Position accuracy determination device, storage medium, and determination method

Abstract

Disclosed is a position accuracy determination device, a storage medium, and a determination method. Provided is a position accuracy determination device capable of determining the state of accuracy of a presumed current position of a moving object. The position accuracy determination device, based on an image of a road feature on a road surface around a moving object at a first time and position information of the road feature on the road surface, estimates a first position of the moving object at the first time, estimates a second position of the moving object at the first time based on a position of the moving object at a second time before the first time and a movement amount and a change in azimuth angle of the moving object from the second time to the first time, inputs the first position and the second position at the first time to a prediction filter, calculates a current position of the moving object at the first time, and determines the state of accuracy of the current position of the moving object based on a first difference between the current position of the moving object at the first time and a position estimated based on a position of the moving object at a time before the first time and a movement amount and a change in azimuth angle of the moving object from the time before the first time to the first time.

Description

Technical Field

[0001] The present disclosure relates to a position accuracy determination device, a non-transitory storage medium storing a computer program for position accuracy determination, and a position accuracy determination method. Background Art

[0002] The vehicle's automated control system generates a navigation route based on the vehicle's current location, destination, and navigation maps. The automated control system uses the map information to estimate the vehicle's current location and guides the vehicle along the navigation route.

[0003] Ideally, an automatic control system can accurately estimate the vehicle's current position. For example, the automatic control system estimates a first estimated position based on an image representing road features on the road surface surrounding the vehicle and the positional information of the road features on a map. Furthermore, the automatic control system estimates a second estimated position based on the vehicle's position at a previous moment and the amount of movement and azimuth change from the previous moment to the current moment. Furthermore, the automatic control system estimates the vehicle's current position based on the first estimated position and the second estimated position (see, for example, Patent Document 1).

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: Japanese Patent Application Laid-Open No. 2016-224802 Summary of the Invention

[0007] The current position of the vehicle estimated based on the first and second estimated positions may not be correct due to factors such as the state of road features shown in the image, errors in the vehicle's movement or azimuth change, or the vehicle slipping. The automatic control system drives the vehicle based on its current position, so if the current position of the vehicle is incorrect, the vehicle may not be able to drive safely.

[0008] Therefore, an object of the present disclosure is to provide a position accuracy determination device capable of determining the accuracy of the estimated current position of a moving object.

[0009] According to one embodiment, a position accuracy determination device is provided. The position accuracy determination device includes: a first position estimation unit for estimating a first position of a mobile object at a first moment based on an image representing road features on a road surface surrounding the mobile object at a first moment and position information of the road features on the road surface; a second position estimation unit for estimating a second position of the mobile object at the first moment based on the position of the mobile object at a second moment before the first moment and the amount of movement and azimuth change of the mobile object from the second moment to the first moment; a calculation unit for inputting the first and second positions at the first moment into a prediction filter to calculate the current position of the mobile object at the first moment; and a determination unit for determining the accuracy of the current position of the mobile object based on a first difference between the current position of the mobile object at the first moment and the position estimated based on the position of the mobile object at a moment before the first moment and the amount of movement and azimuth change of the mobile object from the previous moment to the first moment.

[0010] In addition, in the position accuracy determination device, it is preferred that the determination unit calculates the second difference between the first position and the second position of the moving object at the first moment, and the determination unit determines the accuracy state of the current position of the moving object based on the first difference and the second difference.

[0011] In addition, in the position accuracy judgment device, it is preferred that the judgment unit uses different criteria when judging that the accuracy of the current position of the moving object has changed from a good state to a poor state and when judging that the accuracy of the current position of the moving object has changed from a poor state to a good state.

[0012] Furthermore, in the position accuracy determination device, it is preferable that the determination unit compares the second difference amount with a predetermined reference threshold value to determine the state of accuracy of the current position of the moving object.

[0013] Furthermore, in the position accuracy determination device, it is preferable that the determination unit notify the driver of a request for driving intervention with respect to the moving object via the notification unit, based on the accuracy of the current position of the moving object.

[0014] Furthermore, in the position accuracy determination device, it is preferable that the determination unit uses the second time as the previous time.

