WO2020094323A1 - Method for classifying objects by means of an automatically driving motor vehicle, and automatically driving motor vehicle - Google Patents

Method for classifying objects by means of an automatically driving motor vehicle, and automatically driving motor vehicle Download PDF

Info

Publication number
WO2020094323A1
WO2020094323A1 PCT/EP2019/077313 EP2019077313W WO2020094323A1 WO 2020094323 A1 WO2020094323 A1 WO 2020094323A1 EP 2019077313 W EP2019077313 W EP 2019077313W WO 2020094323 A1 WO2020094323 A1 WO 2020094323A1
Authority
WO
WIPO (PCT)
Prior art keywords
passenger
motor vehicle
objects
automated
stored
Prior art date
Application number
PCT/EP2019/077313
Other languages
German (de)
French (fr)
Inventor
Robin Richter
Original Assignee
Volkswagen Aktiengesellschaft
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Volkswagen Aktiengesellschaft filed Critical Volkswagen Aktiengesellschaft
Publication of WO2020094323A1 publication Critical patent/WO2020094323A1/en

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W50/08—Interaction between the driver and the control system
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/40—Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor
    • G06F18/41—Interactive pattern learning with a human teacher
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2540/00—Input parameters relating to occupants
    • B60W2540/215—Selection or confirmation of options
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2554/00—Input parameters relating to objects
    • B60W2554/40—Dynamic objects, e.g. animals, windblown objects
    • B60W2554/402—Type
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
    • G06F18/24133—Distances to prototypes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
    • G06V10/449—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters
    • G06V10/451—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters with interaction between the filter responses, e.g. cortical complex cells
    • G06V10/454—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]

Definitions

  • the invention relates to a method for classifying objects by means of an automated motor vehicle and an automated motor vehicle.
  • Automated vehicles drive trajectories automatically, whereby the longitudinal as well as the lateral dynamics are regulated automatically.
  • Motor vehicles of this type also have at least one device for recording and classifying objects, the data of which are required, inter alia, for trajectory planning.
  • the device comprises an environmental sensor system and an evaluation unit.
  • the environment sensors consist of several cameras, for example. Lidar, radar and / or ultrasonic sensors can also be used.
  • the data from the environment sensors are then evaluated in the evaluation unit and objects found are classified, for example whether a detected object is a motor vehicle, a pedestrian, a tree or the like.
  • the evaluation unit can be a trained neural network, for example.
  • the object recognition and classification is very reliable. However, there are always situations in which objects or putative objects appear in the data from the environment sensors, for which the neural network, for example, was not learned.
  • the invention is based on the technical problem of improving a method for classifying objects by means of an automated motor vehicle and of creating such an improved automated motor vehicle.
  • the automated motor vehicle having a device for detection and classification of objects, as well as output and input means, has the method step that objects detected by the device, but which cannot be classified or cannot be clearly classified, are brought to the attention of at least one passenger of the automated motor vehicle via the output means.
  • the passenger is asked to classify the detected object, this step also being able to be carried out in advance and once, for example by informing the passenger that objects that are subsequently not classifiable are to be brought to the customer's attention, which objects are then to be classified.
  • the classifications of the at least one passenger are stored.
  • the passenger is asked to enter or confirm attributes of the object. For example, when classifying a restaurant, the passenger can be asked whether he knows the restaurant and can recommend it. With other classifications, however, this can be dispensed with.
  • the objects are preferably visually displayed to the at least one passenger, this then being able to take place in various ways.
  • an image of the object can be displayed on a display unit, the display unit being permanently installed in the motor vehicle or the display of a mobile device.
  • the object can be displayed, for example, on a head-up display.
  • the passenger may also have put on AR glasses (augmented reality), in which case the objects are imported.
  • the data of the objects now additionally classified can be used in a variety of ways.
  • the stored classifications of the objects are read in as training data in the device for recording and classifying in order to learn the evaluation unit in an improved manner.
  • the data can be stored in a digital street map. For example, a crane was recorded in the pictures and classified by the passenger. A construction site can now be entered on the basis of this information. This means that there is information that there are traffic disruptions due to construction vehicles at this point can come. But restaurants and similar objects can also be saved and output as points of interest.
  • the stored classifications with the associated image data can be transmitted to a vehicle-external neural network as training data.
  • a vehicle-external neural network For example, certain objects have been classified as birds. These are of no further interest for the navigation of the automated motor vehicle, but for a neural network by means of which birds are to be acquired from image data.
  • the inputs of the at least one passenger of a monetization device are transmitted and stored. This can create incentives for the passenger to classify the objects. It is also possible to create a game situation for the passengers, where they classify objects in competition with one another.
  • the classifications can also be validated if, for example, several passengers classify the same object. For example, only classifications can be adopted that are classified equally by all or at least the majority of the passengers.
  • Fig. 1 is a schematic block diagram of an automated motor vehicle
  • FIG. 2 shows a flowchart of a method for classifying objects.
  • the motor vehicle 50 has a device 1 for detecting and classifying objects.
  • the device 1 comprises an environmental sensor system 2, an evaluation unit 3 and a memory 4. Furthermore, the motor vehicle 50 has output means 5, input means 6 and one
  • the environment sensor system 2 comprises, for example, a large number of cameras, the data of which are transmitted to the evaluation unit 3.
  • the evaluation unit 3 determines objects in the data and classifies them.
  • the objects relevant to the automated journey such as other motor vehicles or obstacles driving ahead, are transferred to a trajectory planning device 8, the trajectory to be driven then being adapted as a function of the detected and classified objects.
  • the device 1 detects objects that the evaluation unit 3 cannot classify or cannot classify with sufficient certainty, the objects are brought to the attention of at least one passenger of the automated motor vehicle 50 via the output means 5 and the passenger is asked to classify the object.
  • the passenger can then use the input means 6 to classify the object and, if appropriate, enter attributes of the object.
  • the classification carried out is then stored in the memory 4 and the passenger's input is transmitted to the monetization device 7. It should be noted here that both the output means 5 and the input means 6 can be mobile, ie do not have to be an integral part of the motor vehicle.
  • the classifications stored in the memory 4 can then be, for example, as
  • Training data for neural networks are used and / or supplement digital road maps.
  • a first step S1 at least one passenger of the automated motor vehicle 50
  • a third step S3 the passenger classifies the object, the classified objects finally being stored in a step S4.
  • Reference symbol list

