Method, computer program, device, and system for detecting an anomaly in a monitoring region

Abstract

The invention relates to a method for detecting an anomaly in a monitoring region, wherein in the monitoring region a monitoring scene with at least one vehicle (8) can be arranged, with at least one image (7) of the monitoring region, wherein coordinates can be derived from the image (7), wherein the coordinates define the projected vertices of a polyhedron (9) in the image (7), wherein the polyhedron (9) represents the vehicle (8) and defines a polyhedron position, wherein a vehicle orientation of the vehicle (8) can be derived from the image (7), wherein the anomaly is detected on the basis of the polyhedron position and the vehicle orientation.

Description

[0001] Description

[0002] title

[0003] METHOD, COMPUTER PROGRAM, DEVICE AND SYSTEM FOR DETECTING AN ANOMALY IN A SURVEILLANCE AREA

[0004] State of the art

[0005] The invention relates to a method for detecting an anomaly in a surveillance area, a computer program, a detection device and a detection system.

[0006] The monitoring and early detection of anomalies in traffic flow on highways, intersections or parking garages is a fundamental objective of traffic monitoring systems.

[0007] In principle, it is known to detect vehicles using digital image processing and, for example, to calculate their trajectory.

[0008] The document US 2019 / 0355150 A1 discloses a method for detecting a vehicle in an image, wherein first a heatmap is generated from the image for each pixel, wherein the heatmap is a probability for the center of a vehicle, and wherein subsequently a 3D bounding box for the vehicle is determined by a neural network on the basis of the heatmap.

[0009] Disclosure of the invention

[0010] The invention relates to a method for detecting an anomaly having the features of claim 1, a computer program having the features of claim 12, a detection device having the features of claim 13 and a detection system having the features of claim 14. Preferred or advantageous embodiments of the invention emerge from the subclaims, the following description and the attached figures.

[0011] The invention relates to a method for detecting an anomaly in a surveillance area. The method is thus designed to detect the anomaly in the surveillance area, wherein a surveillance scene with at least one vehicle can be arranged in the surveillance area. Multiple vehicles can also be arranged in the surveillance scene.

[0012] The surveillance scene is designed, in particular, as a driving area of ​​vehicles. The surveillance scene can be realized, in particular, as a road, a highway, an intersection, a parking lot, or the like. The at least one vehicle can be designed as a passenger car, a truck, a bus, a trailer, a motorcycle, etc.

[0013] An anomaly is understood to be an incident that violates normal and / or smooth traffic flow in the surveillance scene. The anomaly can be an incident in which the vehicle violates traffic regulations and / or is damaged and / or causes damage to another road user.

[0014] The method uses at least or exactly one image of the surveillance area, whereby coordinates are derived from the image which define the projected corner points of a polyhedron in the image. The coordinates are in particular designed as image coordinates and / or as 2D coordinates and represent points in the image on or in the image plane. In particular, eight corner points are determined which define six side surfaces (front, rear, top, bottom, left, right in the direction of travel). Depending on the geometry of the vehicle, right angles do not necessarily have to be present. However, the polyhedron preferably encloses the vehicle in such a way that no parts of the vehicle protrude beyond the polyhedron, nor is it too large and creates a "gap" between the polyhedron and the vehicle. In particular, the polyhedron has six side surfaces, specifically as a cuboid and particularly preferably as a rectangular cuboid. The polyhedron represents the vehicle.A polyhedron position can be defined based on the polyhedron. The polyhedron position can be, for example, a center of gravity or a center of gravity projected onto the base surface or center point of the polyhedron. Preferably, the polyhedron position is defined as a vertex, center point, or base point of the polyhedron.

[0015] Additionally, the vehicle's orientation is derived from the image. In particular, the front and rear of the vehicle are derived from the image, particularly in relation to the polyhedron. The combination of the vehicle orientation and the polyhedron position can optionally be referred to as the pose of the polyhedron and / or the vehicle.

[0016] It is proposed that the anomaly be detected based on the polyhedron position and the vehicle orientation. In particular, information about the anomaly can be output, transmitted, stored, and / or further processed. For example, data about the detected anomaly can be transmitted to a traffic control system, which is configured to issue a traffic warning and / or traffic control based on the localization and / or classification of the anomaly. Traffic control may involve, for example, changing speed limits and / or closing lanes in the area of ​​the anomaly.

