Filtering Device, Filtering Method, and Program

Abstract

To appropriately remove vehicle sensor data.SOLUTION: A filtering device C32 that executes processing for removing vehicle sensor data indicating a detection result of a vehicle sensor 11A mounted on a probe vehicle 11, comprises: a scoring processing unit C32A that performs scoring of vehicle sensor data; a determination unit C32B that compares a score for the vehicle sensor data with a first threshold value; and a filtering processing unit C32C that executes processing for removing the vehicle sensor data on the basis of a determination result of the determination unit C32B. The scoring processing unit C32A performs scoring of vehicle sensor data on the basis of at least any one of information indicating whether or not the probe vehicle 11 is a vehicle that made a stop sign violation, information indicating whether or not the probe vehicle 11 is a vehicle that breached a limited speed, information indicating whether or not a traffic lane deviation warning is frequently generated on the probe vehicle 11, and information indicating whether or not an inter-vehicle distance between the probe vehicle 11 and a vehicle ahead or a TTC is equal to a second threshold value or less.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a filtering device, a filtering method, and a program.

Background Art

[0002] Patent Document 1 describes a technique for generating a driving route map based on probe information obtained by cameras and various sensors mounted on a vehicle. In the technique described in Patent Document 1, filtering of probe information is performed based on the result of driving diagnosis (scoring result). Patent Document 1 describes that the frequency of sudden acceleration, the frequency of sudden braking, the frequency of sudden left / right turns, and the frequency of left / right wobbling are calculated from the acceleration of the probe information, and scores are given such that the lower the frequency, the higher the score, and the total score is calculated. In the technique described in Patent Document 1, when the total score is less than the passing score, it is estimated that the probe information is due to dangerous driving, and the probe information is not used for generating the driving route map.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Incidentally, in the technique described in Patent Document 1, although it is determined whether or not to exclude probe information (vehicle sensor data indicating the detection result of a vehicle sensor) based on the frequency of sudden acceleration, sudden braking, sudden left / right turns, and left / right wobbling frequency of a probe vehicle, it is not determined whether or not to exclude probe information based on information indicating whether or not the probe vehicle has violated traffic rules. Further, in the technique described in Patent Document 1, it is not determined whether or not to exclude probe information based on information indicating whether or not a lane departure warning frequently occurs in the probe vehicle. Furthermore, in the technique described in Patent Document 1, it is not determined whether or not to exclude probe information based on the inter-vehicle distance between the probe vehicle and the preceding vehicle or TTC (Time To Collision). Therefore, in the technique described in Patent Document 1, there is a possibility that the probe information may not be appropriately excluded.

[0005] In view of the above points, an object of the present disclosure is to provide a filtering device, a filtering method, and a program capable of appropriately excluding vehicle sensor data indicating the detection result of a vehicle sensor mounted on a probe vehicle.

Means for Solving the Problems

[0006] (1) One aspect of the present disclosure is a filtering device that executes a process of excluding vehicle sensor data indicating the detection result of a vehicle sensor mounted on a probe vehicle, the filtering device including a quantification processing unit that quantifies the vehicle sensor data, a determination unit that compares the quantified value of the vehicle sensor data with a first threshold value, and a filtering processing unit that executes a process of excluding the vehicle sensor data based on the determination result of the determination unit, wherein the quantification processing unit quantifies the vehicle sensor data based on at least any one of information indicating whether or not the probe vehicle is a vehicle that has violated a stop rule, information indicating whether or not the probe vehicle is a vehicle that has violated a speed limit, information indicating whether or not a lane departure warning frequently occurs in the probe vehicle, and information indicating whether or not the inter-vehicle distance or TTC between the probe vehicle and the preceding vehicle is equal to or less than a second threshold value.

[0007] (2) In the filtering device of (1), the quantization processing unit quantizes the vehicle sensor data based on information indicating whether the probe vehicle is a vehicle that has violated a temporary stop. When the probe vehicle is a vehicle that has violated a temporary stop, the filtering processing unit may be more likely to exclude the vehicle sensor data than when the probe vehicle is not a vehicle that has violated a temporary stop.

[0008] (3) In the filtering device of (1) or (2), the quantization processing unit quantizes the vehicle sensor data based on information indicating whether the probe vehicle is a vehicle that has violated the speed limit. When the probe vehicle is a vehicle that has violated the speed limit, the filtering processing unit may be more likely to exclude the vehicle sensor data than when the probe vehicle is not a vehicle that has violated the speed limit.

[0009] (4) In any of the filtering devices of (1) to (3), the quantization processing unit quantizes the vehicle sensor data based on information indicating whether the lane departure warning frequently occurs in the probe vehicle. When the lane departure warning frequently occurs in the probe vehicle, the filtering processing unit may be more likely to exclude the vehicle sensor data than when the lane departure warning does not frequently occur in the probe vehicle.

[0010] (5) In any of the filtering devices of (1) to (4), the quantization processing unit quantizes the vehicle sensor data based on information indicating whether the inter-vehicle distance or the TTC is less than or equal to the second threshold. When the inter-vehicle distance or the TTC is less than or equal to the second threshold, the filtering processing unit may be more likely to exclude the vehicle sensor data than when the inter-vehicle distance or the TTC is not less than or equal to the second threshold.

[0011] (6) In any of the filtering devices of (1) to (5), the vehicle sensor includes a front camera that captures an image in front of the probe vehicle, the vehicle sensor data includes an image in front of the probe vehicle captured by the front camera, and when the inter-vehicle distance or the TTC is less than or equal to the second threshold value, only the preceding vehicle may be included in the image in front of the probe vehicle captured by the front camera, or most of the image in front of the probe vehicle captured by the front camera may indicate the preceding vehicle.

[0012] (7) In any of the filtering devices of (1) to (6), the determination unit determines whether the mounting position of the vehicle sensor has changed, and when the mounting position of the vehicle sensor has changed, the filtering processing unit may execute a process of excluding the post-mounting-position-change vehicle sensor data indicating the detection result of the vehicle sensor after the mounting position has changed.

[0013] (8) In the filtering device of (7), the vehicle sensor includes an in-vehicle camera that captures an image indicating the situation outside the probe vehicle, the determination unit determines whether the mounting position of the in-vehicle camera has changed based on the post-mounting-position-change image, which is an image indicating the situation outside the probe vehicle captured by the in-vehicle camera after the mounting position has changed, and the pre-mounting-position-change image, which is an image indicating the situation outside the probe vehicle captured by the in-vehicle camera before the mounting position changes, and when the mounting position of the in-vehicle camera has changed, the filtering processing unit may execute a process of excluding the post-mounting-position-change image.