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Automation & Control Theory (AREA)
  • Evolutionary Computation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Human Computer Interaction (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Artificial Intelligence (AREA)
  • Medical Informatics (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Traffic Control Systems (AREA)
  • Navigation (AREA)

Abstract

The invention relates to a method for classifying objects by means of an automatically driving motor vehicle (50). The automatically driving motor vehicle (50) has a device (1) for detecting and classifying objects and output and input means (5, 6), wherein an object which has been detected but is not classifiable or positively classifiable is brought to the attention of at least one passenger of the automatically driving motor vehicle (50) via the output means (5), and the passenger is required to classify the detected object, the classifications made by the at least one passenger being stored. The invention also relates to an automatically driving motor vehicle (50).

Description

Beschreibung  description
Verfahren zum Klassifizieren von Objekten mittels eines automatisiert fahrenden Method for classifying objects using an automated vehicle
Kraftfahrzeuges und automatisiert fahrendes Kraftfahrzeug  Motor vehicle and automated motor vehicle
Die Erfindung betrifft ein Verfahren zum Klassifizieren von Objekten mittels eines automatisiert fahrenden Kraftfahrzeuges sowie ein automatisiert fahrendes Kraftfahrzeug. The invention relates to a method for classifying objects by means of an automated motor vehicle and an automated motor vehicle.
Automatisiert fahrende Kraftfahrzeuge fahren selbsttätig Trajektorien ab, wobei sowohl die Längs- als auch die Querdynamik automatisiert eingeregelt wird. Derartige Kraftfahrzeuge weisen auch mindestens eine Einrichtung zur Erfassung und Klassifizierung von Objekten auf, deren Daten unter anderem für die Trajektorienplanung benötigt werden. Eine solche Automated vehicles drive trajectories automatically, whereby the longitudinal as well as the lateral dynamics are regulated automatically. Motor vehicles of this type also have at least one device for recording and classifying objects, the data of which are required, inter alia, for trajectory planning. Such
Einrichtung umfasst eine Umfeldsensorik und eine Auswerteeinheit. Die Umfeldsensorik besteht beispielsweise aus mehreren Kameras. Zusätzlich können auch Lidar-, Radar- und/oder Ultraschallsensoren zum Einsatz kommen. Die Daten der Umfeldsensorik werden dann in der Auswerteeinheit ausgewertet und aufgefundene Objekte klassifiziert, beispielsweise ob es sich bei einem erfassten Objekt um ein Kraftfahrzeug, einen Fußgänger, einen Baum oder Ähnliches handelt. Dabei kann die Auswerteeinheit beispielsweise ein trainiertes neuronales Netz sein. Dabei ist die Objekterkennung und Klassifizierung sehr zuverlässig. Allerdings gibt es immer wieder Situationen, wo in den Daten der Umfeldsensorik Objekte oder vermeintliche Objekte auftauchen, auf die beispielsweise das neuronale Netz nicht angelernt wurde. The device comprises an environmental sensor system and an evaluation unit. The environment sensors consist of several cameras, for example. Lidar, radar and / or ultrasonic sensors can also be used. The data from the environment sensors are then evaluated in the evaluation unit and objects found are classified, for example whether a detected object is a motor vehicle, a pedestrian, a tree or the like. The evaluation unit can be a trained neural network, for example. The object recognition and classification is very reliable. However, there are always situations in which objects or putative objects appear in the data from the environment sensors, for which the neural network, for example, was not learned.
Der Erfindung liegt das technische Problem zugrunde, ein Verfahren zum Klassifizieren von Objekten mittels eines automatisiert fahrenden Kraftfahrzeugs zu verbessern sowie ein solches verbessertes automatisiert fahrendes Kraftfahrzeug zu schaffen. The invention is based on the technical problem of improving a method for classifying objects by means of an automated motor vehicle and of creating such an improved automated motor vehicle.
Die Lösung des technischen Problems ergibt sich durch ein Verfahren mit den Merkmalen des Anspruchs 1 sowie ein automatisiert fahrendes Kraftfahrzeug mit den Merkmalen des The solution to the technical problem results from a method with the features of claim 1 and an automated motor vehicle with the features of
