WO2021001171A1

WO2021001171A1 - Method for operating a 3d distance sensor device

Info

Publication number: WO2021001171A1
Application number: PCT/EP2020/067110
Authority: WO
Inventors: Simon Bell; Karl Christoph Goedel; Johannes Richter; Holger Maris Gilbergs
Original assignee: Robert Bosch Gmbh
Priority date: 2019-07-04
Filing date: 2020-06-19
Publication date: 2021-01-07
Also published as: DE102019209846A1

Abstract

The invention relates to a method for operating a 3D distance sensor device, comprising the steps of: - providing a defined number of detected measurement points; - evaluating the detected measurement points with the aid of at least one defined criterion; and - estimating a visibility of the 3D distance sensor device from the evaluated measurement points.

Description

description

title

Method of operating a 3D distance sensor device

The invention relates to a method for operating a 3D distance sensor device. The invention also relates to a 3D distance sensor device. The

The invention also relates to a computer program. The invention also relates to a machine-readable storage medium.

Highly and fully automated vehicles (level 3-5) will be in the next

Years to be found on our streets more and more often. There are various concepts of how such an automated vehicle can be implemented.

All of these approaches require a wide variety of sensors (e.g. video cameras,

□ DAR, radar, ultrasonic sensors, etc.). For safe driving in all

Weather conditions require that all sensors independently provide an estimate of their current visibility.

DE 199 31 825 A1 discloses a device for measuring visibility, especially for motor vehicles, which has at least one optical transmitting element, at least one optical receiving element and a measurement signal evaluation unit which determines a current visibility from the light reflected at one or more different spatial zones.

DE 10 2016 014 549 A1 discloses an autonomous vehicle that contains both a LiDAR sensor and a camera. The LiDAR sensor can work in the range of visible light, in the IR range or another wavelength range. If signs have an IR-reflective print, the IR range can be used.

DE 10 2016 014 549 A1 discloses a method for determining the visual range from a vehicle by means of a LiDAR sensor. EP 2 101 193 A1 discloses a safety sensor and a method for the contactless measurement of positions, distances and / or speeds.

US Pat. No. 6,362,773 B1 discloses a combination of several sensors or an adapted, complex evaluation of raw sensor data. However, these raw data are not available in many sensor architectures of the central sensor fusion and evaluation unit.

Disclosure of the invention

It is an object of the present invention to provide an improved method for operating a 3D distance sensor device.

According to a first aspect, the invention creates a method for operating a 3D distance sensor device, comprising the steps:

Providing a defined number of detected measuring points;

- Evaluation of the detected measuring points with the aid of at least one defined criterion; and

Estimating a visual range of the 3D distance sensor device from the measured points evaluated.

In this way, it can advantageously be recognized whether the range of vision is impaired for the 3D distance sensor device. In this way it can advantageously be quantified how much the visibility is reduced, it being possible to distinguish a reduced visibility from a blockage. The proposed method advantageously represents a method at the system level, i.e. that an already evaluated 3D point cloud of measuring points is used and raw sensor data is therefore not required.

According to a second aspect, the object is achieved with a 3D distance sensor device which is designed to carry out the proposed method.

According to a third aspect, the object is achieved with a computer program. According to a fourth aspect, the object is achieved with a machine-readable storage medium.

Preferred embodiments of the proposed method are the subject of the respective dependent claims.

An advantageous further development of the method is characterized in that measurement points of a geometrically lowest level of the 3D distance sensor device are analyzed, the number of detected measurement points being compared with a defined first threshold value. It is expected that all scan or measurement points of the geometrically lowest level are always available, because a road is arranged in this level, on which the vehicle with the 3D distance sensor device is located, whereby the road is usually only approx. 10m far from the 3D distance sensor device. If this is not the case, appropriate conclusions can be drawn from it.

A further advantageous development of the method is characterized in that in the event that the proportion of the detected measurement points is defined below the first threshold value, a cover glass of the 3D distance sensor device is covered. In this case, a cleaning process of the 3D distance sensor device can be triggered, for example, in order to establish the full range of vision.