[0015] According to another embodiment, a non-transitory storage medium storing a position accuracy determination program is provided. The position accuracy determination program causes a processor to execute: estimating a first position of a mobile object at a first time based on an image representing a road feature on a road surface around the mobile object at the first time and position information of the road feature on the road surface; estimating a second position of the mobile object at the first time based on a position of the mobile object at a second time before the first time and a movement amount and a change amount of a direction angle of the mobile object from the second time to the first time; inputting the first position and the second position at the first time to a prediction filter to calculate a current position of the mobile object at the first time; and determining a state of accuracy of the current position of the mobile object based on a first difference between the current position of the mobile object at the first time and a position estimated based on the position of the mobile object at a time before the first time and the movement amount and the change amount of the direction angle of the mobile object from the time before the first time to the first time.

[0016] In addition, according to another embodiment, a position accuracy determination method is provided. In the position accuracy determination method, a first position of a mobile object at a first time is estimated based on an image representing a road feature on a road surface around the mobile object at the first time and position information of the road feature on the road surface, a second position of the mobile object at the first time is estimated based on a position of the mobile object at a second time before the first time and a movement amount and a change amount of a direction angle of the mobile object from the second time to the first time, the first position and the second position at the first time are inputted to a prediction filter to calculate a current position of the mobile object at the first time, and a state of accuracy of the current position of the mobile object is determined based on a first difference between the current position of the mobile object at the first time and a position estimated based on the position of the mobile object at a time before the first time and the movement amount and the change amount of the direction angle of the mobile object from the time before the first time to the first time.

[0017] The position accuracy determination device according to the present disclosure can determine a state of accuracy of a current position of a mobile object estimated. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a diagram explaining an outline of an action of the position accuracy determination device of the present embodiment.

[0019] Figure 2 is a diagram explaining an outline of an action of the position accuracy determination device of the present embodiment.

[0020] Figure 3 is a diagram explaining an outline of an action of the position accuracy determination device of the present embodiment.

[0021] Figure 4 This is a diagram illustrating a prediction filter.

[0022] Figure 5 This is a diagram illustrating an example of the first difference amount.

[0023] Figure 6 This is a diagram illustrating an example of the second difference amount.

[0024] Figure 7 This is a diagram illustrating an example of determination processing.

[0025] Figure 8 This is an example of the operation flow related to the determination processing of the position estimation device of this embodiment.

[0026] Figure 9 This is a diagram (part 1) explaining a method of determining a condition for state deterioration and a condition for state improvement.

[0027] Figure 10 This is a diagram (part 2) explaining a method of determining a condition for state deterioration and a condition for state improvement.

[0028] (Explanation of Symbols)

[0029] 1: Vehicle control system; 2: Camera; 3: Positioning information receiver; 4: Navigation device; 5: User interface; 5a: Display device; 6: Vehicle speed sensor; 7: Yaw rate sensor; 10: Vehicle; 11: Map information storage device; 12: Position estimation device; 21: Communication interface; 22: Memory; 23: Processor; 230: Position estimation unit; 231: Calculation unit; 232: Determination unit; 13: Object detection device; 14: Driving lane planning device; 15: Driving planning device; 16: Vehicle control device; 17: In-vehicle network. DETAILED DESCRIPTION

[0030] Figure 1 This is a diagram for explaining the outline of the operation of the position accuracy determination device of this embodiment. Figure 1 , an overview of the operations related to the position accuracy determination processing of the position estimation device 12 as an example of the position accuracy determination device disclosed in this specification is described.

[0031] The vehicle 10 travels while estimating its current position at predetermined intervals using the position estimation device 12. The vehicle 10 is an example of a moving object.

[0032] like Figure 1As shown in (A), the position-speculating device 12 speculates the position of the vehicle 10 at the time t1 based on the image representing the road features on the road surface around the vehicle 10 at the time t1 and the position information of the road features on the road surface (for example, map information). Likewise, the position-speculating device 12 speculates the position P1 of the vehicle 10 at the time t2 after a predetermined period from the time t1 (speculated position using road features).

