DE102022129015A1 - METHOD AND DEVICE FOR CONTACTLESS PROVISION OF A FUNCTION IN A MOTOR VEHICLE - Google Patents

Abstract

Method for contactless provision of a function in a motor vehicle, comprising the steps: a) detecting (S2) an approach of a person (120) to the motor vehicle (100); and b) providing (S5) the function depending on the detection according to step a); wherein: (i) ultrasonic signals are transmitted and received; and (ii) the detection according to step a) is carried out using multi-target tracking depending on a Poisson Multi-Bernoulli Mixture Filter and the received ultrasonic signals.

Description

The present invention relates to a method and a device for contactless provision of a function in a motor vehicle, in particular for contactless opening of a door, in particular a tailgate, of a motor vehicle. The present invention further relates to a computer program product.

There are various methods and devices in use that allow a person to open the tailgate of a vehicle without contact. Common methods include detecting a gesture, such as a foot movement, especially a kick towards the bumper, and then triggering the opening of the tailgate. To detect the foot movement, systems with capacitive sensors, such as those from the EN 10 2009 025 212 A1 known, or with radar-based sensors, such as those from the EN 10 2016 220 084 A1 known, used.

From the WO 2021/037052 A1 A method is known in which, in a first step, ultrasonic sensors are used to detect whether a person is within a predetermined distance from the motor vehicle. If this is the case, an image is projected onto the ground and a camera is used to detect the person's reaction to the image. If the reaction corresponds to a predefined gesture, the tailgate is opened.

From the US 2017/009509 A1 A method is known in which proximity detection is first carried out. The proximity detection is carried out using a smart key. In a second step, a gesture is recognized. This can be done with a camera or with the help of infrared or ultrasonic signals.

• Aus der DE 10 2018 121 165 A1 ist ein Verfahren bekannt, das eine verlässliche Abschätzung der Umgebung eines Kraftfahrzeugs bezüglich des Vorliegens sowohl von dynamischen Objekten als auch von statischen Objekten ermöglichen soll. Hierbei wird eine Umfelderfassung der Umgebung des Kraftfahrzeugs mit einem Umfelderfassungssensor, z.B. auch einem Ultraschallsensor, unter Generierung von Messdaten durchgeführt. Die Abschätzung erfolgt mittels eines Bayes-Filters, wobei ein Poisson Multi-Bernoulli Mixture Filter als ein Beispiel für einen Bayes-Filter angeführt wird.

In particular, from the technical field of autonomous driving, different approaches for so-called multi-target tracking are known, in which several objects and their motion states are tracked, as described below as an example:

• From the EN 10 2018 121 165 A1 A method is known that is intended to enable a reliable estimation of the surroundings of a motor vehicle with regard to the presence of both dynamic objects and static objects. In this case, the surroundings of the motor vehicle are detected using an environment detection sensor, e.g. an ultrasonic sensor, and measurement data is generated. The estimation is carried out using a Bayes filter, with a Poisson Multi-Bernoulli Mixture Filter being cited as an example of a Bayes filter.

From the EN 11 2018 007 810 T5 A method and a system for object tracking in automotive applications is known. According to this publication, the detection and interpretation of dynamic objects in the environment are a major challenge for autonomous driving. A labeled multi-Bernoulli filter based on random finite set statistics is proposed to solve the multi-target tracking problem. In addition to radar sensors and optical sensors, especially cameras, ultrasonic or lidar sensors are also proposed.

Against this background, an object of the present invention is to provide an improved method and an improved device for contactless provision of a function in a motor vehicle, in particular for contactless opening of a door, in particular a tailgate, of a motor vehicle.

Accordingly, a method for contactless provision of a function in a motor vehicle is proposed, comprising the steps of: a) detecting a person approaching the motor vehicle; and b) providing the function depending on the detection according to step a); wherein: (i) ultrasonic signals are transmitted and received, which in particular have pulses or pulse sequences; and (ii) the detection according to step a) is carried out using multi-target tracking depending on a Poisson Multi-Bernoulli Mixture Filter and the received ultrasonic signals. In particular, during detection according to step a), at least at times only exactly one ultrasonic signal, in particular having pulses or pulse sequences, is transmitted at the same time.

Certain functions should be provided when the vehicle is switched off. For example, the doors of a motor vehicle should usually be opened when the vehicle is switched off. In order to keep power consumption low when the vehicle is switched off, fast and reliable detection should be possible, for example. In order to enable sufficiently fast detection of the person approaching the vehicle, a Poisson Multi-Bernoulli Mixture Filter is used. This makes it possible to detect the person approaching early and reliably. In particular, during detection, at least temporarily only exactly one ultrasonic signal, in particular comprising pulses or pulse sequences, can be emitted at the same time. For improved spatial resolution, it can be helpful if several ultrasonic sensors emit signals one after the other.

