CN117208000A

CN117208000A - Method for a vehicle for warning a vehicle user against temporary dangerous situations

Info

Publication number: CN117208000A
Application number: CN202310680672.0A
Authority: CN
Inventors: A·迈尔
Original assignee: Volkswagen AG
Current assignee: Volkswagen AG
Priority date: 2022-06-10
Filing date: 2023-06-09
Publication date: 2023-12-12
Also published as: DE102022205932A1

Abstract

A method for alerting a vehicle user of a temporary hazard condition in a vehicle surroundings of a vehicle, comprising: -invoking map data about at least one area of a vehicle surroundings of the vehicle; -detecting at least one potential temporary hazard situation with respect to the vehicle, which is determined based on the invoked map data; -evaluating the determined at least one potential temporary hazard situation based on position-resolved ambient data or data derived therefrom for perceptibility from the perspective of the vehicle, said ambient data being determined by means of at least one sensor device for detecting the vehicle ambient; -determining a risk variable based on the evaluation result, from which a warning signal can be provided for output to the vehicle user by means of an output device in order to warn the user against a temporary risk situation.

Description

Method for a vehicle for warning a vehicle user against temporary dangerous situations

Technical Field

The present invention relates to a method and a warning device for a vehicle for warning a vehicle user of a (particularly temporary) dangerous situation in the vehicle surroundings of the vehicle.

Background

In the case of motor vehicles, there may be a danger situation which is difficult to predict even on unknown roads, despite modern technology. In this way, for example, in the case of a city center, vehicles sometimes come to rest on the roadside so closely in sequence that the exit (although it is not plugged) is almost invisible and therefore there is a risk of accident. Such and similar traffic situations are here the danger that the backend data according to a pure navigation map or even with traffic flow information cannot be covered.

A display for a vehicle with a camera is known from US 10,924,679B2, which is configured for capturing images of the surroundings of the vehicle. Further, the captured image enables the processor to detect objects surrounding the vehicle and detect object properties in order to generate image information. The degree of risk can furthermore be calculated by the application device on the basis of different objects around the vehicle (e.g. another vehicle, a lane, a road surface). Driver assistance information may be output based on the calculated risk level. Furthermore, objects in the blind spot region can be detected by the sensor device and a warning graphic image can be output in the event of a collision hazard.

A rear view mirror on a full-face windscreen head-up display is known from US 8,482,486B2. Furthermore, a measuring system for detecting and locating objects is known, which has an object locating sensor. In this case, the sensor is arranged in the vehicle in a relatively unobstructed position with respect to the field of view in front of the vehicle. The trajectory of the identified object may be analyzed to facilitate analysis of a possible collision based on the trajectory motion with respect to the vehicle. The driver is alerted to the impending collision. Furthermore, it is possible to determine vehicles which are in traffic (whose activated headlights may produce glare conditions) and then determine a suitable projection of the lane markings for the head-up display.

A camera system is known from US 9,511,711B2, which is mounted on the front vehicle end. The camera system is connected to a navigation display, wherein the driver observes a visual depiction of each oncoming traffic on the navigation display. In this case, the camera may be mounted in the area of the front left fender, so that the driver may use the camera system for identifying an oncoming vehicle in case of a driver facing, for example, a blocked line of sight of a parked car, which is blocked for the driver, crossing a road of the current lane. The camera may be equipped with proximity and speed sensors, which for example determine the distance of an oncoming object. Based on the detected traffic, a camera image may be displayed on a display.

Disclosure of Invention

The object of the invention is to remedy the disadvantages known from the prior art and to provide a method for a vehicle and a warning device for warning a vehicle user of a (particularly temporary) dangerous situation in the vehicle surroundings of the vehicle, which enable a more accurate and even improved warning of the temporary dangerous situation and thus a further increase in traffic safety.

According to the invention, this object is achieved by a method and an alarm device for alerting a user of a vehicle to a temporary dangerous situation in the vehicle surroundings of said vehicle and a vehicle according to the invention. Advantageous embodiments and improvements of the invention are the subject matter of the present invention.

The method according to the invention for warning a vehicle user of a (potentially and/or upcoming and/or especially immediate, upcoming), especially temporary dangerous situation in the vehicle surroundings of the vehicle comprises invoking map data about at least one region of the vehicle surroundings of the vehicle. The map data may in particular be navigation data. The user is in particular the driver of the vehicle.

The invocation of map data is to be understood here as the invocation of map data by a storage device, in particular inside the vehicle, and/or the invocation of map data by a storage device, for example based on cloud, outside the vehicle, for example an external server (for example of a map and/or navigation service provider), which external server may be, for example, a backend server.

The dangerous situation may be a dangerous situation, in particular a traffic situation and/or a weather situation and/or an ambient situation, which relates to in particular direct (and/or indirect) traffic safety and/or vehicle safety.

The dangerous situation in the vehicle surroundings of the vehicle may be, for example, a (potentially) relevant and/or occurring dangerous situation in the case of a continued (expected and/or predefined and/or predicted) journey of the vehicle. For example, it may be a potentially dangerous situation that may occur along an intended travel route of the vehicle. For example, it may be a dangerous situation in the area of the vehicle surroundings, which is located in front of the vehicle as seen in the direction of travel of the vehicle, which is preferably detectable and/or detected by a front camera of the vehicle. For example, it may be a rear region of the vehicle surroundings, which is detectable, for example, by a rear camera of the vehicle, especially in the case of an intended reverse drive.

A potentially dangerous situation is to be understood in particular as a possible occurrence and/or imminent (future) thereof. According to a preferred embodiment, a potentially dangerous situation is likewise understood to mean a dangerous situation which is to occur and/or is to be detected in any case and/or with high reliability.

Furthermore, the method according to the invention comprises detecting at least one potential (in particular temporary) dangerous situation with respect to the vehicle, which is determined on the basis of the invoked map data. It is conceivable that information about the thus determined potentially particularly temporary dangerous situation is contained in the invoked map data and/or (likewise) invoked.

Preferably, the method comprises determining at least one potential (in particular temporary) dangerous situation with respect to the vehicle based on the invoked map data. In other words, it is determined and/or determined whether a (pre-defined) dangerous situation is present in the vehicle surroundings and/or occurs and/or is imminent and/or develops, which is important for the vehicle (also referred to as the host vehicle) (in view of its vehicle safety and/or its participation in the case of general traffic safety).

For example, the map data may include information regarding annotated traffic signs, such as overtaking prohibition signs and/or other warning and/or reminder signs. It is conceivable that a (in particular each) traffic sign (in particular a prohibited and/or overtaking prohibited sign) determined from the annotated map data is identified as and/or regarded as a (potential) hazard source and thus as an object triggering a potential hazard situation. Preferably, the (especially each) determined traffic sign causes a determination and/or identification and/or confirmation of a potentially (especially temporary) dangerous situation.

Furthermore, conditions and/or (potential) hazard sources can be predefined (in particular predefined), which indicate or in this way indicate a possible or potential (in particular temporary) hazard situation, and which are preferably identified and/or identified (by the vehicle). Examples of this (for potential sources of danger) may be road curvature (in particular curvature values and preferably exceeding a threshold value which is predefined for road curvature), intersections, exits, (road branches, traffic signals, traffic signs, in particular warning and warning signs, speed regulations, uphill slopes, ramps (in particular exceeding a threshold value which is predefined for ramps) and/or bumps, etc. In particular, in the case of the presence of such a hazard, a classification of the traffic situation or the surrounding situation associated with the hazard as a potential hazard situation (which can be triggered by the respective hazard) can be achieved.

