DE102022002339A1 - Method for configuring and operating a wildlife crossing warning system of a vehicle - Google Patents

Abstract

The invention relates to a method for configuring and operating an animal crossing warning system of a vehicle (1) in a vehicle fleet (2) with an environment sensor system (3) which continuously detects signals, at least when the vehicle (1) is being driven, by means of which an environment of the respective vehicle (1). Vehicle (1) and located in this objects are detected. According to the invention, the detected signals are used to detect an animal next to or on a lane (F) of the respective vehicle (1) and information about a recognized animal species, a recognized trajectory of the animal, a movement speed of the animal and a location and a time of detection of the animal is stored and transmitted to a central computer unit (4) which is data-technically coupled to the respective vehicle (1), - this central computer unit (4) respectively transmitted location-dependent information is aggregated and if a number of transmitted information for next to or animals detected on this roadway (F) exceeds a predetermined threshold value (SW), it is determined that an animal track (T) is involved, and when a vehicle (1) of the vehicle fleet (2) drives on the roadway (F) a warning (H) relating to the presence of an animal track (T) in the vehicle (1) driving on the roadway (F). is given and / or a driving speed of the vehicle (1) is automatically reduced.

Description

The invention relates to a method for configuring and operating a wildlife crossing warning system of a vehicle in a vehicle fleet with an environment sensor system which continuously detects signals, at least when the vehicle is driving, by means of which an environment of the vehicle and objects located in it are detected.

From the DE 10 2004 050 597 A1 a deer crossing warning device for a motor vehicle with sensor systems and/or interfaces to sensor systems is known. The sensor systems are designed to detect game that appears. In addition, the animal crossing warning device has data output systems and/or interfaces to data output systems, with the data output systems providing data that are relevant for assessing the probability of a wild animal crossing. In addition, means are provided which are designed in terms of programming to determine a risk potential for a deer crossing based on the results of the sensor systems and the data provided by the data output systems.

The invention is based on the object of specifying a method for configuring and operating a wildlife crossing warning system of a vehicle with an environment sensor system.

The object is achieved according to the invention by a method which has the features specified in claim 1.

Advantageous configurations of the invention are the subject matter of the dependent claims.

A method for operating a deer crossing warning system of a vehicle in a fleet of vehicles with an environment sensor provides that by means of

Ambient sensor system, at least while the respective vehicle is being driven, signals are continuously detected, based on which an environment of the respective vehicle and objects located in it are detected. According to the invention, an animal is detected next to or on a roadway of the respective vehicle based on the detected signals and information about a recognized animal species, a recognized trajectory of the animal, a movement speed of the animal and a location and time of detection of the animal is stored and sent to a communicated with the respective vehicle of the vehicle fleet coupled in terms of data technology to the central processing unit. This location-dependent information transmitted to the central computer unit is aggregated and if a number of transmitted information for animals detected next to or on this roadway exceeds a predetermined threshold value, it is determined that an animal track is involved. If the lane is driven on by a vehicle of the vehicle fleet, a warning is issued with regard to the presence of an animal track in the vehicle of the vehicle fleet driving on the lane and/or the driving speed of the vehicle is automatically reduced.

By using the method, driving safety can be increased, at least for the vehicles in the vehicle fleet that drive on the respective lane for which information about the presence of an animal track is available. Due to the issued warning notice, the driver's attention can be increased, so that the roadway may be driven on at a reduced driving speed.

Using the method, the risk of injuries to the animals and accidents involving animals on such a roadway can be significantly reduced, and damage to the vehicle can also be avoided.

In addition, the method can contribute to sustainability, based in particular on animal observations.

Exemplary embodiments of the invention are explained in more detail below with reference to drawings.

• 1 schematisch eine Verkehrssituation zur Erfassung einer Tierfährte,
• 2 schematisch eine weitere Verkehrssituation zur Erfassung einer Tierfährte,
• 3 schematisch die weitere Verkehrssituation gemäß 2 und ein im Fahrzeug ausgegebener Warnhinweis und
• 4 schematisch ein Blockschaltbild zur Ausführung eines Verfahrens zum Betrieb eines Wildwechselwarnsystems eines Fahrzeuges.

show:

• 1 schematic of a traffic situation for recording an animal track,
• 2 schematically another traffic situation for recording an animal track,
• 3 schematically the further traffic situation according to 2 and a warning issued in the vehicle and
• 4 schematically shows a block diagram for the execution of a method for operating a wildlife crossing warning system of a vehicle.

