Method and system for calibrating a motion sensor

Abstract

A computer-implemented method for calibrating a motion sensor disposed at a vehicle and configured to output raw motion measurements indicative of a current motion condition of the vehicle includes calculating processed motion measurements by modifying the output raw motion measurements based on different sets of calibration parameters, a proximity of the vehicle is monitored by an environment-aware sensor, an occupancy map set is determined from mapping data acquired by the environment-aware sensor according to a motion model of the vehicle, each processed motion measurement value being used to establish a motion model for one of the occupancy maps, and selecting one of the different calibration parameters for modifying the subsequently output raw motion measurements based on a comparison of the determined occupancy maps.

Description

Technical Field

[0001] The present disclosure relates to methods and systems for calibrating a motion sensor disposed at a vehicle and configured to output raw motion measurements indicative of a current motion condition of the vehicle. Background Art

[0002] Information about the ego-motion of a vehicle can be important in various aspects of driving, particularly for advanced driver assistance systems (ADAS) and autonomous driving applications. Sensors used to collect ego-motion information may require calibration to compensate for manufacturing tolerances and the dependence of sensor output on environmental conditions such as ambient temperature or tire pressure. Using incorrect ego-motion information can be detrimental to various driving applications.

[0003] Therefore, there is a need to provide reliable calibration of motion sensors. Summary of the invention

[0004] The present disclosure provides a computer implemented method, a computer system, a non-transitory computer readable medium and a vehicle according to the independent claims. Embodiments are given in the dependent claims, the description and the drawings.

[0005] In one aspect, the present disclosure relates to a computer-implemented method for calibrating a motion sensor disposed at a vehicle and configured to output raw motion measurements indicating a current motion condition of the vehicle, the method comprising the following steps: calculating processed motion measurements by modifying the output raw motion measurements based on different calibration parameter sets; monitoring the vicinity of the vehicle by at least one environmental perception sensor; determining an occupancy map set by mapping data acquired by the environmental perception sensor according to a motion model of the vehicle, wherein each processed motion measurement is used to establish a motion model for one of the occupancy maps; and selecting one of the different calibration parameters for modifying the subsequently output raw motion measurements based on a comparison of the determined occupancy maps.

[0006] Thus, the environment perception sensor can be used to calibrate the motion sensor. The motion sensor can be calibrated based on the quality of the occupancy map that has been determined using the data provided by the motion sensor. If the occupancy map is of high quality (i.e., shows clear and well-defined objects) it can be assumed that the calibration is good, if the occupancy map is of low quality (i.e., shows smeared or blurred objects instead) the calibration is poor. The disclosed calibration process can be carried out on-board and online, i.e., when the motion sensor is operating. There is no need to move the vehicle to a test bench.

[0007] An occupancy map, also called an occupancy grid or occupancy map grid, is a map of cells that contains information about the occupancy of a cell by some object. A cell can have a fixed width, length, and height. Occupancy maps are widely used in advanced driver assistance systems and autonomous driving applications because an occupancy map can be interpreted as a map of obstacles around a vehicle. Each occupied cell can represent a non-drivable area. The reverse occupancy map can be interpreted as a free space map showing drivable areas.

[0008] The environmental perception sensor may provide reflections from objects in the form of a point cloud. The term "mapping" may refer to the process of evaluating the correspondence between successively acquired point clouds. In order to compensate for the offset of the points due to the self-motion of the vehicle, the position of the points may be adjusted according to a motion model. The motion model may include a constant linear motion with a speed determined by the motion sensor. The motion model may further include a curvilinear component based on the yaw rate provided by a gyroscope arranged at the vehicle. Depending on the application, a more complex motion model may be provided.

[0009] According to an embodiment, the method may comprise determining a sharpness value for each of the determined occupancy maps, and selecting the calibration parameters based on a comparison of the determined sharpness values. The sharpness value may be an indicator of the quality of the occupancy map. That is, the quality of the mapping process and thus the accuracy of the motion measurements may be assessed by the sharpness value. Thus, the quality of the occupancy map may be used to optimize the calibration process.

[0010] According to an embodiment, extreme values ​​of sharpness values ​​may be determined and calibration parameters for determining an occupancy map associated with the extreme values ​​may be selected. In other words, an occupancy map with the highest or lowest sharpness values ​​may be determined and calibration parameters on which a motion model for the occupancy map is based may be selected. Depending on the application, the sharpness values ​​may have a positive or negative correlation with the image sharpness of the occupancy map.