For example, traffic conditions outside of a city such as overtaking operations, road edges covering a cluster of micturition weeds and such as sources of danger (e.g., traffic signs), etc. may be potentially dangerous conditions.

Examples for hazard sources of a city centre, which may represent (potential) particularly temporary hazard situations, are for example traffic lights, intersections, crosswalks and/or speed limits (or so-called speed limits, tempobeguenzing).

Preferably, the detection (in particular the identification) of at least one potentially, in particular temporary, dangerous situation is carried out as a function of the in particular detected vehicle position of the vehicle.

Preferably, the vehicle position of the vehicle (in the case of the method) is determined via GPS (global positioning system, german Globales Positionierungssystem) and/or COMPASS/Beidou (Beidou is a satellite navigation system of china) and/or via GLONASS (abbreviation of global navigation satellite system) and/or a satellite-supported navigation system and/or via a positioning device (inside the vehicle), for example in the case of using odometry data of the vehicle.

It is conceivable to determine only potential dangerous situations within a predetermined and/or determinable and/or determined interval relative to the current and/or relative to future expected and/or predicted (over a predetermined and/or predeterminable time interval) given vehicle position. Preferably, for this purpose the future vehicle position is calculated and/or estimated and/or predicted, in particular based on vehicle data of the vehicle, such as vehicle speed, vehicle acceleration, navigation data, etc. It is conceivable that the map data about the vehicle surroundings in the predefined (local) region relative to the (in particular determined) vehicle position are used only to determine the at least one potentially, in particular temporary, dangerous situation, whereas the map data about the vehicle surroundings outside the predefined (local) region relative to the vehicle position are not used. This results in a more rapid identification of a possible potentially dangerous situation.

Furthermore, the method comprises evaluating the determined at least one potential, in particular temporary, risk situation on the basis of the (in particular position-resolved) surroundings data (which in particular at least partially depict the vehicle surroundings) derived therefrom for these (determined potential, in particular temporary, risk situations) which are determined by means of at least one sensor device (in particular of the vehicle) for detecting the vehicle surroundings, for the perceptibility of the vehicle and/or from the vehicle perspective. In this way, the probability of a potential hazard situation being determined purely on the basis of map data for a vehicle in view of the actual occurrence of the hazard (or hazard situation) in particular due to the imperceptible or too late perception of the potential hazard situation can be estimated in particular.

In other words, it is determined whether the potentially dangerous situation is identifiable from the vehicle's perspective or perceivable (on the visual and/or sensing means). For example, it can be determined whether the hazard triggering the potentially dangerous situation is blocked by another object (optically or visually) in such a way that it is not possible for both the driver and the sensor device of the vehicle to perceive the potential hazard and thus react to it as early as possible and in any case in time.

Preferably, the method comprises evaluating the determined at least one potential, in particular temporary, hazard situation on the basis of the position-resolved ambient data or the data derived therefrom for the presence and/or appearance of a particularly temporary hazard situation for the vehicle. The evaluation (evaluation) of the potential, in particular temporary, dangerous situation as an existing and/or occurring dangerous situation can be directly derived from the evaluation that the potential, in particular temporary, dangerous situation for the vehicle and in particular for the user and/or the sensor device of the vehicle can be perceived (optically and/or visually) for detecting the vehicle surroundings.

In other words, the determined potentially particularly temporary dangerous situation is verified and/or validated for the actual or actual occurrence and/or imminence and/or existence based on the location-resolved ambient data.

In addition, the method includes determining a risk variable (or risk size,) According to which a warning signal can be provided for output to the vehicle user by means of an output device in order to warn the user against particularly temporary dangerous situations.

The proposed method advantageously provides the possibility of alerting the driver to the previously described dangerous situations, for example dangerous situations which cannot be identified solely by means of map data, for example, exit openings which are barely visible. The proposed method provides the advantage, in particular with respect to a vehicle with a camera or another sensor, which can warn the driver only against dangerous situations seen or visible by the camera or another sensor, that the user, in particular the driver, is alerted against dangerous situations/points not seen by both himself and the sensor.

Preferably, each and/or (preferably all) of the method steps of the proposed method, in particular those mentioned previously and/or described below, are carried out and/or carried out by a processor-based warning device (described further below), in particular, preferably as computer-implemented method steps. The warning device may be an integral and/or fixed (in particular, not atraumatically releasable) component of the vehicle. However, it is also possible for the computer-implemented method steps (in particular for data processing) to be carried out locally separately from one another on distributed data processing units, for example processors and/or servers, which are preferably arranged partly inside the vehicle and partly outside the vehicle.

The method steps for processing the ambient data (which are collected and/or determined by the sensor device) are preferably carried out in particular in the vehicle interior and/or at least partially in the sensor device. It is particularly preferred to carry out the determination of the risk variable in the vehicle interior. This accordingly provides the advantage of particularly fast data processing without the necessity of transmission of large amounts of data.

A temporary dangerous situation is to be understood in particular as meaning that it does not exist permanently. In particular, a temporary hazard situation is preferably understood to be a hazard situation for which the map data does not contain information about the existence of the temporary hazard situation.

Thus, for example, temporary dangerous situations may be caused by vehicles parked at the side edges for a short period, as well as short-term weather and/or lighting conditions (e.g. imperceptibility of lane markings in combination with solar radiation in a narrow angular range in the case of puddles only at specific points).

In particular, the proposed method finds application in temporary dangerous situations caused by events and/or objects that limit and/or obstruct and/or block (the vehicle user and/or the sensor device of the vehicle) the visibility and/or perceptibility of only temporarily sustained. Temporary duration is understood here, for example, to mean a month at most, preferably a week at most, preferably a day at most 5, preferably a day at most 1, preferably a day at most 12, preferably a day at most 5, preferably a day at most 2, preferably a day at most 1, preferably a day at most 30, preferably a day at most 10, preferably a day at most 5, preferably a day at most 1, and particularly preferably a day at most 30 seconds.

The (potential) temporary hazard situation is preferably caused by a moving object, preferably another traffic participant, for example a vehicle, preferably in such a way that the perceptibility of the hazard source (in particular visual and/or optical and/or sensor) caused by the moving object (in particular stationary) hazard source (for example a narrow bifurcation and/or intersection) is limited and/or blocked (at least partially or completely) by the vehicle (for example a user, for example a driver, and/or by a sensor device for detecting the vehicle surroundings, in particular by detecting position-resolved surroundings data).

It is conceivable that the moving object itself is (exactly) in motion. Thus, for example, the (host) vehicle may exceed the (traveling) load vehicle and traffic signs, such as overtaking prohibition signs, are blocked by the load vehicle during overtaking operations.

By means of the (previously) identified or determined potential (temporary) danger situation being based on the proposed evaluation or assessment of the (current) surrounding data at least partly depicting the respective surrounding area, it may be advantageous that the danger situation can be identified and reacted to even in (highly) dynamic driving situations (for example by outputting a warning signal and/or in the case of a driver assistance system and/or by the vehicle control device taking place).