[0014] In the filtering device of (9) (7), the vehicle sensor includes a LiDAR having a function of recognizing an object outside the probe vehicle. The determination unit determines whether or not the mounting position of the LiDAR has changed based on the result of object recognition performed by the LiDAR during the calibration of the LiDAR after the mounting position has changed and the result of object recognition performed by the LiDAR during the calibration of the LiDAR before the mounting position changes. When the determination unit determines that the mounting position of the LiDAR has changed, the filtering processing unit may execute a process of excluding the result of object recognition performed by the LiDAR after the determination unit determines that the mounting position of the LiDAR has changed.

[0015] In any of the filtering devices of (10) (1) to (9), the vehicle sensor data not excluded by the filtering processing unit may be used for at least any one of generation of map information, generation of traffic information, and generation of information indicating the deterioration state of the road surface.

[0016] An aspect of the present disclosure is a filtering method in which a filtering device executes a process of excluding vehicle sensor data indicating a detection result of a vehicle sensor mounted on a probe vehicle, the filtering device including a quantization processing step of quantizing the vehicle sensor data, a determination step of comparing the number of points of the vehicle sensor data with a first threshold value, and a filtering processing step of executing a process of excluding the vehicle sensor data based on the determination result in the determination step. In the quantization processing step, the quantization of the vehicle sensor data is performed based on at least any one of information indicating whether the probe vehicle is a vehicle that has violated a stop sign, information indicating whether the probe vehicle is a vehicle that has violated a speed limit, information indicating whether lane departure warnings frequently occur in the probe vehicle, and information indicating whether the inter-vehicle distance or TTC between the probe vehicle and the preceding vehicle is equal to or less than a second threshold value.

[0017] (12) One aspect of the present disclosure is a program for causing a computer constituting a computer or a server device mounted on a probe vehicle to execute a quantification processing step of quantifying vehicle sensor data indicating a detection result of a vehicle sensor mounted on the probe vehicle, a determination step of comparing the quantified value of the vehicle sensor data with a first threshold value, and a filtering processing step of excluding the vehicle sensor data based on the determination result in the determination step. In the quantification processing step, the quantification of the vehicle sensor data is performed based on at least any one of information indicating whether the probe vehicle is a vehicle violating a stop, information indicating whether the probe vehicle is a vehicle violating a speed limit, information indicating whether a lane departure warning frequently occurs in the probe vehicle, and information indicating whether a distance between the probe vehicle and a preceding vehicle or a TTC is equal to or less than a second threshold value.

Advantages of the Invention

[0018] According to the present disclosure, it is possible to appropriately exclude vehicle sensor data indicating a detection result of a vehicle sensor mounted on a probe vehicle.

Brief Description of the Drawings

[0019]

Figure 1

Figure 2

Figure 3

Figure 4

Modes for Carrying Out the Invention

[0020] Hereinafter, with reference to the drawings, embodiments of the filtering device, filtering method, and program of the present disclosure will be described.

[0021] <First Embodiment> FIG. 1 is a diagram showing a first example of a vehicle data utilization system 1 to which a filtering device C32 according to the first embodiment is applied. FIG. 2 is a diagram showing an example of functions of a processor C3 shown in FIG. 1.

[0022] In the examples shown in FIGS. 1 and 2, the vehicle data utilization system 1 is a system for utilizing vehicle sensor data indicating the detection results of a vehicle sensor 11A mounted on a probe vehicle 11. The vehicle data utilization system 1 includes, for example, a probe vehicle 11, a server device 12, and a map information generation device 13. The probe vehicle 11 is a vehicle connected to a network NW and is referred to as a connected car. The probe vehicle 11 includes, for example, a vehicle sensor 11A, a communication unit 11B, a control unit 11C, and a map information unit 11D. The vehicle sensor 11A, the communication unit 11B, the control unit 11C, and the map information unit 11D are connected via an in-vehicle network 11E. The vehicle sensor 11A includes, for example, an in-vehicle camera (e.g., a front camera, a rear camera, a side camera, etc.) that captures an image indicating the external situation of the probe vehicle 11, a radar having a function of recognizing an object outside the probe vehicle 11, LiDAR (Light Detection And Ranging), etc. Further, the vehicle sensor 11A includes various sensors for detecting data related to the driving situation of the probe vehicle 11 (e.g., engine rotation speed, operation situation such as accelerator / brake, vehicle speed, acceleration, shift position, travel distance, position information, etc.). The position information indicating the current position of the probe vehicle 11 is obtained by a GPS (Global Positioning System) module functioning as the vehicle sensor 11A receiving a GPS signal. Furthermore, the vehicle sensor 11A includes various warning devices for outputting warnings, various diagnostic devices for outputting diagnostic results, etc.

[0023] The communication unit 11B is configured by, for example, a DCM (Data Communication Module). The communication unit 11B has a function of transmitting vehicle sensor data indicating the detection result of the vehicle sensor 11A to the server device 12 via the network NW. Further, the communication unit 11B has a function of receiving, for example, map information generated by the map information generation device 13 via the network NW. The map information unit 11D is formed, for example, in a storage such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive) mounted on the probe vehicle 11. The map information possessed by the map information unit 11D includes various types of information such as road structures (road positions, road shapes, lane structures, etc.) and rules.

[0024] The control unit 11C has functions such as causing the communication unit 11B to transmit vehicle sensor data to the server device 12. The control unit 11C is configured by a computer including a communication interface (I / F) C1, a memory C2, and a processor C3. The communication interface C1, the memory C2, and the processor C3 are connected via a signal line C4. The communication interface C1 has an interface circuit for connecting the control unit 11C to the in-vehicle network 11E. The memory C2 is an example of a storage unit and has, for example, a volatile semiconductor memory and a non-volatile semiconductor memory. The memory C2 stores programs and various types of data used in processes executed by the processor C3. The processor C3 executes various processes (for example, a process of causing the communication unit 11B to transmit vehicle sensor data to the server device 12). Further, the processor C3 has a function of controlling, for example, a steering actuator (not shown), a braking actuator (not shown), and a driving actuator (not shown) of the probe vehicle 11.