Anspruchs 8. Weitere vorteilhafte Ausgestaltungen der Erfindung ergeben sich aus den Claim 8. Further advantageous embodiments of the invention result from the
Unteransprüchen. Subclaims.
Das Verfahren zum Klassifizieren von Objekten mittels eines automatisiert fahrenden The method for classifying objects using an automated vehicle
Kraftfahrzeugs, wobei das automatisiert fahrende Kraftfahrzeug eine Einrichtung zur Erfassung und Klassifizierung von Objekten sowie Ausgabe- und Eingabemittel aufweist, weist den Verfahrensschritt auf, dass von der Einrichtung erfasste Objekte, die aber nicht klassifizierbar sind oder nicht eindeutig klassifiziert werden können, über die Ausgabemittel mindestens einem Passagier des automatisiert fahrenden Kraftfahrzeugs zur Kenntnis gebracht werden. Der Passagier wird zur Klassifizierung des erfassten Objekts aufgefordert, wobei dieser Schritt auch vorab und einmalig erfolgen kann, indem beispielsweise dem Passagier mitgeteilt wird, dass diesem nachfolgend nicht klassifizierbare Objekte zur Kenntnis gebracht werden, die dieser nachfolgend klassifizieren soll. Abschließend werden dann die Klassifizierungen des mindestens einen Passagiers abgespeichert. Motor vehicle, the automated motor vehicle having a device for detection and classification of objects, as well as output and input means, has the method step that objects detected by the device, but which cannot be classified or cannot be clearly classified, are brought to the attention of at least one passenger of the automated motor vehicle via the output means. The passenger is asked to classify the detected object, this step also being able to be carried out in advance and once, for example by informing the passenger that objects that are subsequently not classifiable are to be brought to the customer's attention, which objects are then to be classified. Finally, the classifications of the at least one passenger are stored.
In einer Ausführungsform wird in Abhängigkeit der Klassifizierung durch den mindestens einen Passagier der Passagier aufgefordert, Attribute des Objekts einzugeben oder zu bestätigen. Beispielsweise bei Klassifizierung eines Restaurants kann der Passagier angefragt werden, ob er das Restaurant kennt und empfehlen kann. Bei anderen Klassifizierungen kann dies hingegen entbehrlich sein. In one embodiment, depending on the classification by the at least one passenger, the passenger is asked to enter or confirm attributes of the object. For example, when classifying a restaurant, the passenger can be asked whether he knows the restaurant and can recommend it. With other classifications, however, this can be dispensed with.
Vorzugsweise werden die Objekte dem mindestens einen Passagier optisch dargestellt, wobei dies dann verschiedenartig erfolgen kann. Beispielsweise kann ein Bild des Objekts auf einer Anzeigeeinheit dargestellt werden, wobei die Anzeigeeinheit fest im Kraftfahrzeug installiert sein kann oder aber das Display eines Mobil-Gerätes. Die Anzeige des Objekts kann dabei beispielsweise auf einem Head-up-Display erfolgen. Es ist aber auch möglich, die Blickrichtung des Passagiers zu erfassen und optisch mittels einer Projektion auf die Scheibe das Objekt hervorgehoben werden. Alternativ kann der Passagier auch eine AR-Brille (Augmented Reality) aufgesetzt haben, wobei dann die Objekte eingespielt werden. The objects are preferably visually displayed to the at least one passenger, this then being able to take place in various ways. For example, an image of the object can be displayed on a display unit, the display unit being permanently installed in the motor vehicle or the display of a mobile device. The object can be displayed, for example, on a head-up display. However, it is also possible to record the passenger's line of sight and to optically highlight the object by means of a projection onto the pane. Alternatively, the passenger may also have put on AR glasses (augmented reality), in which case the objects are imported.
Die Daten der so nun zusätzlich klassifizierten Objekte können vielfältig eingesetzt werden. The data of the objects now additionally classified can be used in a variety of ways.
In einer Ausführungsform werden die abgespeicherten Klassifizierungen der Objekte als Trainingsdaten in die Einrichtung zur Erfassung und Klassifizierung eingelesen, um so die Auswerteeinheit verbessert anzulernen. In one embodiment, the stored classifications of the objects are read in as training data in the device for recording and classifying in order to learn the evaluation unit in an improved manner.