In the proposed method, precise map material (map data) together with the vehicle position (in particular GPS position) of the vehicle and (ambient) data of predictive sensor means (or sensor devices for detecting the ambient of the vehicle), in particular cameras, are particularly advantageous, which can however also be combined with other sensors (radar, lidar, etc.) in the sense of sensor fusion. From the current GPS location, the vehicle knows where it is and via map material identifies where there must be a narrow exit nearby, intersections where no full view is seen, critical traffic signs (e.g., speed limit, sharp turn warning, overtaking prohibition), etc.

The camera (which may be, for example, a modern front camera in the mirror base of a windscreen) approximately has the same field of view as the driver. As a result of the knowledge of the current vehicle position, in particular in the case of further utilization of the vehicle sensor system and thus of the vehicle trajectory, the vehicle can check via its camera whether the hazard point is visible or, for example, obscured by an object. Different image recognition methods can be used for this purpose. Due to the same viewing angle, it can then be inferred that in the case of an object that is blocking, neither the camera nor the driver can see the actual danger zone. The driver may be alerted: there is a condition where the whole view is not seen and the driver should thus carefully approach the dangerous point.

The proposed method can advantageously alert the user or the driver also in case of such a situation, which is also not perceived by another sensor device with a field of view different from the driver.

Preferably, a sensor device for detecting the vehicle surroundings (of the vehicle) is selected from a set of sensors, preferably capturing and/or collecting and/or generating and/or determining especially position-resolved surroundings data (which at least partially depict the vehicle surroundings) based on which a potential especially temporary hazard situation is evaluated or assessed, said set of sensors comprising (optical) image capturing devices, (color) cameras, front-facing cameras, rear-facing or rear-facing cameras, laterally arranged cameras for detecting lateral surroundings areas of the vehicle, infrared cameras, LIDAR (abbreviation for Light detection andranging oder Light imaging, detection andranging), radar, ultrasonic sensors etc. and combinations thereof. Preferably, the sensor device for detecting the surroundings of the vehicle generates position-resolved, that is to say position-resolved (in particular 2D and/or 3D), sensor data (from the vehicle surroundings of the respective (host) vehicle).

Preferably, the sensor device is a sensor device of a (host) vehicle, which is mounted at or is an integral part of the host vehicle. However, it is also conceivable to receive (by means of the vehicle or a warning device) ambient data, which are detected and/or generated and/or recorded by a sensor device external to the vehicle for detecting the vehicle ambient. Such external sensor devices may be part of the (especially stationary) infrastructure and/or of a (surrounding environment) vehicle other than the vehicle. The (ambient) vehicle and/or the infrastructure may for example transmit the determined ambient data to an external storage device, for example an external back-end server as described above. The ambient data transmitted to the external storage device or the data derived therefrom can then be used to evaluate or evaluate the determined potential, in particular temporary, dangerous situation from the map data (and for this purpose, for example, to the vehicle and/or the warning device).

The (present) vehicle may be in communication with at least one other vehicle (e.g., via Car2Car communication, C2C communication and/or Car2x communication and/or Car2I communication) at least indirectly and preferably directly. The surrounding vehicle can transmit and/or transfer the determined surrounding data and/or data derived therefrom, for example detected objects, at least indirectly (via other traffic participants or infrastructure) and preferably directly to the (host) vehicle.

Preferably, the sensor device for detecting the surroundings of the vehicle is arranged approximately in the middle with respect to the width direction of the vehicle and/or between the windscreen and the rear view mirror and/or in the rear view mirror mount of the windscreen. Preferably, the sensor device for detecting the surroundings of the vehicle is used for at least one further driving assistance task, in particular an automatic emergency braking function and/or an adaptive cruise control and/or as a lane departure warning system, for pedestrian recognition, as a parking aid and/or for traffic sign recognition. The simultaneous use for a plurality of driving assistance tasks provides the advantage of a resource-saving design.

Preferably, a front camera, in particular a single front camera or a stereo camera, is used as a sensor device for detecting the environment surrounding the vehicle. In contrast to the design of several cameras, which leads to costs and furthermore requires installation space, an embodiment in which only the hardware already on board the vehicle (camera, GPS, if appropriate further predictive sensor devices) is particularly advantageous.

A front-end camera module with an integrated processing unit is preferably used as a sensor device for detecting the surroundings of the vehicle, so that advantageously less data has to be transmitted to the data processing device or to the central controller.

It is also conceivable to use a plurality of cameras distributed at the vehicle, in particular for generating a so-called look around, which transmits the captured image data for (joint) processing to a central electronic controller (english Electronic Control Unit, ECU) in the vehicle.

Obviously, the approach based on a front-facing camera (as a sensor device for detecting the surroundings of the vehicle) can also be applied and/or extended to other cameras or sensing devices at the sides and rear of the vehicle.

Preferably, the map data and/or the ambient data or the sensor data are continuously and/or in real time evaluated, so that a highly dynamic response to potentially dangerous situations can advantageously be achieved.

Preferably, the detection or identification of the potential (temporary) dangerous situation triggers the interrogation and/or detection of the surrounding data by the sensor device for detecting the surroundings of the vehicle and/or the detected surrounding data or the data derived therefrom for evaluation or for evaluating and/or processing of the detected and/or identified potential (temporary) dangerous situation. Preferably, an evaluation or assessment of the determined and/or identified potential (temporary) dangerous situation based on the determined ambient data or the data derived therefrom is additionally or alternatively triggered. Preferably, the evaluation or assessment of the determined and/or identified potentially particularly temporary dangerous situation is effected in less than 10 seconds, preferably in less than 5 seconds, preferably in less than 3 seconds, preferably in less than 1 second and particularly preferably in less than half a second.

The map data is preferably used for determining and/or identifying the evaluation of potential, in particular temporary, dangerous situations at regular time intervals and/or position intervals (as a function of the distance travelled). In this case, the distances can be fixedly or predefinable. In this case, the intervals can be dependent on the speed of travel and/or further environmental data, for example, the brightness and/or weather conditions (fog, snowfall) and/or visibility and/or the like.

Preferably, the method comprises generating a warning signal from the determined risk variable, which is based on the result of the evaluation of the potential, particularly temporary, risk situation, which is preferably available for output to the vehicle user by means of an output device in order to warn the user against the particularly temporary risk situation.

The (generated) warning signal is preferably characteristic for the presence of (thus assessed), in particular temporary, dangerous situations and preferably for the (determined) probability with which the dangerous situation is present.

Preferably, the (generated) warning signal is characteristic for an operation recommendation and/or an operation instruction (e.g. a speed and/or a vehicle handling recommendation). Alternatively or additionally, it is also conceivable that the generated warning signal is characteristic for a prompt for an intervention in the control of the vehicle (for example for longitudinal and/or transverse steering of the vehicle).

Furthermore, a classification of dangerous situations can be achieved, so that the driver knows not only that there is a danger, but also what kind of danger is present, for example "note: exit or note: overtaking is prohibited. Typical classification methods are provided for this purpose.

Preferably, the determined and/or validated and/or (in this way) assessed or confirmed risk situation is classified into (exactly) one of a plurality of predefined types of risk situations, in particular according to a number of predefined forms of risk situations (in particular by means of computer-implemented classification methods, in particular based on machine learning models).

Preferably, the (generated) warning signal has a (machine-readable) information and/or data sequence which is characteristic for the classification of particularly temporary dangerous situations and/or for the classification form of dangerous situations.

