Method for verifying a digital environment map for a driver assistance system of a motor vehicle, computing device, driver assistance system and motor vehicle

The method efficiently verifies digital environment maps by determining the correctness of map elements using sensor data probabilities, addressing inefficiencies in existing systems and supporting autonomous driving.

DE102017207441B4Active Publication Date: 2024-12-05AUDI AG
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
DE102017207441
Authority / Receiving Office
DE · DE
Patent Type
Patents
Current Assignee / Owner
Filing Date
2017-05-03
Publication Date
2024-12-05
Estimated Expiration
2037-05-03

AI Technical Summary

Technical Problem

Existing methods for checking digital environment maps in driver assistance systems are inefficient and require complete object recognition from sensor data, contradicting the goal of using digital maps to avoid sensor-based perception.

Method used

A method that checks the correctness of digital environment map elements without recognizing the objects, using sensor data to determine the probability of correctness based on a series of measurements and a sensor model, allowing for efficient and reliable verification.

Benefits of technology

Enables efficient and reliable verification of digital environment maps, reducing the need for costly and inefficient sensor-based perception, thereby supporting autonomous driving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

Method for checking a digital environment map for a driver assistance system (2) of a motor vehicle (1), in which sensor data describing at least one object (6) in an environment (5) of the motor vehicle (1) are received by at least one environment sensor (4) and a map element in the environment map describing the at least one object (5) is checked on the basis of the sensor data, wherein, for checking purposes, an object position describing a position of the at least one object (6) in the environment (5) is determined on the basis of the map element. wherein, without identifying the at least one object (6), it is checked using the sensor data whether the sensor data describes the at least one object (6) at the specified object position, wherein it is checked using the sensor data whether any object (6) is present at the object position without identifying it, characterized by the fact that the sensor data describe a series of measurements taken over time, which are provided by the at least one environmental sensor (4), A probability of correctness is determined based on the sensor data, which describes the probability that the map element is correct. the correctness probability is determined recursively based on a correctness probability for the map element determined at a previous time point in time, and the probability of correctness determined at the previous time is determined based on previously determined sensor data which describe at least one object (6).
Need to check novelty before this filing date? Find Prior Art

Description

DESCRIPTION:

[0001] The invention relates to a method for checking a digital environment map for a driver assistance system of a motor vehicle, in which sensor data describing at least one object in an environment of the motor vehicle is received from at least one environment sensor and, based on the sensor data, a map element in the environment map describing the at least one object is checked. For the checking, an object position describing a position of the at least one object in the environment is determined based on the map element. Furthermore, the present invention relates to a computing device for a driver assistance system. Furthermore, the present invention relates to a driver assistance system and a motor vehicle.

[0002] Driver assistance systems are known from the state of the art and are designed to support the driver in driving a motor vehicle. These driver assistance systems are used, in particular, to maneuver the motor vehicle at least partially autonomously. In future motor vehicles, driver assistance systems will use digital environment maps for environmental perception. This applies in particular to cases where a purely environmental sensor-based solution is not possible or where a purely environmental sensor-based solution is too expensive. In current research vehicles, digital environment maps are used for environmental perception. It is assumed that the content of these digital environment maps accurately describes the real environment of the motor vehicle.

[0003] If the content of the digital environmental maps is assumed to be correct, this content or the digital environmental map cannot be used for autonomous driving, as autonomous driving is a safety-critical application. Methods are known from the prior art in which a digital environmental map is checked using sensor data provided by at least one environmental sensor. The sensor data is first used to detect objects in the vicinity of the vehicle, and then a check is carried out to determine whether these detected objects are also present in the digital environmental map. This means that the objects must first be fully recorded or detected based on the sensor data in order to be able to compare them with map elements that describe these objects.However, this approach is contradictory, since the use of digital environmental maps aims to avoid complete recognition of objects or environmental perception based on sensor data.

[0004] To this end, DE 10 2007 006 870 B4 describes a map evaluation method for adjusting a reliability level of map data in a storage unit in a vehicle for providing information used in a vehicle controller that controls an in-vehicle device in the vehicle. In the method, an absolute position of a vehicle is detected. Furthermore, the detected absolute position is subjected to a map mapping process for obtaining a vehicle position based on a road map indicated by the map data stored in the storage unit, thereby changing the detected absolute position to a vehicle position on the map based on the road map to match the road map.