[0017] In particular, the method may comprise a step of outputting data about the detected anomaly.

[0018] It is one consideration of the invention that by detecting the coordinates of the projected corner points of the polyhedron, the detection is independent of the specific vehicle type and can therefore be used in a wide range of cases. It is also simpler in terms of data technology to only process the pose of a polyhedron and not to use a large number of support points of the vehicle. It is particularly preferred if the polyhedron completely encloses the vehicle and / or the vehicle is completely arranged in the polyhedron, including any add-on parts. This defines an interfering contour of the vehicle, so that interaction with another object only actually needs to be assumed in the event of a collision with the interfering contour. In particular, this can ensure that if two polyhedra are collision-free, the associated vehicles also do not have a collision.

[0019] In a preferred implementation of the invention, the coordinates are configured as p3D coordinates. These are the projections of the eight (3D) vertices of the polyhedron enclosing the vehicle into the image. Such p3D coordinates can be easily derived from the image using a variety of methods.

[0020] The p3D coordinates, as a frame projection, are the projection into the image of a polyhedron, in particular a cuboid (imagined in the real world), that directly surrounds the object, formed as a 3D frame. The 3D frame and / or the polyhedron is formed in particular by straight line segments. "Immediately surrounding" is to be understood as meaning that a surface defined by the imaginary 3D frame or polyhedron (e.g., a polyhedron defined by the frame) surrounds the vehicle as closely / as possible, for example, with the smallest possible volume. Respective surfaces and edges of the surface spanned by the 3D frame or polyhedron touch the surface of the object. In other words, a so-called "3D bounding box" is determined as a 3D frame or polyhedron. The 3D frame is an enveloping body, preferably in the form of a rectangular cuboid.

[0021] It is possible that the anomaly is already detected and / or exclusively at the coordinate level, in particular the 3D coordinates, and thus based on 2D coordinates. By implementing the method purely in the two-dimensional realm, it can work particularly effectively and quickly and is capable of quickly processing even complex images with a large number of objects, particularly in real time (processing time corresponds at most to the time between two frames), when the images are presented as frames of a video stream of, for example, 25 or more fps (frames per second). For example, with a 25 fps video stream, only one twenty-fifth of a second is available for the detection of all displayed objects and all desired information, including the frame projections.

[0022] In further developments of the invention, the coordinates, in particular the p3D coordinates, are converted into 3D coordinates either via a camera model or via other means, so that the anomaly can be determined on the basis of 3D information based on the polyhedron position and the vehicle orientation.

[0023] In one possible embodiment of the invention, the anomaly is detected based on exactly one image. This makes the method very fast, as only a single image needs to be evaluated. Nevertheless, the combination of polyhedron position and vehicle orientation already has a very high degree of predictive power for a large number of anomalies and is sufficient for a final evaluation. In particular, the method is designed to perform single-frame detection. In particular, the anomalies are static or statically detectable anomalies.

[0024] For example, the anomaly detected on the basis of just one image can be a vehicle driving in the wrong direction, a vehicle parked illegally, in particular a vehicle parked illegally on the hard shoulder, and / or a vehicle involved in an accident. In the case of a vehicle driving in the wrong direction, the vehicle orientation does not correspond to the orientation of other vehicles and / or the surveillance scene. In the case of an illegal parking condition, the polyhedron position and / or the vehicle orientation do not correspond to the possible parking spaces in the surveillance scene. In the case of an accident condition, the polyhedron position and / or the vehicle orientation can be oriented in such a way that they cannot be assigned to a normal driving condition. For example, if the vehicle is upright and the vehicle orientation is therefore upwards, an anomaly can be assumed. In particular, it can also be determined whether a vehicle, for example,has tipped over or is detected in an atypical pose. These anomalies can all be detected from a single image. In an alternative or development of the invention, at least or exactly two vehicles are detected in the image, wherein the in particular static anomaly is designed as a common accident state. For example, two polyhedra for the vehicles can be derived from the image, which are arranged in an overlapping manner. In this case, it can be assumed that these two vehicles have been involved in a common accident. In particular, an accident state, in particular a collision / driving into one another, between two or more vehicles can be recognized by the intersection of the polyhedra. This type of detection is very generic and applicable to many cases. This is particularly interesting because for such rare events there is not a large amount of example data available to train a special algorithm (e.g. neural network).This anomaly can be detected from a single image.

