] It is particularly preferred if the determined actual lateral distances are stored over time and extrapolated to predict an expected actual lateral distance at the moment the second vehicle is reached. This creates a dynamic system that allows the lateral distance required for the actual overtaking maneuver to be gradually achieved depending on the front distance and can thus be used for trajectory planning in automated vehicles.

[0030] Furthermore, in a further embodiment, target lateral distances are stored in the computing unit depending on the distance determined by the radar unit for single-lane and dual-lane vehicles. If the predicted expected actual lateral distance is less than the permissible target lateral distance, a warning signal is issued to the driver of the first vehicle, or the vehicle intervenes independently to correct the situation by increasing the actual lateral distance or by braking. Thus, the overtaking maneuver is continuously monitored even before the first overtaking vehicle pulls out. The necessary data can then be determined and compared over the entire duration of the overtaking maneuver initiated by this, in order to ensure a smooth overtaking maneuver.

[0031] When correcting or issuing the warning signal to the driver, the current speed of the vehicle can also be advantageously taken into account, so that at higher speeds, the overtaking maneuver can be initiated with a greater distance. This also means that the difference in speed between the first and second vehicles is taken into account, which is crucial for the timing and optimal driving path until the second vehicle is reached, and thus the actual lateral distance required there.

[0032] A method is thus proposed for determining and maintaining a lateral distance between a first vehicle and a second vehicle during an overtaking maneuver. This method ensures that a required safe lateral distance is maintained when overtaking for different vehicle types, speeds, and locations. The data obtained during the calculation are verified to exclude errors in determining the lateral distance. By predicting the expected lateral distance upon reaching the vehicle in front, the driver or the vehicle's automatic control system can intervene in a timely manner to ensure a smooth overtaking maneuver.

[0033] An embodiment of the method according to the invention is described below with reference to the figures.

[0034] Figure 1 shows a schematic diagram in plan view, in which the distances and angles to be determined are entered.

[0035] Figure 1 shows a first vehicle 10 with a width B. It is a two-track vehicle in the form of a car. A second vehicle 12 in the form of a bicycle is located in front of this first vehicle 10 in the direction of travel. If the driver of vehicle 10 now wishes to overtake the second vehicle 12, they must maintain a required lateral distance in accordance with legal requirements while overtaking. This distance depends on both their own speed and their location, in particular whether they are within or outside a built-up area, as well as on the type of second vehicle 12 traveling in front.

[0036] For this purpose, the first vehicle 10 has a distance measuring unit 14 directed in the direction of travel or straight forward, which is connected to a computing unit 15 and continuously measures a distance E(t) from the second vehicle 12. This distance E(t) is determined over time and stored in the computing unit 15.

[0037] Furthermore, a digital camera 16 is arranged in the first vehicle 10, the camera axis X of which also points straight forward and thus in the direction of travel and which can be designed as a stereo camera 16.1, a light field camera or a mono camera 16.2 and is connected to an image processing unit 18, which also has a data connection to the computing unit 15.

[0038] When using a stereo camera 16.1, the different perspective created by the stereo camera 16.1 relative to the second vehicle 12 traveling ahead can generate a three-dimensional image in the image processing unit 18, for example using computer stereo vision. This image also makes it possible to calculate sizes, lengths, and thus distances. Thus, a distance E'(t) can be determined in the image processing unit 18 using the stereo camera 16.1. In addition, the available image data also makes it possible to determine with good accuracy an actual offset A(t) of a side surface Z of the vehicle 12 traveling ahead, which is furthest towards the center of the road, from the camera axis X from the available image data, which usually points forward, i.e., in the direction of travel. From this actual offset A(t), an actual lateral distance S(t) to the second vehicle 12 can also be calculated.This actual lateral distance S(t) is determined from the image data in the image processing unit 18 via the angle a(t) between the camera axis X, in this case the axis which is arranged centrally between the two apertures of the stereo camera 16.1, and the second vehicle 12.

[0039] Alternatively, a mono camera 16.2 can be used to calculate the actual lateral distance S(t). The distances E'(t) and the actual lateral distances S(t) can either be determined using known estimation algorithms or calculated using the focal length f and the angle β, which can be determined quite accurately using the focal length f. Nevertheless, these estimates are usually not as accurate as the distance determined using a stereo camera or light-field camera.