Another advantageous development of the method provides that in the event that the proportion of detected measuring points is defined above the first threshold value, an evaluation of a defined, calibrated distance histogram of the measuring points is carried out, a decay rate of the frequency of the measuring points being determined over the distance . The distance histogram is evaluated in this way in order to be able to make more precise statements about the visual range. For example, reduced visibility due to fog or rain can be determined in the distance histogram.

Another advantageous development of the method provides that in the event that the proportion of the detected measuring points is defined below a second threshold value of the decay rate, a defined high visual range of the 3D distance sensor device is detected. This is determined by the fact that the distance histogram slowly declines at a high visibility range.

Another advantageous development of the method is characterized in that in the event that the proportion of the detected measuring points is defined below a second threshold value of the decay rate, an evaluation of a total number of the measuring points is carried out. Here, in particular, the influence of fog on the visibility is determined, which is particularly irrelevant for short distances, in which case the total number of measurement points is evaluated.

Another advantageous embodiment of the method is characterized in that when the total number of measurement points falls below a third threshold value, a defined low visual range of the 3D distance sensor device is determined and, when the third threshold value of the total number of measurement points is exceeded, the 3D distance sensor device is blocked from sight is detected. In this way, suitable knowledge can be obtained depending on the number of measuring points. The distance histogram declines quickly, but there are a large number of measurement points in the 3D point cloud, which can be an indication of a blockage in the view of the 3D sensor device.

A further advantageous embodiment of the method is characterized in that a driving function of a vehicle is adapted from the estimated visual range and / or cleaning of the 3D distance sensor device is initiated. As a result, suitable measures can be taken to adapt to the reduced visibility of the 3D distance sensor device.

Another advantageous embodiment of the method is characterized in that the method is carried out in normal operation of the 3D distance sensor device. In this way, it is advantageously possible to determine the visual range of the 3D distance sensor device in quasi real time.

Another advantageous embodiment of the method is characterized in that the method is carried out within the 3D distance sensor device or on an external computer device. That way one can "Intelligent 3D distance sensor" can be implemented or the computing power required for this can be suitably outsourced.

The invention is described in detail below with further features and advantages using two figures. The figures are intended in particular to clarify the principles essential to the invention and are not necessarily drawn to scale.

In the figures shows:

Fig. 1 is a basic flowchart of a proposed method for operating a 3D distance sensor device; and

2 shows an exemplary distance histogram which can be analyzed within the framework of the proposed method.

Description of embodiments

A core idea of the present invention is in particular to provide an improved method for operating or analyzing a 3D distance sensor device.

One advantage of the proposed method is that, for example, a central evaluation unit of a 3D distance sensor can determine or estimate the visual range, with additional sensors advantageously not being required for this.

Estimating the visibility of a sensor in all weather conditions is necessary in order to enable a partially or fully autonomous vehicle to be driven safely.

For example, the speed must be adjusted to the current field of vision of the sensors. However, the range of vision of a sensor can be severely impaired by various atmospheric weather influences such as rain, fog, snowfall, etc. A distinction must be made as to whether the performance of the sensor is based on such atmospheric influences or due to statistically distributed blockages on the sensor cover, such as raindrops, dirt, etc. is impaired. The proposed method also makes it possible to identify situations in which the view of the 3D distance sensor is blocked by one or more objects (e.g. a vehicle parked in a garage) and therefore no statement is made about the current visibility of the 3D distance sensor can be.

The proposed method for operating a 3D distance sensor device based on a 3D point cloud is based on a combination of several usable parameters. In the following, a 3D point cloud is understood to be a three-dimensional cluster of measurement points that are recorded using a three-dimensional distance sensor (e.g. LiDAR, radar sensor, etc.). The number of points in the environment detected by the 3D distance sensor is primarily used. The method is described below by way of example for a 3D LiDAR distance sensor arranged on a vehicle, but is advantageously also suitable for improved operation of a radar sensor.

1 shows a basic sequence of a proposed method. In a step 100, the 3D distance sensor device provides a 3D point cloud with a defined number of measurement points in the environment.