[0025] In a further development or alternative of the invention, the anomaly can be detected based on a time series and / or series of images. In particular, the vehicle orientation and the polyhedron position are determined for a plurality or for each of the images in the time series. In this further development or alternative, an anomaly can be inferred based on the change in the polyhedron position and / or the vehicle orientation. In particular, the anomalies are dynamic or dynamically detectable anomalies. The use of time series or video data as image series naturally has the advantage over individual images that hypotheses can be tested / confirmed over time. However, there are also cases / anomalies that can be resolved exclusively by adding the temporal component.

[0026] An example of these anomalies is the evasion of the vehicle, particularly in relation to another vehicle and / or the evasion of one polyhedron in front of another polyhedron.

[0027] Another example of this anomaly is the swaying of a trailer from a vehicle, where either the vehicle and trailer form a common polyhedron, with the polyhedron changing with the swaying, making the anomaly detectable, or the vehicle and trailer form two polyhedrons, with the relative position of the two polyhedrons changing with the swaying, making the anomaly detectable. For towing vehicle-trailer combinations, it could thus be detected, especially without calibration, whether the trailer is swaying, i.e., whether its orientation relative to a towing vehicle changes periodically.

[0028] For example, it is possible to detect a trailer that has become detached from the towing vehicle. One would first determine that the corresponding bounding boxes, as polyhedra of the two vehicle parts, are moving closely behind one another, at a constant distance, and in the same trajectory (possibly even taking into account the modeling of typical steered and unsteered axles), and that this changes at a certain point in time, thus indicating that the trailer has been lost.

[0029] The detection of tailgating represents another important piece of information / anomaly, which occurs through the approach of two polyhedra on a common or at least adjacent trajectory.

[0030] In one possible development of the invention, the anomaly is detected based on an environmental model. The environmental model can, for example, comprise road map information and / or traffic flow information. With knowledge of the road map information, parking on the hard shoulder, for example, can be detected. Based on the traffic flow information, a vehicle's sudden evasion of an obstacle can be detected. In particular, detection based on the environmental model can be performed in exactly one image and / or as a static anomaly. Alternatively, detection can be performed in the time series and / or image series and / or as a dynamic anomaly.

[0031] The following anomalies can be detected from exactly one image, taking the environment model into account:

[0032] A wrong-way driving situation can be detected based on the vehicle's orientation of the polyhedron and the polyhedron's position, particularly the vehicle's location / localization, in the environment model, especially on a map. Similarly, taking the environment model into account, it can be determined whether a vehicle is on the hard shoulder or off the road. Finally, vehicles parked sideways can also be detected based on the vehicle's direction and polyhedron position in the environment model.

[0033] The following anomalies can be detected from a time series and / or image series, taking into account the environmental model:

[0034] By incorporating the environmental model, particularly map data, it would also be possible to detect anomalies such as the evasion of vehicles, for example, to avoid lost cargo on a road, such as a highway. This could be achieved, for example, by analyzing the polyhedron's driving trajectories and comparing them with the lane geometries from the environmental model. Sudden steering maneuvers or sudden braking could also be detected. This approach could also be used, for example, to automatically detect construction sites or other road closures or diversions. Finally, it could also be detected whether a vehicle has left the road and, for example, is in a ditch. For this, it is important to know the history of the driving trajectory.

[0035] In an alternative embodiment, the anomaly is detected without the need for an environment model, in particular independent of the environment and / or lane, whereby the pose of the vehicle and / or the trajectory of the vehicle is sufficient to detect the anomaly. For example, a dynamic anomaly can be detected by a sharp deceleration of the vehicle.

[0036] In a preferred development of the invention, the coordinates and / or the vehicle orientation and / or the anomaly are derived or detected using artificial intelligence. In particular, the artificial intelligence is embodied as a neural network. In particular, it is a deep neural network (DNN). The output of the artificial intelligence can, for example, comprise the image with a bounding box or the plotted coordinates for locating the vehicle and / or a classification of the anomaly. Optionally, the classification can include a probability for the classification.

[0037] A further subject matter of the invention is a computer program, wherein the computer program is designed and / or configured to carry out, apply and / or implement the method according to the invention and / or the steps of the method when executed.

[0038] A further subject of the invention is a machine-readable storage medium, wherein the computer program is stored on the storage medium.