Alternativ oder zusätzlich können die Daten in einer digitalen Straßenkarte abgespeichert werden. Beispielsweise wurde in den Bildern ein Kran erfasst und vom Passagier klassifiziert. Anhand dieser Information kann nun eine Baustelle eingetragen werden. Damit liegt eine Information vor, dass es an dieser Stelle zu Verkehrsstörungen aufgrund von Baufahrzeugen kommen kann. Aber auch Restaurants und ähnliche Objekte können abgespeichert werden und als Points of Interest ausgegeben werden. Alternatively or additionally, the data can be stored in a digital street map. For example, a crane was recorded in the pictures and classified by the passenger. A construction site can now be entered on the basis of this information. This means that there is information that there are traffic disruptions due to construction vehicles at this point can come. But restaurants and similar objects can also be saved and output as points of interest.
Alternativ oder zusätzlich können die abgespeicherten Klassifizierungen mit den zugehörigen Bilddaten einem fahrzeug-externen neuronalen Netz als Trainingsdaten übermittelt werden. Beispielsweise wurden bestimmte Objekte als Vögel klassifiziert. Diese sind für die Navigation des automatisiert fahrenden Kraftfahrzeugs nicht weiter von Interesse, jedoch für ein neuronales Netzwerk, mittels dessen aus Bilddaten Vögel erfasst werden sollen. Alternatively or additionally, the stored classifications with the associated image data can be transmitted to a vehicle-external neural network as training data. For example, certain objects have been classified as birds. These are of no further interest for the navigation of the automated motor vehicle, but for a neural network by means of which birds are to be acquired from image data.
In einer weiteren Ausführungsform werden die Eingaben des mindestens einen Passagiers einer Monetarisierungs-Einrichtung übermittelt und abgespeichert. Hierdurch können Anreize für den Passagier geschaffen werden, die Objekte zu klassifizieren. Auch ist es möglich, eine Spielsituation für die Passagiere zu schaffen, wo diese im Wettbewerb miteinander Objekte klassifizieren. Dabei kann auch eine Validierung der Klassifizierungen vorgenommen werden, wenn beispielsweise mehrere Passagiere das gleiche Objekt klassifizieren. So können beispielsweise nur Klassifizierungen übernommen werden, die von allen oder zumindest der Mehrheit der Passagiere gleich klassifiziert werden. In a further embodiment, the inputs of the at least one passenger of a monetization device are transmitted and stored. This can create incentives for the passenger to classify the objects. It is also possible to create a game situation for the passengers, where they classify objects in competition with one another. The classifications can also be validated if, for example, several passengers classify the same object. For example, only classifications can be adopted that are classified equally by all or at least the majority of the passengers.
Hinsichtlich der Ausgestaltung des automatisiert fahrenden Kraftfahrzeugs wird vollinhaltlich auf die vorangegangenen Ausführungen Bezug genommen. With regard to the design of the automated motor vehicle, reference is made in full to the preceding statements.
Die Erfindung wird nachfolgend anhand eines bevorzugten Ausführungsbeispiels näher erläutert. Die Figuren zeigen: The invention is explained in more detail below on the basis of a preferred exemplary embodiment. The figures show:
Fig. 1 ein schematisches Blockschaltbild eines automatisiert fahrenden Kraftfahrzeugs und Fig. 1 is a schematic block diagram of an automated motor vehicle and
Fig. 2 ein Flussdiagramm eines Verfahrens zum Klassifizieren von Objekten. 2 shows a flowchart of a method for classifying objects.
In der Fig. 1 ist schematisch ein automatisiert fahrendes Kraftfahrzeug 50 dargestellt. Das Kraftfahrzeug 50 weist eine Einrichtung 1 zur Erfassung und Klassifizierung von Objekten auf. Die Einrichtung 1 umfasst eine Umfeldsensorik 2, eine Auswerteeinheit 3 und einen Speicher 4. Weiter weist das Kraftfahrzeug 50 Ausgabemittel 5, Eingabemittel 6 sowie eine An automated motor vehicle 50 is shown schematically in FIG. 1. The motor vehicle 50 has a device 1 for detecting and classifying objects. The device 1 comprises an environmental sensor system 2, an evaluation unit 3 and a memory 4. Furthermore, the motor vehicle 50 has output means 5, input means 6 and one