Additionally or alternatively, the (generated) warning signal may likewise have a (machine-readable) information and/or data sequence, which includes map data that are critical for the determination of a potentially particularly temporary dangerous situation and/or ambient data (and/or further relevant sensor data) or data derived therefrom that are critical for the evaluation of a potentially particularly temporary dangerous situation.

Preferably, a (generated) warning signal of the output device (especially of the vehicle) is provided for output to the user in order to warn it against especially temporary dangerous situations. The output device is preferably an output device fixedly (i.e. not nondestructively releasable) arranged at the vehicle. Here, the output device may be selected from a group of output devices including an acoustic output device (for outputting an acoustic warning), an optical output device (e.g., a display, a dashboard, an augmented reality head-up display, etc.), and combinations thereof.

In a preferred method, ambient data is checked for determining the perceptibility of a potentially particularly temporary dangerous situation for at least one particularly temporary and/or moving object which is suitable for influencing the perceptibility of the potentially particularly temporary dangerous situation and/or its perceptibility (on the vision and/or sensor) of the potentially particularly temporary dangerous situation. This provides the advantage that the ambient data is checked specifically for objects that may form (actual) dangerous situations due to the influence of perceptibility from potentially dangerous situations.

Suitability for influencing the perceptibility can be understood as a placement between a sufficiently large and/or small area in the surroundings of the vehicle, in particular between a hazard source which causes a potentially dangerous situation and the vehicle and/or the region of the vehicle (for example in which the user and/or at least one sensor device for detecting the surroundings of the vehicle is arranged).

The perceptible effect may be understood, for example, as an object blocking and/or blocking a certain region of the field of view of the user and/or of at least one sensor device for detecting the surroundings of the vehicle in such a way that the field of view thus blocked and/or blocked no longer or no longer sufficiently comprises at least one hazard source triggering a hazard situation and is therefore no longer perceptible.

It is preferably determined whether the object is a moving object. In this way, a (particularly predicted and/or determined) movement of the moving object (e.g. along a determined and/or predicted trajectory) can be taken into account when determining the risk variable. It can thus be considered, for example, whether a source of danger triggering a potentially dangerous situation is perceived to be possible for the driver at a later (yet still timely) moment. If, for example, a vehicle parked at the lateral edge, which obstructs the line of sight towards the hazard source, is recognized at a still sufficient distance (in particular by the warning device), it is indicated by flashing that the vehicle is parked out and is away, the traffic situation previously considered as a potential temporary hazard situation can now be evaluated as a "non-hazardous" or no longer a (potential) hazard situation.

Preferably, at least one potentially dangerous situation is determined and/or evaluated from the determined vehicle position of the vehicle. The geometric arrangement of the vehicle (via the warning device) with respect to a potential, in particular temporary, hazard situation and in particular with respect to the hazard source triggering the hazard situation and/or the relative hazard cause (which leads to the determination of the traffic situation as a potential hazard situation), for example objects and/or events that influence the perceptibility, can thus be determined in an advantageous manner.

This geometrical arrangement can in turn be used to evaluate more precisely: whether a potentially especially temporary hazard situation is a (real) especially temporary hazard situation and/or is used to more accurately determine the hazard variable.

As mentioned above, the vehicle position is preferably determined by a position determining device (in particular of the vehicle), preferably by means of GPS and/or differential GPS. Preferably, in addition to or alternatively to the determination of the vehicle position, a so-called "synchronous localization and mapping" (SLAM) method may be used, in which markers from the vehicle surroundings, such as landmarks and/or artificial and/or natural buildings or objects, are used for localization of the vehicle. Thereby a more accurate positioning of the vehicle can be achieved.

Preferably, at least one (in particular temporary and/or moving) object is identified by means of an object identification method (based on the environmental data). Object recognition of objects located in front of and/or beside the vehicle (in the direction of travel of the vehicle) and/or in a predefined and/or determined region of the vehicle surroundings (in particular as a function of the direction of travel and/or of the expected travel route and/or as a function of the speed of travel) can thus be carried out, for example.

Preferably, the ambient data is processed by means of image recognition methods (for example SIFT (abbreviation of SIFT stands for "Scale InvariantFeature Transformation") and/or image recognition methods based on deep learning and/or on machine learning) and/or is used for detecting especially moving and/or temporary objects.

Preferably, the position, in particular the relative position, of the (determined or recognized) object with respect to the vehicle position and/or with respect to the (map) position of the hazard source triggering the hazard situation is determined on the basis of the determined or recognized object and on the basis of the map data and/or the ambient data and in particular on the basis of the vehicle position. It is preferably determined therefrom (or from the geometric relationship) whether the object obstructs the view of the hazard situation and/or hazard source (as seen by the vehicle).

In a further preferred method, the potential hazard situation is associated with at least one hazard situation location, in particular with respect to the map data, and the evaluation of the potential hazard situation is carried out as a function of the at least one associated hazard situation location and as a function of the determined vehicle location. Preferably, the potential hazard situation is associated with the location of (at least) one hazard source triggering the hazard situation (e.g., traffic control traffic signs and/or narrow branches). For this purpose, it is possible (in particular on the basis of map data) to determine areas and/or positions which (if they are in principle perceptible or visible) are sufficient and/or suitable for detecting dangerous situations. It is thus advantageously possible to determine precisely with respect to, for example, the determined vehicle position (as precisely as possible, as described above), whether a predefined (or each) sensor device is used to detect the vehicle surroundings of the vehicle and/or whether a user, in particular the driver, can perceive a dangerous situation (in principle) from a geometric relationship.

Preferably, 2D and/or 3D maps with hazard locations or with potential, in particular temporary hazard situations and/or with (thus assessed), in particular temporary hazard situations are generated on the basis of (respectively) associated hazard situation locations, and/or 2D and/or 3D maps with these hazard situations are supplemented on the basis of (respectively) associated hazard situation locations. The map may be provided to a navigation device.

Preferably, a (in particular 2D and/or 3D) surroundings model can be generated from (potential) hazard situations, in particular from the location of the associated (potential) hazard situation, in particular based on the (identified and/or determined) object and based on the map data.

Further advantageously, the (generated) warning signal may (at least in part) comprise (in particular additionally) a (machine-readable) information and/or data sequence from the map, so that it may for example advantageously be output to the user and/or may for example be used for navigation in the case of vehicle control. Preferably, the (generated) warning signal (especially additionally) comprises (machine-readable) information and/or a data sequence relative to the generated (especially 2D and/or 3D) environmental model, so that it can advantageously be used for output and/or for vehicle control (for example in the case of a navigation device).

The critical surroundings are preferably determined, for example, by means of a predefined distance from the vehicle and/or taking into account the viewing distance (for example of the front camera) and/or the typical viewing distance (respectively and/or each) for detecting the surroundings of the vehicle and/or the user. Preferably, only map data about this critical surrounding area is considered for determining potential (especially temporary) dangerous situations.

In a further preferred method, at least one field of view (in particular by the user or the field of view of the respective sensor device) is selected from at least one in particular predefined and/or determined field of view of the sensor device of the vehicle, in particular of the user and/or for detecting the surroundings of the vehicle, based on the map data and preferably on the relative positioning of the vehicle with respect to the potentially dangerous situation, on the basis of which an evaluation and in particular verification of the determined potentially dangerous situation can be carried out and/or carried out. This provides the advantage that not all ambient data collected by the sensor device for detecting the vehicle surroundings have to be evaluated. This leads to an accelerated assay.