A Poisson Multi-Bernoulli Mixture Filter belongs to the class of Bayesian filters. Such mathematical matic filters and their application in tracking objects, especially in multi-target tracking, are known per se and are therefore not described in detail here. Basically, the use of a Bayes filter is a recursive, probabilistic method for estimating probability distributions for given observations or measurements. A statistical evaluation is therefore carried out using a so-called "multi-target Bayes filter", the implementation of which with a Poisson Multi-Bernoulli Mixture Filter can deliver reliable results for the task to be solved here with comparatively little computational effort.

The distance of an object to an ultrasonic sensor can be determined using known methods by measuring the transit time of the ultrasonic signal and the reflection by the object. Ultrasonic sensors have an ultrasonic transmitter and an ultrasonic receiver. The data from an ultrasonic sensor can be evaluated at different levels, e.g. with ASICs, central or stand-alone control units. With two ultrasonic sensors with overlapping measuring ranges, the positions of objects can be determined using the known methods of trilateration. If the ultrasonic sensors transmit at different times, this reduces the accuracy of the location determination for a moving object.

Echoes or peaks are detected using the ultrasonic signals that are sent and received. These echoes or peaks are assigned to existing or new objects that are represented by Bernoulli filters in the Poisson Multi-Bernoulli Mixture Filter. This can be done without a plausibility check, as incorrectly detected objects are eliminated during tracking. This means that objects located at the edge of the detection range of the ultrasonic sensors can also be quickly detected.

The objects are then tracked 2-dimensionally in the Poisson Multi-Bernoulli Mixture Filter. For a new object, this initially results in increased uncertainty with regard to its existence and exact location. After a few measurement cycles, i.e. when several ultrasonic signals have been emitted and corresponding reflections from the object have been received, this uncertainty is reduced.

The 2-dimensional movement of the objects with the 2-dimensional information about location and movement speed is tracked in the Poisson Multi-Bernoulli Mixture Filter and used to assign detected echoes or objects from further measurement cycles to the existing or new objects represented by Bernoulli filters in the Poisson Multi-Bernoulli Mixture Filter.

Because the Poisson Multi-Bernoulli Mixture Filter is able to track many objects simultaneously and to examine different hypotheses for each object simultaneously, it is well suited to deal with many disturbances in the environment and with objects with difficult-to-predict movement behavior, such as pedestrians.

While detecting the approach of a person in step a), one sensor may be active to detect the movement of an object and then activate multiple sensors, or multiple sensors may alternately emit the ultrasonic signals.

Ultrasonic sensors send out ultrasonic signals with a frequency or pitch of, for example, 40 to 60 kHz. Pulses or pulse sequences can be modulated onto this frequency as a carrier for a signal. In particular, several pulses can be sent one after the other as a pulse sequence as an ultrasonic signal. The pitch of the ultrasonic signal, i.e. the carrier frequency, can also be changed in an ultrasonic signal or between two ultrasonic signals. For ultrasonic measurements, the signals are usually sent at regular intervals with a period T. Depending on the required temporal resolution, ultrasonic sensors can send out signals with a suitable repetition rate, in particular of the pulses or pulse sequences, the reciprocal of the period T.

In one embodiment, prior to step a), a map of an environment of the parked motor vehicle is generated using ultrasonic signals; and the detection according to step a) is carried out using multi-target tracking depending on the Poisson Multi-Bernoulli Mixture Filter, the received ultrasonic signals and the generated map.

The resulting map of the environment can be used and processed as additional information in the step of detecting a person approaching the door. The parking position can be taken up, for example, by engaging the P position of an automatic transmission, setting the parking brake, engaging a gear when the vehicle is switched off, all of this in conjunction with detecting that the wheels are stationary using wheel speed sensors.

When using a few ultrasonic sensors, especially a maximum of two ultrasonic sensors, mapping the environment is particularly helpful because 2-dimensional tracking is supported by a priori knowledge. With a single ultrasonic sensor, only a distance from an object can initially be determined with certainty. Over several measurement cycles, an estimate can still be made using mapping and the Poisson Multi-Bernoulli Mixture Filter and the object can be tracked.

According to embodiments, at least temporarily, exactly one ultrasonic sensor is used simultaneously for transmitting and/or receiving the ultrasonic signals.