Monetarisierungs-Einrichtung 7 auf. Die Umfeldsensorik 2 umfasst beispielsweise eine Vielzahl von Kameras, deren Daten an die Auswerteeinheit 3 übermittelt werden. Die Auswerteeinheit 3 ermittelt in den Daten Objekte und klassifiziert diese. Die für die automatisierte Fahrt relevanten Objekte wie beispielsweise vorausfahrende andere Kraftfahrzeuge oder Hindernisse werden an eine Trajektorienplanungs- Einrichtung 8 übergeben, wobei dann in Abhängigkeit der erfassten und klassifizierten Objekte die abzufahrende Trajektorie angepasst wird. Monetization facility 7. The environment sensor system 2 comprises, for example, a large number of cameras, the data of which are transmitted to the evaluation unit 3. The evaluation unit 3 determines objects in the data and classifies them. The objects relevant to the automated journey, such as other motor vehicles or obstacles driving ahead, are transferred to a trajectory planning device 8, the trajectory to be driven then being adapted as a function of the detected and classified objects.
Erfasst die Einrichtung 1 Objekte, die die Auswerteeinheit 3 nicht klassifizieren kann oder nicht ausreichend sicher klassifizieren kann, so werden die Objekte über das Ausgabemittel 5 mindestens einem Passagier des automatisiert fahrenden Kraftfahrzeugs 50 zur Kenntnis gebracht und der Passagier aufgefordert, das Objekt zu klassifizieren. Über das Eingabemittel 6 kann dann der Passagier das Objekt klassifizieren sowie gegebenenfalls Attribute des Objekts eingeben. Die vorgenommene Klassifizierung wird dann im Speicher 4 abgespeichert und die Eingabe des Passagiers an die Monetarisierungs-Einrichtung 7 übermittelt. Dabei sei angemerkt, dass sowohl die Ausgabemittel 5 als auch die Eingabemittel 6 mobil sein können, also nicht fester Bestandteil des Kraftfahrzeugs sein müssen. If the device 1 detects objects that the evaluation unit 3 cannot classify or cannot classify with sufficient certainty, the objects are brought to the attention of at least one passenger of the automated motor vehicle 50 via the output means 5 and the passenger is asked to classify the object. The passenger can then use the input means 6 to classify the object and, if appropriate, enter attributes of the object. The classification carried out is then stored in the memory 4 and the passenger's input is transmitted to the monetization device 7. It should be noted here that both the output means 5 and the input means 6 can be mobile, ie do not have to be an integral part of the motor vehicle.
Die im Speicher 4 abgespeicherten Klassifizierungen können dann beispielsweise als The classifications stored in the memory 4 can then be, for example, as
Trainingsdaten für neuronale Netze (allgemein deep machine learning) verwendet werden und/oder digitale Straßenkarten ergänzen. Training data for neural networks (generally deep machine learning) are used and / or supplement digital road maps.
In der Fig. 2 ist ein Flussdiagramm des Verfahrens dargestellt. Dabei wird in einem ersten Schritt S1 mindestens ein Passagier des automatisiert fahrenden Kraftfahrzeugs 50 2 shows a flowchart of the method. In a first step S1, at least one passenger of the automated motor vehicle 50
aufgefordert, unbekannte Objekte zu klassifizieren. In einem zweiten Schritt S2 werden erfasste, aber nicht klassifizierbare Objekte dem Passagier zur Kenntnis gebracht, asked to classify unknown objects. In a second step S2, detected but not classifiable objects are brought to the attention of the passenger,
vorzugsweise optisch bzw. visuell dargestellt. In einem dritten Schritt S3 klassifiziert der Passagier das Objekt, wobei schließlich die klassifizierten Objekte in einem Schritt S4 abgespeichert werden. Bezugszeichenliste preferably represented optically or visually. In a third step S3, the passenger classifies the object, the classified objects finally being stored in a step S4. Reference symbol list
Einrichtung Facility
Umfeldsensorik  Environment sensors
Auswerteeinheit  Evaluation unit
Speicher  Storage
Ausgabemittel  Output means
Eingabemittel  Input means
Monetarisierungs-Einrichtung  Monetization facility
Trajektionenplanungs-Einrichtung  Trajectory planning facility
automatisiert fahrendes Kraftfahrzeug automated motor vehicle
erster Schritt first step
zweiter Schritt second step
dritter Schritt Third step
vierter Schritt fourth step