In a step 101, an analysis of the number of points of the measurement points of a geometrically lowest level of the 3D distance sensor device is carried out. The geometrically lowest level of a 3D distance sensor device typically covers a road at a short distance from the vehicle, measurement data provided in this way therefore often also being used to identify open spaces. The short distance means that all measuring points of this plane should be present in the 3D point cloud even if the visibility is limited.

This parameter therefore does not represent a measure of the visual range of the 3D distance sensor device, but it does allow, for example, a blockage of vision on the cover of the 3D distance sensor device to be recognized. In a step 102, a total number of measuring points of the geometrically lowest level is checked, with the case that the total number of measuring points defines one falls below the first threshold value SW1 defined, it is determined that a blockage of a cover of the 3D distance sensor device is present. For example, this can be caused by a deposit (eg dirt, snow, ice, etc.) on the sensor cover, which is why a cleaning process of the 3D distance sensor device is initiated in this case.

If the total number of measurement points of the geometrically lowest level of the 3D point cloud exceeds the first threshold value SW1 in a defined manner, an analysis of a decay rate of a distance histogram of the entire 3D point cloud explained in more detail in connection with FIG. 2 is carried out in a step 110.

To analyze the distance histogram of the entire 3D point cloud, all points of the 3D point cloud are shown in the distance histogram, which represents a graph that shows the frequency of the measurement points of the 3D point cloud as a function of distance. A suitable mathematical function can be provided for this data, an exponentially decreasing function being selected in the present case, which is characterized by the two parameters amplitude and decay rate.

These two parameters are influenced by different processes. The amplitude is influenced by blockages on the sensor cover of the 3D distance sensor device. The rate of decay is influenced by the atmospheric visibility, since a small visibility only affects the detection of objects that are far away, but not the detection of objects arranged close to the 3D distance sensor device. Furthermore, the decay rate is influenced by the structure of the landscape or by a blockage of the 3D distance sensor device's view by other objects (e.g. a garage, location behind a truck, etc.). As a result, the distance histogram allows a value for the visual range of the 3D distance sensor device to be specified based on the decay rate.

In a step 111 it is checked whether the decay rate of the measuring points of the distance histogram has a certain second threshold value SW2, in which case in the case that this is less than the defined second threshold value SW2 in In a step 112, a defined high visibility range for the 3D distance sensor device is determined.

If, on the other hand, the decay rate of the distance histogram is greater than the second threshold value SW2, this means that, for example, the 3D distance sensor device may have a low visibility or an object blockage, which is examined in more detail below in a step 120. An analysis of the total number of points of the measuring points of the 3D point cloud is carried out in step 121.

In a step 122 it is determined that the total number of points mentioned for the measurement points of the 3D point cloud is less than a defined third threshold value SW3, whereby a low visibility range (e.g. due to fog, rain, snowfall, etc.) is determined for the 3D distance sensor device . This can be the case, for example, in fog, in which case there is hardly any restriction in the field of view of the 3D sensor device for short distances, but a total number of measuring points can be considerably reduced for greater distances.

In a step 123 it is established that the named total number of points of the measurement points of the 3D point cloud is greater than the defined third threshold value SW3, in which case a visual blockage of the 3D distance sensor device is established. This situation can exist, for example, when the vehicle equipped with the 3D distance sensor device is parked in front of a wall.

For a complete decision-making regarding the visual range estimation of the 3D distance sensor device, an evaluation of all of the aforementioned three threshold values SW1, SW2 and SW3 is necessary, which is why separate detailed analyzes are carried out in steps 101, 110 and 120 for this purpose. The decision can be made using the decision tree from FIG. 1. A quantitative value of the visual range can be determined by a decay rate of the distance histogram that is quantified by means of a calibration process. In this way, the distance histogram can advantageously also be used without the other two threshold values. Advantageously, only the distance histogram can be evaluated for the estimation of the visual range, so that the detailed analyzes of steps 101, 120 can also be omitted. This advantageously saves computing time and in this way selected criteria or all of the proposed criteria for checking the visual range of the 3D distance sensor device can be checked.