[0005] Furthermore, DE 10 2013 015 145 B4 discloses a method for computer-assisted testing and evaluation of the quality of data from a digital map present in a data format and stored digitally. The data is intended to be incorporated into applications of navigation systems and / or driver assistance systems. In the method, a set of quality criteria describing the quality of the digital map, which have associated subcriteria subdivided at one or more levels, is evaluated using quality measures. The evaluation of the quality measure and / or the digital map is carried out depending on the requirements of various applications using the data from the digital map.

[0006] Furthermore, DE 10 2008 053 ​​531 A1 describes an evaluation module for evaluating data from a digital map for a vehicle. The evaluation is based on the map's own quality and measurement data from the vehicle's environmental sensors. The connected driver assistance systems and safety systems of the vehicle use the digital map data based on the evaluation.

[0007] Furthermore, US 9,321,461 B1 describes a method in which data identifying an object detected in a vehicle environment is received by a computing device, wherein the data contains location coordinates for the object. Furthermore, a corresponding feature is identified by the computing device from pre-stored map information based on the location coordinates and a map position of the corresponding feature. The corresponding feature is defined as a curve and is associated with a marker that identifies a type.

[0008] Furthermore, DE 10 2015 220 695 A1 discloses a method for evaluating the content of a map, wherein the map contains at least one first environmental feature. Furthermore, a second environmental feature is detected by at least one sensor of a vehicle. Furthermore, the at least one first environmental feature contained in the map is compared with the at least one detected second environmental feature, and the content of the map is evaluated based on the comparison performed.

[0009] The object of the present invention is to provide a reliable method for checking a digital environment map for a driver assistance system of a motor vehicle, which method can be carried out more efficiently. Furthermore, a corresponding computing device, a driver assistance system, and a motor vehicle are to be provided.

[0010] This object is achieved according to the invention by a method, a computing device, a driver assistance system, and a motor vehicle having the features according to the respective independent claims. Advantageous developments of the present invention are the subject of the dependent claims.

[0011] A method according to the invention is used to check a digital environment map for a driver assistance system of a motor vehicle. Sensor data describing at least one object in the environment of the motor vehicle is received from at least one environment sensor. Based on the sensor data, a map element in the environment map describing the at least one object is checked, wherein for the check based on the map element an object position describing a position of the at least one object in the environment is determined. In this case, it is provided that, without detecting the at least one object, a check is carried out on the basis of the sensor data to determine whether the sensor data describe the at least one object at the determined object position. In this case, a check is carried out on the basis of the sensor data to determine whether any object is present at the object position, without identifying it.

[0012] The method is intended to check a digital environmental map or digital map for a driver assistance system. The method can be carried out using a computing device of the driver assistance system. The digital environmental map is available to this computing device. The environmental map can be referred to as an a priori map. The digital environmental map describes the surroundings of the motor vehicle in digital form. The digital environmental map contains several map elements, each of which describes individual objects in the environment. The digital environmental map can describe, for example, streets, buildings, vegetation (plants, trees, bushes, and the like), development elements (curbs, guard rails, and the like), and / or traffic elements (traffic signs, traffic lights, and the like) as map elements. The computing device can receive the sensor data provided by the at least one environmental sensor.The environmental sensor can be, for example, a camera, a radar sensor, a lidar sensor, and / or an ultrasonic sensor. It can also be provided that sensor data is provided by several environmental sensors of the motor vehicle and transmitted to the computing device.

[0013] The sensor data is now to be used to check whether the map of the surroundings is correct or whether the individual map elements stored in the map of the surroundings are correct. To do this, an object position within the surroundings of the motor vehicle is assigned to the respective map elements in the map of the surroundings. For this purpose, the current position of the motor vehicle can be determined. This can be done using a satellite-based positioning system and / or odometry. This can then be used to determine the position of the motor vehicle in the digital map. The sensor data describes at least one object in the surroundings and the relative position of the object to the sensor or the motor vehicle. This can be used to check whether the object position determined on the basis of the digital map of the surroundings matches the position of the object determined using the sensor data.

