DE102020000694A1 - Method for the detection of people in the vicinity of a vehicle - Google Patents

Abstract

The invention relates to a method for detecting people (P1, P2) in the vicinity of a vehicle (1). According to the invention, a person (P1, P2) in the vicinity of the vehicle (1) is detected by the person (P1, P2) detected radio transmitters and / or radio receivers. Data captured by means of the at least one camera (8 to 11) and / or by means of the at least one radio sensor (2 to 5) are compared with data from a navigation system of the vehicle (1) and / or data from a digital map.

Description

The invention relates to a method for detecting people in the vicinity of a vehicle.

From the EP 2 571 005 A1 A vehicle environment monitoring device for detecting people in camera images with an object extraction unit, a physical-body evaluation unit and a distance calculation unit is known. The camera images are captured using an infrared camera.

The invention is based on the object of specifying an improved method for detecting people in the vicinity of a vehicle.

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

Advantageous embodiments of the invention are the subject of the subclaims.

In the method for detecting people in the vicinity of a vehicle, according to the invention, a person in the vicinity of the vehicle is detected by means of at least one vehicle's own camera and / or by means of radio transmitters and / or radio receivers carried by the person carried along by at least one vehicle's own radio sensor, wherein by means of the at least one camera and / or by means of the at least one radio sensor, data acquired are compared with data from a navigation system of the vehicle and / or data from a digital map.

Due to the comparison of the data, the method enables a significant increase in the reliability of the detection of people in the vicinity of a vehicle. Thus, a reliability of vehicle systems using such detection results can be increased and an accident probability can be reduced.

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

• 1 schematisch eine Verkehrssituation mit einem Fahrzeug und mehreren in einer Umgebung des Fahrzeugs befindlichen Personen,
• 2 schematisch einen Ablauf eines Verfahrens zur Detektion der Personen in der Umgebung eines Fahrzeugs gemäß 1 anhand von Funksignalen,
• 3 schematisch eine Verkehrssituation mit einem Fahrzeug und einer in einer Umgebung des Fahrzeugs befindlichen Person,
• 4 schematisch einen Ablauf eines Verfahrens zur Detektion der Person in der Umgebung eines Fahrzeugs gemäß 3 anhand von Daten einer Wärmebildkamera,
• 5 schematisch eine Verkehrssituation mit einem Fahrzeug und einer in einer Umgebung des Fahrzeugs befindlichen Person und
• 6 schematisch einen Ablauf eines Verfahrens zur Detektion der Person in der Umgebung eines Fahrzeugs gemäß 5 anhand von Daten einer an einem Fahrzeugaußenspiegel angeordneten Kamera.

Show:

• 1 schematically a traffic situation with a vehicle and several people in the vicinity of the vehicle,
• 2 schematically shows a sequence of a method for detecting people in the vicinity of a vehicle according to FIG 1 based on radio signals,
• 3 schematically, a traffic situation with a vehicle and a person in the vicinity of the vehicle,
• 4th schematically shows a sequence of a method for detecting the person in the vicinity of a vehicle according to FIG 3 based on data from a thermal imaging camera,
• 5 schematically, a traffic situation with a vehicle and a person in the vicinity of the vehicle and
• 6th schematically shows a sequence of a method for detecting the person in the vicinity of a vehicle according to FIG 5 based on data from a camera arranged on a vehicle exterior mirror.

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

In 1 is a traffic situation with a vehicle 1 and several in a vicinity of the vehicle 1 people P1 , P2 shown. The person P1 is a pedestrian, the person in the illustrated embodiment P2 a cyclist.

The vehicle 1 comprises a plurality of radio sensors 2 to 5 , with each wireless sensor 2 to 5 each comprises at least one radio transmitter and radio receiver. There is at least one radio sensor on each side of the vehicle 2 to 5 arranged.

The people P1 , P2 also each have a wireless sensor 6th , 7th with itself, with the wireless sensors 6th , 7th each comprise a radio transmitter and / or radio receiver. The radio sensors 6th , 7th are for example in one each of the persons P1 , P2 Any mobile device carried along, for example a smartphone, a so-called wearable or other device, is integrated.

Both the radio sensors 2 to 5 of the vehicle 1 as well as the radio sensors 6th , 7th of the people P1 , P2 work in one possible embodiment of the so-called Bluetooth industry standard, for example the Bluetooth industry standard 5.1 .