[0033] In addition, the position-speculating device 12 speculates the position P2 of the vehicle 10 at the time t2 based on the position of the vehicle 10 at the time t1 and the amount of movement and the amount of change in the azimuth angle of the vehicle 10 from the time t1 to the time t2 (speculated position based on dead reckoning).

[0034] Furthermore, the position-speculating device 12 inputs the position P1 and the position P2 of the vehicle 10 at the time t2 to a prediction filter (for example, a Kalman filter) and calculates the current position P0 of the vehicle 10 at the time t2 (speculated position using a prediction filter). This current position P0 is used for controlling the driving of the vehicle 10. In addition, the current position P0 is also used for calculating the next second speculated position (speculated position based on dead reckoning).

[0035] The position-speculating device 12 calculates the amount of divergence D between the speculated position P2 and the current position P0 of the vehicle 10 at the time t2. In addition, in the example shown in (A), as the amount of divergence D, a lateral amount of divergence in the lateral direction orthogonal to the advancing direction of the vehicle 10 is shown. Figure 1

[0036] As the cause of the generation of the amount of divergence D, for example, the following can be cited. (1) abnormality of a camera that takes an image representing the road features on the road surface around the vehicle 10, (2) unclearness of the road features on the road surface around the vehicle 10 (for example, intermittent lane dividers), (3) inconsistency between a high-precision map representing the position information of the road features on the road surface and the current terrain, (4) abnormality of a speed sensor used for calculating the amount of movement of the vehicle 10, (5) abnormality of a yaw rate sensor used for calculating the amount of change in the azimuth angle of the vehicle 10, (6) sideways sliding of the vehicle 10, and the like.

[0037] The position-speculating device 12 determines the state of the precision of the current position of the vehicle 10 based on the amount of divergence D. In Figure 1 ​In the example shown in (B), the accuracy of the current position of vehicle 10 is determined as Normal 1 (Normal), Normal 2 (Hands-off), Abnormal 1 (Hands-on), and Abnormal 2 (Transition Demand: TD). The accuracy of the current position of vehicle 10 is determined as Normal 1 (Normal), with the accuracy decreasing in order of Normal 2 (Hands-off), Abnormal 1 (Hands-on), and Abnormal 2 (TD).

[0038] When the accuracy of the current position of the vehicle 10 is determined to be normal 1 or normal 2, the driving of the vehicle 10 is safely and automatically controlled based on the current position of the vehicle 10 estimated by the position estimation device 12 .

[0039] On the other hand, when the accuracy of the current position of the vehicle 10 is determined to be abnormal 1 or abnormal 2, there is a possibility that the driving of the vehicle 10 cannot be safely automatically controlled based on the current position of the vehicle 10 estimated by the position estimation device 12.

[0040] Therefore, when the position estimation device 12 determines that the accuracy of the current position of the vehicle 10 is abnormal 1 (Hands-on), it notifies the driver of a grip request to grip the steering wheel. Furthermore, when the position estimation device 12 determines that the accuracy of the current position of the vehicle 10 is abnormal 2 (TD), it notifies the driver of a control change request to change the driving of the vehicle 10 from automatic control to manual control.

[0041] For example, if a lane marking, an example of a road feature, is not correctly displayed on the image due to a camera malfunction, the accuracy of the vehicle 10's current position may be determined to be abnormal 1 or 2. Furthermore, if the lane marking is intermittent and cannot be correctly recognized from the image, the accuracy of the vehicle 10's current position may also be determined to be abnormal 1 or 2. Therefore, the accuracy of the vehicle 10's current position is affected by factors other than hardware, including sensors.

[0042] In this way, the position estimation device 12 can determine the accuracy of the estimated current position of the vehicle 10 , and thus the determination result can serve as information for safely driving the vehicle 10 .

[0043] Figure 2This is a schematic diagram of a vehicle equipped with a vehicle control system 1 including a position estimation device 12 according to this embodiment. The vehicle control system 1 includes a camera 2, a positioning information receiver 3, a navigation device 4, a user interface (UI) 5, a vehicle speed sensor 6, a yaw rate sensor 7, a map information storage device 11, a position estimation device 12, an object detection device 13, a lane planning device 14, a driving planning device 15, and a vehicle control device 16. Furthermore, the vehicle control system 1 may include a distance measuring sensor (not shown) such as a LiDAR sensor for measuring the distance to objects around the vehicle 10.