Preferably, the field of view in which the sensor device for detecting the surroundings of the vehicle, and in particular the camera, must see dangerous situations, for example exits and/or traffic signs, is determined. In particular, the intended field of view of the front camera is determined so that the exit opening is visible.

Preferably, at least one sensor device (preferably depending on the form of the sensor device and/or the geometric position of the field of view of the sensor device) is selected from a plurality of sensor devices for detecting the surroundings of the vehicle on the basis of the map data and preferably on the basis of the relative positioning of the vehicle with respect to the potentially dangerous situation, on the basis of the surrounding environment data recorded by the sensor devices it being possible to evaluate and/or in particular to verify the determined potentially dangerous situation.

In a further preferred method, the provision of the warning signal is performed on the basis of in particular determined perception variables which are characteristic for the predicted and/or determined perception by the user of in particular potentially dangerous situations. This provides the advantage that the warning signal can be output matching the actual or possible perception by the user.

Preferably, the sensor variable is determined (in particular by a sensor device of the vehicle), for example by means of a gaze tracking device and/or by means of an internal camera of the vehicle. It can be confirmed here whether the direction of the user's line of sight, in particular at a specific time in which a (potentially) dangerous situation can be seen, can be seen in a direction in which a (potentially) dangerous situation can be seen. It is also conceivable to observe the behavior of the user and to infer, for example, from a change in the behavior of the user (e.g. braking) or else from a change in the non-behavior of the user (in particular by means of a warning device), whether the user perceives a dangerous situation.

Preferably, the use of a vehicle interior camera and/or a combination with a vehicle interior camera is possible in order to control and/or determine, in particular, whether a user, for example a driver, sees in a dangerous area. If not, it may then also be alerted, although the sensing device may see a hazard.

The perceived variable may be a predicted variable which in particular shows the probability that the user has perceived a (potential) dangerous situation.

Preferably, not only is the current "seeing" of the dangerous situation by the camera considered, but the previous time window is considered.

A comparison of the determined time period (or time point) with the ambient data associated with the time period (or time point) is preferably made, in which the user is looking in the direction of the dangerous situation and in particular in the case of the viewing direction the dangerous situation is already visually perceptible, in order to check from the ambient data whether in the time period (or time point) a (potential) dangerous situation or a dangerous source triggering a (potential) dangerous situation can already be perceived by the user.

This provides the advantage, for example, that it is conceivable that the user, in a smaller period of time, is looking in the opposite direction, although for example a narrow bifurcation (for example between two parked vehicles) can be perceived by the user. On the other hand, it is contemplated that these shorter periods of time may actually be available to the user in order to facilitate the perception of dangerous situations. In the latter case, for example, a warning signal need not be output.

In a further preferred method, the determination of the at least one field of view is carried out from the at least one field of view as a function of the in particular determined accuracy in the case of determining the vehicle position and/or the dangerous situation position. This provides the advantage that, for example, the size of the field of view can be selected such that a buffer (buffer) is included so that inaccuracies in the overall position determination can be taken into account.

In a further preferred method, different fields of view and/or fields of view between the user and the sensor device collecting the surrounding data (for detecting the vehicle surroundings) are taken into account in evaluating the potentially particularly temporary dangerous situation. This provides the advantage that different fields of view between, for example, the (front) camera and the driver can be compensated thereby.

The evaluation of the potential, in particular temporary, data stored on a storage device, in particular in the interior of the vehicle, which is characteristic for the (in particular user-specific) field of view and/or field of view of the user and/or for the field of view and/or field of view of a sensor device for detecting the surroundings of the vehicle, which determines the surroundings data, is preferably carried out.

In a further preferred method, the determined risk situations are classified according to a predefined type of risk situation. Preferably, the variable characteristic for the risk level is determined on the basis of a predefined and/or predefinable priority according to the classification type of the risk situation. In particular, the evaluation of the risk situation or the priority is carried out as a function of the severity, that is to say the level of risk.

Preferably, the variables characteristic for the risk level are determined from the current and/or planned vehicle trajectory and/or the maximally expected collision partner (load vehicle/passenger vehicle/person, direction therefrom, possible speed and/or predicted trajectory of the collision partner). Thus, the severity of the possible collision is determined taking into account the current vehicle trajectory and estimating the largest expected collision opponent (load vehicle/passenger vehicle/person, direction from there, possible speed of the collision opponent).

Ambient data (temperature, coefficient of friction of the road, precipitation) can also be taken into account when determining variables characteristic for the risk level, as these can influence the severity of the accident.

In particular, risk variables characteristic for the risk are determined, for example, on the basis of the risk probability and on the basis of variables characteristic for the risk level, preferably the product of the risk probability and the severity, wherein a (in particular audio-visual) warning of the driver is preferably carried out when a risk threshold to be defined is exceeded.

In a further preferred method, the provision of the warning signal is carried out on the basis of an empirical variable determined specifically for the vehicle user, said empirical variable being characteristic for the user's experience with respect to the determined temporary dangerous situation. This provides the advantage that it can be taken into account whether the driver knows the path. In the case of a known path, it can therefore be assumed that a defined sign or a difficult course of the road is known to the driver and therefore the threshold value for warning against dangerous situations is increased.

Empirical variables may be determined based on the number, whether, and how often the travel path and/or the vehicle surroundings (by the user) are travelled. Here, the time of the previous travel may be considered, so that it may be considered how long the previous travel has been performed.

In determining the empirical variables, not only the same but also similar ambient conditions and/or traffic conditions and/or dangerous conditions (and possible consequences or developments thereof) can be considered.

In a further preferred method, at least partially automated and preferably fully automated control of a certain vehicle function of the vehicle, preferably activation of at least one occupant safety system, in particular of the occupant protection system and/or of the vehicle safety system and/or of the vehicle control, is effected as a function of the determined risk variable.

As a further reaction (in particular in accordance with the identified risk), the vehicle and/or the warning device can also intervene in the vehicle control and, for example, can brake or "pretension" the safety system, for example, active occupant protection, in a preventive manner.

Furthermore, the invention relates to a warning device for a vehicle for warning a vehicle user against temporary dangerous situations in the vehicle surroundings of the vehicle. The warning device is configured and/or set and determines map data for invoking at least one region of the vehicle surroundings with respect to the vehicle and detects at least one potential temporary hazard situation with respect to the vehicle, which is determined on the basis of the invoked map data, preferably as a function of the vehicle position, in particular as determined, of the vehicle.

The warning device is preferably adapted and configured to determine and/or be configured to determine at least one potentially temporary hazard situation with respect to the vehicle on the basis of the invoked map data, preferably from the vehicle's in particular determined vehicle position.

According to the invention, the warning device is configured and/or set and determines at least one potential temporary dangerous situation determined on the basis of position-resolved ambient data (which is determined by means of at least one sensor device of the vehicle in order to detect the vehicle ambient environment in particular) or derived therefrom for the evaluation of the data from the angle of the vehicle, and determines a dangerous variable on the basis of the evaluation event, from which a warning signal can be provided for output to the vehicle user by means of an output device in order to warn the user against the temporary dangerous situation.