[0040] In the simplest form of the method according to the invention, the camera 16 thus serves both to determine an angle a(t) and the distance E(t) by setting E'(t) equal to E(t) and thus as a distance measuring unit 14. Different cameras can also be used, via which a verification of the distance measurement can be carried out by comparing the determined values ​​E(t) and E'(t) with one another in the computing unit 15.

[0041] Alternatively, to determine the distance E(t) in a known manner, a radar unit is used as the distance measuring unit 14, which in Figure 1 has the same line of sight X as the camera 16. The values ​​of this radar unit can either be used alone to determine the distance E(t), since these are usually very accurate, or they can be used in addition to the values ​​for the distance E'(t) determined via the image data of the camera 16 to verify it.

[0042] Naturally, the distance measuring unit 14 and also the camera 16 always detect all objects 0 present in the detected area, so that a multitude of data over time (E(t), E'(t), a(t), O(i), S(t)) can be determined and stored for the measurement by the camera 16 and for the measurement by the radar unit 14. The data used for this purpose are cyclically updated by subsequent measurements. In the process, the objects O1, O2, etc. are also identified in the image processing unit 18 or the computing unit 15, i.e., assigned to a vehicle type and, for example, a direction of travel.For example, based on the available views of the camera 16, the computing unit 15 or the image processing unit 18 can identify whether the vehicle 12 is a single-track or a dual-track vehicle based on the width of the vehicle 12, and the data from the radar unit 14 and the vehicle's own speed can also be used to determine the direction in which the detected object 0 is moving, i.e., whether it is an oncoming vehicle 12 or a vehicle traveling in the same direction. However, the image processing unit 18 can also assign additional identification parameters, resulting in a list of detected objects of the type [(0'1, E'1, a1, Typ1), (0'2, E'2, a2, Typ2),....)].

[0043] When determining the distance and angle to a cyclist traveling ahead as a second vehicle 12, when using a radar unit as the distance measuring unit 14, it must be taken into account in this comparison that the angle a(t) from the image processing unit 18 and the distance E(t) from the radar measurement are directed to the same point. If a 3D radar measurement is used, the distance can actually be measured to the side surface Z of the bicycle furthest towards the center of the road, which is usually formed by the left handlebar grip. Thus, the side surface Z of the camera 16 furthest towards the center of the road also corresponds to that of the radar unit 14. With a 2D radar unit, however, it may be necessary to correct the position of the point either during image analysis or during distance determination by the radar unit 14. This is because the 2D radar unit 14 usually identifies a shortest distance.Accordingly, the reference point of the camera 16 in the image is the side surface Z of the second vehicle 12 facing furthest towards the road center, which is used in the calculation, while the radar measurement identifies the rear area of ​​the vehicle 12. For two-lane vehicles 12 traveling ahead, this is usually approximately the same, whereas for a bicycle as the second vehicle 12, the radar measurement of the distance measuring unit 14 determines a distance to the nearest area pointing towards the first vehicle 10, for example to the tail light or rear tire, which, however, appears further to the right in the image for the camera 16 than, for example, the handlebar part of the bicycle projecting to the left.Accordingly, either the second vehicle 12 must be identified as a bicycle, and the rear part of the bicycle must be identified in the image processing unit 18 and the computing unit 15, and the angle measurement on the image must then be performed at this position in order to be able to compare the distances E(t) and E'(t), or the distance from the radar measurement must be corrected accordingly, for example, by adding a usual distance between the rear part of the bicycle and the handlebars. To calculate the correct actual lateral distance S(t), in the first case, a usual width difference between the middle part of the bicycle and the part protruding to the left must be taken into account and added.