Between the active measurement cycles in which the ultrasonic signals are sent and/or received, there are signal pauses in which the ultrasonic sensors do not send and/or receive. The pauses serve to separate the repeating ultrasonic signals, in particular pulses or pulse sequences, from one another so that transit time measurements can be made for the ultrasonic signals and their reflections. In order for the ultrasonic sensors to use as little power as possible, these signal pauses should be as long as possible and the active measurement cycles in which the ultrasonic signals, in particular the pulses or pulse sequences, are sent should be as short as possible. The ultrasonic sensors can, for example, send ultrasonic signals in less than 10% of the period T or in less than 5% of the period T and send no signals in the rest of the period. However, if the time interval between the transmitted ultrasonic signals, in particular the pulses or pulse sequences, becomes too large, the temporal resolution decreases. An object can then approach the motor vehicle without it being able to be detected during the approach time.

According to one embodiment, the repetition rate of the ultrasound signals, in particular of the pulses or pulse sequences, is at least temporarily between 0.5 and 5 Hz, in particular between 1 and 3 Hz and/or at least temporarily between 3 and 20 Hz, in particular between 7 and 13 Hz.

The signal pauses have, at least temporarily, a duration of between 250 ms and 2000 ms, in particular between 500 ms and 1000 ms.

According to one embodiment, the measuring cycles have a duration between 20 ms and 60 ms, in particular between 30 ms and 50 ms.

A measurement cycle that can detect objects up to a distance of 4 m can typically take 25 ms due to the propagation speed and approximately 40 ms with data processing.

If several ultrasonic sensors are used, they do not have to transmit to each other at the same time or at a specific interval.

According to one embodiment, a birth model of the Poisson Multi-Bernoulli Mixture Filter is designed such that for objects newly detected using the received ultrasound signals, the associated probability of existence is location-dependent and at least partially increases with increasing distance of the location of the newly discovered object from the ultrasound sensor.

By designing the birth model of the Poisson Multi-Bernoulli Mixture Filter in this way, early tracking of objects can be carried out from the detection limit of the ultrasonic sensors.

According to one embodiment, a Murty's algorithm is used to assign the objects detected using the received ultrasound signals to the Bernoulli filters or to the birth model of the Poisson Multi-Bernoulli Mixture Filter, which assigns several objects detected using the received ultrasound signals to a Bernoulli filter.

Ultrasonic sensors usually have overlapping fields of view. Therefore, the method presented here should be able to assign several objects detected in the ultrasonic sensor data to a Bernoulli filter.

According to one embodiment, a gesture of the person is recognized between step a) and step b) using ultrasound signals, and the provision according to step b) takes place depending on the recognition according to step a) and the recognized gesture.

By requiring an additional gesture, the security of correctly providing the function can be further increased.

The gesture can include an arm movement, a foot movement towards a bumper of the motor vehicle, a step movement forwards and backwards or a step movement to the side and back. For example, when using at least two ultrasonic sensors, a step movement forwards and backwards or a step movement to the side and back can be easily detected and distinguished. Other movement patterns and gestures are of course also possible.

According to one embodiment, the motor vehicle has at least one ultrasonic sensor which is designed to be used for parking aid, wherein the number of ultrasonic signals according to (i) are transmitted and/or received by means of the at least one ultrasonic sensor.

Most modern vehicles already have ultrasonic sensors for parking assistance. If these sensors can be used to provide a contactless function, such as opening the trunk, there are no additional costs for additional sensors. For example, vehicles often have between four and six ultrasonic sensors in the rear area for parking assistance.

According to one embodiment, step a) and/or step b) is only carried out if or while a smart key belonging to the motor vehicle is in communication with the motor vehicle and the motor vehicle is in the park position.

For example, opening the door of the motor vehicle while driving should be ruled out. In addition, the ultrasonic sensors of the motor vehicle may be required for other functions such as parking assistance or parking space searches while driving. It is therefore advantageous if the proposed method for contactless provision of a function is only used in embodiments when the motor vehicle is in the parking position. The parking position can be determined, for example, by the P position of an automatic transmission, an applied parking brake, a switched-off motor vehicle with a gear engaged, all of this also in conjunction with detection of a standstill of the wheels using wheel speed sensors.

In addition, in embodiments, unauthorized persons should not be able to trigger the contactless provision of a function, e.g. neither intentionally nor inadvertently open the door of the motor vehicle. The risk of unauthorized persons opening the door can be reduced if a smart key is in communication with a control unit of the motor vehicle. Smart key includes a variety of motor vehicle access controls, such as radio keys, RFID cards and RFID transponders in keys, but also mobile devices that can serve as keys for the motor vehicle.