[0025] In the examples shown in FIGS. 1 and 2, the control unit 11C is constituted by, for example, an automatic driving ECU (Electronic Control Unit), and can control the probe vehicle 11 at a driving control level of level 3 according to the definition of SAE (Society of Automotive Engineers), that is, a driving control level that does not require operation of the steering actuator, braking actuator, and driving actuator by the driver and monitoring around the probe vehicle 11. Further, the control unit 11C has a function of controlling the probe vehicle 11 at a driving control level in which the driver participates in the driving of the probe vehicle 11, for example, a driving control level of levels 0 to 2 according to the definition of SAE. In addition, the control unit 11C has a function of outputting, for example, a lane departure warning or the like to an HMI (Human Machine Interface) (not shown). In other examples, the control unit 11C may not have a function of controlling the probe vehicle 11 at a driving control level of level 3 according to the definition of SAE. That is, the probe vehicle 11 may be a manually driven vehicle instead of an autonomous vehicle. In still other examples, the control unit 11C may not have a function of controlling the steering actuator, braking actuator, and driving actuator of the probe vehicle 11. That is, an ECU having a function of controlling the steering actuator, braking actuator, and driving actuator of the probe vehicle 11 may be provided in the probe vehicle 11 separately from the control unit 11C having a function of causing the communication unit 11B to transmit vehicle sensor data to the server device 12.

[0026] In the examples shown in FIGS. 1 and 2, the processor C3 has a function as a collection device C31 and a function as a filtering device C32. The collection device C31 collects vehicle sensor data indicating the detection result of the vehicle sensor 11A from the vehicle sensor 11A. On the other hand, the vehicle sensor data collected from the vehicle sensor 11A by the collection device C31 includes, for example, vehicle sensor data that is not suitable for being utilized in the map information generation device 13 or the like. In view of this point, in the examples shown in FIGS. 1 and 2, the processor C3 has a function as a filtering device C32 that executes a process of excluding vehicle sensor data indicating the detection result of the vehicle sensor 11A. The filtering device C32 includes a digitization processing unit C32A, a determination unit C32B, and a filtering processing unit C32C.

[0027] The digitization processing unit C32A digitizes the vehicle sensor data collected from the vehicle sensor 11A by the collection device C31. Specifically, the digitization processing unit C32A digitizes the vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated a stop. When the probe vehicle 11 is a vehicle that has violated a stop, the digitization processing unit C32A assigns a lower score to the vehicle sensor data than when the probe vehicle 11 is not a vehicle that has violated a stop. In the examples shown in FIGS. 1 and 2, the digitization processing unit C32A recognizes a stop location on the road where the probe vehicle 11 is traveling based on an image in front of the probe vehicle 11 captured by the front camera functioning as the vehicle sensor 11A. Further, when the probe vehicle 11 passes through the stop location, the digitization processing unit C32A determines whether the probe vehicle 11 has violated a stop based on the vehicle speed of the probe vehicle 11 detected by the vehicle speed sensor functioning as the vehicle sensor 11A. The vehicle sensor data to be determined whether to be excluded includes an image including the stop location in front of the probe vehicle 11 captured by the front camera and the vehicle speed of the probe vehicle 11 detected by the vehicle speed sensor. In other examples, the digitization processing unit C32A may determine whether the probe vehicle 11 has violated a stop based on the position information of the probe vehicle 11 obtained by the GPS unit functioning as the vehicle sensor 11A and the map information of the map information unit 11D. In this example, the vehicle sensor data to be determined whether to be excluded includes the position information of the probe vehicle 11 obtained by the GPS unit (including the speed information of the probe vehicle 11).

[0028] Also, in the examples shown in FIGS. 1 and 2, the scoring processing unit C32A performs scoring of vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated the speed limit. When the probe vehicle 11 is a vehicle that has violated the speed limit, the scoring processing unit C32A assigns a lower score to the vehicle sensor data than when the probe vehicle 11 is not a vehicle that has violated the speed limit. In the examples shown in FIGS. 1 and 2, the scoring processing unit C32A recognizes the speed limit of the road on which the probe vehicle 11 is traveling based on an image in front of the probe vehicle 11 captured by the front camera functioning as the vehicle sensor 11A. Further, when the probe vehicle 11 is traveling on a road where the speed limit is set, the scoring processing unit C32A determines whether the probe vehicle 11 has violated the speed limit based on the vehicle speed of the probe vehicle 11 detected by the vehicle speed sensor functioning as the vehicle sensor 11A. The vehicle sensor data to be determined whether to be excluded includes an image including a speed limit sign or indication in front of the probe vehicle 11 captured by the front camera and the vehicle speed of the probe vehicle 11 detected by the vehicle speed sensor. In other examples, the scoring processing unit C32A may recognize the speed limit based on the position information of the probe vehicle 11 obtained by the GPS unit functioning as the vehicle sensor 11A and the map information of the map information unit 11D, and determine whether the probe vehicle 11 has violated the speed limit. In this example, the vehicle sensor data to be determined whether to be excluded includes the position information of the probe vehicle 11 obtained by the GPS unit (including the speed information of the probe vehicle 11).

[0029] Furthermore, in the examples shown in FIGS. 1 and 2, the scoring processing unit C32A performs scoring of vehicle sensor data based on information indicating whether lane departure warnings frequently occur in the probe vehicle 11. When lane departure warnings frequently occur in the probe vehicle 11, the scoring processing unit C32A assigns a lower score to the vehicle sensor data than when lane departure warnings do not frequently occur in the probe vehicle 11. In the examples shown in FIGS. 1 and 2, the scoring processing unit C32A determines whether or not lane departure warnings frequently occur in the probe vehicle 11 based on the output history of the lane departure warnings by the HMI. In other examples, the vehicle sensor 11A includes, for example, an in-vehicle camera that detects a lane departure warning output by the HMI, and the scoring processing unit C32A may determine whether or not lane departure warnings frequently occur in the probe vehicle 11 based on the detection result of the lane departure warning by the in-vehicle camera or the like.

