DE102013202463A1 - Method for determining movement model of vulnerable road user i.e. motorized road user e.g. electrical bicycle riders, involves determining predicted position of vulnerable road user by motion model of vulnerable road user - Google Patents

Abstract

The method involves determining whether a vulnerable road user (VV) is in environment of a vehicle (1). Position of the vulnerable road user is determined if the vulnerable road user is in the environment of the vehicle. Movement influence characteristics are determined. A motion model of the vulnerable road user is determined based on the calculated motion parameters influence and the determined position of the vulnerable road user. Predicted position of the vulnerable road user is determined by the motion model of the vulnerable road user. An independent claim is also included for a device for determining a movement model of a vulnerable road user.

Description

The invention relates to a method and a device for determining a movement model of a vulnerable road user.

Driver assistance and safety systems in vehicles often use dynamic models of participating road users for the prediction. This can lead to standing traffic participants possibly not being taken into account in the prediction.

The object on which the invention is based is, on the one hand, to provide a method and, on the other hand, a device which contributes to a precise movement model of a vulnerable road user being determined.

The object is solved by the features of the independent claims. Advantageous embodiments of the invention are characterized in the subclaims.

The invention is characterized on the one hand by a method and on the other hand by a corresponding device for determining a movement model of a vulnerable road user. It is determined, in particular by means of an environment detection, whether a vulnerable road user is located in an environment of a vehicle. If it has been determined that a vulnerable road user is in the vicinity of the vehicle, the position of the vulnerable road user is determined. Motion influence characteristics are determined that are representative of a given situation and / or a given device and / or a predetermined significant point that may have an influence on a movement of the detected vulnerable road user. Depending on the determined movement influencing parameters and the determined position of the vulnerable road user, the movement model of the vulnerable road user is determined, by means of which a predicted position of the vulnerable road user can be determined.

In one possible implementation, the movement influencing parameters are determined in particular on the basis of an environment detection by the vehicle. By the environment detection objects are recognized in the environment of the vehicle, the objects are assigned to certain behaviors of the vulnerable road user. The position of the vulnerable road user in relation to the objects is associated with a possible behavior of the vulnerable road user. The behaviors may depend on the position of the vulnerable road user in relation to the object. For example, if the vulnerable road user is on the side of the road opposite a bus stop, there is a certain likelihood that the vulnerable road user will cross the road to get to the bus stop. This behavior can be expressed by a motion vector and a likelihood of the vulnerable road user moving according to the motion vector. Motion vector and associated probability then form the motion influence parameter.

Vulnerable road users in this context mean non-motorized or light motorized road users, such as pedestrians, skateboarders, cyclists, electric cyclists, pedelec drivers and the like. The vulnerable road user can be determined for example by means of vehicle sensors. Alternatively or additionally, it is also possible for the vulnerable road user to carry a transmitter, for example in his mobile telephone and / or in clothing and / or in other entrained objects and / or appliances, such as a biosensing transmitter, for example, and the vehicle has a corresponding receiver by means of which the vulnerable road user can be detected. Alternatively or additionally, it is also possible that the vulnerable road user is detected via vehicle-to-vehicle communication from another vehicle. The movement influencing parameters can be determined, for example, by means of suitable vehicle sensors and / or from data provided, for example, a navigation device and / or a server and / or another vehicle.

In this way, it is easily possible, after detection of a vulnerable road user, to take into account determined movement influence parameters in the determination of the movement model. Thus, for example, a movement model can be determined for stationary and / or moving vulnerable road users. Thus, a precise movement model for vulnerable road users can be determined, by means of which a predicted position of the vulnerable road user can be determined.

According to an advantageous embodiment, a physical orientation of the vulnerable road user is determined and the movement model additionally determined depending on the determined physical orientation. With the physical orientation in this context, the orientation of the vulnerable road users in the environment of the vehicle, or the Meant alignment in the world. For example, the physical orientation of legs of a pedestrian can be an indication of which direction he wants to move and / or the orientation of the face an indication of which direction he wants to move. In a cyclist, for example, the orientation of the front wheel with respect to the rear wheel may be an indication of which direction the road user wants to move. The physical orientations can thus be used, where appropriate, to determine a very precise movement model. In this case, the physical orientation can be determined, for example, by means of suitable vehicle sensors, such as, for example, by means of a camera and / or a stereo camera and / or by means of laser sensors.

