Method for evaluating the reliability of probe data

Abstract

To enable more accurate evaluation of probe data than conventional methods.SOLUTION: A computer-executed method for evaluating the reliability of probe data includes steps of: acquiring the probe data including at least a position, a speed, and a direction of travel of a vehicle; determining a first reliability regarding the positional accuracy in the direction of travel and a second reliability regarding the positional accuracy in the lateral direction based on the speed of the vehicle in the probe data; and associating the first reliability and the second reliability with the probe data.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present disclosure relates to a method for evaluating the reliability of probe data.

Background Art

[0002] The accuracy of a map generated and updated based on probe data greatly depends on the accuracy of the probe data.

[0003] Patent Document 1 discloses setting reliability according to a driving lane in a system that acquires probe data indicating a driving situation when a vehicle travels on a target road section. Patent Document 2 discloses setting the reliability of probe data based on a vehicle driving pattern. [[ID=1...]

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Summary of the Invention

Problems to be Solved by the Invention

[0005] [[ID=5)] One aspect of the present disclosure aims to provide a technique capable of more appropriately evaluating the accuracy of probe data than in the prior art.

Means for Solving the Problems

[0006] One aspect of the present disclosure is a reliability evaluation method performed by a computer for evaluating the reliability of probe data, comprising the steps of: acquiring probe data including at least the position, speed, and direction of travel of a vehicle; determining a first reliability for position accuracy in the direction of travel and a second reliability for position accuracy in the lateral direction based on the vehicle speed in the probe data; and associating the first reliability and the second reliability with respect to the probe data. [Effects of the Invention]

[0007] According to the aspects of this disclosure, the accuracy of probe data can be evaluated more appropriately than in the conventional method. [Brief explanation of the drawing]

[0008] [Figure 1] A diagram showing the configuration of the information processing device according to the embodiment. [Figure 2] A flowchart showing the processing flow according to the embodiment. [Figure 3] A diagram illustrating the reliability of the direction of travel. [Figure 4] A diagram illustrating the reliability of lateral position. [Figure 5] A diagram illustrating the relationship between probe data and confidence level. [Modes for carrying out the invention]

[0009] This disclosure relates to a system for generating map data based on probe data obtained from sensors mounted on a vehicle, and more particularly to a technology for evaluating the reliability of probe data. Probe data is data obtained from sensors such as cameras, LiDAR sensors, radar sensors, and sonar mounted on a vehicle. High-performance sensors can provide highly reliable probe data. While high-performance sensors can be obtained, cost constraints make it difficult to equip many vehicles with them. Therefore, if lower-performance sensors are used, it is expected that reliable and unreliable probe data will be mixed together. In order to successfully extract reliable probe data and generate map data, it is necessary to accurately evaluate the reliability of the probe data.

[0010] (Overall structure) Figure 1 shows the configuration of an information processing device 100 in one embodiment. The information processing device 100 may be functionally distributed and deployed on the cloud side, the vehicle side, or both. The information processing device 100 acquires probe data, evaluates the reliability of each probe data, and generates map data from the probe data, taking the reliability into consideration. The information processing device 100 can be considered as either a probe data reliability evaluation device or a map data generation device.

[0011] The information processing device 100 is comprised of a computer comprising a processor 110 and a memory 120. The memory 120 non-temporarily stores computer programs that cause the processor 110 to function as a probe data acquisition unit 111, a reliability evaluation unit 112, a storage unit 113, and a map generation unit 114.

[0012] Figure 2 is a flowchart showing the processing flow performed by the information processing device.

[0013] In step S201, the probe data acquisition unit 111 acquires probe data obtained from sensors such as cameras and radar mounted on the vehicle. The probe data acquisition unit 111 may acquire probe data directly from the vehicle, or it may acquire probe data from an information server that stores probe data obtained from the vehicle.

[0014] The probe data includes information about the vehicle and information about the surrounding environment of the vehicle. Examples of information about the vehicle include vehicle attitude (behavior) information such as the position, traveling direction, speed, acceleration, and yaw rate of the vehicle. Information about the surrounding environment of the vehicle includes the type, position, shape, and other attributes of road features such as lane lines, traffic signals, and signs.

[0015] In step S202, the reliability evaluation unit 112 calculates the reliability for each of the probe data acquired by the probe data acquisition unit 111. Here, the reliability evaluation unit 112 calculates a plurality of reliabilities with different viewpoints for one piece of probe data. Examples of the plurality of reliabilities include the reliability of the position of the vehicle along the traveling direction (first reliability) and the reliability of the position of the vehicle along the lateral direction (direction orthogonal to the traveling direction) (second reliability). The traveling direction can also be expressed as the direction along the road or lane.