[0030] Also, in the examples shown in FIGS. 1 and 2, the scoring processing unit C32A performs scoring of the vehicle sensor data based on information indicating whether or not the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than a second threshold value. The scoring processing unit C32A assigns a lower score to the vehicle sensor data when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than the second threshold value than when it is not equal to or less than the second threshold value. In the examples shown in FIGS. 1 and 2, the second threshold value is set such that only the preceding vehicle is included in the image of the front of the probe vehicle 11 captured by the front camera when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than the second threshold value. The scoring processing unit C32A determines that the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than the second threshold value when only the preceding vehicle is included in the image of the front of the probe vehicle 11 captured by the front camera, and determines that the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is not equal to or less than the second threshold value when the image of the front of the probe vehicle 11 captured by the front camera includes the preceding vehicle and other objects (for example, scenery, etc.). The vehicle sensor data to be determined whether or not to be excluded includes, for example, an image including the preceding vehicle in front of the probe vehicle 11 captured by the front camera. In another example, the second threshold may be set such that when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than the second threshold, a region of a predetermined ratio or more of the image in front of the probe vehicle 11 captured by the front camera shows the preceding vehicle. In this example, when a region of a predetermined ratio or more of the image in front of the probe vehicle 11 captured by the front camera shows the preceding vehicle, the scoring processing unit C32A determines that the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than the second threshold, and when a region of less than the predetermined ratio of the image in front of the probe vehicle 11 captured by the front camera shows the preceding vehicle, determines that the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is not equal to or less than the second threshold. Also in this example, the vehicle sensor data to be determined whether to be excluded includes an image including the preceding vehicle in front of the probe vehicle 11 captured by the front camera.

[0031] In the examples shown in FIGS. 1 and 2, the determination unit C32B compares the score of the vehicle sensor data given by the scoring processing unit C32A with the first threshold. Specifically, the determination unit C32B determines whether the score of the vehicle sensor data given by the scoring processing unit C32A is equal to or less than the first threshold. The filtering processing unit C32C executes a process of excluding the vehicle sensor data based on the determination result of the determination unit C32B. When the determination unit C32B determines that the score of the vehicle sensor data given by the scoring processing unit C32A is higher than the first threshold, the filtering processing unit C32C does not exclude the vehicle sensor data (specifically, the vehicle sensor data to be determined whether to be excluded as described above). On the other hand, when the determination unit C32B determines that the score of the vehicle sensor data given by the scoring processing unit C32A is equal to or less than the first threshold, the filtering processing unit C32C excludes the vehicle sensor data (the vehicle sensor data to be determined whether to be excluded). The communication unit 11B transmits the vehicle sensor data not excluded by the filtering processing unit C32C to the server device 12 via the network NW.

[0032] That is, in the examples shown in FIGS. 1 and 2, when the probe vehicle 11 is a vehicle that violates a stop sign, the filtering processing unit C32C is more likely to exclude vehicle sensor data than when the probe vehicle 11 is not a vehicle that violates a stop sign. Also, when the probe vehicle 11 is a vehicle that violates the speed limit, the filtering processing unit C32C is more likely to exclude vehicle sensor data than when the probe vehicle 11 is not a vehicle that violates the speed limit. Furthermore, when lane departure warnings frequently occur in the probe vehicle 11, the filtering processing unit C32C is more likely to exclude vehicle sensor data than when lane departure warnings do not frequently occur in the probe vehicle 11. Also, when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is less than or equal to the second threshold value, the filtering processing unit C32C is more likely to exclude vehicle sensor data than when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is not less than or equal to the second threshold value.

[0033] In a modification of the filtering device C32 of the first embodiment, when the probe vehicle 11 is a vehicle that violates a stop sign, the point scoring processing unit C32A assigns a higher score to the vehicle sensor data than when the probe vehicle 11 is not a vehicle that violates a stop sign. Also, when the probe vehicle 11 is a vehicle that violates the speed limit, the point scoring processing unit C32A assigns a higher score to the vehicle sensor data than when the probe vehicle 11 is not a vehicle that violates the speed limit. Furthermore, when lane departure warnings frequently occur in the probe vehicle 11, the point scoring processing unit C32A assigns a higher score to the vehicle sensor data than when lane departure warnings do not frequently occur in the probe vehicle 11. Also, when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is less than or equal to the second threshold value, the point scoring processing unit C32A assigns a higher score to the vehicle sensor data than when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is not less than or equal to the second threshold value. Also, in a modification of the filtering device C32 of the first embodiment, the determination unit C32B determines whether the score of the vehicle sensor data assigned by the scoring processing unit C32A is equal to or greater than a third threshold value. The filtering processing unit C32C does not exclude the vehicle sensor data when the determination unit C32B determines that the score of the vehicle sensor data assigned by the scoring processing unit C32A is lower than the third threshold value. On the other hand, the filtering processing unit C32C excludes the vehicle sensor data when the determination unit C32B determines that the score of the vehicle sensor data assigned by the scoring processing unit C32A is equal to or greater than the third threshold value. Also in a modification of the filtering device C32 of the first embodiment, the communication unit 11B transmits the vehicle sensor data not excluded by the filtering processing unit C32C to the server device 12 via the network NW.

[0034] In the example shown in FIGS. 1 and 2, the determination unit C32B determines whether the attachment position of the vehicle sensor 11A has changed. Specifically, the determination unit C32B determines whether the attachment position of the in-vehicle camera has changed based on a post-attachment position change image, which is an image showing the external situation of the probe vehicle 11 captured by the in-vehicle camera as the vehicle sensor 11A after the attachment position has changed (for example, during driving of the probe vehicle 11 by the driver of the probe vehicle 11), and a pre-attachment position change image, which is an image showing the external situation of the probe vehicle 11 captured by the in-vehicle camera before the attachment position has changed (for example, during inspection before shipment of the probe vehicle 11). The determination unit C32B determines whether the attachment position of the in-vehicle camera has changed by using a known image comparison technique. For example, when a modification is made to lower the vehicle height of the probe vehicle 11, the determination unit C32B determines that the attachment position of the in-vehicle camera has changed. Further, the determination unit C32B determines whether the mounting position of the LiDAR as the vehicle sensor 11A has changed based on the result of object recognition performed by the LiDAR during the calibration of the LiDAR as the vehicle sensor 11A after the mounting position has changed (for example, during the vehicle inspection of the probe vehicle 11) and the result of object recognition performed by the LiDAR during the calibration of the LiDAR before the mounting position has changed (for example, during the pre - shipment inspection of the probe vehicle 11). The determination unit C32B determines whether the mounting position of the LiDAR has changed by using a known point cloud comparison technique. For example, when modifications such as to the fender or tire of the probe vehicle 11 are made, the determination unit C32B determines that the mounting position of the LiDAR has changed.