According to a further advantageous embodiment, movements of the vulnerable road user are determined and the movement model additionally determined depending on the determined movements. Such movements may be, for example, changes in the direction of the head, which may be, for example, an indication of a road crossing. Alternatively or additionally, movement sequences, such as stepping or stopping in change, can be an indication of a precise determination of the movement model. The movements can thus be used, where appropriate, to determine a very precise movement model. The movements can be determined, for example, by means of suitable vehicle sensors, such as, for example, by means of a camera.

According to a further advantageous embodiment, historical movement data of the vulnerable road user are provided and the movement model is additionally determined as a function of the historical movement data. The historical movement data may be, for example, older determined movement models of the vulnerable road user and / or older detected movements of the vulnerable road user and / or older determined movement models and / or older detected movements of another vulnerable road user in the same location. In this way, for example, a crossing behavior of a road of a vulnerable road user may be taken into account, depending on the time and / or date when determining the movement model. By taking into account the historical movement data, a very precise movement model can thus possibly be determined.

It is thus possible by additional information to the vulnerable road user itself, in particular the physical orientation, the movement and / or historical movement data of the vulnerable road user to improve the movement model. This can be done, in particular, by confirming or refuting assumptions on which the determination of the motion influence parameters is based. It is also possible for the movement influencing parameters to be adjusted based on the information. In the bus stop example above, this may mean that the likelihood associated with the motion vector is adjusted depending on whether the additional information indicates that the vulnerable road user will cross the road or not.

According to a further advantageous embodiment, calendar data of the vulnerable road user are provided and the movement model is additionally determined as a function of the calendar data. Such calendar data can be provided for example by a smartphone of the vulnerable road user and / or by a server. For example, you can have information about the appointments of the vulnerable road user. In this way, if appropriate, the motion model can be determined very precisely by taking account of calendar data.

According to a further advantageous embodiment, the motion influence parameters are determined as a function of map data of a digital map. From map data of a digital map, for example, intersections and / or pedestrian crossings and / or bus stops and similar situations and / or devices and / or significant points can be determined which may have an influence on a movement of the vulnerable road user.

According to a further advantageous embodiment, the movement influence parameters are determined as a function of a measurement signal of at least one environmental sensor. Such environmental sensors can be, for example, cameras and / or ultrasound sensors and / or similar sensors, by means of which situations and / or devices and / or significant points can be easily determined, which can have an influence on a movement of the vulnerable road user, such as parked Cars, buses and / or other road users and the like.

The invention is also based on the following considerations:
Driver assistance and safety functions mainly use dynamic models of the participating road users for the prediction of a situation and the calculation of a collision risk. Essential data that are incorporated into the models are position, speed and orientation.

The prediction of a situation based solely on dynamic models underestimates the criticality of a situation where a vulnerable road user stands, and thus has speed v = 0 in the dynamic model but intends to move.

In most cases, the assumption of uniform motion is additionally made in this case (that is, acceleration = 0). This occurs, for example, when people intend to cross the road behind buses at stops or between parked vehicles. The people stand first and then cross the street.

Thus, the use of environmental, localization, Inertialsensorik on vehicles or road users, communication systems, digital maps, biosensors and other electronic systems and to detect situations that force vulnerable road users to stand still advantageous.

• a) aus dem Verhalten des verletzlichen Verkehrsteilnehmers, beispielsweise: – Zuwendung zur Straße – Gesichtserkennung in hochauflösender Kamerasensorik mit nachfolgender Kopfrichtungserkennung – Wechsel der Kopfrichtung zur Beobachtung der Verkehrs – Bewegungsabläufe, wie Schritt und Stopp im Wechsel – Vorderrad eines Fahrrads
• b) aus dem Kontext von anderen statischen oder dynamischen Objekten, beispielsweise: – Bus an Haltestelle – Parkende Fahrzeuge – Ladezonen – Ampelphasen – Fußgängerübergang
• c) aus dem zeitlichen Kontext, beispielsweise: – Busfahrplan – Überquerungsverhalten einer Person wochentags- und tageszeitabhängig täglich zur gleichen Uhrzeit – Anlieferzeiten in Ladenstraßen

The intention to cross the road can be concluded:

• a) from the behavior of the vulnerable road user, for example: - turning to the road - face recognition in high-resolution camera sensors with subsequent head direction detection - change of head direction to observe the traffic - movements, such as step and stop in the change - front wheel of a bicycle
• b) from the context of other static or dynamic objects, for example: - bus to stop - parking vehicles - loading zones - traffic light phases - pedestrian crossing
• c) from the temporal context, for example: - bus timetable - a person's day-to-day and daily-day-time crossing behavior at the same time of day - delivery times in shopping streets

The advantage is based on an enrichment of dynamic models to behavioral models and consideration of the context, to achieve an improved situation prediction or to allow in previously unaddressable situations in the first place security functions.