[0011] According to an embodiment, the sharpness value may be determined according to the following sharpness metric:

[0012]

[0013] Where S (f) is the sharpness value of the occupancy map determined based on the calibration factor f, and is the occupancy probability at position x,y in the occupancy map determined based on the calibration factor f. This sharpness measure penalizes probability values ​​close to the extreme values ​​0 and 1 less than values ​​close to 0.5. This explains why high-quality occupancy maps have many extremes, i.e., accumulation of occupancy probabilities at locations where objects are present and accumulation of no probabilities at locations where objects are not present, and why low-quality occupancy maps have distributed occupancy probabilities, i.e., a "smeared" appearance.

[0014] According to an embodiment, the vicinity of a vehicle may be monitored while the vehicle is moving.

[0015] According to an embodiment, the motion sensor may be a mileage sensor. The mileage sensor may be mounted to a wheel or axle of a vehicle. The mileage sensor may sense the rotation speed of the wheel of the vehicle and determine the speed of the vehicle based on the sensed rotation speed and the stored tire diameter value. Changes in air pressure may result in incorrect speed values. The output speed value may be corrected according to the calibration of the present disclosure. According to another embodiment, the motion sensor is a gyroscope that outputs the yaw rate of the vehicle.

[0016] According to an embodiment, the raw motion measurement may represent the vehicle speed.The raw motion measurement may also represent the component of the vehicle speed in a predefined direction.

[0017] According to an embodiment, the environment perception sensor may be a radar sensor, a lidar sensor, or a camera. Such a sensor is capable of sensing the environment of the vehicle. The point cloud may correspond to a continuous scan or frame of the environment perception sensor. In the context of the present disclosure, the object points output by the environment perception sensor correspond to detected surface points of objects (e.g., stationary objects) present in the vicinity of the vehicle.

[0018] According to an embodiment, different calibration parameter sets include a plurality of different scaling factors, and processed motion measurements are calculated by multiplying the raw motion measurements with the different scaling factors.

[0019] According to an embodiment, different calibration parameter sets include a plurality of different offset values, and processed motion measurement values ​​are calculated by adding the different offset values ​​to the raw motion measurement values.

[0020] Different calibration parameter sets may include multiple different scaling factors and multiple different bias values. For example, different calibration parameter sets may include matrix-like or tensor-like structures of different scaling factors and biases.

[0021] Modifying the raw motion measurements using a scale factor and a bias corresponds to a linear sensor model. However, depending on the application, a more complex sensor model may be provided for the motion sensor. For example, the sensor model of the motion sensor may be based on a polynomial function. Accordingly, three or more calibration parameters may be required to calibrate the motion sensor.

[0022] According to an embodiment, the different calibration parameters are distributed over a range of values ​​extending from the desired value (eg using a predefined step size) in a positive direction and in a negative direction. The size of the range may be defined based on the application and / or the estimation accuracy.

[0023] According to an embodiment, data acquired by the environment perception sensors are collected over a predetermined number of time intervals, and each occupancy map is determined by overlaying the collected data in a spatial representation of the vicinity of the vehicle. A presence probability of an object at a defined location may be calculated based on the overlay data.

[0024] In another aspect, the present disclosure relates to a computer system configured to perform some or all of the steps of the computer-implemented methods described herein.

[0025] The computer system may include a processing unit, at least one storage unit and at least one non-transient data storage. The non-transient data storage and / or memory unit may include a computer program for instructing a computer to perform several or all steps or aspects of the computer-implemented method described herein.

[0026] In another aspect, the present disclosure relates to a non-transitory computer-readable medium comprising instructions for performing some or all steps or aspects of the computer-implemented methods described herein. The computer-readable medium may be configured as: an optical medium, such as a compact disk (CD) or a digital versatile disk (DVD); a magnetic medium, such as a hard disk drive (HDD); a solid-state drive (SSD); a read-only memory (ROM), such as flash memory; or the like. In addition, the computer-readable medium may be configured as a data storage accessible via a data connection, such as an Internet connection. The computer-readable medium may be, for example, an online data repository or a cloud storage.

[0027] The present disclosure also relates to a computer program for instructing a computer to execute several or all steps or aspects of the computer-implemented method described herein.

[0028] The present disclosure also relates to a vehicle, such as a motor vehicle or an automobile, comprising the computer system, motion sensors, and environmental perception sensors described herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Exemplary embodiments and functionality of the present disclosure are described herein in conjunction with the following drawings, which schematically illustrate:

[0030] Figure 1 A top view of a vehicle in a traffic space.