It is thus also proposed in the case of the warning device according to the invention that map material, in particular (ambient environment) data of the (predicted) sensor device (sensor means) in addition to the vehicle position, for example the GPS position of the vehicle, is used to identify potential, in particular temporary, dangerous situations in the (nearby) environment of the vehicle from the map material, and that it is checked by means of the ambient environment data of the (predicted) sensor device whether the identified dangerous situations can be perceived in this way by the driver of the vehicle and/or the sensor device.

Such a warning device (similar to the method presented above) offers the advantage that it functions even in the case of temporary events, for example in the case of vehicles parked briefly on the road, and also when there is no accident. Furthermore, the warning device (similar to the method) acts with an already existing sensor device and can be operated on an already existing controller/central computer of the vehicle. Furthermore, it is advantageous if the warning device also functions (similar to the method) when a hazard (for example a poorly visible exit opening) is not identified. In addition, the generated warning signal may be output to the user by means of an augmented reality head-up display of the vehicle.

Preferably, the warning device is configured, adapted and/or defined for carrying out the above-described method and all of the method steps already described above in connection with the method, alone or in combination with one another. Instead, the method can be equipped with all the features described in the case of warning devices, alone or in combination with one another.

The warning device may have the above-described sensor device for the (position-resolved) detection of the vehicle surroundings and/or be connected to the sensor device for the (position-resolved) detection of the vehicle surroundings for data exchange or for data reception.

The warning device may have at least one or more of the above-described output devices for outputting the (generated) warning signal and/or be connected to a corresponding output device of the vehicle for data exchange or data reception.

The invention further relates to a vehicle, in particular a motor vehicle, comprising the warning device according to one embodiment. In particular, the vehicle may be a (motorized) road vehicle. The vehicle preferably has at least one sensor device (described above) for the (position-resolved) detection of the surroundings of the vehicle and/or an output device (described above in particular) for outputting a (generated) warning signal.

The vehicle may be a motor vehicle, in particular a motor vehicle controlled by the driver himself ("driver only"), a semi-autonomous, autonomous (for example autonomous class 3 or 4 or 5 (SAE J3016 standard)), or an autonomous motor vehicle. Here, the 5-level autonomous driving level means a fully automatic driving vehicle.

The proposed approach (method and/or warning device) may also be utilized in an automated/autonomous driving situation. But in the case of manual driving is warned against critical or dangerous situations and for example low speeds are recommended, so that in the case of automated/autonomous driving the vehicle can be driven slowly through the dangerous situation.

Likewise, the vehicle may be an unmanned transport system. Here, the vehicle may be controlled by the driver or autonomously driven. Furthermore, the vehicle may also be in the form of an aerial taxi, an aircraft and another transport means or another vehicle, such as an aerial, water or rail vehicle, in addition to the road vehicle.

Furthermore, the invention relates to a computer program or a computer program product, comprising program means, in particular program code, which represent or encode at least some, preferably all, of the method steps of the method according to the invention and preferably one of the described preferred embodiments, and which are designed for implementation by a processor device.

The invention further relates to a data storage device on which at least one embodiment of the computer program according to the invention or a preferred embodiment of the computer program is stored.

Furthermore, the invention relates to a feature of a machine-readable warning signal (in particular as described above) and/or of a machine-readable data sequence, which is preferably generated and/or can be generated by the method described above, in particular according to a preferred embodiment.

The warning signal is preferably characteristic for the above-described risk variable (in particular determined by the warning device) and/or is preferably generated as a function of the risk variable, and/or is transmitted and/or determined by the computer program, and/or is provided for transmission and/or output and/or for controlling at least one vehicle function based on the warning signal.

Preferably, the (machine-readable) warning signal and/or the data sequence (machine-readable) may in particular additionally comprise information and/or data sequences which are characteristic for the risk probability (above) and/or for the variable characterizing the risk level (above) and/or for the risk level (above) and/or the empirical variable. This provides the advantage that, on the basis of these variables, control of at least one vehicle function, for example lateral and/or longitudinal control, and/or an additional preventive function of the vehicle to be implemented for avoiding danger, can be achieved. Thus, for example, a further more detailed execution of the measurement data based on these variables can take place.

The invention is described in connection with temporary dangerous situations. The method and the device can also be applied to permanent dangerous situations which are not displayed in particular in map data.

Drawings

Further advantages and embodiments result from the following figures:

wherein:

FIG. 1 shows a schematic illustration of a traffic situation illustrating a use possibility of a method for alerting a vehicle user according to one embodiment;

FIG. 2 shows a schematic illustration for illustrating traffic conditions of another use possibility of a method for alerting a vehicle user according to another embodiment;

Fig. 3 shows a schematic illustration of a warning device according to an embodiment of the invention; and is also provided with

Fig. 4 and 5 show schematic illustrations for illustrating the use of a warning device for a vehicle according to an embodiment of the invention or for illustrating the traffic situation of a method according to an embodiment of the invention.

Detailed Description

Fig. 1 shows a schematic illustration of a traffic situation in which a vehicle 1 is moving on a road 2, wherein a narrow exit opening 4 is not visible to the driver, although the vehicle 5 is approaching there.

While driving, modern technology is difficult to predict on unknown roads, even though dangerous situations can always occur. As a result, as shown in fig. 1, the vehicle 1 is traveling normally on the road 2, while the other vehicles 6 are parked next to each other at the road edge.

The exit opening 4 (although it is not blocked) can be so difficult to identify here as shown in fig. 1 that a vehicle approaching from this exit opening 4 cannot be seen by the driver of the vehicle 1 on the normal road 2 and there is thus a risk of an accident.

For example, in the case of a delivery vehicle, which, although kept on the right-hand lane with an activated warning flashing device, may obstruct the view towards the actual hazard. This applies similarly to traffic situations in other urban centers, in which, for example, traffic lights, intersections, crosswalks or speed limits are not or hardly identifiable due to poor visibility.

However, such dangerous situations may also exist outside the traffic of the city center. Thus, an impending cut-in operation is shown in fig. 2 where approaching Pkw (passenger vehicle) 10 passes at Lkw/bus 8 (Lkw abbreviation for load-carrying vehicle). However, the overtaking prohibition flag 12 used for the passenger vehicle driver is not recognizable because the load vehicle/bus 8 blocks the view V of the passenger vehicle driver and/or the position-resolved sensor devices for detecting the vehicle surroundings of the vehicle 1. Here, reference numeral V denotes a field of view of the vehicle 1. The field of view of the vehicle 1 may here be a field of view of a user of the vehicle 1, for example a field of view (and in particular a field of view) of a driver, and/or a field of view of (optical) sensor devices for (optically) detecting position-resolved ambient data of the vehicle surroundings of the vehicle 1.

Fig. 2 illustrates an exemplary traffic situation in which the vehicle 1 tries to cut-in a load vehicle/bus on, for example, a (rural) road 2, however the cut-in prohibition flag 12 cannot be recognized for the driver because his line of sight is blocked.

A similar situation may also occur, for example, in the case of climbing a hill or a road side where weeds are clustered, so that the driver cannot correctly estimate the situation in front of his vehicle.

The example shown above means in particular that it cannot be recognized via a generic technique. Pure map information is not helpful here, since it cannot or can only be used to describe the dynamics of the situation, i.e. for example vehicles which obstruct the line of sight.

In the prior art, in addition to road signs and, for example, narrow exits, there are, in particular, means for preventing accidents, for example, by means of Car-2-x/Car-2-Car warning in the event of a fault (liegenbliber) or, for example, via a backend service provider warning with traffic flow information.