[0035] When the mounting position of the vehicle sensor 11A has changed, the filtering processing unit C32C executes a process of excluding the post - mounting - position - change vehicle sensor data indicating the detection result of the vehicle sensor 11A after the mounting position has changed. Specifically, when the determination unit C32B determines that the mounting position of the in - vehicle camera has changed, the filtering processing unit C32C executes a process of excluding the post - mounting - position - change image as the vehicle sensor data. In addition, when the determination unit C32B determines that the mounting position of the LiDAR as the vehicle sensor 11A has changed, the filtering processing unit C32C executes a process of excluding the post - mounting - position - change vehicle sensor data, which is the result of object recognition performed by the LiDAR after the determination unit C32B has determined that the mounting position of the LiDAR has changed.

[0036] In the example shown in FIGS. 1 and 2, the server device 12 has a function of collecting vehicle sensor data obtained in the probe vehicle 11 (specifically, vehicle sensor data not excluded by the filtering processing unit C32C). The server device 12 is configured by a computer having, for example, a CPU (Central Processing Unit), a memory, a storage, a NIC (Network Interface Card), etc. The server device 12 includes a communication unit 12A, a storage unit 12B, and a control unit 12C. For example, the NIC functions as the communication unit 12A. The communication unit 12A receives vehicle sensor data transmitted by the communication unit 11B of the probe vehicle 11 via the network NW. Also, the communication unit 12A transmits the vehicle sensor data collected by the server device 12 to the map information generation device 13 via the network NW in response to a request from, for example, the map information generation device 13 or the like.

[0037] For example, a memory such as a RAM (Random Access Memory) and a ROM (Read Only Memory) and a storage such as an HDD and an SSD function as the storage unit 12B. The storage unit 12B stores the vehicle sensor data received by the communication unit 12A (vehicle sensor data not excluded by the filtering processing unit C32C of the probe vehicle 11). For example, the CPU functions as the control unit 12C. The control unit 12C executes processes such as causing the communication unit 12A to receive vehicle sensor data (vehicle sensor data not excluded by the filtering processing unit C32C) transmitted by, for example, the communication unit 11B of the probe vehicle 11, and causing the storage unit 12B to store the vehicle sensor data received by the communication unit 12A. Also, the control unit 12C executes a process of causing the communication unit 12A to transmit vehicle sensor data to the map information generation device 13 via the network NW in response to a request from, for example, the map information generation device 13 or the like.

[0038] The map information generation device 13 has a function of generating map information based on the vehicle sensor data obtained in the probe vehicle 11. For example, the map information generation device 13 generates map information indicating roads that can actually be traveled by using vehicle sensor data (vehicle sensor data not excluded by the filtering processing unit C32C) indicating the roads that the probe vehicle 11 has actually traveled in the most recent 24 hours.

[0039] As described above, in the first example of the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, the vehicle sensor data not excluded by the filtering processing unit C32C is used for generating map information in the map information generation device 13. On the other hand, in the second example of the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, the vehicle sensor data not excluded by the filtering processing unit C32C may be used for generating traffic information in a traffic information generation device (not shown). In the second example of the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, as the probe vehicle 11 travels, the traffic information generation device generates traffic information regarding the road on which the probe vehicle 11 has traveled. By adding the traffic information generated by the traffic information generation device to the traffic information of the main roads by the road traffic information communication system, it becomes possible to provide route guidance based on richer real-time information for a wide range of roads. In addition, by analyzing the traffic information for a certain period, the traffic information generation device can also be expected to be used for solving regional issues such as traffic congestion reduction and tourism promotion.

[0040] In the third example of the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, the vehicle sensor data not excluded by the filtering processing unit C32C may be used for generating information indicating the deterioration state of the road surface in a road surface deterioration state information generation device (not shown). This is because grasping the state of the road not only maintains the convenience of life but also leads to providing safety and peace of mind such as preventing accidents in advance and ensuring evacuation routes in the event of disasters such as earthquakes. In road maintenance management, local governments, as road administrators, need to grasp in advance the roads to be maintained through daily patrols, regular inspections, etc. By applying the technology of analyzing vehicle sensor data (vehicle sensor data not excluded by the filtering processing unit C32C) obtained from the probe vehicle 11, the deterioration state of the road surface can be quantified, and it becomes possible to consider the plan, priority, etc. of the maintenance inspection work. As a result, the costs and labor required for local governments to inspect roads are reduced, and with the progress of road maintenance, repairs, etc., it is expected to lead to a rich mobility society.

[0041] FIG. 3 is a flowchart for explaining an example of the processing executed by the filtering device C32 of the first embodiment. In the example shown in FIG. 3, in step S11, the scoring processing unit C32A scores the vehicle sensor data (vehicle sensor data indicating the detection result of the vehicle sensor 11A) collected from the vehicle sensor 11A by the collection device C31. Specifically, the scoring processing unit C32A scores the vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated a stop rule. In addition, the scoring processing unit C32A scores the vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated the speed limit. Furthermore, the scoring processing unit C32A scores the vehicle sensor data based on information indicating whether lane departure warnings frequently occur in the probe vehicle 11. Also, the scoring processing unit C32A scores the vehicle sensor data based on information indicating whether the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is less than or equal to a second threshold value. In other examples, the scoring processing unit C32A may score the vehicle sensor data based on at least any one of information indicating whether the probe vehicle 11 is a vehicle that has violated a stop rule, information indicating whether the probe vehicle 11 is a vehicle that has violated the speed limit, information indicating whether lane departure warnings frequently occur in the probe vehicle 11, and information indicating whether the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is less than or equal to a second threshold value.

[0042] In the example shown in FIG. 3, in step S12, the determination unit C32B compares the score of the vehicle sensor data attached in step S11 with the first threshold value. Specifically, the determination unit C32B determines whether the score of the vehicle sensor data attached in step S11 is less than or equal to the first threshold value. If the score of the vehicle sensor data is less than or equal to the first threshold value, the process proceeds to step S13. On the other hand, if the score of the vehicle sensor data is not less than or equal to the first threshold value, the routine shown in FIG. 3 ends. In step S13, the filtering processing unit C32C executes a process of excluding the vehicle sensor data.