In addition, the power consumption of the proposed method can be reduced if the step of detecting a person approaching the door is only started if a smart key is authorized to provide the function without contact. Ultrasonic signals are only emitted if it is certain that an authorized person could approach. In embodiments, the proposed method can allow the smart key to be inside or outside the motor vehicle or only allow the smart key to be outside the motor vehicle. In the first case, for example, a passenger without the smart key can open a door, e.g. a trunk door, while the driver is in the motor vehicle with his smart key.

The proposed method is preferably used for the contactless provision of a function for opening a trunk door or tailgate. However, the proposed method for the contactless opening of a door of a motor vehicle can also be used to open a driver's door, a passenger door, a front door, a rear door, a frunk door or frunk flap or a door for a power connection or for a fuel filler opening. "Frunk" stands for "front trunk" and refers to the storage compartment, especially in electric cars, at the front of the vehicle.

According to embodiments, the function according to step b) comprises the contactless folding away or folding in of a mirror or extending of a trailer hitch.

When opening a door without contact, a target area can be defined for the door, where a person is expected to be when they want to open the door. If an object remains in the target area at the door, the object's movement history can be evaluated and it can be determined how the object moved into the target area. Based on defined criteria, e.g. whether the object moved towards the target area in a straight line or with changes in direction within defined limits and at a uniform speed, it can then be decided whether a person has approached the door. Corresponding criteria can also be defined for the contactless provision of other functions.

Furthermore, a device for contactless provision of a function in a motor vehicle is proposed, comprising: a detection unit for detecting a person approaching the motor vehicle; and a provision unit for providing the function depending on the detection using the first detection unit; wherein the detection unit is set up to transmit and receive an ultrasonic signal, which in particular has pulses or pulse sequences, and wherein: (i) ultrasonic signals are transmitted and received; and (ii) the detection according to step a) takes place using multi-target tracking depending on a Poisson Multi-Bernoulli Mixture Filter and the number of ultrasonic signals received. In particular During detection according to step a), at least temporarily only exactly one ultrasonic signal, in particular comprising pulses or pulse sequences, is emitted simultaneously.

The units described here, for example the detection unit or the provision unit, can be implemented by hardware and/or software. For example, the units can be implemented on a microprocessor together with associated storage means. The microprocessor together with storage means can, for example, be formed on a central control unit of the motor vehicle.

According to one embodiment, a motor vehicle has the device for contactless provision of a function.

Furthermore, a computer program product is proposed which comprises instructions which, when the program is executed by a computer, cause the computer to carry out the method described above.

A computer program product, such as a computer program means, can be provided or delivered, for example, as a storage medium, such as a memory card, USB stick, CD-ROM, DVD, or in the form of a downloadable file from a server in a network. This can be done, for example, in a wireless communications network by transmitting a corresponding file with the computer program product or the computer program means.

The embodiments and features described for the proposed device apply accordingly to the proposed method and, conversely, the features of the method also apply to the device.

Further possible implementations of the invention also include combinations of features or embodiments described above or below with respect to the exemplary embodiments that are not explicitly mentioned. The person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the invention.

• 1 zeigt eine schematische Ansicht eines Kraftfahrzeugs, an das sich eine Person annähert;
• 2 a) und b) zeigen unterschiedliche Annäherungspfade einer Person an die Türe eines Kraftfahrzeugs;
• 3 zeigt eine schematische Ansicht eines Ausführungsbeispiels, bei dem sich eine Person an ein Kraftfahrzeug annähert und nur ein Ultraschallsensor aktiv ist;
• 4 zeigt ein Flussdiagramm für ein Ausführungsbeispiel eines Verfahrens zum berührungslosen Öffnen einer Türe eines Kraftfahrzeugs;
• 5 zeigt schematisch vier Beispiele von Pulsen oder Pulsfolgen, die von einem Ultraschallsensor gesendet werden können; und
• 6 zeigt schematisch eine Vorrichtung zur Durchführung des Verfahrens zum berührungslosen Öffnen einer Türe eines Kraftfahrzeugs gemäß einem Ausführungsbeispiel.

Further advantageous embodiments and aspects of the invention are the subject of the dependent claims and the embodiments of the invention described below. The invention is explained in more detail below using preferred embodiments with reference to the attached figures.

• 1 shows a schematic view of a motor vehicle being approached by a person;
• 2 a) and b) show different approach paths of a person to the door of a motor vehicle;
• 3 shows a schematic view of an embodiment in which a person approaches a motor vehicle and only one ultrasonic sensor is active;
• 4 shows a flow chart for an embodiment of a method for contactless opening of a door of a motor vehicle;
• 5 shows schematically four examples of pulses or pulse sequences that can be sent by an ultrasonic sensor; and
• 6 shows schematically a device for carrying out the method for contactless opening of a door of a motor vehicle according to an embodiment.