The distance histograms are each suitably pre-calibrated for a specific type of 3D distance sensor device, whereby a specific decay rate is assigned to a visual range of the 3D distance sensor device. Due to the large number of different types of 3D sensor devices, it is not possible to specify generally valid numerical values for visibility, distances, numbers of measuring points of the 3D point cloud and threshold values SW1, SW2, SW3.

2 shows an exemplary distance histogram, the distance d being plotted on the x-axis and a logarithmic frequency H of the measuring points of the 3D point cloud on the y-axis. You can see two courses A, B of measuring points of a 3D point cloud. The course A represents a high visibility range with a low exponential decay rate of the number of measuring points over the distance d and the course B a low visibility range with a high exponential decay rate of the number of measuring points over the distance d. The second threshold value SW2 corresponds to the decay rate or the slope of the curves A, B.

The proposed 3D distance sensor device can advantageously be used to detect the environment in highly and fully automated vehicles (level 3-5).

The proposed method can preferably be carried out in a defined time pattern of the 3D distance sensor device, quasi in real time (eg with a cycle time of approx. 10 frames / s, ie with approx. 10 Hz) in a background operation. It can be provided that the proposed method is implemented as software running in the 3D distance sensor device. Alternatively, it can also be provided that the software for executing the proposed method runs on an external computer device, which can be arranged in the cloud, for example.

Suitable measures can be derived from the determined visibility, e.g. an adaptation of one or more driver assistance functions of the

Vehicle, a speed of the vehicle, etc.

The proposed method is preferably carried out in normal operation of the 3D distance sensor device, in which the 3D distance sensor device is arranged in or on a highly or fully automated vehicle and is used to detect the surroundings when the vehicle is in operation. By determining the visual range of the 3D distance sensor device quasi in real time, a useful value of the 3D distance sensor device is significantly increased. Although the invention in connection with an optoelectronic 3D-

Scanner in the form of a LiDAR sensor for a motor vehicle was explained, it is for example also conceivable to use the proposed method for other 3D distance sensor devices, e.g. for a radar sensor. The person skilled in the art will thus recognize that a large number of modifications are possible without departing from the essence of the invention.

Claims

Expectations

1. A method for operating a 3D distance sensor device, comprising the steps:

Providing a defined number of detected measuring points;

Estimating a visual range of the 3D distance sensor device from the evaluated measuring points.

2. The method according to claim 1, wherein measuring points of a geometrically lowest level of the 3D distance sensor device are analyzed, the number of detected measuring points being compared with a defined first threshold value (SW1).

3. The method according to claim 2, wherein in the event that the proportion of the detected measurement points is defined below the first threshold value (SW1), an obscuration of a cover glass of the 3D distance sensor device is determined.

4. The method according to any one of the preceding claims, wherein an evaluation of a defined calibrated distance histogram of the measurement points is performed, wherein a decay rate of a frequency of the measurement points is determined over the distance.

5. The method according to any one of the preceding claims, wherein in the event that the proportion of the detected measurement points is defined below a second threshold value (SW2) of the decay rate, a defined high visibility of the 3D distance sensor device is determined.

6. The method according to claim 5, wherein in the case that the proportion of the detected measurement points is defined below a second threshold value (SW2) Decay rate is, an evaluation of a total number of the measuring points is carried out.

7. The method according to claim 6, wherein when falling below a third threshold value (SW3) of the total number of measurement points, a defined low visibility range of the 3D distance sensor device is determined and when exceeding the third threshold value (SW3) of the total number of measurement points, a blockage of vision 3D distance sensor device is detected.

8. The method according to any one of the preceding claims, wherein from the

estimated visual range a driving function of a vehicle is adapted and / or cleaning of the 3D distance sensor device is initiated.

9. The method according to any one of the preceding claims, wherein the method is carried out in normal operation of the 3D distance sensor device.

10. The method according to any one of the preceding claims, wherein the method is carried out within the 3D distance sensor device or on an external computer device.

11. 3D distance sensor device that is set up to carry out a method according to one of claims 1 to 10.

12. Computer program comprising instructions that are used in the execution of the

Computer program causing a computer to execute a method according to one of claims 1 to 10.

13. Machine-readable storage medium on which the computer program according to claim 12 is stored.