[0014] According to the invention, it is provided that a check is carried out to determine whether the sensor data describe the at least one object at the specific object position, without the at least one object being recognized on the basis of the sensor data. The sensor data is generally used to check whether any object is present at the object position, without identifying or characterizing it in more detail. If, for example, the map element in the surroundings map indicates that the end of a roadway is at the object position, the sensor data can be used to check whether the sensor data for this object position can describe such an end of the roadway. This means that, in particular, it is not provided that the at least one object is classified or recognized on the basis of the sensor data.Based on the digital environment map or the map element in the digital environment map, the object position of the object in the environment can be determined. The sensor data is then only used to check whether there is actually an object at this object position. This means that it is not necessary for at least one object to be first detected and then compared with the object described by the map element in the digital environment map. In the present case, only the correctness of the digital environment map or the map element is checked using the sensor data. This means that it is not necessary to perceive the object corresponding to the map element using the sensor data. The present invention is based on the finding that for many problems it is easier to check whether a solution to the problem is correct rather than calculating the solution.This enables an efficient yet reliable review of the digital environmental map.

[0015] The sensor data describes a series of chronologically consecutive measurements provided by the at least one environmental sensor. The series of measurements thus describes multiple measurements or measured values ​​that are recorded consecutively by the environmental sensor. In particular, it is intended that the respective measurements each describe the at least one object. Thus, multiple measurements are provided, each of which describes the object. This enables reliable detection of the object based on the sensor data.

[0016] Using the sensor data, a probability of correctness is determined, which describes the probability with which the map element is correct. A probability distribution can be determined on the basis of the sensor data or the series of measurements. This probability distribution can indicate the probability that the map element present in the environmental map is correct. This probability distribution for the chronologically successive measurements can, for example, be a normal distribution. On the basis of this normal distribution, variables such as the mean and / or the variance can then be determined in order to determine the probability of correctness. In particular, it is provided that the probability of correctness describes a conditional probability, which indicates that the map element is correct under the assumption that the sensor data or the series of measurements is taken into account.For this purpose, a Boolean variable can be used, which takes the value "true" if the map element is correct and takes the value "false" if the map element is incorrect. Preferably, the probability of correctness is determined based on a Bayesian filter. This allows the probability of correctness to be determined in a simple and reliable manner.

[0017] The probability of correctness is determined recursively based on a probability of correctness for the map element determined at a previous point in time. It is preferably provided that the probability of correctness determined at the previous point in time is determined based on predetermined sensor data describing the at least one object. Sensor data or a series of measurements can be continuously provided in the method. The probability of correctness can then be determined based on each series of measurements. If a probability of correctness for a map element was already determined at an earlier point in time, this can be taken into account to determine the current probability of correctness. This allows the probability of correctness to be determined more reliably and with less effort.

[0018] Furthermore, it is advantageous if the probability of correctness is determined based on a sensor model that describes the at least one environmental sensor. As already explained, different types of environmental sensors can be used. These environmental sensors differ in terms of the measurement principle and / or measurement accuracy. In particular, the sensor model can describe a spatial uncertainty in the detection of the at least one object. Based on the sensor model, a measurement probability can be determined, which is then used to determine the probability of correctness. If the sensor model is taken into account when determining the probability of correctness, the digital environmental map can be precisely verified.

[0019] In a further embodiment, a decision signal is output based on the probability of correctness, which describes whether the driver assistance system should be maneuvered depending on the map element. Depending on the application or design of the driver assistance system, it can be checked whether the determined probability of correctness is sufficient for the driver assistance system. If the driver assistance system is used to maneuver the motor vehicle autonomously, the required probability of correctness can be decided, for example, based on the autonomy level of the driver assistance system. If the probability of correctness is considered sufficient, the map element can be assumed to be a correct representation of the environment or the object, and the driver assistance system can thus be operated based on the map element.

[0020] The invention also includes the computing device for a driver assistance system of a motor vehicle. The computing device has a processor device configured to carry out an embodiment of the method according to the invention. For this purpose, the processor device can have at least one microprocessor and / or at least one microcontroller. Furthermore, the processor device can have program code configured to carry out the embodiment of the method according to the invention when executed by the processor device. The program code can be stored in a data memory of the processor device.