In the illustrated embodiment, the vehicle is moving 1 , for example at night and / or when visibility is poor, towards an intersection where pedestrians (person P1 ) and on which the cyclist (person P2 ) approaches. The person P1 is on a sidewalk parallel to a priority road and the person P2 on a cycle path that runs parallel to the main road. Both people P1 , P2 intend to cross the lane on which the vehicle is located 1 emotional.

In road traffic, it is important to record the presence of all road users, correctly assess their speeds and react appropriately. It is often difficult, especially in poor lighting conditions, obscuration by bushes, buildings, etc. or because of curves or in road intersection areas, other road users, such as. B. pedestrians, cyclists, etc. to recognize. This creates a hazard and is the cause of serious accidents.

To avoid this, there is an automated recognition of the people P1 , P2 in the vehicle 1 provided while moving the vehicle 1 by means of the radio sensors 2 to 5 the surroundings of the vehicle 1 is continuously scanned and for radio sensors 6th , 7th , especially radio receivers, in the vicinity of the vehicle 1 is sought. Thus, the people can P1 , P2 are recorded. In this case, when at least one radio receiver is detected, a bearing can be carried out on the same and a trajectory of the radio receiver can be determined on the basis of the bearing. Depending on the trajectory of the radio receiver, a collision probability between the vehicle becomes 1 and a person carrying the radio receiver P1 , P2 determined and appropriate measures can be initiated to prevent a collision.

This takes advantage of the fact that currently the majority of people P1 , P2 a mobile device with radio sensors 6th , 7th carries with it. The vehicle's own radio sensors are also used in the vehicle environment 2 to 5 Used for coupling with mobile devices, such as smartphones, for example to implement remote-controlled and / or remotely monitored parking assistance systems. Here is, for example, when using the Bluetooth industry standard 5.1 , also a relatively precise transmission and reception radio direction finding by means of a so-called Angle-of-Departure (AoD for short) and Angle-of-Arrival (AoA for short) can be realized.

2 shows a possible sequence of a method for detecting people P1 , P2 around the vehicle 1 according to 1 based on radio signals.

After the method has been started, a first branch V1 checks whether a vehicle system is connected to one by means of the radio sensors 2 to 5 carried out monitoring of the surroundings of the vehicle 1 is active. If this is not the case (represented by a no branch N1), the method is ended.

If the vehicle system is active (represented by a yes branch J1), it is checked in a second branch V2 whether the vehicle is 1 moves, i.e. has a speed greater than zero. If this is not the case (represented by a no branch N2), the method is started again.

Is the vehicle 1 in motion (represented by a yes branch J2), a scanning process is started in an operation 01, in which the vehicle system continuously searches for radio sensors 6th , 7th , especially radio receivers, in the vicinity of the vehicle 1 is sought.

In a branch V3 it is checked whether in the vicinity of the vehicle 1 Radio receivers are detected. If this is not the case (represented by a no branch N3), the operation O1, that is to say the scanning process, is carried out further.

On the other hand, it is in the vicinity of the vehicle 1 If at least one radio receiver is detected (represented by a yes branch J3), radio location or direction finding is carried out in a further operation 02.

In a branch V4 it is checked whether the location or bearing was successful. If this is not the case (represented by a no branch N4), corresponding driver information is provided within the vehicle in a further operation 03 1 issued. The driver information includes, for example, a message about the detection of a radio receiver in the vicinity of the vehicle 1 . This is useful, for example, when driving overland at night, since only a small number of pedestrians or cyclists can be expected here. A vehicle driver can thus be provided with a valuable advance warning if, for example, people are at the edge of a country road due to a vehicle breakdown. In this context there is the possibility of using the at least one radio sensor 2 to 5 recorded data with data from a navigation system of the vehicle 1 and / or to match data on a digital map. Thus, depending on a current position of the vehicle 1 , for example in a city center or on a country road, different responses are given to the vehicle driver.

If, on the other hand, the locating or direction finding was successful (represented by a yes branch J4), a trajectory of a carrier of the radio receiver is created in a further operation 04 on the basis of the determined data, that is to say according to 1 the person P1 and / or the person P2 , determined.