In the figures, identical or functionally equivalent elements have been given the same reference numerals unless otherwise stated.

1 shows a schematic view of a motor vehicle 100 from a bird's eye view, which a person 120 is approaching. A tailgate closes the trunk of the motor vehicle 100 as a trunk door 110. The motor vehicle 100 is equipped with six ultrasonic sensors 131 to 136 of a parking assistance system in the rear area. The respective measuring direction of the ultrasonic sensors 131 to 136 is shown schematically in dotted lines, whereby the limitation does not represent the maximum detection range 145. The ultrasonic sensors 131 to 136 have an ultrasonic transmitter and a receiver. The data generated by the transmission and reception of ultrasonic signals from the ultrasonic sensors 131 to 136 are evaluated and forwarded at various levels. For example, a sensor-related evaluation for the respective sensor takes place with the help of ASICs. The data from the ultrasonic sensors 131 to 136 are then sent to a central or stand-alone control unit and evaluated there.

If the motor vehicle 100 is reversing, all ultrasonic sensors 131 to 136 are active and measure with a repetition rate of the ultrasonic signals, in particular the pulses or pulse sequences, so that both static and dynamic obstacles are quickly detected and the driver can be warned of the obstacles. If the motor vehicle 100 is put into the parking position, e.g. by engaging gear position P and switching off of the engine, the ultrasonic sensors 131 to 136 are switched off.

The person 120 now approaches the motor vehicle 100. The person 120 carries a smart key 150 with them. This can be a so-called fob, a key chain with a passive transponder in the form of an RFID token, or a card with an RFID chip, or can also be implemented via a mobile phone. By bringing the smart key 150 into the reception range of the motor vehicle 100 for the smart key 150, the ultrasonic sensors 133 and 134 are activated. These alternately send ultrasonic signals in order to detect an approach to the trunk door 110. The ultrasonic sensors 131, 132, 135 and 136 are inactive. The two ultrasonic sensors 133 and 134 operate with a measuring cycle of approximately 40 ms in which the ultrasonic signals are sent and received, with a carrier frequency or pitch of the ultrasonic signal between 40 and 60 kHz. Between these measurement cycles, the ultrasonic sensors 133 and 134 do not send any signals for 950 ms each. The signals from the two ultrasonic sensors 133 and 134 are offset from each other by 500 ms, so that an ultrasonic signal is sent every 500 ms. For a parking assistance system, this repetition rate with a half-second interval between the ultrasonic signals would be too low. In the situation of the parked motor vehicle, however, this low repetition rate of the ultrasonic signals reduces the energy consumption of the ultrasonic sensors 133 and 134 and is sufficient to detect the movement of the person 120.

As soon as the person 120 enters the detection area 145 of the two ultrasonic sensors 133 or 134, he or she can be recognized as an object. In 1 the person 120 is in the area of the measuring directions of both ultrasonic sensors 133 and 134 and is moving in the direction of the arrow. The position of the person 120 can be determined with the two ultrasonic sensors 133 and 134 by trilateration.

Reflections found in the ultrasonic sensor data of the ultrasonic sensors 133 and 134 are assigned, with their position information, to existing or new objects that are represented by Bernoulli filters in the Poisson Multi-Bernoulli Mixture Filter. This assignment takes place, with the corresponding higher uncertainty, already at the edge of the detection range 145 of the ultrasonic sensors 133 and 134. The birth model of the Poisson Multi-Bernoulli Mixture Filter is set up in such a way that the probability of discovering new objects at the edge of the detection range 145 of the ultrasonic sensors 133 and 134 is higher than closer to the ultrasonic sensors 133 and 134. This ensures that objects can be detected early. Incorrectly detected objects are eliminated again by the Poisson Multi-Bernoulli Mixture Filter during tracking on the basis of further ultrasonic sensor data.

The objects are then tracked 2-dimensionally in the Poisson Multi-Bernoulli Mixture Filter. For a new object, this initially results in increased uncertainty with regard to its existence and the exact location and speed of movement. After a few measurement cycles, i.e. when several ultrasonic signals have been emitted by the ultrasonic sensors 133 and 134 and reflections from the object have been received, this uncertainty is reduced.