[0016] Before explaining the method for calculating the reliability, the tendency obtained by experiments regarding the accuracy of the probe data will be explained. In the experiment, two types of sensors were mounted on the vehicle to collect probe data. One sensor is a relatively low-cost sensor assumed to be used for collecting probe data in actual operation, and the other sensor is a high-precision sensor for verification. Since the former sensor is the one used to actually acquire the probe data, the data obtained from the former sensor is referred to as probe data here. On the other hand, the data obtained from the latter sensor is referred to as verification data. The verification data can be regarded as accurate and thus can be used to evaluate the error of the probe data.

[0017] As a result of the experiment, it was found that although the amount of error of the vehicle position varies according to the speed of the vehicle with respect to both the traveling direction position and the lateral direction position, the correlation between the speed and the error shows a different tendency between the traveling direction position and the lateral direction position.

[0018] Regarding the accuracy of the forward direction position, first, it was found that when the vehicle speed is close to 0 km / h, for example, when it is 0 to 1.5 km / h, the accuracy extremely decreases. After that, as the vehicle speed increases, the error decreases and the accuracy improves, and it was found that a high-precision position with little error can be obtained at a certain speed or higher. For example, at 5 km / h or higher, results that meet the required accuracy were obtained. On the other hand, it was also found that when exceeding a certain speed threshold, for example, 30 km / h, the error becomes larger as the speed increases.

[0019] Regarding the accuracy of the lateral position, it was found that when the vehicle speed is close to 0 km / h, the accuracy is about the same up to a certain speed. The position accuracy in this speed range is not sufficient, but the error is not so large that it cannot be used. The range in which this approximately the same accuracy can be obtained was, for example, the range of 0 to 10 km / h. After that, as the vehicle speed increases, the error decreases and the accuracy improves, and it was found that a high-precision position with little error can be obtained at a certain speed or higher. For example, at 15 km / h or higher, results that meet the required accuracy were obtained. Furthermore, the error did not increase at 15 km / h or higher.

[0020] Based on the above experimental results, in this embodiment, the reliability of the forward direction position and the lateral position is determined as follows. In this embodiment, the minimum value of the reliability is 0 and the maximum value is 1.

[0021] Figure 3 illustrates the reliability of the direction of travel position. As shown in the figure, the reliability is set to 0 when the vehicle speed is less than the first reliability level (e.g., 1.5 km / h). When the vehicle speed is greater than the first threshold and less than the second threshold (e.g., 5 km / h), the reliability increases monotonically from 0 to 1. When the vehicle speed is greater than the second threshold and less than the third threshold (e.g., 30 km / h), the reliability is set to 1. When the vehicle speed is greater than the third threshold, the reliability decreases monotonically. Here, if the allowable error differs depending on the type of road the vehicle is traveling on, the rate of decrease in reliability may be changed. For example, in Figure 3, when the vehicle speed is greater than the third threshold, the rate of decrease in reliability for highways (302) is set to be smaller than the rate of decrease for general roads (301). For general roads, the rate of decrease is such that the reliability becomes 0 at a vehicle speed of 80 km / h, but for highways, the rate of decrease is such that the reliability becomes 0 at a vehicle speed of 120 km / h. Furthermore, the rate of decrease in reliability when the vehicle speed is greater than the third threshold may be the same regardless of the type of road the vehicle is traveling on.

[0022] Figure 4 illustrates the confidence level of the lateral position. As shown in the figure, when the vehicle speed is less than the fourth threshold (e.g., 10 km / h), the confidence level is set to a fixed value (e.g., 0.3). When the vehicle speed is greater than the fourth threshold but less than the fifth threshold (e.g., 15 km / h), the confidence level increases monotonically in a linear fashion. When the vehicle speed is greater than the fifth threshold, the confidence level is set to 1. The confidence level of the lateral position should be set to increase monotonically (broadly monotonically) with respect to the vehicle speed.

[0023] The threshold values ​​and specific confidence values ​​shown here are merely examples and may be modified as appropriate depending on the required positional accuracy and the accuracy of the sensors used. Furthermore, while examples of linear monotonically increasing or decreasing confidence are shown when it is monotonically increasing or decreasing according to vehicle speed, it is not necessary to change it linearly if it is simply monotonically increasing or decreasing. Quadratic functions, higher-order polynomial functions, exponential functions, or stepwise monotonically increasing / decreasing functions can be used.