[0021] The computing device can have a verification module and a decision module. The environmental map and the sensor data can be fed to the verification module. The verification module can then determine the probability of correctness. This probability of correctness can then be transmitted to the decision module. The decision module can then output the decision signal, which describes whether the driver assistance system should be maneuvered depending on the digital environmental map, a region of the digital environmental map, and / or the map element. The verification module and the decision module can be provided, for example, by separate control units. It can also be provided that the computing device is provided by a common control unit.

[0022] A driver assistance system according to the invention for a motor vehicle comprises a computing device according to the invention. The driver assistance system is preferably configured to maneuver the motor vehicle at least partially autonomously. This means that the driver assistance system can intervene in a steering system, a drive motor, and / or a braking system of the motor vehicle. In particular, the driver assistance system can assume longitudinal and / or lateral guidance of the motor vehicle. By checking the digital environmental map, it can be used by driver assistance systems or autonomous motor vehicles. The environmental map can be used for environmental recognition if environmental sensor-based environmental recognition is not possible or too expensive. In this way, the costs for autonomous vehicles can be reduced, and the new functionalities can be introduced to the market earlier.

[0023] A motor vehicle according to the invention comprises a driver assistance system according to the invention. The motor vehicle can be designed as a passenger car or a commercial vehicle.

[0024] The preferred embodiments presented in connection with the method according to the invention and their advantages apply accordingly to the computing device according to the invention, the driver assistance system according to the invention and the motor vehicle according to the invention.

[0025] Exemplary embodiments of the invention are described below. Shown are: Fig. 1 is a schematic representation of a motor vehicle having a driver assistance system with a computing device; and Fig. 2 a schematic representation of the computing device, which comprises a verification module and a decision module.

[0026] The exemplary embodiments explained below are preferred embodiments of the invention. In the exemplary embodiments, the described components of the embodiments each represent individual, independently considered features of the invention, which also further develop the invention independently of one another and are thus also to be considered as components of the invention, either individually or in a combination other than that shown. Furthermore, the described embodiments can also be supplemented by further features of the invention already described.

[0027] In the figures, functionally identical elements are provided with the same reference numerals.

[0028] Fig. 1 shows a schematic representation of a motor vehicle 1 according to one embodiment of the invention. The motor vehicle 1, which is embodied here as a passenger car, comprises a driver assistance system 2. The driver assistance system 2 serves to assist a driver in driving the motor vehicle 1. In particular, the driver assistance system 2 can maneuver the motor vehicle 1 autonomously. For this purpose, the driver assistance system 2 can intervene in a longitudinal and / or lateral guidance of the motor vehicle 1.

[0029] The driver assistance system 2 comprises a computing device 3 and an environment sensor 4, which is connected to the computing device 3 for data transmission. The environment sensor 4 is embodied as a camera.

[0030] The environment sensor 4 can provide sensor data that describe an environment 5 of the motor vehicle 1. In particular, the sensor data can describe at least one object 6 in the environment 5. In the present case, an object 6 is shown as an example that is located in the environment 5 of the motor vehicle 1.

[0031] Fig.2 shows a schematic representation of the computing device 3. With the help of the computing device 3, a digital environmental map can be checked, which describes the environment 5 of the motor vehicle 1. The computing device 3 comprises an input 7 or an interface, via which the sensor data from the environmental sensor 4 can be received. In addition, the computing device 3 comprises a memory 8, on which the digital environmental map is stored. It can also be provided that the digital environmental map is received from a server. Alternatively, the digital environmental map can be received from another vehicle via vehicle-to-vehicle communication, or the digital environmental map can be received from an infrastructure facility via vehicle-to-infrastructure communication.

[0032] The computing device 3 comprises a verification module 9, which serves to verify the digital environment map. On the one hand, the digital environment map is available to the verification module 9. For this purpose, the digital environment map can be transferred from the memory 8 to the verification module 9. On the other hand, the verification module 9 has the sensor data provided by the environment sensor 4. The sensor data describes a measurement series mt. The measurement series includes a plurality of temporally consecutive measurements m t = [z 0, z1,...,z t ], whereby the measurements are provided by the environment sensor 4.

