DE102019004075A1 - Method for determining a relevance of an object in an environment of a motor vehicle by means of a driver assistance system and driver assistance system - Google Patents

Abstract

The invention relates to a method for determining a relevance of an object (16, 20) in an environment (28) of a motor vehicle (10) for the motor vehicle (10) by means of a driver assistance system (12) of the motor vehicle (10), in which the Detection device (22) of the driver assistance system (12) detects object (16, 20) in the environment (28) of the motor vehicle (10) with an object stored in a storage device (24) of the driver assistance system (12) by means of an electronic computing device (26) of the Driver assistance system (12) is compared and, depending on the comparison, the object (16, 20) is identified by means of the electronic computing device (26), with a direction of view of a driver (32) of the motor vehicle (12) of the driver assistance system (12) being detected by a viewing direction. 10) is recorded and the relevance for the identified object (16, 20) is determined as a function of the recorded viewing direction. The invention also relates to a driver assistance system (12).

Description

The invention relates to a method for determining a relevance of an object in an environment of a motor vehicle for the motor vehicle by means of a driver assistance system of the motor vehicle according to the preamble of claim 1. Furthermore, the invention relates to a driver assistance system.

In the situation assessment of a driver assistance system, which can be designed in particular as a driving and safety system, the relevant ones have to be extracted from a large number of objects. It is known that when driving the motor vehicle, the human being focuses, among other things, on the objects which are of relevant relevance for carrying out the driving task. However, the intuitive human knowledge of object relevance cannot be transferred directly to a software system. The labeling known from the prior art, which characterizes the objects, of object data for automated driving and safety systems in motor vehicles is very complex and cost-intensive and often has to be carried out by hand.

Furthermore describes the DE 101 10 275 A1 a method for the identification of patterns that are available as stored information. A labeling proposal is automatically generated for the sample. This proposal can be generated separately from the actual labeling process, for example when the method for generating the proposal is very time-consuming. However, it is also possible to generate the proposal during the labeling process, for example when the method is able to generate the proposals in real time or almost in real time. As a result, the labeling suggestions for the observer are generated quickly and representative object recognition is made possible.

The object of the present invention is to provide a method and a driver assistance system by means of which the relevance of an object in the surroundings of a motor vehicle can be determined in an improved manner.

This object is achieved by a method and by a driver assistance system in accordance with the independent patent claims. Advantageous embodiments are specified in the subclaims.

One aspect of the invention relates to a method for determining a relevance of an object in an environment of a motor vehicle for the motor vehicle by means of a driver assistance system of the motor vehicle, in which the object detected with a detection device of the driver assistance system is stored in the surroundings of the motor vehicle with one stored in a storage device of the driver assistance system Object is compared by means of an electronic computing device of the driver assistance system and, depending on the comparison, the object is identified by means of the electronic computing device.

It is provided that a viewing direction of a driver of a motor vehicle is detected by means of a viewing direction detection device of the driver assistance system and the relevance for the identified object is determined as a function of the detected viewing direction.

In particular, it is provided that, for example, the driver assistance system is installed with the viewing direction detection device in a development vehicle or in a customer vehicle. In the motor vehicle there are also common detection devices for, for example, at least partially autonomous driving, such as a radar device, a lidar device, a camera device or an ultrasound device. The driver assistance system is then installed in the motor vehicle with the memory device, which records vehicle data, viewing direction data and object data of the detection device in particular. This data can, for example, be transmitted to a central server for further evaluation via a mobile radio device.

In a pre-processing step, the objects are labeled as relevant in the viewing direction. Objects in the driver's line of sight are highly relevant. Objects with high relevance are used with a high priority to trigger vehicle reactions, such as warnings, braking, evasive maneuvers of the driving and safety systems.

This enables object relevance to be realized on the basis of the driver assistance system. Furthermore, the data can be labeled automatically, that is, identified and processed with the electronic computing device. In particular, the invention thus helps to improve current driving and safety systems with real data.

According to an advantageous embodiment, the electronic computing device for identifying the object and for determining the relevance is taught in by means of machine learning.

It has also proven to be advantageous if, by means of a monitored learning process for machine learning the electronic computing device is taught.

In a further advantageous embodiment, the object is identified in an image generated by the detection device.

It is also advantageous if a current driving situation of the motor vehicle is taken into account in order to determine the relevance for the object.

In a further advantageous embodiment, a control signal is generated with the relevance for another driver assistance system of the motor vehicle for use of the relevance information as a function of the determined relevance by means of the electronic computing device.