Claims

Patentansprüche Claims
1. Verfahren zum Klassifizieren von Objekten mittels eines automatisiert fahrenden 1. Method for classifying objects using an automated driving
Kraftfahrzeugs (50), wobei das automatisiert fahrende Kraftfahrzeug (50) eine Einrichtung (1 ) zur Erfassung und Klassifizierung von Objekten sowie Ausgabe- und Eingabemittel (5, 6) aufweist,  Motor vehicle (50), the automated motor vehicle (50) having a device (1) for detecting and classifying objects and output and input means (5, 6),
dadurch gekennzeichnet, dass  characterized in that
das erfasste, aber nicht klassifizierbare oder nicht eindeutig klassifizierbare Objekt über die Ausgabemittel (5) mindestens einem Passagier des automatisiert fahrenden  the detected but not classifiable or not clearly classifiable object via the output means (5) of at least one passenger of the automated vehicle
Kraftfahrzeugs (50) zur Kenntnis gebracht werden, wobei der Passagier zur  Motor vehicle (50) are brought to the attention of the passenger
Klassifizierung des erfassten Objekts aufgefordert wird, wobei die Klassifizierungen des mindestens einen Passagiers abgespeichert werden.  Classification of the detected object is requested, the classifications of the at least one passenger being stored.
2. Verfahren nach Anspruch 1 , dadurch gekennzeichnet, dass in Abhängigkeit der 2. The method according to claim 1, characterized in that depending on the
Klassifizierung durch den mindestens einen Passagier der Passagier aufgefordert wird, Attribute des Objekts einzugeben oder zu bestätigen.  Classification by the at least one passenger asking the passenger to enter or confirm attributes of the object.
3. Verfahren nach Anspruch 1 oder 2, dadurch gekennzeichnet, dass die Objekte dem 3. The method according to claim 1 or 2, characterized in that the objects
mindestens einen Passagier optisch dargestellt werden.  at least one passenger can be visualized.
4. Verfahren nach einem der vorangegangenen Ansprüche, dadurch gekennzeichnet, dass die abgespeicherten Klassifizierungen als Trainingsdaten in die Einrichtung (1 ) zur Erfassung und Klassifizierung eingelesen werden. 4. The method according to any one of the preceding claims, characterized in that the stored classifications are read as training data in the device (1) for detection and classification.
5. Verfahren nach einem der vorangegangenen Ansprüche, dadurch gekennzeichnet, dass die abgespeicherten Klassifizierungen in einer digitalen Straßenkarte abgespeichert werden. 5. The method according to any one of the preceding claims, characterized in that the stored classifications are stored in a digital road map.
6. Verfahren nach einem der vorangegangenen Ansprüche, dadurch gekennzeichnet, dass die abgespeicherten Klassifizierungen mit den zugehörigen Bilddaten einem fahrzeug- externen neuronalen Netz als Trainingsdaten übermittelt werden. 6. The method according to any one of the preceding claims, characterized in that the stored classifications with the associated image data are transmitted to a vehicle-external neural network as training data.