[0044] If a verification of the distance measurements is carried out, for example, only the data of the distance measurements for which a difference between E(t) and E'(t) is below a threshold value of, for example, 5% of E(t) are used to calculate the actual lateral distance S(t). If this is the case, the presumably less accurate measurement E'(t) can be replaced by the more accurate measurement E(t), resulting in the object entries (0'1, El, a1, Typl) for the second vehicle 12. Regardless of any verification performed, the corresponding actual lateral distance S can then be calculated in the computing unit 15 for each relevant object from E(t) and a(t) and the vehicle geometry from S(t) = E(t) * sin (a(t)) - B / 2, where B / 2 represents the distance of the camera to the right outer side of the vehicle. In the case of 2D radar measurement, half the width b / 2 of the bicycle must be subtracted.As already described above, it is important to ensure that E(t) and the angle a(t) point to the same reference point. This means that for a bicycle, half the width of the bicycle (b / 2) must be subtracted if the reference point is at the rear of the bicycle. When using 3D radar sensors, the distance to the left outer side of the bicycle, in particular the left part of the handlebars or the cyclist's hand, can be determined directly. This means that the image analysis and the radar analysis refer to the same point, thus determining the angle a directly from the image, eliminating the need for further deductions when calculating the lateral distance.

[0045] Both the determined shortest distances E(t), which typically change during overtaking and are continuously recorded, for example, 10 times per second, and the calculation of the actual lateral distance S(t) are performed cyclically. During the overtaking maneuver, these data change due to the changing distances E(t) and S(t) caused by steering movements and different speeds. This results in a curve S(t) that also allows the calculation of a continuously updated prediction of the actual lateral distance S(t) when vehicle 10 and vehicle 12 are at the same height. This can be achieved, for example, by extrapolation.

[0046] Thus, for example, drivers can be continuously informed by displaying a current actual lateral distance or the actual lateral distance to be expected upon reaching vehicle 12 if this is less than a stored, prescribed target lateral distance Ssoii(t), and can react accordingly if the target lateral distance according to the road traffic regulations is not maintained. For autonomous vehicles, a consistent overtaking maneuver with safe distances can also be automatically calculated in this way. In addition, position measurement using the navigation device or existing sign recognition can be used to take into account whether vehicles 10, 12 are moving within or outside of a built-up area, so that the lateral distance can be correctly set depending on the position of the vehicles.

[0047] The described method for determining and maintaining a lateral distance between a first vehicle and a second vehicle during an overtaking maneuver can ensure that a prescribed safety distance can be correctly calculated and verified for different vehicle types.

[0048] It should be clear that the claimed method can be implemented in different ways. For example, either a single camera can be used to determine the distances and angles and, consequently, the lateral distance, or multiple cameras can be used, allowing verification of the calculated distance. Alternatively, a radar unit can be used to measure the distance, the measured values ​​of which can either be used directly to calculate the lateral distance as a distance or to verify the determined distance, since radar measurements are very accurate. Especially when using a monocamera, whose values ​​are usually less accurate when determining the distance because they are estimates, it is recommended to use a radar unit as the distance measuring unit to achieve greater accuracy.

Claims

P A T E N T A N S P R Ü C H E 1. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver, comprising a distance measuring unit (14) directed in the direction of travel of the first vehicle (10), a camera (16) directed in the direction of travel of the first vehicle (10), an image processing unit (18) and a computing unit (15), characterized in that an actual distance E(t) of the first vehicle (10) to the second vehicle (12) is measured continuously or cyclically via the distance measuring unit (14),an image of the second vehicle (12) is continuously recorded via the camera (16), and image data is cyclically determined in the image processing unit (18) and transmitted to the computing unit, and in the computing unit (15) using the determined actual distance E(t) and the determined image data, a current actual offset A(t) of a side surface (Z) of the second vehicle (12) facing furthest towards the road centre to the camera axis (X) is determined, and from this actual offset A(t) in the computing unit (15) an actual lateral distance S(t) is calculated.

2. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 1, characterized in that in the computing unit (15) the determined actual lateral distances S(t) are compared with stored target lateral distances Ssoii.

3. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) from a second vehicle (12) during an overtaking maneuver according to claim 1 or 2, characterized in that an angle a(t) between the camera axis (X) and the side surface (Z) of the second vehicle (12) facing the road center is calculated in the computing unit (15) from the continuously recorded image data of the image processing unit (18), and the actual lateral distance S(t) is continuously calculated therefrom in the computing unit (15) by multiplying the actual distance E(t) of the side surface (Z) of the second vehicle (12) facing the road center, determined by means of the distance measuring unit (14), by the sine of the calculated angle a(t), and subtracting a lateral distance (B) of the camera (16) from a side of the first vehicle (10) facing the road edge.

4. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 3, characterized in that a stereo camera (16.1) or a light field camera is used to determine the angle a(t).

5. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 3, characterized in that a mono camera (16.2) and optical deflection means are used to determine the angle a(t).

6. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 3, characterized in that a mono camera (16.2) and the focal length (f) of the camera (16) are used to determine the angle a(t).

7. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, characterized in that the distance measuring unit (14) is a radar unit.

8. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of claims 1 to 7, characterized in that the distance measuring unit (14) is the camera (16), which is also used for angle determination.

9. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 8, characterized in that the distance E'(t) to the second vehicle (12) is calculated from two successive images of the mono camera (16.1) on the basis of the difference in size of an object (0) in the image, the size of which is known.

10. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during overtaking according to claims 7 and 8, characterized in that a second actual distance E'(t) to the second vehicle (12) is calculated from the data of the image processing unit (18) in the computing unit (15) and is compared and verified with the first actual distance E(t) determined by means of the radar unit (14) at an identical time (t).

11. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 10, characterized in that when the actual distances E(t), E'(t) are determined to be the same, a correct measurement is concluded from the image processing unit (18) and the radar unit (14) and this is used to calculate the actual lateral distance S(t).

12. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, characterized in that the calculated actual lateral distance S(t) to the second vehicle (12) is displayed in the first vehicle (10).

13. Method for determining and maintaining a lateral distance of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 1, characterized in that the images of a mono camera (16.2) are used and in the image processing unit (18) an offset of the side surface (Z) of the second vehicle (12) facing the middle of the road to the camera axis (X) is measured on the image from image data and in the computing unit (15) with reference images from a characteristic map in which different distances E(t) the actual offsets are compared to the determined side offsets stored in the image, and from this comparison the actual offset A(t) is determined.

14. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 13, characterized in that the determined difference is displayed or processed as a scaled deviation in the first vehicle (10).

15. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, characterized in that the second vehicle (12) is identified as a single-track or two-track vehicle from the data of the image processing unit (18) via the computing unit (15).

16. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, characterized in that a navigation device or traffic sign recognition is used to determine whether the first vehicle (10) and the second vehicle (12) are located inside or outside a built-up area.

17. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, characterized in that the determined actual lateral distances S(t) are stored over time and extrapolated in order to predict an expected actual lateral distance S(t) for the moment of reaching the second vehicle (12).

18. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to claim 17, characterized in that target lateral distances Ssoii for single-track and two-track vehicles (12) are stored in the computing unit (15), so that if the predicted expected actual lateral distance S(t) is less than the permissible target lateral distance Ssoii, a warning signal is output to a driver of the first vehicle (10) or the first vehicle (10) intervenes independently to correct the situation by increasing the actual lateral distance S(t) or by braking.

19. Method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, characterized in that the speed of the first vehicle (10) is taken into account in the correction or the warning signal.

20. Device for carrying out the method for determining and maintaining a lateral distance S(t) of a first vehicle (10) to a second vehicle (12) during an overtaking maneuver according to one of the preceding claims, with a distance measuring unit (14) directed in the direction of travel of the first vehicle (10), a camera directed in the direction of travel of the first vehicle (10) an image processing unit (18) and a computing unit (15), characterized in that the distance measuring unit (14) is used to determine an actual distance E(t) of the first vehicle (10) from the second vehicle (12), the camera (16) is used to continuously record an image of the second vehicle (12), the image processing unit (18) is used to determine and transmit image data, and the computing unit (15) is connected to the image processing unit for receiving the image data and is used to calculate an actual offset A(t) of a side surface (Z) of the second vehicle (12) facing furthest towards the road centre from the camera axis (X) from the image data and the actual distance E(t) determined via the distance measuring unit and to calculate an actual lateral distance S(t) from the determined actual offset A(t).

Images