[0039] A further subject matter of the invention is a detection device with a data processing device, wherein the data processing device is configured to carry out the previously described method and / or implement the computer program. The data processing device is configured, for example, as a computer or as a cloud. The detection device comprises an input interface for receiving the image, wherein the input interface is connected to the data processing device for data processing and transfers the image to the data processing device. Furthermore, the detection device comprises an output interface for outputting the detected anomaly, wherein the output interface is connected to the data processing device for data processing in order to receive and output the aforementioned data.

[0040] Optionally, the detection device has a camera for recording the image, wherein the camera is connected to the input interface for data purposes. Alternatively or additionally, the detection device has an output device, such as a screen, a display or the like, wherein the output device is connected to the output interface for data purposes and is designed to output the anomaly. For example, the image can be displayed on the output device with a drawn-in boundary box of the anomaly and / or information for classifying the anomaly. Preferably, the detection device is designed as a real-time system, wherein the runtime between the acquisition and / or recording of the image and the output of the said data is less than 1 s, preferably less than 0.1 s.

[0041] A further subject of the invention is a traffic monitoring system, wherein the traffic monitoring system comprises the detection device and the camera. Furthermore, the traffic monitoring system has a traffic control device, wherein the traffic control device is configured to implement a traffic warning and / or traffic control based on the localization and / or classification of the anomaly. During traffic control, for example, speed limits are changed or lanes are closed in the area of ​​the anomaly.

[0042] The advantages over other methods vary depending on the type of anomaly. What they have in common, however, is that the method efficiently and quickly detects anomalies based on very compact and generic features / metadata.

[0043] A short processing time (e.g. from just one image) allows (1) a corresponding output (such as an alarm) to be triggered quickly. On the other hand, (2) many video streams can also be analyzed by just one central device, for example by analyzing an image from a different camera one after the other (multiplexing). This would not be possible (with the same resource requirements) if time series / video data had to be analyzed each time. A high level of generalization is advantageous so that the system (1) can be used in different regions of the world without adaptation, without, for example, having to train a new specific detector for special vehicle classes, and so that the system is also (2) future-proof (e.g. when new vehicle models come onto the market). Since many of the anomalies described here occur only very rarely, it is advantageous to use a detection algorithm that generalizes as well as possible, e.g.Using simple rules, the intersection of the vehicle-encompassing polyhedron / bounding box can be evaluated to detect an accident. Finally, the use of the p3D detector (or an equivalent alternative) eliminates the need for additional information, such as camera calibration or other knowledge about the scene (distances and surfaces, lanes, etc.), in many places, making setup significantly simpler and less error-prone.

[0044] Further features, advantages, and effects of the invention will become apparent from the following description of a preferred embodiment of the invention and the accompanying figures. These show:

[0045] Figure 1 is a schematic block diagram of a traffic monitoring system;

[0046] Figure 2 shows a flowchart for a method for operating the traffic monitoring system;

[0047] Figure 3 is a schematic representation of an accident, with the vehicles represented by polyhedra;

[0048] Figure 4 is a schematic representation of a traffic scene, with the vehicles partially represented by polyhedra;

[0049] Figure 5 is a schematic representation of a horizontally transverse vehicle, the vehicle being represented by a polyhedron;

[0050] Figure 6 is a schematic representation of a vertically transverse vehicle, wherein the vehicle is represented by a polyhedron.

[0051] Figure 1 shows a schematic block diagram of a traffic monitoring system 1 for monitoring traffic on roads, parking lots, highways, etc. as an embodiment of the invention.

[0052] The traffic monitoring system 1 comprises a detection device 2 for detecting an anomaly in a monitoring area, wherein a monitoring scene with at least one vehicle can be arranged in the monitoring area, so that the monitoring scene includes the traffic. The detection device 2 is designed as a digital data processing device, such as a computer, a server, a client, a cloud, etc. The detection device 2 is designed to implement the method according to Figure 2. In particular, a computer program that implements the method is executed on the detection device 2.