[0043] In the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied as described above, when the probe vehicle 11 is a vehicle that violates a stop sign, the vehicle sensor data indicating the detection result of the vehicle sensor 11A mounted on the probe vehicle 11 is more likely to be excluded than when the probe vehicle 11 is not a vehicle that violates a stop sign. Also, when the probe vehicle 11 is a vehicle that violates the speed limit, the vehicle sensor data is more likely to be excluded than when the probe vehicle 11 is not a vehicle that violates the speed limit. Further, when lane departure warnings frequently occur in the probe vehicle 11, the vehicle sensor data is more likely to be excluded than when lane departure warnings do not frequently occur in the probe vehicle 11. Also, when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is less than or equal to the second threshold value, the vehicle sensor data is more likely to be excluded than when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is not less than or equal to the second threshold value. That is, in the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, vehicle sensor data that should not be utilized in the map information generation device 13 or the like is excluded by the filtering device C32 according to the characteristics of the driver of the probe vehicle 11. Therefore, in the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, it is possible to generate map information or the like that reflects highly reliable vehicle sensor data (vehicle sensor data not excluded by the filtering processing unit C32C).

[0044] Also, in the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, when the attachment position of the vehicle sensor 11A changes, the vehicle sensor data after the attachment position change, which indicates the detection result of the vehicle sensor 11A after the attachment position change, is excluded. That is, in the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, vehicle sensor data (vehicle sensor data after the attachment position change) that should not be utilized in the map information generation device 13 or the like due to the characteristics of the owner of the probe vehicle 11 or the like is excluded by the filtering device C32. Therefore, in the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, it is possible to generate map information or the like that reflects highly reliable vehicle sensor data (vehicle sensor data that does not include vehicle sensor data after the attachment position change).

[0045] The processing by the determination unit C32B or the like of the above-described filtering device C32 may be executed in real time (that is, at the timing when the vehicle sensor data is collected by the collection device C31), or may be executed every time the probe vehicle 11 travels a certain distance.

[0046] As described above, in the example shown in FIGS. 1 and 2, the map information generation device 13 generates map information using the vehicle sensor data that has not been excluded by the filtering processing unit C32C (that is, the vehicle sensor data after the processing by the determination unit C32B or the like). However, in other examples (for example, an example where highly accurate map information is not required), the map information generation device 13 may generate map information using the vehicle sensor data before the processing by the determination unit C32B or the like.

[0047] In the examples shown in FIGS. 1 and 2, in order to set a first threshold value used for comparison with the number of points of vehicle sensor data, first, a plurality of probe vehicles 11 capable of collecting vehicle sensor data (probe information) are sampled. Next, points are assigned to the vehicle sensor data of each of the sampled plurality of probe vehicles 11. Further, the first threshold value is set by using the deviation or the like of the distribution of the number of points of the plurality of vehicle sensor data. When the point conversion processing unit C32A performs point conversion of the vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated the speed limit, the point conversion processing unit C32A changes the weight according to the value by which the probe vehicle 11 has exceeded the speed limit, and assigns points to the vehicle sensor data. Specifically, the point conversion processing unit C32A makes the points deducted from the vehicle sensor data of the probe vehicle 11 that has exceeded the speed limit by 20 km / h higher than twice the points deducted from the vehicle sensor data of the probe vehicle 11 that has exceeded the speed limit by 10 km / h. In addition, when the point conversion processing unit C32A performs point conversion of the vehicle sensor data based on information indicating whether lane departure warnings frequently occur in the probe vehicle 11, the point conversion processing unit C32A increases the points deducted according to the time (for example, proportionally) when the lane departure warning is output. When the time when the lane departure warning is output is short, since the points deducted from the vehicle sensor data are low, the points of the vehicle sensor data do not fall below the first threshold value, and the filtering processing unit C32C does not exclude the vehicle sensor data.

[0048] Even when the filtering processing unit C32C executes the process of excluding the vehicle sensor data of the probe vehicle 11, after the filtering processing unit C32C executes the process of excluding the vehicle sensor data of the probe vehicle 11, when the state where the points of the vehicle sensor data of the probe vehicle 11 are not below the first threshold value continues for a predetermined time or more, the vehicle sensor data of the probe vehicle 11 is not excluded by the filtering processing unit C32C and is reused for generating map information in the map information generation device 13.

[0049] <Second Embodiment> The vehicle data utilization system 1 to which the filtering device 2C2 of the second embodiment is applied is configured in the same manner as the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment described above is applied, except for the points described later.

[0050] FIG. 4 is a diagram showing an example of functions of the control unit 12C of the server device 12 of the vehicle data utilization system 1 to which the filtering device 2C2 of the second embodiment is applied. As described above, in the first example (the examples shown in FIGS. 1 and 2) of the vehicle data utilization system 1 to which the filtering device C32 of the first embodiment is applied, the filtering device C32 is provided in the probe vehicle 11. On the other hand, in the first example (the examples shown in FIGS. 1 and 4) of the vehicle data utilization system 1 to which the filtering device 2C2 of the second embodiment is applied, the filtering device 2C2 is provided in the server device 12.

[0051] In the examples shown in FIGS. 1 and 4, unlike the examples shown in FIGS. 1 and 2, the processor C3 does not have the functions as a collection device and a filtering device. In the examples shown in FIGS. 1 and 4, instead, the control unit 12C of the server device 12 has the functions as a collection device 2C1 and a filtering device 2C2.

[0052] In the examples shown in FIGS. 1 and 4, the communication unit 11B of the probe vehicle 11 transmits vehicle sensor data indicating the detection result of the vehicle sensor 11A to the server device 12 via the network NW, the communication unit 12A of the server device 12 receives the vehicle sensor data transmitted by the communication unit 11B of the probe vehicle 11, and the collection device 2C1 of the control unit 12C of the server device 12 collects the vehicle sensor data received by the communication unit 12A of the server device 12. On the other hand, the vehicle sensor data collected by the collection device 2C1 also includes vehicle sensor data that is not suitable for being utilized, for example, in the map information generation device 13 or the like. In view of this point, in the examples shown in FIGS. 1 and 4, the control unit 12C of the server device 12 has a function as a filtering device 2C2 that executes a process of excluding vehicle sensor data. The filtering device 2C2 includes a digitization processing unit 2C2A, a determination unit 2C2B, and a filtering processing unit 2C2C.