[0031] Figure 2 Illustration of the sharpness values ​​of the occupancy map depending on the calibration factor.

[0032] Figure 3 A flow chart of a method for calibrating a motion sensor according to various embodiments is shown.

[0033] Figure 4A computer system having a plurality of computer hardware components configured to perform the steps of a computer-implemented method for calibrating a motion sensor according to various embodiments. DETAILED DESCRIPTION

[0034] Figure 1 A motor vehicle or host vehicle 11 is depicted in a top view including a computer system 12. A motion sensor 13 and an environment perception sensor 15 are arranged at the host vehicle 11 and connected to the computer system 12. The motion sensor 13 is configured to provide ego-motion information in the form of raw motion measurements, such as a current speed v of the host vehicle 11 in a travel direction 17. According to an embodiment, the motion sensor 13 may be an odometer sensor.

[0035] The environmental perception sensor 15 may be a radar (radio detection and ranging) sensor, a lidar (light detection and ranging) sensor and / or a camera, such as a time-of-flight camera. By means of the environmental perception sensor 15, objects 19 present in the environment of the host vehicle 11, such as trees, pillars, pedestrians or other vehicles, may be detected. Additional environmental perception sensors (not shown) may be arranged at the host vehicle 11 to cover a wider area. Furthermore, an environmental perception system comprising a combination of different types of perception sensors may be provided. The computer system 12 receives data from the motion sensor 13 and the environmental perception sensor 15, and provides an "advanced driver assistance" function or an autonomous driving function based on an evaluation of the received data.

[0036] According to various embodiments, an occupancy map around the host vehicle 11 may be determined by accumulating readings from the environment perception sensors 15 and converting the accumulated sensor readings into probabilities over time. The occupancy map may include a two-dimensional or three-dimensional grid of cells, with an occupancy probability or free space probability assigned to each cell, as is known in the art. The occupancy probabilities may be updated after each data acquisition process, i.e., after each frame or scan of the environment perception sensors 15. The occupancy map shows obstacles in the form of stationary objects 19, and therefore also shows drivable space, i.e., space without obstacles.

[0037] When the host vehicle 11 moves, the data received from the environment perception sensor 15 must be mapped according to the motion model. In other words, it is necessary to offset the object reflections provided by the environment perception sensor in a specific time frame relative to the object reflections of an earlier time frame to account for the ego-motion of the host vehicle 11. In order to build the motion model, the output of the motion sensor 13 is used. However, the raw speed value v as the output of the motion sensor 13 depends on manufacturing tolerances and environmental conditions and therefore needs to be calibrated. This is done by calculating the processed speed value In a simplified form that has proven sufficient for many applications, the calibration can be based on a scaling factor f and a constant offset b according to the following formula:

[0038]

[0039] Where v is the original speed value output by the motion sensor 13, is the processed velocity value calculated by the computer system 12, f is the scaling factor and b is the bias. For f=1 and b=0, the processed velocity is equal to the original velocity v.

[0040] To calibrate the motion sensor 13, calibration parameters f and b may be estimated. According to various embodiments, a point cloud acquired by the environment perception sensor 15 is collected and stored in the memory of the computer system 12 for a given number of timestamps or time frames, such as 10 time frames. The point cloud is mapped via an occupancy grid mapping (OGM) process with free parameters f and b to generate an occupancy map where x and y are Figure 1 The two-dimensional space coordinates shown.

[0041] Different calibration parameter sets f, b can be determined based on predefined value ranges. The value ranges can extend around the desired value. For example, the following value ranges can be selected:

[0042] f=[0.8;1.2]; b=[-0.1;0.1]; Δs=0.01

[0043] where Δs is the step size. The resulting calibration parameter set includes multiple pairs of f,b.

[0044] Based on the calibration parameter set, the occupancy grid atlas For each occupancy grid map The sharpness value is determined according to the following sharpness metric:

[0045]

[0046] Where S (f，b) is the sharpness value of the occupancy map determined based on the calibration factors f and b, and is the occupancy probability at position x,y in the occupancy map determined based on the calibration factors f and b. This sharpness measure penalizes probability values ​​close to the extreme values ​​0 and 1 less than values ​​close to 0.5. This explains why high-quality occupancy maps have many extremes, i.e., accumulation of occupancy probabilities at locations where objects are present and accumulation of no probabilities at locations where objects are not present, and why low-quality occupancy maps have distributed occupancy probabilities, i.e., a "smeared" appearance.