A target area 140 can be defined for a door 110, where a person 120 is expected to be when they want to open the door 110. The target area 140 is a zone with a distance of 0.5 to 1 m from the rear of the motor vehicle and approximately 1 m wide. The target area 140 is detected by the ultrasonic sensors 133 and 134. If an object remains in the target area 140 on the door 110, the movement history of the object can be evaluated using the Poisson Multi-Bernoulli Mixture Filter and it can be determined how the object moved into the target area 140. According to defined criteria, e.g. whether the object has moved towards the target area in a straight line or with changes in direction within defined limits and at a uniform speed, it can then be decided whether a person 120 has approached the door 110 who wants to open the door 110.

If no approach of the person 120 to the trunk door 120 is detected, the method for contactless opening of the trunk door 120 is terminated after a predefined period of time and can be restarted, e.g. by a locking process using the smart key 150 or by disconnecting and re-establishing the communication between the smart key 150 and the motor vehicle 100.

If the approach of a person 120 to the trunk door 110 is detected, the door 110 can be opened or an opening gesture is waited for.

If the opening gesture is defined as a foot movement from the target area 140 in the direction of the right rear light 160, the ultrasonic sensor 134 may be sufficient to detect this opening gesture. Since the opening gesture alone does not decide whether the trunk door 110 is opened, but the approach of the person 120 was detected beforehand, the detection of the opening gesture alone can prevent more false detections. without leading to an increased occurrence of incorrect openings in the overall process.

If the movement of the foot of the person 120 from the target area 140 in the direction of the right rear light 160 is recognized as an opening gesture from the data of the ultrasonic sensor 134, the control unit sends a control command to an actuator 190 to open the trunk door 110. In 1 The trunk door 110 is designed as a tailgate and swings open by spring force after the trunk lock is released.

The 2 a and b show different movement paths 146 of a person 120 to the door 110 of a motor vehicle 100. In both cases, the path runs within the detection range 145 of the ultrasonic sensors 133 and 134. In 2 a a linear movement into the target area 140 is visible. If such an approach to the trunk door 110 is detected, the trunk door 110 can be opened directly based on the detected approach or an opening gesture can be waited for. 2 B shows a course of approach and movement away that is less clear. In this case, at least one additional opening gesture would be necessary to open the trunk 110.

3 shows an embodiment in which the approach of the person 120 to the trunk door 110 is to be detected at least temporarily with only one active ultrasonic sensor 134. Since the ultrasonic data of the one ultrasonic sensor 134 initially only provide distances to the ultrasonic sensor 134, the data is enriched with a priori knowledge. When the motor vehicle 100 was brought into the parking position, a mapping of the surroundings of the motor vehicle 100 was carried out with the help of all ultrasonic sensors 131 to 136. The position of objects that reflect ultrasound was determined and stored with the help of several ultrasonic sensors 131 to 136. In 3 the mapped objects are shown hatched. Curbs 161 enclose a tree 162 and there is a lamppost 163 on a sidewalk. By mapping the environment, with the help of the Poisson Multi-Bernoulli Mixture Filter and with the ultrasound data generated by the one ultrasound sensor 134, a sufficiently accurate 2-dimensional estimate of the position of objects can be made, so that the approach of the person 120 to the trunk door 110 can also be detected with just one ultrasound sensor 134. If the approach is assumed to be probable with the help of the one sensor 134, further or all ultrasound sensors 131 to 136 can be activated in order to detect the approach. Here too, the mapping can be used to support this. Alternatively, several ultrasound sensors 131 to 136 can be active and the mapping can be used as additional support.

4 shows a flow chart for a method for contactless opening of a trunk door 110 of a motor vehicle 100 according to an embodiment.

In a first step S1, the method for contactless opening of a trunk door 110 of a motor vehicle 100 is activated. It must be impossible for the trunk door 110 to open while driving due to an incorrectly performed method. Furthermore, it should be impossible for an unauthorized person 120 to open the trunk door 110. In addition, the ultrasonic sensors 131 to 136 are required for other purposes depending on the situation, e.g. for parking assistance.

To activate the method for contactless opening of the trunk door 110 of the motor vehicle 100, conditions are therefore set that must be met in order for the method to run. In S1, it is therefore checked whether the motor vehicle 100 is in the park position. For this purpose, wheel rotation sensors are evaluated and it is checked whether an automatic transmission is in position P and the parking brake is applied. It is also checked whether the motor vehicle 100 should be able to be opened. For this purpose, it is checked whether a smart key 150 belonging to the motor vehicle 100 is in or near the motor vehicle in communication with a control unit of the motor vehicle 100. Only if the motor vehicle is not locked or ready to be opened can it be expected that an authorized person 120 will attempt to open the trunk door 110. The conditions must be met to start the method for contactless opening of the trunk door 110 of the motor vehicle 100 and remain met during the method. If, for example, the parking brake is released and a gear is engaged, the procedure is terminated.