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

the 1 and 2 each show a traffic situation for recording an animal track T. In 3 is an in 2 shown further traffic situation and an embodiment of an output of a warning H in a vehicle 1 and in 4 FIG. 12 is a block diagram for carrying out a method for operating a wildlife crossing warning system of a vehicle 1 belonging to a vehicle fleet 2 is shown.

It is usually known that accidents involving wildlife, including elk accidents, are comparatively frequent and can even be fatal for a driver of a vehicle 1, since the body of an animal can roll through a windshield of the vehicle 1 into the vehicle interior in the event of a collision with the vehicle 1, in particular due to long legs .

A brake assistant of the vehicle 1 takes animals into account, with detection of an animal, in particular in the case of camera-based detection, often based on lighting conditions, a comparatively long braking distance, in particular on slippery roads, for example due to snow, or due to physical limits, for example if the animal suddenly appears in front of the Vehicle 1 is running, can fail.

A method for configuring and operating the deer crossing warning system of a vehicle 1 is described below, which takes into account information about animal tracks T, historically determined animal movements and animal migration and documents the animal movements.

In particular, the method provides animal track detection, with an evaluation of detected signals from an environmental sensor system 3 of the vehicle 1 being expanded to the effect of detecting animal tracks T crossing a roadway F and so-called beaten paths. One focus is on those wild animals that are comparatively often involved in wildlife accidents.

In 1 a traffic situation with a lane F is shown, on which a vehicle 1 is driving and an animal track T runs tangentially to the lane F. In 2 Another traffic situation is shown, with the animal track T crossing the roadway F and the vehicle 1 driving towards the animal track T.

The vehicle 1 belongs to a vehicle fleet 2, the vehicles 1 of the vehicle fleet 2 having an in 4 Central processing unit 4 shown as an example and greatly simplified are coupled.

If vehicle 1 is driving on roadway F, with one or more animals being or being detected on or next to roadway F on the basis of detected signals from environmental sensors 3, information about a recognized animal species, its trajectory, a movement speed and a location and a time of detection is stored on the vehicle side and transmitted to the central processing unit 4 .

The environmental sensor system 3 of the vehicle 1 has a number of sensors 3.1 to 3.n arranged in and/or on the vehicle 1. The environmental sensor system 3 includes sensors 3.1 to 3.n, which are designed, for example, as a camera, based on lidar, based on radar, based on infrared, based on ultrasound, as a microphone, etc.

• - ein Ort mit vergleichsweise häufigem Wildwechsel bestimmt werden kann;
• - anhand der Trajektorien eine jeweilige Tierbewegung und Richtung bestimmt werden können;
• - anhand der Tierbewegung eine Bewegungsgeschwindigkeit abgeleitet werden kann;
• - auf eine Gefahr von Nachzüglern einer Rotte oder Herde hingewiesen werden kann;
• - anhand der Uhrzeit ermittelt werden kann, beispielsweise zur Dämmerung, zu welchen Zeiten mit verhältnismäßig häufigem Wildwechsel gerechnet werden muss und
• - erkannte oder bekannte Tierfährten T anhand der vorliegenden Informationen bestätigt oder auch verworfen werden können.

Furthermore, the method provides that the information transmitted with regard to a respective lane F is aggregated so that:

• - a place with comparatively frequent deer crossing can be determined;
• - A respective animal movement and direction can be determined on the basis of the trajectories;
• - A movement speed can be derived from the movement of the animal;
• - a danger of latecomers in a group or herd can be pointed out;
• - it can be determined based on the time, for example at dusk, at which times relatively frequent game crossings are to be expected and
• - Recognized or known animal tracks T can be confirmed or discarded based on the available information.

A recognition of the animal species can be specified by supplementing expert knowledge, such as size, color, shape of the animals, reflection of the eyes of the animals and/or sounds emitted by the animals.

Static methods, signal processing methods and machine learning methods, in particular computer-aided vision, can be used to identify, ie to classify, the animals.

Data determined for animal tracks T relating to a respective lane F are made available to the vehicle fleet 2 by means of the central computer unit 4 for the parameterization of the deer crossing warning system of the respective vehicle 1 .

By determining the animal species using recorded signals from detected animals, conclusions can be drawn about wild animal observation, in particular with regard to a population or loss.