7. Verfahren nach einem der vorangegangenen Ansprüche, dadurch gekennzeichnet, dass Eingaben des mindestens einen Passagiers einer Monetarisierungs-Einrichtung (7) übermittelt und von dieser abgespeichert werden. 7. The method according to any one of the preceding claims, characterized in that inputs of the at least one passenger of a monetization device (7) are transmitted and stored by the latter.
8. Automatisiert fahrendes Kraftfahrzeug (50), umfassend eine Einrichtung (1 ) zur Erfassung und Klassifizierung von Objekten sowie Ausgabe- und Eingabemittel (5, 6), 8. Automated motor vehicle (50), comprising a device (1) for detecting and classifying objects and output and input means (5, 6),
dadurch gekennzeichnet, dass  characterized in that
die Einrichtung (1 ) zur Erfassung und Klassifizierung von Objekten derart ausgebildet ist, dass erfasste, aber nicht klassifizierbare oder nicht eindeutig klassifizierbare Objekte über die Ausgabemittel (5) mindestens einem Passagier des automatisiert fahrenden  the device (1) for detecting and classifying objects is designed in such a way that detected, but not classifiable or not clearly classifiable, objects via the output means (5) of at least one passenger of the automated vehicle
Kraftfahrzeugs zur Kenntnis gebracht werden, wobei der Passagier zur Klassifizierung des erfassten Objekts aufgefordert wird, wobei die Einrichtung (1 ) weiter derart ausgebildet ist, dass die Klassifizierungen des mindestens einen Passagiers  Motor vehicle are brought to the knowledge, the passenger being asked to classify the detected object, the device (1) being further designed such that the classifications of the at least one passenger
abgespeichert werden.  can be saved.
9. Austomatisiert fahrendes Kraftfahrzeug nach Anspruch 8, dadurch gekennzeichnet, dass die Einrichtung (1 ) weiter derart ausgebildet ist, dass diese in Abhängigkeit der 9. Automated motor vehicle according to claim 8, characterized in that the device (1) is further designed such that it is a function of
Klassifizierung durch den mindestens einen Passagier den Passagier auffordert, Attribute des Objekts einzugeben oder zu bestätigen.  Classification by the at least one passenger requests the passenger to enter or confirm attributes of the object.
10. Automatisiert fahrendes Kraftfahrzeug nach Anspruch 8 oder 9, dadurch gekennzeichnet, dass die Einrichtung (1 ) zur Erfassung und Klassifizierung von Objekten mit einer Monetarisierungs-Einrichtung (7) datentechnisch verbunden ist. 10. Automated motor vehicle according to claim 8 or 9, characterized in that the device (1) for detecting and classifying objects with a monetization device (7) is connected in terms of data technology.
PCT/EP2019/077313 2018-11-09 2019-10-09 Method for classifying objects by means of an automatically driving motor vehicle, and automatically driving motor vehicle WO2020094323A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102018219125.5A DE102018219125A1 (en) 2018-11-09 2018-11-09 Method for classifying objects using an automated motor vehicle and automated motor vehicle
DE102018219125.5 2018-11-09

Publications (1)

Publication Number Publication Date
WO2020094323A1 true WO2020094323A1 (en) 2020-05-14

Family

ID=68210803

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2019/077313 WO2020094323A1 (en) 2018-11-09 2019-10-09 Method for classifying objects by means of an automatically driving motor vehicle, and automatically driving motor vehicle

Country Status (2)