• – Zum Schutz von Personen, die vor und hinter Bussen die Straße überqueren wollen
• – Zum Schutz von Personen, die zwischen parkenden Fahrzeugen die Straße überqueren wollen

This can be used as follows:

• - To protect people who want to cross the street in front of and behind buses
• - To protect people who want to cross the street between parked vehicles

Embodiments of the invention are explained in more detail with reference to the schematic drawings. Show it:

1 a vehicle and

2 a flowchart.

Elements of the same construction or function are identified across the figures by the same reference numeral.

1 shows a vehicle 1 , The vehicle 1 For example, at least one camera has KAM. The at least one camera KAM is, for example, at the front of the vehicle 1 and / or on one of the side mirrors of the vehicle 1 arranged. The vehicle 1 also has a control device SV.

The control device SV has a computing unit, a data and program memory and at least one interface. By means of the interface, the control device SV is signal-technically coupled to the camera KAM, for example, for receiving image data of the camera KAM. The control device SV is, for example, additionally coupled to other environmental sensors and / or other vehicle sensors for receiving data. Furthermore, the control device SV can also be coupled to a server and / or with a navigation device and / or with other vehicles for transmitting and receiving data.

Further shows 1 a vulnerable road user VV. Such vulnerable road users VV are, for example, non-motorized or light motorized road users, such as pedestrians, skateboarders, cyclists, e-bike riders, Pedelecfahrer and the like. The vulnerable road user VV has, for example, a transmitter SE, for example in a smartphone and / or in his clothing. The transmitter SE sends, for example, a signal that can be received by a receiver E that the vehicle 1 having. The control device SV is, for example, additionally signal-coupled to the receiver E, which receives the signal of the transmitter SE, for transmitting and receiving data.

The control device SV can also be used as a device for determining a movement model a vulnerable road user.

For determining a movement model BM of the vulnerable road user VV, at least one program is preferably stored in the data and program memory of the control device SV. The program is described below with reference to the flowchart of 2 explained in more detail.

The program that is executed in the control device SV is started in a step S1, in which variables can be initialized if necessary.

In a step S2, it is determined whether a vulnerable road user VV is in an environment of the vehicle 1 located. The vulnerable road user VV can, for example, by means of the camera KAM of the vehicle 1 and / or by means of other sensors of the vehicle 1 be detected. Alternatively or additionally, the vulnerable road user VV can be detected by means of the transmitter SE. Alternatively or additionally, the vulnerable road user VV can also be detected via vehicle-to-vehicle communication from another vehicle. It is determined that no vulnerable road users VV in the vicinity of the vehicle 1 is the program is continued in a step S17. It is determined that a vulnerable road user VV in the vicinity of the vehicle 1 is, then the program is continued in a step S3.

In step S3, the position POS of the vulnerable road user VV is determined. The POS position may be related to the vehicle, for example 1 and / or in relation to the world. The position POS can be determined, for example, by means of the same method by means of which the vulnerable road user VV was detected.

In a step S4, movement influencing parameters BEK that are representative of a predetermined situation and / or a predetermined device and / or a predetermined significant point that may have an influence on a movement of the detected vulnerable road user VV are determined. Such situations, devices and / or points are, for example, whether there is a bus stop in the area and / or whether there is a bus at the bus stop and / or a bus timetable of a bus and / or if parked vehicles are in the vicinity and / or if a loading zone is in the area and / or if there is a traffic light in the area and / or if a traffic light is switched to red or green and / or if a pedestrian crossing is nearby and / or if a single person is away from a group of people is positioned. If the motion influence parameters BEK are determined, the time and / or the date may also be taken into account if necessary. The movement influencing parameters BEK can be determined, for example, by means of a measuring signal of an environmental sensor, for example by means of image data from the camera KAM and / or from other vehicle sensors of the vehicle 1 be determined. Alternatively or additionally, they can be determined depending on map data of a digital map, which are provided for example by a navigation device. Alternatively or additionally, they can also be determined as a function of further data which are provided, for example, by a server and / or another vehicle by means of radio communication.