If the conditions for the method for contactless opening of the trunk door 110 of the motor vehicle 100 are met, it is detected in step S2 whether a person 120 is approaching the trunk door 110. For this purpose, the ultrasonic sensors 133 and 134 send out signals. Using the ultrasonic sensor data, objects in the detection range of the ultrasonic sensors 133 and 134 are then tracked in 2-dimensional form using the Poisson Multi Bernoulli Mixture Filter. If an object reaches the target area 140, the history of the movement data from the Poisson Multi Bernoulli Mixture Filter is used to determine the movement path 146, whether it is an approach approach to the door 110 or whether it is a movement past or away.

If the approach of a person 120 is detected in the movement path 146 of the object, the trunk door 110 can be opened or other conditions for opening are checked.

In a step S3, the ultrasonic sensors 133 and 134 are used to check whether the person 120 remains standing in the target area 140. This can also be done using the Poisson Multi Bernoulli Mixture Filter. If it has been determined in S3 that the person 120 remains standing, a predefined opening gesture can be recognized in a step S4. For example, it can be recognized from the data of the ultrasonic sensors 133 and 134 that the person 120 moves forwards and backwards in the target area 140.

If a predefined opening gesture is detected in step S4, a control signal is sent to an actuator 190 in step S5 to open the trunk door 110. The actuator 190 can, for example, be an electric motor that unlocks and opens a trunk door 110.

5 shows schematically four examples of ultrasonic signals as pulses or pulse sequences that can be sent by an ultrasonic sensor. In 5 a A square pulse is modulated onto the carrier frequency of the ultrasound, e.g. 50 kHz. The square pulse is re-emitted after a period T. The repetition rate of the ultrasound signal, in the example of the 5 of the pulses or pulse sequences, is the reciprocal of the period T. In 5 b the transmitted pulse has a modulation of the amplitude and is also transmitted with the repetition rate 1/T. In 5c a pulse sequence is shown that is sent with a repetition rate of 1/T, the ultrasound signal is the combination of the rectangular pulse with the pulse with variable amplitude, the signal pause is the time until the next rectangular pulse. In 5 days In addition to the amplitude of the pulse, the carrier frequency is also modulated, but the repetition rate 1/T is the same as in the other examples. In all four examples, the signal portion of the period is less than 50%, i.e. most of the time no signal is sent. For the proposed method, it is advantageous if the signal portion of the period is even lower so that the power consumption of the ultrasonic sensor is reduced, e.g. less than 90%. However, the repetition rate of the pulses or pulse sequence cannot be arbitrarily reduced, since otherwise it would be possible to approach the motor vehicle between the signals without being detected. The duration of the measurement sequence within which an ultrasonic sensor sends and receives signals can, in embodiments, be between 20 and 60 ms, in particular between 30 and 50 ms. The duration of the signal pause can, in embodiments, be between 250 ms and 2000 ms, in particular between 500 ms and 1000 ms.

6 shows schematically a device for carrying out the method for contactless opening of a door 110 of a motor vehicle 100 according to an embodiment.

Four ultrasonic sensors 131 to 134 are each provided with an ASIC 200 close to the sensor for preprocessing the sensor data and for sending the sensor data via a data bus 210 to a computer 170, whereby the computer 170 can be part of a motor vehicle control unit. The ultrasonic sensors 131 to 134 measure their measuring range with ultrasonic signals of a first repetition rate, in particular the pulses or pulse sequences, of 1 Hz each with a measuring cycle of 50 ms and a signal pause of 950 ms. The signals of the ultrasonic sensors 131 to 134 are shifted against each other so that every 250 ms one of the ultrasonic sensors sends out an ultrasonic signal. A first module 171 of the computer 170 now processes the sensor data in such a way that it detects reflections in the ultrasonic sensor data and generates information about the possible position of objects from the ultrasonic sensor data and makes this data available to a second module 172 (here also "recognition unit"). The second module 172 carries out multi-target tracking using a Poisson Multi-Bernoulli Mixture Filter and detects an approach of a person 120 to the door 110. To do this, the second module 172 evaluates trajectories from the history of the Poisson Multi-Bernoulli Mixture Filter for objects that move into a target area 140. If the trajectories correspond to predetermined or learned parameters for persons 120 who approach the door 110 in order to open the door 110, then a third module 173 generates a control command that is transmitted via a control line 180 to an actuator 190 that opens a door 110 and ends the method for contactless opening of a door 110 of a motor vehicle 100. The third module 173, the control line 180 and the actuator 190 are collectively referred to here as a "provisioning unit". The first, second and third modules 171, 172, 173 can in particular be designed as an "embedded system", i.e. as an application-specific combination of hardware and software.