[0053] The detection device 2 has an input interface 3, wherein the input interface is connected for data purposes to a camera 4, wherein the camera 4 forms a component of the traffic monitoring system 1. Multiple cameras 4 may also be present. The camera 4 is directed at the surveillance area with the surveillance scene and can record at least one image or series of images showing the surveillance area and / or the surveillance scene. Furthermore, the traffic monitoring system 1, in particular the detection device 2, has an output interface 5, wherein the output interface 5 is connected for data purposes to a traffic control device 6. The traffic control device 6 receives data from the detection device 2 and implements measures to direct traffic. For example, speed limits can be implemented or changed, lanes can be closed or opened, or warnings can be issued.

[0054] Figure 2 shows a schematic flow diagram of a method for detecting an anomaly in a surveillance area. In step 100, an image of the surveillance area is captured. In step 200, coordinates of a polyhedron, in particular a cuboid, are derived from the image. The coordinates define the projected vertices of the polyhedron in the image. In particular, the coordinates are configured as 2D coordinates / or image coordinates. The coordinates can, in particular, be configured as so-called p3D coordinates or equivalent coordinates. The polyhedron represents the vehicle and defines a polyhedron position.

[0055] In particular, the vehicle is arranged within the polyhedron and / or the polyhedron forms a tight or even smallest possible envelope for the entire vehicle. In step 300, a vehicle orientation is derived from the image. Steps 200 and 300 can also be implemented as a single step. The vehicle orientation is also defined with respect to the polyhedron, so that a pose is derived from the polyhedron. In step 400, the anomaly is detected based on the polyhedron position and the vehicle orientation.

[0056] Steps 200, 300, and optionally 400 can also be implemented as a single step. In particular, these steps can be implemented by artificial intelligence, in particular by a neural network.

[0057] The anomalies to be detected are primarily deviations from normal traffic and / or violations of traffic regulations. A variety of use cases for anomaly detection are presented below.

[0058] The use cases can be classified in different ways: On the one hand, they can be static anomalies, which can be detected from just one image. On the other hand, they can be dynamic anomalies, which require time series and / or image series to detect them. Another classification arises from the fact that certain anomalies can be detected using an environmental model, while others can be detected without an environmental model.

[0059] Figure 3 shows a highly schematic image 7 with two vehicles 8, wherein the vehicles are each arranged in a polyhedron 9, in particular designed as a cuboid, wherein the polyhedra 9 are each defined by eight vertices, wherein the coordinates of the vertices are designed as p3D coordinates, i.e. coordinates of the polyhedra 9 in the 3D space, which are projected onto the image plane of the image 7.

[0060] Figure 4 shows, in another image 7, a surveillance area with a surveillance scene. Two vehicles 8 are marked in the surveillance scene, forming an anomaly. A polyhedron 9 is drawn only for one of the vehicles 8. Figure 5 shows, in another image 7, a vehicle 8 with an enclosing polyhedron 9, where the polyhedron 9 is oriented transversely to a road course and / or traffic flow known from an environmental model.

[0061] Figure 6 shows in a further image 7 a vehicle 7 with an enclosing polyhedron 9, wherein the polyhedron is arranged at a 45° angle to the ground and / or in the image 7.

[0062] Use case: Anomaly: Accident condition

[0063] A collision between two vehicles can be detected as a static anomaly because the corresponding polyhedra overlap, as shown in Figure 3. Alternatively, it can be detected as a dynamic anomaly because the polyhedra "collapse into each other." An environment model is not necessary.

[0064] A vehicle accident can be detected as a static anomaly. Without an environment model, the vehicle's orientation can at least be used to detect whether the vehicle is, for example, upright or in a different, non-drivable or atypical vehicle orientation, as shown in Figure 6. With the environment model, the vehicle's orientation can be used to detect whether the vehicle is perpendicular to a direction of travel or is located in a "non-drivable area," such as a ditch, hard shoulder, etc., as shown in Figure 5.

[0065] An accident involving one or more vehicles can also be detected as a dynamic anomaly if the vehicle and / or vehicles decelerate in an accidental manner. This dynamic anomaly can be detected without an environment model.

[0066] Use case: Anomaly: Ghost driving

[0067] On the one hand, this anomaly can be detected as a static anomaly as long as an environment model is used, in particular by comparing the vehicle orientation with map information of the environment model, as shown in Figure 4 on the right side with vehicle 8, which wants to enter as a wrong-way driver.

[0068] Use case: Anomaly “unusual driving behavior”

[0069] Sudden deceleration can be detected as a dynamic anomaly with or without an environmental model. A vehicle swerving or a deviation from normal traffic flow can be detected as a dynamic anomaly by taking the environmental model into account, including lanes and / or normal traffic flow. This includes, for example, swerving around an obstacle or swerving onto a hard shoulder.