[0033] Verification module 9 will check whether the digital environment map or a part of it is correct. A Boolean random variable C is used for this purpose, which can take the value "true" or "false." If the random variable C is true, this means that the part of the environment map is correct. If the random variable C is false, this means that the part of the environment map is incorrect. Verification module 9 can then be used to determine a correctness probability, which describes the probability that the specific part of the environment map is correct, taking all measurements into account: p(C=true|mt).

[0034] To determine the probability of correctness, a series of measurements m t-1 which was determined at a previous point in time. This allows the correctness probability to be determined recursively: p(C=true|mt)=η⋅p(zt|C=true)⋅p(C=true|mt−1).

[0035] Here, η describes a normalization constant for the probability p(z t |m t-1 ). This means that if the last measurement z t was received, the correctness probability is calculated using a correctness probability p(C = true|m t-1 ) can be determined based on a previously determined series of measurements. In addition, a measurement probability p(z t | C = true), which is determined based on a sensor model for the environment sensor 4. The sensor model can be determined depending on the measuring principle of the environment sensor 4 and / or the spatial uncertainty of the environment sensor 4 with regard to the detection of the object 6.

[0036] The determined correctness probability p(C = true|m t) is output by the verification module 9 and transmitted to the decision module 10. The probability of correctness thus serves as the input for the decision module 10. Depending on the design of the driver assistance system 2, a decision can then be made as to whether the probability of correctness is sufficient for the area of ​​the surrounding map or the map element to be assumed to be correct. If the probability of correctness is considered sufficient, the driver assistance system 2 can be operated based on the surrounding map or the area of ​​the surrounding map.

[0037] Overall, the examples demonstrate how the invention can reliably and efficiently verify the digital environment map. If the environment map is assumed to be correct based on the probability of correctness, it can be used for environment perception. Thus, purely environment sensor-based environment perception can be at least partially dispensed with.

Claims

[1] Method for checking a digital environment map for a driver assistance system (2) of a motor vehicle (1), in which sensor data describing at least one object (6) in an environment (5) of the motor vehicle (1) are received by at least one environment sensor (4) and, based on the sensor data, a map element in the environment map describing the at least one object (5) is checked, wherein, for checking, based on the map element, an object position describing a position of the at least one object (6) in the environment (5) is determined, wherein, without detecting the at least one object (6), the sensor data is used to check whether the sensor data describe the at least one object (6) at the specific object position, wherein the sensor data is used to check whether any object (6) is present at the object position without identifying it, characterized by , that the sensor data describe a series of measurements with temporally successive measurements provided by the at least one environmental sensor (4), based on the sensor data, a probability of correctness is determined, which describes the probability with which the map element is correct, the correctness probability is determined recursively based on a correctness probability for the map element determined at a previous point in time, and the probability of correctness determined at the previous point in time is determined on the basis of previously determined sensor data which describe the at least one object (6). [2] Method according to claim 1, characterized by that the probability of correctness is determined based on a sensor model which describes the at least one environmental sensor (4). [3] Method according to one of claims 1 or 2, characterized bythat based on the probability of correctness a decision signal is output which describes whether the driver assistance system (2) is to be maneuvered depending on the map element. [4] Computing device (3) for a driver assistance system (2) of a motor vehicle (1), wherein the computing device (3) is configured to carry out a method according to one of the preceding claims. [5] Driver assistance system (2) for a motor vehicle (1) with a computing device (3) according to claim 4, wherein the driver assistance system (2) is configured for at least partially autonomous maneuvering of the motor vehicle (1). [6] Motor vehicle (1) with a driver assistance system (2) according to claim 5.

Citation Information

Patent Citations

  • Map rating system and map rating procedure

    DE102007006870B4

  • verification of digital cards

    DE102008053531A1

  • Methods for testing and evaluating the quality of digital maps

    DE102013015145B4

  • Method and device for evaluating the content of a card

    DE102015220695A1

  • Change detection using curve alignment

    US9321461B1