Although the present invention has been described using exemplary embodiments, it can be modified in many ways.

LIST OF REFERENCE SYMBOLS

100

Motor vehicle

110

Trunk door

120

person

131 to 136

Ultrasonic sensors

140

Target area

145

Detection range

146

Movement path

150

Smart Key

160

right tail light

161

curb

162

Tree

163

Lamppost

170

computer

171

first module

172

second module

173

third module

180

Control line

190

Actuator

200

Sensor-near ASIC

210

Data bus

QUOTES INCLUDED IN THE DESCRIPTION

This list of 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 accepts no liability for any errors or omissions.

Cited patent literature

• DE 102009025212 A1 [0002]
• DE 102016220084 A1 [0002]
• WO 2021/037052 A1 [0003]
• US 2017009509 A1 [0004]
• DE 102018121165 A1 [0005]
• EN 112018007810 T5 [0006]

Claims (14)

Method for contactless provision of a function in a motor vehicle (100) with the steps: a) detecting (S2) an approach of a person (120) to the motor vehicle (100); and b) providing (S5) the function depending on the detection according to step a); wherein: (i) ultrasonic signals are sent and received; and (ii) the detection according to step a) is carried out using multi-target tracking depending on a Poisson Multi-Bernoulli Mixture Filter and the received ultrasonic signals. Procedure according to Claim 1 , wherein: before step a), a map of an environment of the parked motor vehicle (100) is generated using ultrasonic signals; and the detection according to step a) is carried out using multi-target tracking depending on the Poisson Multi-Bernoulli Mixture Filter, the received ultrasonic signals and the generated map. Procedure according to Claim 1 or 2 , wherein, at least temporarily, exactly 1 ultrasonic sensor (131-136) is used simultaneously for transmitting and/or receiving the ultrasonic signals. Method according to one of the preceding claims, wherein the repetition rate of the emitted ultrasonic signals is at least temporarily between 0.5 and 5 Hz, in particular between 1 and 3 Hz and/or at least temporarily between 3 and 20 Hz, in particular between 7 and 13 Hz. Method according to one of the preceding claims, wherein a birth model of the Poisson Multi-Bernoulli Mixture Filter is designed such that for objects newly detected using the received ultrasound signals, the associated probability of existence is location-dependent and at least partially increases with increasing distance of the location of the newly discovered object from the ultrasound sensor (131-136). Method according to one of the preceding claims, wherein a modified Murty's algorithm is used to assign the objects detected using the received ultrasound signals to the Bernoulli filters or to the birth model of the Poisson Multi-Bernoulli Mixture Filter, which assigns several objects detected using the received ultrasound signals to a Bernoulli filter. Method according to one of the preceding claims, wherein: between step a) and step b) a gesture of the person (120) is recognized using ultrasound signals (S4); and the provision according to step b) takes place depending on the recognition according to step a) and the recognized gesture. Procedure according to Claim 7 , wherein the gesture comprises an arm movement, a foot movement towards a bumper of the motor vehicle (100), a stepping movement forwards and backwards, or a stepping movement sideways and backwards. Method according to one of the preceding claims, wherein the motor vehicle (100) has at least one ultrasonic sensor (131 - 136) which is designed to be used for a parking aid, wherein the ultrasonic signals according to (i) are transmitted and/or received with the aid of the at least one ultrasonic sensor (131 - 136). Method according to one of the preceding claims, wherein step a) and/or step b) is carried out when or while a smart key (150) belonging to the motor vehicle (100) is in communication with the motor vehicle (100) and the motor vehicle (100) is in the parking position. Method according to one of the preceding claims, wherein the function according to step b) comprises a contactless opening of a driver's door, a passenger door, a front door, a rear door, a trunk door (110) or tailgate, a front door or front flap, or a door for a power connection or for a fuel filler opening. Device for contactless provision of a function in a motor vehicle (100), comprising: a detection unit (172) for detecting an approach of a person (120) to the motor vehicle (100); and a provision unit (173, 180, 190) for providing the function depending on the detection using the first detection unit; wherein the detection unit (172) is set up to transmit and receive an ultrasonic signal and where: (i) an ultrasonic signal is transmitted and received; and (ii) the detection according to step a) is carried out using multi-target tracking depending on a Poisson Multi-Bernoulli Mixture Filter and the received ultrasonic signals. Motor vehicle (100) with a device according to Claim 12 . Computer program product comprising instructions which, when the program is executed by a computer (170), cause the computer (170) to carry out the method according to one of the Claims 1 - 11 to execute.

Images