[0047] Figure 2The sharpness value S is shown as an example. (f，b) (vertical axis) depends on the scaling factor f (horizontal axis). The individually calculated values ​​are plotted as points 23. In addition, a fitting curve 24 is shown. The sharpness value S (f，b) The minimum value 25 of the fitted curve 24 (which may be the lowest point 23 or the minimum value) defines a value which may be referred to as f opt A similar process is performed to find the optimal value of b, which can be called b opt Through the two-dimensional minimum search, two optimal values ​​f can be found in a common process opt 、b opt .

[0048] Optimal value f opt and b opt can be selected as a calibration parameter for modifying the raw velocity value v that is subsequently output. In other words, f opt and b opt The current calibration parameters may be determined and may be stored in the memory of the computer system 12. The estimated parameters may be used by any application modules that rely on the ego-motion information from the motion sensor 13 and may be re-estimated whenever necessary.

[0049] Figure 3 A diagram is shown for calibration Figure 1 4 is a flow chart 400 of a method of the motion sensor 13 shown in FIG.

[0050] At 402, raw motion measurements may be received from a motion sensor. At 404, processed motion measurements may be calculated by modifying the output raw motion measurements based on different sets of calibration parameters. At 406, a vicinity of a vehicle is monitored by an environment perception sensor 13. At 408, a set of occupancy maps is determined by mapping data acquired by the environment perception sensor according to a motion model of the vehicle, wherein each processed motion measurement is used to establish a motion model for one of the occupancy maps. At 410, one of the different calibration parameters is selected for modifying the subsequently output raw motion measurements based on a comparison of the determined occupancy maps.

[0051] Figure 3 The calibration method illustrated in may be performed on a regular time basis when an error signal is received and / or after a repair process has been performed.

[0052] According to various embodiments, the method may further include determining a sharpness value for each of the determined occupancy maps, and selecting the calibration parameter based on a comparison of the determined sharpness values.

[0053] According to various embodiments, the method may further include determining an extreme value of the sharpness value, and selecting calibration parameters for determining an occupancy map associated with the extreme value.

[0054] According to various embodiments, the method may further include determining the sharpness value according to the following sharpness metric:

[0055]

[0056] Where S (f) is the sharpness value of the occupancy map determined based on the calibration factor f, and is the occupancy probability of position x,y in the occupancy map determined based on the calibration factor f.

[0057] According to various embodiments, the method may further include monitoring the vicinity of the vehicle while the vehicle is moving.

[0058] According to an embodiment, the motion sensor may be a mileage sensor.

[0059] According to various embodiments, the raw motion measurements may represent vehicle speed.

[0060] According to various embodiments, the environment perception sensor may be a radar sensor, a lidar sensor, or a camera.

[0061] According to various embodiments, different calibration parameter sets may include a plurality of different scaling factors, and the method may further include calculating processed motion measurement values ​​by multiplying the raw motion measurement values ​​with the different scaling factors.

[0062] According to various embodiments, different calibration parameter sets may include a plurality of different offsets, and the method may further include calculating processed motion measurement values ​​by adding the different offsets to the raw motion measurement values.

[0063] According to various embodiments, the different calibration parameters may be distributed over a range of values ​​extending in a positive direction and in a negative direction from the desired value.

[0064] According to various embodiments, the method may further include collecting data acquired by the environment perception sensors over a predetermined number of time intervals, and determining each occupancy map by superimposing the collected data in a spatial representation of the vicinity of the vehicle.

[0065] Each of steps 402, 404, 406, 408, and 410, and further steps described above, may be performed by computer hardware components.

[0066] Figure 4A computer system 500 having a plurality of computer hardware components configured to perform the steps of a computer-implemented method for calibrating a motion sensor according to various embodiments is shown. The computer system 500 may include a processor 502, a memory 504, and a non-transitory data store 506. A sensor system 508 may be provided as part of the computer system 500 (e.g., Figure 4 ), or may be provided outside the computer system 500. The sensor system 508 may include a motion sensor 13 and an environmental perception sensor 15, such as Figure 1 shown.

[0067] Processor 502 may execute instructions provided in memory 504. Non-transitory data storage 506 may store computer programs including instructions that may be transferred to memory 504 and then executed by processor 502. Sensor system 508 may be used to capture vehicle speed and for sensing objects in the vicinity of the vehicle.