The data determined by means of the central processing unit 4 based on the transmitted information from the vehicles 1 of the vehicle fleet 2 can be provided to third parties 6, such as, for example be made available to a forestry office and/or a road planning office.

The determination of the presence of an animal track T, which affects the roadway F, i.e. in particular a section of it, is based on the detected signals of at least one camera, a radar-based sensor, a sensor of a different design, for example based on infrared, and a microphone. The respectively detected signals are used to animal tracks T and trails that cross or touch the roadway F, as in the 1 and 2 is shown to recognize.

Relevant animal species are, in particular, animals that lead to wild animal accidents comparatively frequently, such as wild boar, red deer, foxes, wolves, cats, kangaroos and monkeys.

Animals which, due to their body size, can lead to a particularly serious collision with vehicle 1 are, for example, the "Big Five" in Africa and elks in Scandinavia, etc.

In particular, the animal tracks T are recognized on the basis of tracks on surfaces such as earth and snow. So-called beaten paths, for example in meadows and/or bushes, are recognized by damaged, kinked and/or deformed vegetation, animal excrement, etc.

A frequency of such features M indicating animal tracks T and footpaths can also be used to draw conclusions about a number of animals that were recently on or near the respective roadway F.

In other words, a vehicle sensor system, in particular the environmental sensor system 3 of the respective vehicle 1 of the vehicle fleet 2 , is used to detect animal tracks T while the vehicle 1 is being driven. For example, this expands the scope of the deer crossing warning system and/or another assistance system of the vehicle 1, for example the brake assistant.

Furthermore, the method provides that the presence of an animal track T in the area of a roadway F is derived using available historical data.

Assistance systems of the vehicle 1 , for example the deer crossing warning system, map the surroundings of the vehicle 1 with the aid of detected signals from the surroundings sensors 3 . The detected signals of individual sensors 3.1 to 3.n of the environmental sensors 3 are pre-processed and combined via a sensor fusion SF. One output of the sensor fusion SF is object data O for each recognized object, according to the present method for each recognized animal.

In particular, the deer crossing warning system is expanded in such a way that specified animal species are recognized on the basis of the detected signals from the environmental sensors 3 . In this way, object data O, including a probability W for a classification of the animals, can be derived for each recognized animal, in particular after the sensor fusion SF.

For an application, among other things, an animal species, a number of animals, the trajectories of the animals, a movement speed of the animals and a location and a time, extended by animal tracks T, can be determined.

• - Orte mit vergleichsweise häufigem Wildwechsel bestimmt werden;
• - aus den Trajektorien Tierbewegungen und Richtungen bestimmt werden;
• - auf eine Gefahr von Nachzüglern, beispielsweise Jungtiere, hingewiesen werden;
• - anhand der Uhrzeit bestimmte Zeiten mit häufigem Wildwechsel bestimmt werden, beispielsweise zur Dämmerung und
• - erkannte Tierfährten T, beispielsweise Spuren basierend, mit Hilfe der Informationen bestätigt oder verworfen werden.

If this information about the time of vehicles 1 of the vehicle fleet 2 is aggregated using the central processing unit 4:

• - Locations with comparatively frequent deer crossings are determined;
• - animal movements and directions are determined from the trajectories;
• - be informed of the danger of stragglers, such as young animals;
• - Based on the time, certain times with frequent game crossings are determined, for example at dusk and
• - Recognized animal tracks T, for example based on tracks, are confirmed or rejected with the help of the information.

In order to identify animal species, the method provides for animal species to be identified on the basis of detected signals from imaging sensors 3.1 to 3.n and the microphone of the environmental sensor system 3 using a machine learning method, in particular a multi-layer learning method (deep learning).

For this purpose, a so-called supervised learning approach is trained in relation to schematized characteristics M of animals. Detected signals from the environmental sensors 3 are combined via the sensor fusion SF, so that object data O are available.

In addition, the object data O are supplemented by expert knowledge. For example, animal-specific characteristics M, so-called features, can be derived from the detected signals of the environmental sensors 3 for relevant animal species. These features M, such as size, color, shape, locomotion, track type, properties a track, etc., are then extracted from the detected signals and combined with a

Database DB compared with stored animal characteristics in terms of matches. Thus, for each feature M, one obtains a probability W for agreement with a respective animal species.