In a branch V5, depending on the trajectory of the radio receiver, it is checked whether there is a likelihood of a collision between the vehicle 1 and the person carrying the radio receiver P1 , P2 exceeds a specified value. If this is the case (represented by a yes branch J5), the vehicle is automatically braked in an emergency in an operation 05 1 executed. If this is not the case (represented by a no branch N5), however, a minimum level of a collision probability is determined, in a further operation 06 a warning within the vehicle 1 issued, which warns a vehicle driver of the risk of collision. In this context there is the possibility of using the at least one radio sensor 2 to 5 recorded data with data from a navigation system of the vehicle 1 and / or to match data on a digital map. Thus, depending on a current position of the vehicle 1 , for example in a city center or on a country road, different levels can be integrated in a system reaction.

The method thus makes it possible to prevent accidents involving pedestrians, cyclists and other people P1 , P2 avoid and increase road safety. In this way, in particular, early detection of other road users is made possible, even in poor light conditions or when they are covered by other objects. With the help of the additional data, the vehicle system is able to record all road users and, if necessary, to warn the vehicle driver if a dangerous situation arises. In individual cases, under certain boundary conditions, the vehicle system can initiate emergency braking based on the data in order to avoid a collision. Thus, the vehicle system and the method carried out by means of it represent support for the vehicle driver and help to avoid accidents and, in particular, to better protect pedestrians and cyclists in traffic.

In 3 is a traffic situation with a vehicle 1 and one in a vicinity of the vehicle 1 located person P2 shown. The person P2 is a cyclist in the illustrated embodiment.

The vehicle 1 comprises a plurality of cameras designed as thermal imaging cameras 8th , 9 .

In the illustrated embodiment, the vehicle is moving 1 towards an intersection, where the cyclists (person P2 ) approaches. The person P2 is on a cycle path running parallel to the main road and intends to cross the lane on which the vehicle is located 1 emotional. The cyclist has the right of way which the vehicle 1 must grant.

In today's traffic, cyclists often bear the highest risk of accidents. The reason for this is that the speed of cyclists is often underestimated or that cyclists are only recognized very late. Crossing cycle paths are often difficult to see for vehicle drivers and a cycle lane is driven into without crossing cyclists being expected. In addition, cyclists often only participate in road traffic with insufficient lighting at dusk or at night.

To grant the cyclist right of way in the illustrated embodiment, it is necessary that the vehicle 1 first stops in front of the bike path. This is often not known to vehicle drivers, so they drive into the cycle path and only stop at the edge of the actual road to check whether other vehicles are coming from the left or right. If the traffic situation is also poorly illuminated, for example during twilight and / or when there is poor or no lighting on the bicycle, a very dangerous traffic situation also arises due to the sometimes very high speed of the cyclist.

In order to significantly reduce the risk potential for various road users in such a traffic situation, it is provided that the cameras designed as thermal imaging cameras 8th , 9 the surroundings of the vehicle 1 is monitored. The cameras 8th , 9 are in the front of the vehicle 1 installs and monitors vehicle flanks in one detection area E1 , E2 . Using the cameras 8th , 9 objects with a corresponding heat signature can be detected by humans or other living beings and can be recognized as such. Thus, the cyclist (person P2 ) can be recognized automatically and an evaluation of the recorded data can be used to determine whether and at what speed the cyclist is in the vehicle 1 approaching. Thus, there can be a likelihood of collision between the vehicle 1 and the person P2 can be determined and appropriate measures can be initiated to prevent a collision.

4th shows a possible sequence of a method for detecting the person P2 around the vehicle 1 according to 3 on the basis of cameras designed as thermal imaging cameras 8th , 9 collected data.

After the method has been started, a first branch V1 checks whether a vehicle system is connected to one by means of the cameras 8th , 9 carried out monitoring of the surroundings of the vehicle 1 is active. If this is not the case (shown by a no branch N1), the method is ended.

If the vehicle system is active (represented by a yes branch J1), it is checked in a second branch V2 whether the vehicle is 1 moves, i.e. has a speed greater than zero. If this is not the case (represented by a no branch N2), the method is started again.

Is the vehicle 1 in motion (represented by a yes branch J2), a scanning process is started in an operation O1 in which the vehicle system evaluates the data from the cameras 8th , 9 continuously for objects in the vicinity of the vehicle 1 is sought. This creates heat signatures of objects in the vicinity of the vehicle 1 detected, whereby it is determined by means of the heat signatures whether the associated object is a living being. An automatic evaluation of a size and / or distance of the corresponding object from the vehicle takes place here 1 .