Country Link
DE (1) DE102018219125A1 (en)
WO (1) WO2020094323A1 (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11875551B2 (en) * 2020-06-09 2024-01-16 Navbirswagen Aktiengesellschaft Collecting and processing data from vehicles
DE102021002918B4 (en) 2021-06-07 2023-04-06 Mercedes-Benz Group AG Method for detecting objects that are safety-relevant for a vehicle

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170308092A1 (en) * 2014-10-11 2017-10-26 Audi Ag Method for operating an automatically driven, driverless motor vehicle and monitoring system
US20180307925A1 (en) * 2017-04-20 2018-10-25 GM Global Technology Operations LLC Systems and methods for traffic signal light detection

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102012220146A1 (en) * 2012-11-06 2014-05-22 Robert Bosch Gmbh Method for characterizing driving behavior of driver of e.g. motor car, involves obtaining accumulation information of trends over deviation time and providing accumulation information for characterizing driving behavior
DE102013102087A1 (en) * 2013-03-04 2014-09-04 Conti Temic Microelectronic Gmbh Method for operating a driver assistance system of a vehicle
DE102014004675A1 (en) * 2014-03-31 2015-10-01 Audi Ag Gesture evaluation system, gesture evaluation method and vehicle
DE102014214507A1 (en) * 2014-07-24 2016-01-28 Bayerische Motoren Werke Aktiengesellschaft Method for creating an environment model of a vehicle
DE102015007493B4 (en) * 2015-06-11 2021-02-25 Audi Ag Method for training a decision algorithm and a motor vehicle used in a motor vehicle

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170308092A1 (en) * 2014-10-11 2017-10-26 Audi Ag Method for operating an automatically driven, driverless motor vehicle and monitoring system
US20180307925A1 (en) * 2017-04-20 2018-10-25 GM Global Technology Operations LLC Systems and methods for traffic signal light detection

Also Published As

Publication number Publication date
DE102018219125A1 (en) 2020-05-14

Similar Documents

Publication Publication Date Title
DE112012006226B4 (en) Driver assistance device
DE102012205316B4 (en) Navigation system and display method thereof
WO2015149971A1 (en) Method for analysing a traffic situation in an area surrounding a vehicle
DE102010039634A1 (en) Arrangement and method for traffic sign recognition
DE102018214510A1 (en) Parking ticket generation on the street
DE102018203583B4 (en) Method, driver assistance system and motor vehicle for the prediction of a position or a trajectory by means of a graph-based environment model
DE102016210534A1 (en) Method for classifying an environment of a vehicle
DE102015207123A1 (en) Driving assistance apparatus and method
DE102008041679A1 (en) Method for environment recognition for navigation system in car, involves storing data of object or feature in storage, and classifying object or feature by comparison of data after visual inspection of object or feature
DE102018116036A1 (en) Training a deep convolutional neural network for individual routes
WO2019007605A1 (en) Method for verifying a digital map in a more highly automated vehicle, corresponding device and computer program
DE102017209347A1 (en) Method and device for controlling a vehicle
WO2020094323A1 (en) Method for classifying objects by means of an automatically driving motor vehicle, and automatically driving motor vehicle
DE102017207441A1 (en) Method for checking a digital environment map for a driver assistance system of a motor vehicle, computing device, driver assistance system and motor vehicle
DE102017206344A1 (en) Driver assistance system for a vehicle
DE102014013298A1 (en) Method for operating a vehicle and driver assistance system
DE102016122200A1 (en) Forming a rescue lane considering their necessity
DE102018206743A1 (en) A method of operating a driver assistance system of an ego vehicle having at least one environment sensor for detecting an environment of the ego vehicle, computer-readable medium, system, and vehicle
DE102008043756A1 (en) Traffic sign information providing method for e.g. driver assistance system in vehicle, involves storing traffic sign data, assigning stored data to actual position of vehicle, and providing traffic sign information based on assigned data
DE102022002337B3 (en) Method for determining and providing a maximum permissible speed for vehicles and use of the method
DE102016214599A1 (en) Method for detecting traffic signs in motor vehicles
DE102018214506A1 (en) Method for further developing a driver assistance system and driver assistance system for a vehicle
DE102018112164A1 (en) A method of operating a driving support system of a vehicle to identify a particular driving situation and driving support system
DE102017011023A1 (en) Method for learning user behavior
DE102022102782A1 (en) Driver assistance system and method for operating a driver assistance system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19786556

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19786556

Country of ref document: EP

Kind code of ref document: A1