[0068] The processor 502, the memory 504 and the non-transitory data storage 506 may be coupled to each other, for example, via an electrical connection 510 (such as, for example, a cable or a computer bus) or via any other suitable electrical connection to exchange electrical signals. The sensor system 508 may be coupled to the computer system 500, for example, via an external interface, or may be provided as part of the computer system (in other words: inside the computer system, for example coupled via the electrical connection 510).

[0069] The terms “coupling” or “connection” are intended to include a direct “coupling” (eg, via a physical link) or a direct “connection” as well as an indirect “coupling” or an indirect “connection” (eg, via a logical link), respectively.

[0070] It should be understood that what is described with respect to one of the above methods may be similarly applicable to the computer system 500 .

[0071] Reference numerals list

[0072] 11 Main means of transportation

[0073] 12 Computer Systems

[0074] 13 Motion Sensor

[0075] 15 Environmental Perception Sensors

[0076] 17 Driving Directions

[0077] 19 Objects

[0078] 23 o'clock

[0079] 24 Fitting curve

[0080] 25 Minimum

[0081] 400 is a flow chart showing a method for calibrating a motion sensor

[0082] 402 A step of receiving raw motion measurements from a motion sensor disposed at a vehicle

[0083] 404 Step of calculating processed motion measurements by modifying the raw motion measurements based on a different set of calibration parameters

[0084] 406 Steps for monitoring the vicinity of a vehicle using an environment sensing sensor

[0085] 408 Step of determining an occupancy atlas using mapping data acquired by an environment perception sensor

[0086] 410 Step of selecting one of the different calibration parameters based on the comparison of the determined occupancy maps

[0087] 500 Computer system according to various embodiments

[0088] 502 processor

[0089] 504 Memory

[0090] 506 Non-transient data storage

[0091] 508 Sensor System

[0092] 510 connected.

Claims

1. A computer-implemented method for calibrating a motion sensor (13) arranged at a vehicle (11) and configured to output raw motion measurements indicative of a current motion condition of the vehicle (11), The method comprises: - calculating processed motion measurements by modifying the output raw motion measurements based on a different set of calibration parameters, - monitoring the vicinity of the vehicle (11) by means of at least one environment perception sensor (15), - determining a set of occupancy maps by mapping data acquired by the environment perception sensor (15) according to a motion model of the vehicle (11), wherein each processed motion measurement value is used to establish the motion model for one of the occupancy maps, and - Based on the comparison of the determined occupancy maps, selecting one of the different calibration parameters for modifying a subsequently outputted raw motion measurement value.

2. The computer-implemented method of claim 1, It is characterized in that A sharpness value is determined for each of the determined occupancy maps, and the calibration parameters are selected based on a comparison of the determined sharpness values.

3. The computer-implemented method of claim 2, It is characterized in that An extreme value (25) of the sharpness value is determined, and the calibration parameters for determining the occupancy map associated with the extreme value are selected.

4. The computer-implemented method of claim 3, It is characterized in that The sharpness value is determined according to the following sharpness metric: Where S (f) is the sharpness value of the occupancy map determined based on the calibration factor f, and is the occupancy probability of position x, y in the occupancy map determined based on the calibration factor f.

5. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that The vicinity of the vehicle (11) is monitored while the vehicle (11) is moving.

6. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that The motion sensor (13) is a mileage sensor.

7. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that The raw motion measurements represent vehicle speed.

8. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that The environment perception sensor (15) is a radar sensor, a lidar sensor or a camera.

9. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that The different calibration parameter sets include a plurality of different scale factors, and the processed motion measurements are calculated by multiplying the raw motion measurements by the different scale factors.

10. The computer-implemented method of at least one of claims 1 to 4, The different calibration parameter sets include a plurality of different offset values, and the processed motion measurement values ​​are calculated by adding the different offset values ​​to the raw motion measurement values.

11. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that The different calibration parameters are distributed over a value range extending from the expected value in a positive direction and in a negative direction.

12. The computer-implemented method of at least one of claims 1 to 4, It is characterized in that Data acquired by the environment perception sensors (15) are collected over a predetermined number of time intervals and each of the occupancy maps is determined by superimposing the collected data in a spatial representation of the vicinity of the vehicle (11).

13. A computer system (500) configured to carry out the computer-implemented method of at least one of claims 1 to 12.

14. A non-transitory computer-readable medium comprising instructions for carrying out the computer-implemented method of at least one of claims 1 to 12.

15. A vehicle (11) comprising the computer system (500) according to claim 13, the motion sensor (13) and the environment perception sensor (15).

Images