In a branch V3 it is checked whether this is a living being or person P2 detected object has a minimum size. If this is not the case (represented by a no branch N3), the operation O1, that is to say the scanning process, is carried out further.

If, on the other hand, the detected object has a size that at least corresponds to the minimum size, an approach and a speed of the object are determined in a further operation 02. If, for example, an increasing enlargement of the object is detected in the field of view, the vehicle system can estimate an approach speed based thereon.

In a branch V4, a check is made as a function of the approach speed to determine whether there is a collision between the vehicle 1 and the person P2 is likely. If this is not the case (represented by a no branch N4), the operation O1, that is to say the scanning process, is carried out further. In addition, in a manner not shown in detail, corresponding driver information is provided within the vehicle in a further operation 1 issued. The driver information includes, for example, a message about the detection of a person P2 around the vehicle 1 . This is useful, for example, when driving overland at night, since only a small number of pedestrians or cyclists can be expected here. A vehicle driver can thus be provided with a valuable advance warning if, for example, people are at the edge of a country road due to a vehicle breakdown. In this context there is the possibility of using the cameras 8th , 9 recorded data with data from a navigation system of the vehicle 1 and / or to match data on a digital map. Thus, depending on a current position of the vehicle 1 , for example in a city center or on a country road, different responses are given to the vehicle driver.

On the other hand, there is a collision between the vehicle 1 and the person P2 likely or exceeds a collision probability between the vehicle 1 and the person P2 a predetermined value (represented by a yes branch J4), an emergency braking of the vehicle is automatically performed in an operation 03 1 executed. The process is then ended.

Using the method, it is thus possible to inform a vehicle driver that a cyclist is crossing. The vehicle system actively points out any dangerous situation that may arise and thus actively helps the vehicle driver to correctly grasp the current traffic situation and to react appropriately. Furthermore, the vehicle system is designed to determine approach speeds and, based thereon, trigger emergency braking. Furthermore, the vehicle system enables cyclists to recognize them early in various situations, even in the case of very poor lighting or poor or missing lighting.

In 5 is a traffic situation with a vehicle 1 and one in a vicinity of the vehicle 1 located person P2 shown. The person P2 is a cyclist in the illustrated embodiment.

The vehicle 1 includes a plurality of cameras 10 , 11 , which are each on a vehicle exterior mirror 12 , 13 are arranged. The cameras 10 , 11 are part of what is known as a surround view system for displaying the vehicle 1 and sharing an environment of the vehicle 1 from a bird's eye view. The cameras 10 , 11 each have a detection area E3 , E4 on.

In the illustrated embodiment, the vehicle is moving 1 towards an intersection, where the cyclists (person P2 ) approaches. The person P2 is on a cycle path running parallel to the main road and intends to cross the lane on which the vehicle is located 1 emotional. The cyclist has the right of way which the vehicle 1 must grant.

Vehicle drivers often recognize cyclists too late at road crossings and bike paths, and serious accidents can easily occur in these traffic situations. Such traffic situations are particularly critical at dusk and at night.

In order to significantly reduce the risk potential for various road users in such a traffic situation, it is provided that the cameras 10 , 11 the surroundings of the vehicle 1 is monitored. The captured camera images are continuously evaluated for light sources recognized from the side. If a light source is detected at the side, there is the possibility that this is an approaching cyclist and a corresponding warning message can be output to the driver. To increase the robustness of the vehicle system and to avoid false reports, the cameras 10 , 11 recorded data with data from a navigation system of the vehicle 1 and / or matched data from a digital map. The detection of lateral light sources is compared with the current traffic situation. A current position of the vehicle 1 allows an exact forecast of a traffic situation ahead using map data available in the navigation system. If an intersection situation with crossing cycle paths is recognized and a lateral light source is recognized at a certain distance from the intersection, an approaching cyclist can be assumed with great certainty and a warning message can be output.

6th shows a possible sequence of a method for detecting the person P2 around the vehicle 1 according to 5 by means of the in the vehicle exterior mirrors 12 , 13 arranged cameras 10 , 11 collected data.

After the method has been started, a first branch V1 checks whether a vehicle system is connected to one by means of the cameras 10 , 11 carried out monitoring of the surroundings of the vehicle 1 is active. If this is not the case (represented by a no branch N1), the method is ended.