[0053] The digitization processing unit 2C2A digitizes the vehicle sensor data collected by the collection device 2C1. Specifically, the digitization processing unit 2C2A digitizes the vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated a stop sign. When the probe vehicle 11 is a vehicle that has violated a stop sign, the digitization processing unit 2C2A assigns a lower score to the vehicle sensor data than when the probe vehicle 11 is not a vehicle that has violated a stop sign. In addition, the digitization processing unit 2C2A digitizes the vehicle sensor data based on information indicating whether the probe vehicle 11 is a vehicle that has violated the speed limit. When the probe vehicle 11 is a vehicle that has violated the speed limit, the digitization processing unit 2C2A assigns a lower score to the vehicle sensor data than when the probe vehicle 11 is not a vehicle that has violated the speed limit. Furthermore, the digitization processing unit 2C2A digitizes the vehicle sensor data based on information indicating whether lane departure warnings frequently occur in the probe vehicle 11. When lane departure warnings frequently occur in the probe vehicle 11, the digitization processing unit 2C2A assigns a lower score to the vehicle sensor data than when lane departure warnings do not frequently occur in the probe vehicle 11. In addition, the digitization processing unit 2C2A digitizes the vehicle sensor data based on information indicating whether the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than a second threshold value. When the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is equal to or less than the second threshold value, the digitization processing unit 2C2A assigns a lower score to the vehicle sensor data than when the inter-vehicle distance or TTC between the probe vehicle 11 and the preceding vehicle is not equal to or less than the second threshold value.

[0054] In the examples shown in FIGS. 1 and 4, the determination unit 2C2B compares the score of the vehicle sensor data assigned by the scoring processing unit 2C2A with the first threshold value. Specifically, the determination unit 2C2B determines whether the score of the vehicle sensor data assigned by the scoring processing unit 2C2A is less than or equal to the first threshold value. The filtering processing unit 2C2C executes a process of excluding the vehicle sensor data based on the determination result of the determination unit 2C2B. When the determination unit 2C2B determines that the score of the vehicle sensor data assigned by the scoring processing unit 2C2A is higher than the first threshold value, the filtering processing unit 2C2C does not exclude the vehicle sensor data. On the other hand, when the determination unit 2C2B determines that the score of the vehicle sensor data assigned by the scoring processing unit 2C2A is less than or equal to the first threshold value, the filtering processing unit 2C2C excludes the vehicle sensor data.

[0055] Also, in the examples shown in FIGS. 1 and 4, the determination unit 2C2B determines whether the mounting position of the vehicle sensor 11A has changed. When the mounting position of the vehicle sensor 11A has changed, the filtering processing unit 2C2C executes a process of excluding the vehicle sensor data after the mounting position change indicating the detection result of the vehicle sensor 11A after the mounting position change. Specifically, when the determination unit 2C2B determines that the mounting position of the in-vehicle camera has changed, the filtering processing unit 2C2C executes a process of excluding the image after the mounting position change as the vehicle sensor data. Also, when the determination unit 2C2B determines that the mounting position of the LiDAR as the vehicle sensor 11A has changed, the filtering processing unit 2C2C executes a process of excluding the vehicle sensor data after the mounting position change, which is the result of object recognition performed by the LiDAR after the determination unit 2C2B determines that the mounting position of the LiDAR has changed.

[0056] The storage unit 12B of the server device 12 stores the vehicle sensor data that has not been excluded by the filtering processing unit 2C2C. In addition, the control unit 12C of the server device 12 executes a process of causing the communication unit 12A to transmit vehicle sensor data (vehicle sensor data not excluded by the filtering processing unit 2C2C) to the map information generation device 13 via the network NW in response to a request from, for example, the map information generation device 13 or the like. The map information generation device 13 generates map information by using vehicle sensor data (vehicle sensor data not excluded by the filtering processing unit 2C2C).

[0057] As described above, in the first example of the vehicle data utilization system 1 to which the filtering device 2C2 of the second embodiment is applied, the vehicle sensor data not excluded by the filtering processing unit 2C2C is used for generating map information in the map information generation device 13. On the other hand, in the second example of the vehicle data utilization system 1 to which the filtering device 2C2 of the second embodiment is applied, the vehicle sensor data not excluded by the filtering processing unit 2C2C may be used for generating traffic information in a traffic information generation device (not shown). In the third example of the vehicle data utilization system 1 to which the filtering device 2C2 of the second embodiment is applied, the vehicle sensor data not excluded by the filtering processing unit 2C2C may be used for generating information indicating the deterioration state of the road surface in a road surface deterioration state information generation device (not shown).

[0058] As described above, the embodiments of the filtering device, the filtering method, and the program of the present disclosure have been described with reference to the drawings. However, the filtering device, the filtering method, and the program of the present disclosure are not limited to the above-described embodiments, and can be appropriately modified without departing from the spirit of the present disclosure. The configurations of the respective examples of the above-described embodiments may be combined as appropriate. In each of the above-described embodiments, the processing performed in the filtering devices C32 and 2C2 has been described as software processing performed by executing a program. However, the processing performed in the filtering devices C32 and 2C2 may be processing performed by hardware. Alternatively, the processing performed in the filtering devices C32 and 2C2 may be processing that combines both software and hardware. Further, a program stored in the memory C2 of the control unit 11C (a program that realizes the functions of the processor C3 of the control unit 11C) or a program stored in the storage unit 12B of the server device 12 (a program that realizes the functions of the control unit 12C of the server device 12) may be recorded and provided, distributed, etc. on a computer-readable storage medium such as a semiconductor memory, a magnetic recording medium, an optical recording medium, etc.

Explanation of Signs

[0059] 1 Vehicle data utilization system 11 Probe vehicle 11A Vehicle sensor 11B Communication unit 11C Control unit C1 Communication interface C2 Memory C3 Processor C31 Collection device C32 Filtering device C32A Quantization processing unit C32B Determination unit C32C Filtering processing unit 11D Map information unit 11E In-vehicle network 12 Server device 12A Communication unit 12B Storage unit 12C Control unit 2C1 Collection device 2C2 Filtering device 2C2A Quantization processing unit 2C2B Determination unit 2C2C Filtering processing unit 13 Map Information Generation Device

Claims

1. A filtering device that executes a process of excluding vehicle sensor data indicating detection results of vehicle sensors mounted on a probe vehicle, a digitization processing unit that digitizes the vehicle sensor data, a determination unit that compares the digitized value of the vehicle sensor data with a first threshold value, and a filtering processing unit that executes a process of excluding the vehicle sensor data based on the determination result of the determination unit, wherein the digitization processing unit digitizes the vehicle sensor data based on at least any one of information indicating whether the probe vehicle is a vehicle that has violated a stop prohibition, information indicating whether the probe vehicle is a vehicle that has violated a speed limit, information indicating whether lane departure warnings frequently occur in the probe vehicle, and information indicating whether a distance between the probe vehicle and a preceding vehicle or TTC (Time To Collision) is less than or equal to a second threshold value; the vehicle sensor includes a front camera that captures an image in front of the probe vehicle, the vehicle sensor data includes an image in front of the probe vehicle captured by the front camera, and when the distance between the vehicles or the TTC is less than or equal to the second threshold value, the filtering device is configured such that only the preceding vehicle is included in the image in front of the probe vehicle captured by the front camera, or most of the image in front of the probe vehicle captured by the front camera shows the preceding vehicle.