It is known that animal eyes, especially in the dark, reflect light, ie also headlights, so that this feature M can increase an animal recognition rate of signals detected by a camera of the environmental sensor system 3 at night.

Animal sounds, for example calls, can be very loud and can be recognized by means of known methods for computer-bound processing of natural-language information that are implemented in the vehicle.

In addition, animal species occur with different geographical frequencies, which means that individual animal species can be largely excluded from detection depending on a geographical area.

Statistical methods, methods of signal processing and methods of machine learning, in particular computer-aided vision, can be used for the animal-specific features M mentioned above. A schematized training data set with these features M is required for this.

Determined animal-specific object data O are then added to the object, ie the animal, and individual probabilities W with regard to the features M or the object recognition are combined to form an overall probability GW for the recognition of an animal, for example by averaging.

If it is determined that the overall probability GW exceeds a predetermined threshold value SW, this animal-specific object data O is stored SPO.

The wildlife crossing warning system and/or at least one other suitable assistance system of the vehicle 1 of the vehicle fleet 2 can be parameterized using the currently recognized animal tracks T and the animal tracks T determined using historical data.

Provision can be made for the driving speed of the vehicle 1 to be automatically reduced when driving on the roadway F, in particular in the area of a detected animal track T, for a so-called dashcam to be activated and/or for a warning H to be output visually, as in 3 is shown. For example, an image captured by a camera and output on a display unit of the vehicle 1 can be displayed with an animal track T superimposed on it.

In addition or as an alternative to the optical output of the warning H, it can also be output acoustically and/or haptically.

In one embodiment of the method, it is provided that not only animal species but also individuals are recognized with a certain probability W.

If vehicles 1 of a vehicle fleet 2 collect the above-mentioned information, individual identification can be carried out on the vehicle side and/or by means of the central computer unit 4 . As a result, individual animals can be assigned a unique identification as recognized individuals, so that conclusions can be drawn about wildlife observation, a population, a loss of animals, but also animal behavior, for example that wild boars regularly walk a certain route.

In a further embodiment, data can also be aggregated over longer periods of time. Thus, a disappearance or an increased occurrence of animals in certain regions can be shown. Such information may be used, for example, in the context of corporate social responsibility. The data generated are made available to third parties 6, for example by means of the central computer unit 4. For example, a forest office or individual foresters can access the data and thus receive information about a particular animal population.

Cities, in particular certain offices, can also use this data and information, for example to reconcile animal migration routes with road construction.

Furthermore, the determined animal tracks T can be stored in an animal track card 5 .

Reference List

1

vehicle

2

vehicle fleet

3

environmental sensors

3.1 to 3.n

sensor

4

central processing unit

5

animal tracking card

6

third party

DB

Database

f

roadway

GW

overall probability

H

warning notice

M

feature

O

object data

SF

sensor fusion

SPO

Storage of object data

SW

threshold

T

animal track

W

probability

QUOTES INCLUDED IN DESCRIPTION

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

Patent Literature Cited

• DE 102004050597 A1 [0002]

Claims (3)

Method for configuring and operating a wildlife crossing warning system of a respective vehicle (1) of a vehicle fleet (2) with an environment sensor system (3) which continuously detects signals, at least when the respective vehicle (1) is being driven, based on which an environment of the respective vehicle (1) and objects located in this are detected, characterized in that - based on the detected signals, an animal is detected next to or on a lane (F) of the respective vehicle (1) and information about a recognized animal species, a recognized trajectory of the animal, a movement speed of the animal and a location and time of detection of the animal are stored and transmitted to a central computer unit (4) which is data-technically linked to the respective vehicle (1) of the vehicle fleet (2), - these are transmitted to the central computer unit (4) in each case Location-based information is aggregated and when a number is transmitted information for animals detected next to or on this roadway (F) exceeds a predetermined threshold value (SW), it is determined that it is an animal track (T), and - when a vehicle (1) drives on the roadway (F) of the Vehicle fleet (2) a warning (H) relating to the presence of an animal track (T) in the vehicle (1) of the vehicle fleet (2) driving on the lane (F) is issued and/or the driving speed of the vehicle (1) is automatically reduced becomes. procedure after claim 1 , characterized in that a respective animal is identified on the basis of a size, a color and/or sounds emitted. procedure after claim 1 or 2 , characterized in that the determined animal track (T) is stored in an animal track card (5).

Images