If the vehicle system is active (represented by a yes branch J1), it is checked in a second branch V2 whether the vehicle is 1 moves, i.e. has a speed greater than zero. If this is not the case (represented by a no branch N2), the method is started again.

Is the vehicle 1 in motion (represented by a yes branch J2), a scanning process is started in operation 01 in which the vehicle system evaluates the data from the cameras 10 , 11 continuously for light sources in the vicinity of the vehicle 1 is sought.

In a branch V3 it is checked whether it is to the side of the vehicle 1 a light source is detected. If this is not the case (represented by a no branch N3), the operation O1, that is to say the scanning process, is carried out further.

On the other hand, it is to the side of the vehicle 1 If a light source is detected, corresponding driver information is generated within the vehicle in a further operation O2 1 issued. The driver information includes, for example, a message about the detection of an approaching cyclist (person P2 ) in the vicinity of the vehicle 1 . In this context there is the possibility of using the cameras 10 , 11 collected data according to the description 5 with data from the vehicle's navigation system 1 and / or to match data on the digital map.

In a further refinement, which is not detailed, it can be provided that when a predetermined probability of collision is exceeded between the vehicle 1 and the person P2 automatic emergency braking of the vehicle 1 is performed.

It is thus possible by means of the method to support a vehicle driver, particularly in poor lighting conditions, in mastering complex traffic situations better and in avoiding accidents with cyclists. The hardware required for this, i.e. the cameras 10 , 11 and a required evaluation, is in many vehicles 1 already available. In particular, the vehicle driver is supported by the vehicle system in recognizing approaching cyclists who are crossing the vehicle's lane early at night or at dusk. Alternatively or additionally, the vehicle system automatically triggers an emergency braking.

List of reference symbols

1

vehicle

2 to 7

Radio sensor

8 to 11

camera

12, 13

Vehicle exterior mirror

E1 to E4

Detection area

J1 to J5

Yes branch

N1 to N5

No branch

O1 to 06

surgery

P1, P2

person

V1 to V5

branch

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed 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

• EP 2571005 A1 [0002]

Claims (9)

Method for the detection of people (P1, P2) in the surroundings of a vehicle (1), characterized in that - a person (P1, P2) located in the surroundings of the vehicle (1) by means of at least one vehicle's own camera (8 to 11) and / or is detected by means of at least one vehicle's own radio sensor (2 to 5) recorded by the person (P1, P2) carried radio transmitters and / or radio receivers and - by means of the at least one camera (8 to 11) and / or by means of the at least a radio sensor (2 to 5) recorded data can be compared with data from a navigation system of the vehicle (1) and / or data from a digital map. Procedure according to Claim 1 , characterized in that - while the vehicle (1) is moving, the surroundings of the vehicle (1) are continuously scanned by means of several radio sensors (2 to 5) and radio receivers in the vicinity of the vehicle (1) are searched for. Procedure according to Claim 2 , characterized in that when at least one radio receiver is detected, driver information is output within the vehicle (1). Procedure according to Claim 2 or 3 , characterized in that - when at least one radio receiver is detected, a bearing is carried out for the same and - a trajectory of the radio receiver is determined on the basis of the bearing. Procedure according to Claim 4 , characterized in that a collision probability between the vehicle (1) and a person (P1, P2) carrying the radio transmitter is determined as a function of the trajectory of the radio receiver. Method according to one of the preceding claims, characterized in that - at least one camera (10, 11) arranged on a vehicle exterior mirror is used as the camera (10, 11), - light sources in the vicinity of the vehicle by means of the camera (10, 11) 1) are detected, - with at least one light source detected laterally next to the vehicle (1), an approach of a cyclist is concluded and corresponding driver information is output within the vehicle (1). Method according to one of the preceding claims, characterized in that - at least one thermal imaging camera is used as the camera (8, 9), - while the vehicle (1) is moving, the thermal imaging camera continuously detects heat signatures of objects in the vicinity of the vehicle (1) and - based on the heat signatures it is determined whether the associated object is a living being. Procedure according to Claim 7 , characterized in that - an object recognized as a living being is observed and - a collision probability between the vehicle (1) and the object is determined as a function of the results of the observation. Method according to one of the preceding claims, characterized in that when a predetermined value of a collision probability between the vehicle (1) and a person (P1, P2) is exceeded, corresponding driver information is output within the vehicle (1) and / or automatic Emergency braking is carried out.

Images