A further aspect of the invention relates to a driver assistance system for a motor vehicle with a detection device, with a storage device and with an electronic computing device, the driver assistance system being designed to carry out a method according to the preceding aspect. In particular, the method is carried out using the driver assistance system.

Yet another aspect of the invention relates to a motor vehicle with a driver assistance system according to the previous aspect. The motor vehicle is designed in particular as a passenger car, truck, bus or flight taxi.

Further advantages, features and details of the invention emerge from the following description of a preferred exemplary embodiment and from the drawing. The features and combinations of features mentioned above in the description as well as the features and combinations of features mentioned below in the description of the figures and / or shown alone in the single figure can be used not only in the combination specified in each case, but also in other combinations or on their own, without the frame to leave the invention.

The single figure shows a schematic side view of a motor vehicle with an embodiment of a driver assistance system.

In the figure, elements that are the same or have the same function are provided with the same reference symbols.

The figure shows a motor vehicle in a schematic side view 10 with an embodiment of a driver assistance system 12 , The car 10 is on a first lane 14 , On the first lane 14 is located in particular from the front of the motor vehicle 10 another motor vehicle 16 , The other motor vehicle 16 forms an object in the present exemplary embodiment. The figure also shows that on another lane 18 yet another motor vehicle 20 you can see. Another motor vehicle 20 especially forms another object. The roadway 14 and the further lane 18 are in particular independent of each other. In other words, there is another motor vehicle 20 not on the same lane 14 how the motor vehicle 10 ,

The driver assistance system 12 has a detection device 22 , a storage device 24 and an electronic computing device 26 on.

In the process of determining relevance of an object 16 . 20 in an environment 28 for the motor vehicle 10 using the driver assistance system 12 is with the detection device 22 of the driver assistance system 12 the object 16 . 20 in the neighborhood 28 captured, and the captured object 16 . 20 will with that in the storage device 24 of the driver assistance system 12 stored object 16 . 20 by means of the electronic computing device 26 of the driver assistance system 12 compared, and depending on the comparison, the object 16 . 20 by means of the electronic computing device 26 characterized.

It is provided that the driver assistance system 12 a line of sight detection device 30 has and by means of the viewing direction detection device 30 a driver's line of sight 32 of the motor vehicle 10 the relevance for the marked object is recorded and, depending on the direction of view recorded 16 . 20 is determined.

The object 16 . 20 is designed as a motor vehicle in the present exemplary embodiment. Alternatively or in addition, the object can be 16 . 20 also about road users in general, for example about pedestrians or cyclists and / or standing objects 16 . 20 , such as guardrails, house walls, bricks, traffic bars, traffic lights, traffic signs or garbage cans. In particular, a large number of objects can also be involved in a detection 16 . 20 be captured, in other words the capture is not on two objects 16 . 20 limited.

In the present exemplary embodiment it can be seen in particular that the further motor vehicle 16 on the same lane 14 located like the motor vehicle 10 , The other motor vehicle 16 a higher relevance is thus assigned than the further motor vehicle 20 , which is in the further lane 18 located. In particular, this is because the driver 32 the direction of view of the other motor vehicle 16 mainly aimed in contrast to the other motor vehicle 20 , The relevance can therefore be the other motor vehicle 16 assigned higher than the relevance of another motor vehicle 20 , The relevance assignment based on the lane 14 is to be understood purely as an example. Other relevance criteria can also arise.

In particular, it is provided that the electronic computing device by means of machine learning 26 to identify the object 16 . 20 and being taught to determine relevance. In particular, this can be carried out by means of a monitored learning process for machine learning. The supervised learning or the supervised learning process can also be referred to as "supervised learning".

Furthermore, it can in particular be provided that the object 16 . 20 in one by the detection device 22 generated image is marked. Should it be at the registration facility 22 For example, the object can be a radar sensor and / or a lidar sensor and / or an ultrasonic sensor 16 . 20 be marked in a corresponding generated point cloud. This can be done in a 2-D point cloud or in a 3-D point cloud.

Furthermore, provision can be made for determining the relevance for the object 16 . 20 a current driving situation of the motor vehicle 10 is taken into account. For example, it can be taken into account on which lane 14 . 18 the motor vehicle 10 located. Furthermore, it can be taken into account, for example, whether it is the lane 14 for example, a highway with high speeds.

Furthermore, it has proven to be advantageous if, depending on the specific relevance, using the electronic computing device 26 a control signal with relevance information for the relevance for a further driver assistance system 34 of the motor vehicle 10 is generated for using the relevance information.