The signs of the road are sometimes unreliable because they are difficult to read or they are easily missing.

Car-2-X/Car-2-Car is generally premised on an accident/malfunction because only before the hazard point is warned. The backend server, which is warned mainly in the case of slow moving traffic, congestion or accidents, behaves similarly, however also with a certain inertia and is thus not as dynamic as necessary. In the case of a difficult-to-see exit, it may therefore be difficult to see for only a few seconds or minutes, since the other vehicle is only parked for a short time and is blocked from view.

Car-2-X/Car-2-Car is not considered here because no accident or malfunction occurs in the described scenario, and thus no warnable vehicle is present. The backend service provider, which depicts e.g. dynamic congestion information or accident situations, is likewise not helpful, as there is not necessarily a need for congestion as is evident at fig. 2. In addition, these backend providers sometimes also have algorithms that cause too late to be alerted via the service, for example, before a temporarily parked vehicle that is blocked from view.

In order to solve this problem, it is therefore proposed here in particular that the vehicle detects not only static data (all preceding map information) but also dynamic data (for example objects blocking the line of sight) and, after the risk prediction, warns the driver against a possible dangerous situation. In particular, the sensitivity of the driver by means of the warning is selected as the preferred response, since the vehicle likewise cannot detect the situation itself completely, for example because it likewise cannot recognize other approaching objects, and therefore likewise a very reliable prediction of the situation is not possible.

Fig. 3 shows a schematic illustration of a warning device 10 for a vehicle 1 and/or a system for warning a driver of a vehicle 1 according to an embodiment of the invention. The processor-based warning device 10 or the system can have a new or existing controller, which is equipped in particular with and/or with a novel algorithm for warning against temporary hazards. Preferably, the controller and/or the warning device 10 is connected to a sensor or a sensor arrangement 20 or a plurality of sensors or a plurality of (each other and/or spaced from) sensor arrangements 20, preferably (first) predictive sensors, continuously evaluate the data and preferably issue a warning, in particular an audiovisual warning, in the event of a hazard. Preferably, the warning device 10 and/or the control unit evaluates or evaluates whether the driver is unable to detect a (current) dangerous situation and warns in an audiovisual manner in the dangerous situation.

Preferably, for warning purposes, an augmented reality head-up display is used, which preferably graphically highlights the cause of the emergency or danger situation and thereby also deflects the driver's attention to the possible danger.

It is conceivable that the warning device 10 is configured as a controller.

As regards the sensing means to be used, it is preferred that the (GPS) position of the (own) vehicle 1 and preferably the associated map data, in particular, is regarded as a ground fact via a position detection device, in particular of the vehicle, in particular via GPS (abbreviation for global positioning system, globales Positionsbestimmungssystem). Reference numeral 30 denotes a (detected) own vehicle location with (related) map data. Advantageously, the map data is considered 100% correct here as a simplification. In particular, the high quality based on current map data further assumes that navigation maps have achieved a very high degree of accuracy and detail and are also becoming better due to the high dependence of navigation maps in the field of autopilot and the increasingly networked intelligent vehicles on roads.

In a preferred embodiment, the map data contains (directly) annotated traffic signs. In a further or alternative embodiment, the vehicle recognizes (further) conditions indicating a possible dangerous situation, depending on the road curvature, the structure of the intersection, the exit opening, etc.

Preferably, the vehicle 1 has an image detection device, in particular a (front) camera 22, which detects an occurrence in the vehicle surroundings of the vehicle 1, in particular in front of the vehicle 1. Even when the image detection device, in particular the camera, does not have exactly the same angle of view as the driver due to its installation position in the vehicle 1, in a preferred embodiment it is assumed in principle that the camera and the driver see or do not see exactly the same.

Particularly preferably, algorithms of a preferred design are used (in addition to the sensor device) in order to identify and evaluate or evaluate the dangerous situation.

In an exemplary embodiment, the vehicle continuously captures the situation in front of the vehicle during driving by means of a front camera and attempts to identify objects in front of its own vehicle and beside the vehicle (for example at the roadside) by means of (especially typical) image identification methods, preferably using scale-invariant feature transform (SIFT) or object image identification based on deep learning.

Preferably, further sensors or sensor devices, such as radar 24, lidar 28, ultrasound 26 and further cameras, are used in an assisted manner, which preferably in the case of sensor fusion can produce a more complete, virtual image of the surroundings.

Preferably, the vehicle 1 and/or the warning device 10 (furthermore) detects its own vehicle position on the road as precisely as possible via GPS, possibly also via differential GPS, and further vehicle data 32 of the vehicle 1 (for example data characteristic for speed, acceleration and/or angle), possibly also in combination with so-called "synchronous positioning and mapping" (acronym SLAM) via a camera.

Preferably, the vehicle 1 and/or the warning device 10 are adapted and determined for determining annotated traffic signs and/or dangerous roads/exits/intersections at a certain distance in front of the vehicle 1 by means of navigation data and/or map data. This distance should ideally be as high as possible, but is limited in particular by the field width of the front camera and should therefore be selected appropriately in order to reduce the computational effort.

Since the vehicle location of the vehicle 1 is known, as well as the keypoints and their positions extracted from the navigation data, it is possible to position the vehicle itself, the keypoints and the identified objects relative to each other in 2D or also in the 3D plane in consideration of the altitude information.

Furthermore, environmental data 34, such as the brightness, temperature, etc. of the outside world, may be considered in assessing whether a (potentially) severe hazard situation exists.

The warning device 10 or the controller, preferably with an algorithm for warning against temporary hazards, is preferably configured for continuous querying of the sensor data and evaluation or assessment of the situation.

In case of a dangerous situation, preferably initiated and/or triggered by the warning device 10 or the controller, a warning signal is output by means of the output device 40, in particular the acoustic warning device 36 and/or the optical warning via the dashboard 38 of the vehicle 1 and/or via the augmented reality head-up display 42.

Fig. 4 and 5 each show a traffic situation in which a vehicle 1 on a road 2 has a warning device 10 according to an embodiment of the invention and a traffic situation of a potentially dangerous situation in the surroundings of the vehicle. The vehicle position P or the vehicle position P of the vehicle 1 (in fig. 4 and 5, respectively) is known and preferably determined as precisely as possible.

In addition to the vehicle position of the vehicle 1, the location PG of the (potential) hazard point and/or hazard cause that (precisely and/or as verified) leads to and/or triggers and/or causes the potential hazard situation is likewise known. In this case, the location PG can be associated with a (measured) hazard point and/or hazard cause, in particular of a potentially dangerous situation.

In fig. 4, the location PG of the hazard point may be given by the (geographical) location of the surrounding vehicle approaching the (e.g. narrow) exit of the (host) vehicle 1, so as to be in the area shown in the figure. As shown in fig. 4, the view of the exit, particularly the view of the surrounding vehicles 5 approaching from the exit, is blocked by the vehicles 6 stopped at the roadside of the road 2, and thus a potentially dangerous situation is generated from the blocked view. Reference numeral PG may indicate a geographic position of an exit determined from map data. The position PG of the exit and of the object (for example a parked vehicle) is preferably known. The determination of the viewing field is preferably performed in that the camera of the vehicle 1 has to see the exit opening. The reference sign PV may denote a desired field of view, for example from a front camera of the vehicle 1, in order to be able to see the exit opening.