2. The digitization processing unit digitizes the vehicle sensor data based on information indicating whether the probe vehicle is a vehicle that has violated a stop prohibition, and when the probe vehicle is a vehicle that has violated a stop prohibition, the filtering processing unit is configured to more easily exclude the vehicle sensor data than when the probe vehicle is not a vehicle that has violated a stop prohibition. The filtering device according to claim 1.

3. The point conversion processing unit performs point conversion of the vehicle sensor data based on information indicating whether the probe vehicle is a vehicle that violates the speed limit, The filtering processing unit is more likely to exclude the vehicle sensor data when the probe vehicle is a vehicle that violates the speed limit than when the probe vehicle is not a vehicle that violates the speed limit. The filtering device according to claim 1.

4. The point conversion processing unit performs point conversion of the vehicle sensor data based on information indicating whether the lane departure warning frequently occurs in the probe vehicle, The filtering processing unit is more likely to exclude the vehicle sensor data when the lane departure warning frequently occurs in the probe vehicle than when the lane departure warning does not frequently occur in the probe vehicle. The filtering device according to claim 1.

5. The point conversion processing unit performs point conversion of the vehicle sensor data based on information indicating whether the inter-vehicle distance or the TTC is less than or equal to the second threshold value, The filtering processing unit is more likely to exclude the vehicle sensor data when the inter-vehicle distance or the TTC is less than or equal to the second threshold value than when the inter-vehicle distance or the TTC is not less than or equal to the second threshold value. The filtering device according to claim 1.

6. The determination unit determines whether the mounting position of the vehicle sensor has changed, The filtering processing unit executes a process of excluding the vehicle sensor data after the change in the mounting position, which indicates the detection result of the vehicle sensor after the change in the mounting position, when the mounting position of the vehicle sensor has changed. The filtering device according to claim 1.

7. The vehicle sensor includes an in-vehicle camera that captures an image indicating the external situation of the probe vehicle, The determination unit determines whether the mounting position of the in-vehicle camera has changed based on an image after the mounting position has changed, which is an image showing the external situation of the probe vehicle captured by the in-vehicle camera, and an image before the mounting position has changed, which is an image showing the external situation of the probe vehicle captured by the in-vehicle camera before the mounting position changes. The filtering processing unit executes a process of excluding the image after the mounting position change when the mounting position of the in-vehicle camera has changed. The filtering device according to claim 6.

8. The vehicle sensor includes a LiDAR having a function of recognizing an object outside the probe vehicle. The determination unit determines whether the mounting position of the LiDAR has changed based on the result of object recognition performed by the LiDAR during calibration of the LiDAR after the mounting position has changed and the result of object recognition performed by the LiDAR during calibration of the LiDAR before the mounting position changes. When the determination unit determines that the mounting position of the LiDAR has changed, the filtering processing unit executes a process of excluding the result of object recognition performed by the LiDAR after the determination unit determines that the mounting position of the LiDAR has changed. The filtering device according to claim 6.

9. The vehicle sensor data not excluded by the filtering processing unit is used for at least one of generation of map information, generation of traffic information, and generation of information indicating the deterioration state of the road surface. The filtering device according to claim 1.

10. When only the preceding vehicle is included in the image in front of the probe vehicle captured by the front camera or when a region of a predetermined ratio or more of the image in front of the probe vehicle captured by the front camera indicates the preceding vehicle, the inter-vehicle distance or the TTC between the probe vehicle and the preceding vehicle is set as the second threshold value. The filtering device according to claim 1.

11. A filtering method in which a filtering device executes a process of excluding vehicle sensor data indicating detection results of vehicle sensors mounted on a probe vehicle, a quantization processing step in which the filtering device quantizes the vehicle sensor data, a determination step in which the filtering device compares the number of points of the vehicle sensor data with a first threshold value, and a filtering processing step in which the filtering device executes a process of excluding the vehicle sensor data based on the determination result in the determination step, in the quantization processing step, quantization of the vehicle sensor data is performed based on at least any one of information indicating whether the probe vehicle is a vehicle that has violated a stop sign, information indicating whether the probe vehicle is a vehicle that has violated a speed limit, information indicating whether lane departure warnings frequently occur in the probe vehicle, and information indicating whether the inter-vehicle distance or TTC between the probe vehicle and the preceding vehicle is less than or equal to a second threshold value, the vehicle sensor includes a front camera that captures an image in front of the probe vehicle, the vehicle sensor data includes an image in front of the probe vehicle captured by the front camera, and when the inter-vehicle distance or the TTC is less than or equal to the second threshold value, only the preceding vehicle is included in the image in front of the probe vehicle captured by the front camera, or most of the image in front of the probe vehicle captured by the front camera shows the preceding vehicle. Filtering method.

12. In a computer constituting a computer or server device mounted on a probe vehicle, a quantization processing step of quantizing vehicle sensor data indicating detection results of vehicle sensors mounted on the probe vehicle, a determination step of comparing the number of points of the vehicle sensor data with a first threshold value, A program for executing a filtering process step of excluding the vehicle sensor data based on the determination result in the determination step, In the point conversion process step, Based on at least any one of information indicating whether the probe vehicle is a vehicle that has violated a temporary stop, information indicating whether the probe vehicle is a vehicle that has violated a speed limit, information indicating whether lane departure warnings frequently occur in the probe vehicle, and information indicating whether the inter-vehicle distance or TTC between the probe vehicle and the preceding vehicle is equal to or less than a second threshold value, the vehicle sensor data is converted into points. The vehicle sensor includes a front camera that captures an image in front of the probe vehicle. The vehicle sensor data includes an image in front of the probe vehicle captured by the front camera. A program in which when the inter-vehicle distance or the TTC is equal to or less than the second threshold value, only the preceding vehicle is included in the image in front of the probe vehicle captured by the front camera, or most of the image in front of the probe vehicle captured by the front camera shows the preceding vehicle.

Images