In particular, it is therefore provided that in the motor vehicle 10 the line of sight detection device 30 is co-ordered. In the motor vehicle 10 there are common sensors, in particular the detection device 22 , which are designed in particular for automated driving and can be designed, for example, as a radar device, lidar device, camera device or ultrasonic sensor device. In the motor vehicle 10 becomes the driver assistance system 12 with the storage device 24 installed, in particular vehicle data, viewing direction data and object data of the detection device 22 records. The car 10 has the option of sending the data to a central server for evaluation, for example using WLAN or mobile radio. Alternatively, the evaluation, as also shown in the present exemplary embodiment, can take place within the electronic computing device 26 be performed.

In a pre-processing step, the objects 16 . 20 labeled as relevant in the direction of view, marked in other words. By means of machine learning, which can be designed, for example, as “artificial intelligence” or “deep learning”, an algorithm is trained with the help of “supervised learning”, which, based on the collected training data, depends on the driving situation of the objects 16 . 20 assigns a relevance measure. objects 16 . 20 in the driver's line of sight 32 get a high relevance. objects 16 . 20 With high relevance, vehicle reactions such as warnings, braking and evasive maneuvers of the further driver assistance system are used with high priority 24 trigger.
This enables the driver assistance system to implement human knowledge of object relevance 12 is used and transferred. The data are automatically labeled and can be processed with "Supervised Learning". In this way, it can be realized that current driving and safety systems can be improved with the real data as a further driver assistance system.

In other words, a method for object detection using image information obtained using environmental sensors is proposed, training data being generated for machine learning, objects relevant for object detection 16 . 20 marked, especially labeled, in image data. It is envisaged that initially those objects 16 . 20 labeled as relevant in the image data, which is in the driver's line of sight 32 are located. Subsequently, objects contained in the training data are monitored by means of monitored machine learning 16 . 20 , related to a particular driving situation, assigned a relevance measure. objects 16 . 20 in the driver's line of sight 32 get high relevance. The use of the objects 16 . 20 is implemented for situation assessment in order to take actions from the further driver assistance system depending on the assigned relevance measure 34 trigger.

Overall, the invention shows a relevance assessment of objects 16 . 20 in driving and safety systems, evaluating the driver's line of sight 32 ,

LIST OF REFERENCE NUMBERS

10

motor vehicle

12

Driver assistance system

14

roadway

16

another motor vehicle

18

further lane

20

yet another motor vehicle

22

detector

24

memory device

26

electronic computing device

28

Surroundings

30

Sight detector

32

driver

34

further driver assistance system

QUOTES INCLUDE IN THE DESCRIPTION

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

Patent literature cited

• DE 10110275 A1 [0003]

Claims (7)

Method for determining a relevance of an object (16, 20) in an environment (28) of a motor vehicle (10) for the motor vehicle (10) by means of a driver assistance system (12) of the motor vehicle (10), which comprises a detection device (22) Object (16, 20) of the driver assistance system (12) detected in the surroundings (28) of the motor vehicle (10) with an object stored in a memory device (24) of the driver assistance system (12) by means of an electronic computing device (26) of the driver assistance system (12) is compared and, depending on the comparison, the object (16, 20) is identified by means of the electronic computing device (26), characterized in that by means of a viewing direction detection device (30) of the driver assistance system (12) a viewing direction of a driver (32) of the motor vehicle (10 ) is detected and the relevance for the identified object (16, 20) is determined depending on the detected viewing direction. Procedure according to Claim 1 , characterized in that the electronic computing device (26) for identifying the object (16, 20) and for determining the relevance is taught by means of machine learning. Procedure according to Claim 2 , characterized in that the electronic computing device (26) is taught in by means of a monitored learning process for machine learning. Method according to one of the preceding claims, characterized in that the object (16, 20) is identified in an image generated by the detection device (22). Method according to one of the preceding claims, characterized in that a current driving situation of the motor vehicle (10) is taken into account in order to determine the relevance for the object (16, 20). Method according to one of the preceding claims, characterized in that, depending on the determined relevance, a control signal with relevance information for the relevance for a further driver assistance system (34) of the motor vehicle (10) for using the relevance information is generated by means of the electronic computing device (26). Driver assistance system (12) for a motor vehicle (10) with a detection device (22), with a memory device (24), with a line of sight detection device (30) and with an electronic computing device (26), the driver assistance system (12) for performing a method according to one of the Claims 1 to 6 is trained.

Images