In fig. 5, the location PG of the dangerous situation may be given by the location of the traffic sign 12, which represents overtaking prohibition. Here, reference numeral PG may represent a geographic location of a traffic sign determined according to map data.

In fig. 5, the dangerous situation is derived from a bus/load vehicle 8 in front, which obscures the view to the traffic sign 12 from the view of the vehicle 1. Preferably, the vehicle position or vehicle position P of the vehicle 1, the position PG of the traffic sign 12 and the object are known. The determination of the field of view in which the camera of the vehicle 1 must see the traffic sign 12 is preferably effected. In fig. 5, reference numeral PV denotes an expected field of view of a sensor device for detecting the surroundings of the vehicle (here, a front camera of the vehicle 1) so that the sign 12 can be seen.

As shown in fig. 4, the vehicle 1 can now precisely determine in which field of view of the front camera PV the critical point/hazard situation is visible to the driver as a result of this relative positioning. The size of this field of view can be calculated directly from the position, but is loaded with buffers in order to compensate for inaccuracies in the overall positioning determination as well as slightly different fields of view of the camera and the driver.

The field of view thus defined, which can be easily projected into the 2D plane of a single image frame, can then be used to attempt to identify markers or road branches or the like by means of image recognition methods (SIFT, based on deep learning). The generated surroundings model is also helpful here, since it can be recognized so quickly whether an object obstructing the line of sight is between the vehicle and the dangerous situation.

Based on which a risk assessment is performed. In this case, not only is the current "view" of the dangerous situation by the camera head approached, but also the time window in front of it is considered. The probability assessment is derived on how long the dangerous situation is visible to the driver, i.e. the driver may actually ignore the dangerous situation. Additionally, the severity, i.e., the level of danger, is evaluated, wherein a fixed evaluation table (e.g., a poorly visible exit, which is worse than the neglected overtaking prohibition flag) can be used for this purpose.

Another way might be to determine the severity of a possible collision taking into account the current vehicle trajectory and estimating the largest expected collision opponent (load vehicle/passenger vehicle/person, direction from which it comes, possible speed of the collision opponent).

Environmental data (temperature, coefficient of friction of the road, precipitation) can also be considered here, since these can influence the severity of the accident.

In particular, the risk is determined, for example, from the product of the risk probability and the severity, wherein an audible warning of the driver is preferably carried out in the event of a risk threshold value to be defined being exceeded.

Alternatively, it is naturally likewise possible to pay attention to only the probability of danger and to warn if the minimum probability is exceeded.

Applicant reserves the requirement to claim all features disclosed in the application document as essential to the invention if they are new relative to the prior art, alone or in combination. It is furthermore pointed out that in the various figures the following features are likewise described, which may be advantageous per se. Those skilled in the art will immediately recognize that certain features depicted in the drawings may be advantageous even in the absence of additional features from the drawings. Furthermore, the skilled person realizes that advantages can also be obtained by combining a plurality of features which are shown in each or in different figures.

List of reference numerals:

1 vehicle

2 road

6 parked vehicle

8 load vehicle/bus

10 warning device

12 overtaking prohibition sign

20 sensor device

22 (front) camera

24 radar sensor device

26 ultrasonic sensor device

28 lidar sensor device

30 GPS location with map data

32 own vehicle data

34 environmental data

40 output device

36 acoustic output device

38 instrument panel

42 augmented reality head-up display

V field of view

P vehicle position

Location of PG dangerous situations

Expected field of view of a PV sensor apparatus

Claims

1. A method for alerting a user of a vehicle (1) of a temporary dangerous situation in a vehicle surroundings of the vehicle (1), comprising:

-invoking map data about at least one area of a vehicle surroundings of the vehicle (1);

-detecting at least one potential temporary hazard situation determined on the basis of the invoked map data with respect to the vehicle (1), preferably as a function of the in particular determined vehicle position of the vehicle (1);

-evaluating the determined at least one potential temporary dangerous situation based on position-resolved ambient data or data derived therefrom for perceptibility from the perspective of the vehicle (1), the ambient data being determined by means of at least one sensor device of the vehicle (1), in particular, for detecting the vehicle ambient;

-determining a risk variable based on the evaluation result, from which a warning signal can be provided for output by means of an output device to a user of the vehicle (1) in order to warn the user against a particularly potential temporary risk situation.

2. Method according to claim 1, characterized in that for determining the perceptibility of the potentially temporary dangerous situation, the surrounding environment data is checked for the presence of at least one temporary and/or moving object adapted to influence the perceptibility of the potentially temporary dangerous situation.

3. Method according to the preceding claim, characterized in that at least one potentially dangerous situation is determined and/or evaluated from the determined vehicle position (P) of the vehicle (1).

4. Method according to the preceding claim, characterized in that the potential hazard situation is associated with at least one hazard situation location, in particular with respect to the map data, and that the evaluation of the potential hazard situation is performed as a function of at least one associated hazard situation location (PG) and as a function of the determined vehicle location (P).

5. Method according to any one of the preceding claims, characterized in that at least one field of view (PV) is selected from at least one, in particular predefined and/or determined field of view (V) of the vehicle (1), in particular of the user and/or of a sensor device (20) for detecting the vehicle surroundings, based on the map data and preferably on the relative positioning of the vehicle (1) with respect to the potentially dangerous situation, on the basis of which field of view an evaluation and in particular verification of the determined potentially dangerous situation can be carried out and/or carried out.

6. Method according to any of the preceding claims, characterized in that the provision of the warning signal is performed on the basis of in particular determined perception variables which are characteristic for the predicted and/or determined perception by the user of in particular potentially dangerous situations.

7. Method according to the preceding claim, characterized in that the determination of the at least one field of view (PV) is performed from the at least one field of view (V) as a function of the accuracy in particular determined when determining the vehicle position (P) and/or the hazard situation Position (PV).

8. Method according to any of the preceding claims, characterized in that the provision of the warning signal is performed on the basis of determined empirical variables specific to the user of the vehicle (1), which are characteristic for the user's experience with respect to the determined temporary dangerous situation.

9. The method according to any of the preceding claims, characterized in that a different field of view between the user and a sensor device (20) collecting the ambient data is considered in evaluating the potential temporary dangerous situation.

10. Method according to any of the preceding claims, characterized in that the determined risk situations are classified according to a predefined type of risk situations and that a variable characteristic for the risk level is determined based on a predefined priority according to the classification type of risk situation.

11. A warning device (10) for a vehicle (1) for warning a user of the vehicle (1) of a temporary hazard situation in the vehicle surroundings of the vehicle (1), wherein the warning device (10) is configured and/or set and determines for invoking map data of at least one region of the vehicle surroundings of the vehicle (1) and detects at least one potential temporary hazard situation with respect to the vehicle (1) which is determined on the basis of the invoked map data, preferably as a function of a particularly determined vehicle position of the vehicle (1),

it is characterized in that the method comprises the steps of,

the warning device (10) is configured and/or set and determines a warning signal for outputting to a user of the vehicle (1) by means of an output device in order to warn the user of the temporary dangerous situation, based on the location-resolved ambient data or the data derived therefrom for the perceptibility from the perspective of the vehicle (1), which are determined by means of at least one sensor device for detecting the vehicle ambient environment, in particular of the vehicle (1), and based on the result of the evaluation a risk variable is determined.

12. Vehicle (1), in particular motor vehicle, comprising a warning device (10) according to the preceding claim.