In a step S7, a physical orientation PA of the vulnerable road user VV is determined. The physical orientation PA is, for example, the orientation of the vulnerable road user VV with respect to the world. For example, an orientation of the legs of the vulnerable road user VV is determined and / or an orientation of the face and / or an orientation of a front wheel of a bicycle with respect to a rear wheel of the bicycle.

In an optional step S9, movements BE of the vulnerable road user VV are determined. Such movements BE, for example, by means of the camera KAM of the vehicle 1 and / or by means of other suitable sensors of the vehicle 1 be determined. Such movements BE, for example, change the direction of the head and / or movements such as step and stop in the change.

In an optional step S11, historical movement data HD of the vulnerable road user VV is provided. Such historical movement data HD are, for example, older determined movement models BM and / or older detected movements of the vulnerable road user VV. Such historical movement data HD can be obtained, for example, from a server by means of a wireless radio link and / or from a data memory of the vehicle 1 and / or provided by the data memory of the control device SV.

In an optional step S13, calendar data KD of the vulnerable road user VV is provided. The calendar data KD are provided, for example, by means of a wireless radio link from a smartphone of the vulnerable road user VV and / or from a server. Such calendar data KD have, for example, appointments of the vulnerable road user VV.

In a step S15, the movement model BM is determined. The movement model BM is determined as a function of the determined position POS and as a function of the determined movement influence parameters BEK. In addition, the movement model BM can be determined as a function of the ascertained movements BE of the vulnerable road user VV and / or as a function of the historical movement data HD and / or as a function of the calendar data KD.

By means of the movement model BM, a predicted position of the vulnerable road user VV can be determined.

In step S17, the program is ended and may optionally be started again in step S1.

By means of the movement model BM, for example, a predicted position can be determined, if appropriate, with an associated probability. For example, by means of the movement model BM, it can be determined, for example, where the injured road user VV is at a given future time with what probability. This can be, for example, for safety functions of the vehicle 1 be used.

LIST OF REFERENCE NUMBERS

1

vehicle

BE

movements

BEK

Movement influence parameters

BM

movement model

e

receiver

HD

historical movement data

CAME

camera

KD

Calendar data

PA

physical alignment

POS

position

VV

vulnerable road users

SE

transmitter

SV

control device

Claims (9)

Method for determining a movement model (BM) of a vulnerable road user (VV) in which - it is determined whether a vulnerable road user (VV) is in an environment of a vehicle (VV) 1 ) - if it has been determined that a vulnerable road user (VV) is present in the vicinity of the vehicle ( 1 ), the position (POS) of the vulnerable road user (VV) is determined, - motion influence parameters (BEK) are determined, which are representative of a given situation and / or a given device and / or a predetermined significant point that influences a movement of the detected vulnerable road user (VV) may have, - depending on the determined Bewegungseinflusskenngrößen (BEK) and the determined position (POS) of the vulnerable road user (VV) the movement model (BM) of the vulnerable road user (VV) is determined by means of a predicted position of the vulnerable road user (VV) can be determined. The method of claim 1, wherein a physical orientation (PA) of the vulnerable road user (VV) is determined and the movement model (BM) is determined in addition depending on the determined physical orientation (PA). Method according to Claim 1 or 2, in which movements (BE) of the vulnerable road user (VV) are determined and the movement model (BM) is additionally determined as a function of the movements (BE) determined. Method according to one of the preceding claims, in which historical movement data (HB) of the vulnerable road user (VV) are provided and the movement model (BM) is additionally determined as a function of the historical movement data (HB). Method according to one of the preceding claims, in which calendar data (KD) of the vulnerable road user (VV) are provided and the movement model (BM) is additionally determined as a function of the calendar data (KD). Method according to one of the preceding claims, in which the motion influence parameters (BEK) are determined as a function of map data (KD) of a digital map. Method according to one of the preceding claims, in which the movement influence parameters (BEK) are determined as a function of a measurement signal of at least one environmental sensor. Method according to one of the preceding claims, wherein the movement influence parameters (BEK) are determined as a function of objects detected in the surroundings of the vehicle and the position of the vulnerable road user in relation to these objects, wherein the objects are each assigned specific behaviors of the vulnerable road user. Device for determining a movement model (BM) of a vulnerable road user (VV), the device being designed to be Method according to one of Claims 1 to 8.

Images