[0070] Use case: Anomaly illegal parking condition

[0071] Illegal parking, for example, on a hard shoulder, can be detected as a static anomaly using an environmental model, as shown on the left side of Figure 4, where vehicle 8 is parked on the hard shoulder. It is also possible to detect this illegal parking as a dynamic anomaly by evaluating the image sequence to determine whether vehicle 8 has been stationary for an extended period.

[0072] Use case: Anomaly trailer problem or loss

[0073] This anomaly can be detected particularly easily as a dynamic anomaly by tracking the trajectories of the towing vehicle and trailer, represented by corresponding polyhedra, and in the case of unusual behavior such as periodic rocking of the trailer, a trailer problem is concluded, and in the case of a change in the distance, a trailer loss is concluded as an anomaly.

[0074] Use case: tailgating / pushing anomaly

[0075] This anomaly can be detected as a dynamic anomaly by evaluating the trajectories of successive vehicles, represented by polyhedra, and if the distance between the polyhedra is too small, it is concluded that there is close driving / pushing.

[0076] Use case: Anomaly Avoidance of Obstacle

[0077] This anomaly can be detected as a dynamic anomaly by evaluating the trajectories of vehicles in a lane according to the environment model. If the trajectories of the polyhedra deviate from the lanes, an obstacle (lost cargo, accident, etc.) in the lane can be inferred. Sudden steering maneuvers or sudden braking could also be detected. This approach could also be used, for example, to automatically detect construction sites or other road closures or diversions. Finally, it could also be detected whether a vehicle has left the road and, for example, is in a ditch. For this, it is important to know the history of the vehicle's trajectory.

Claims

Claims 1 . Method for detecting an anomaly in a surveillance area, wherein a surveillance scene with at least one vehicle (8) can be arranged in the surveillance area, wherein coordinates are derived from at least one image (7) of the surveillance area, wherein the coordinates define the projected corner points of a polyhedron (9) in the image (7), wherein the polyhedron (9) represents the vehicle (8) and defines a polyhedron position, wherein a vehicle orientation of the vehicle (8) is derived from the image (7), wherein the anomaly is detected on the basis of the polyhedron position and the vehicle orientation.

2. Method according to claim 1, characterized in that the polyhedron (9) encloses the vehicle (8) 3. Method according to claim 1 or 2, characterized in that p3D coordinates are determined as coordinates.

4. Method according to one of the preceding claims, characterized in that the anomaly is detected on the basis of exactly one image (7).

5. Method according to one of the preceding claims, characterized in that the anomaly is designed as a ghost drive of the vehicle (8), as an illegally parked state of the vehicle (8) and / or as an accident state of the vehicle (8).

6. Method according to one of the preceding claims, characterized in that at least or exactly two vehicles (8) are detected in the image (7), wherein the anomaly is formed as a common accident condition.

7. Method according to one of the preceding claims 1 to 3, 5 or 6, characterized in that the anomaly is detected on the basis of a time series of images (7).

8. Method according to one of the preceding claims, characterized in that the anomaly is detected taking into account an environmental model.

9. Method according to one of the preceding claims 1 to 7, characterized in that the anomaly is detected without an environment model.

10. Method according to one of the preceding claims, characterized in that the coordinates and / or the vehicle orientation and / or the anomaly are derived or detected via a Cl.

11. A computer program, wherein the computer program is designed and / or configured to execute, apply and / or implement the method according to one of claims 1 to 10 when executed.

12. A machine-readable storage medium, wherein the computer program according to claim 11 is stored on the storage medium.

13. Detection device (2) for detecting an anomaly, characterized in that the detection device (2) is designed to carry out the method according to one of the preceding claims, wherein the detection device (2) has an input interface (3) for receiving at least one image (7) and / or an output interface (5) for outputting data about the detected anomaly.

14. Traffic monitoring system (1) for monitoring traffic, with the detection device (2) according to claim 13, with a camera (4) for recording at least one image (7), wherein the camera (4) is connected to the input interface (3) for data processing, and with a traffic control device (6) for issuing a traffic warning and / or Traffic control information, wherein the traffic control device (6) is connected to the output interface (5) for data purposes.

Images