[0024] In step S203, the storage unit 113 stores the calculated confidence score in association with the probe data from which it was calculated. Figure 5 shows an example of the association. As shown in the figure, the identifier (data ID) that identifies the probe data and the confidence scores for the forward position and lateral position are used. The confidence level is associated and stored. Alternatively, instead of associating the data ID of the probe data with the confidence level, the association may be made by adding the confidence level to the probe data. Furthermore, the data may be stored inside the information processing device 100, in the information server from which the probe data was collected, or in other devices.

[0025] In step S204, the map generation unit 114 generates map data from the probe data, taking into account the reliability of the probe data. The map generation unit 114 integrates information from multiple probe data to generate information for a single feature. In this process, probe data with a reliability below a standard value is excluded, or weighting is applied according to the reliability before generating information for the feature from the probe data. The reliability of the probe data includes reliability for the position in the direction of travel and reliability for the position in the lateral direction. It is possible to focus on the reliability of only one of these, or to focus on the integrated reliability obtained by integrating (e.g., averaging) these reliabilitys, or to apply different criteria to each reliability. Furthermore, different criteria may be applied depending on the feature being calculated. The location and size information of the feature is determined based on the location information of the probe car and the relative position and size of the feature relative to the probe car. Therefore, the positional accuracy of the probe data is important for generating highly accurate map data. It is possible that the required accuracy for the position in the direction of travel and the lateral direction differs depending on the type of feature. For example, for one feature, accuracy in the lateral position is important; for another, accuracy in the direction of travel position is important; and for yet another, accuracy in both positions may be important. In this embodiment, the reliability (accuracy) of the direction of travel and lateral position is calculated for each probe data, so it becomes possible to use appropriate probe data according to the target feature information, and as a result, it becomes possible to generate high-quality map data.

[0026] (Effects of the embodiment) According to this embodiment, multiple confidence levels from different perspectives are calculated for a single probe data set. Therefore, processing can be performed based on an appropriate confidence level depending on the application of the probe data, enabling the generation of high-quality map data. Furthermore, this embodiment proposes a method for calculating confidence levels based on novel insights into the direction of travel position and lateral position in the probe data, thereby enabling a more appropriate evaluation of the reliability of the probe data than before, and ultimately enabling the generation of high-quality map data.

[0027] (Other embodiments) The embodiments described above are merely examples, and this disclosure may be modified as appropriate without departing from its essence.

[0028] In the above embodiment, the information processing device 100 performs both the reliability evaluation of probe data and map generation, but the reliability evaluation and map generation may be performed by different devices or systems.

[0029] The present disclosure can also be realized by supplying a computer program implementing the functions described in the embodiments above to a computer, and having one or more processors in the computer read and execute the program. Such a computer program may be provided to the computer by a non-temporary computer-readable storage medium that can be connected to the computer's system bus, or it may be provided to the computer via a network. Non-temporary computer-readable storage mediums include, for example, any type of disk such as magnetic disks (floppy disks, hard disk drives (HDDs), etc.), optical disks (CD-ROMs, DVDs, Blu-ray discs, etc.), read-only memory (ROM), random access memory (RAM), EPROM, EEPROM, magnetic cards, flash memory, optical cards, and any type of medium suitable for storing electronic instructions. [Explanation of symbols]

[0030] 100: Information processing device 110: Processor 111: Probe data acquisition unit 112: Reliability evaluation unit 113: Memory Unit 114: Map Generation Unit

Claims

1. A reliability evaluation method performed by a computer to evaluate the reliability of probe data, The steps include acquiring probe data that includes at least the vehicle's position, speed, and direction of travel, The steps include determining a first confidence level regarding the position accuracy in the direction of travel and a second confidence level regarding the position accuracy in the lateral direction based on the vehicle speed in the probe data, The steps include associating the first confidence level and the second confidence level with the probe data, The first confidence level is, If the vehicle speed is less than the first threshold, the value is 0. If the vehicle speed is greater than the first threshold and less than the second threshold, it increases monotonically. If the vehicle speed is greater than the second threshold and less than the third threshold, the value is 1. If the vehicle speed is greater than the third threshold, it decreases monotonically. It was decided that, Includes, The second confidence level is determined to increase monotonically with respect to the vehicle speed. Confidence assessment method.

2. The rate of decrease in the first reliability may be calculated using different reliability values ​​depending on the type of road on which the vehicle is traveling. When the vehicle speed is greater than the third threshold, the rate of reduction when the vehicle is traveling on an expressway is smaller than the rate of reduction when the vehicle is traveling on a regular road. The reliability evaluation method according to feature 1.

3. The second confidence level is, If the vehicle speed is less than the fourth threshold, it is a constant. If the vehicle speed is greater than the fourth threshold and less than the fifth threshold, it increases monotonically. If the vehicle speed is greater than the fifth threshold, the value is 1. It is decided that The reliability evaluation method according to feature 1.

Images