[0044] The camera 2, the positioning information receiver 3, the navigation device 4, the UI 5, the vehicle speed sensor 6, the yaw rate sensor 7, the map information storage device 11, the position estimation device 12, the object detection device 13, the driving lane planning device 14, the driving planning device 15, and the vehicle control device 16 are communicatively connected via an in-vehicle network 17 that complies with a standard such as a controller area network.

[0045] Camera 2 is an example of an imaging unit installed in vehicle 10. Camera 2 is mounted on vehicle 10 so as to face forward. Camera 2 captures a camera image representing the environment of a predetermined area in front of vehicle 10 at a predetermined interval. The camera image can depict the road within the predetermined area in front of vehicle 10 and road features such as lane markings on the road surface. Camera 2 includes a two-dimensional detector composed of an array of photoelectric conversion elements sensitive to visible light, such as CCDs or C-MOSs, and an imaging optical system that forms an image of the area to be captured on the two-dimensional detector.

[0046] Each time the camera 2 captures a camera image, it outputs the camera image and the time at which the camera image was captured to the position estimation device 12 and the object detection device 13 via the in-vehicle network 17. The camera image is used by the position estimation device 12 to estimate the position of the vehicle 10. Furthermore, the camera image is used by the object detection device 13 to detect other objects around the vehicle 10.

[0047] The positioning information receiver 3 outputs positioning information indicating the current position of the vehicle 10. For example, the positioning information receiver 3 may be a GNSS receiver. Whenever the positioning information receiver 3 acquires positioning information at a predetermined reception cycle, it outputs the positioning information and the time at which the positioning information was acquired to the navigation device 4 and the map information storage device 11.

[0048] The navigation device 4 generates a navigation route from the current position of the vehicle 10 to the destination location in response to a driver's request based on navigation map information, the destination location of the vehicle 10 input from the UI 5, and positioning information indicating the current location of the vehicle 10 input from the positioning information receiver 3. The navigation route includes information regarding locations such as right turns, left turns, merges, and divergences. The navigation device 4 generates a new navigation route for the vehicle 10 when a new destination location is set or when the current location of the vehicle 10 deviates from the navigation route. Each time the navigation device 4 generates a navigation route, it outputs the route to the position estimation device 12 and the driving lane planning device 14 via the in-vehicle network 17.

[0049] UI5 is an example of a notification unit. UI5 is controlled by the navigation device 4 and the position estimation device 12, and notifies the driver of vehicle 10 driving information, grip requests, or control change requests. Vehicle 10 driving information includes the vehicle's current location, navigation route, and other information related to the vehicle's current and future paths. A grip request requires the driver to grip the steering wheel. A control change request requires the driver to change the driving of vehicle 10 from automatic control to manual control. In addition, UI5 generates operation signals corresponding to the driver's operation on vehicle 10. In order to display driving information, UI5 has a display device 5a such as a liquid crystal display or a touch panel. In addition, UI5 may also have an audio output device (not shown) for notifying the driver of driving information. In addition, UI5 has an input device such as a touch panel or operation buttons as an input device for inputting operation information from the driver to vehicle 10. Examples of operation information include the destination, route, vehicle speed, and other vehicle 10 control information. UI5 outputs the input operation information to the navigation device 4 and the vehicle control device 16 via the in-vehicle network 17.

[0050] The vehicle speed sensor 6 detects vehicle speed information of the vehicle 10 and outputs the vehicle speed information and the time at which the vehicle speed information was acquired to the position estimation device 12, etc., via the in-vehicle network 17. The vehicle speed sensor 6 is attached to, for example, an axle (not shown), detects the rotational speed of the axle, and outputs a pulse signal proportional to the rotational speed.

[0051] The yaw rate sensor 7 detects the yaw rate of the vehicle 10 and outputs the yaw rate and the yaw rate information acquisition time to the position estimation device 12 via the in-vehicle network 17. As the yaw rate sensor 7, an acceleration sensor such as a gyroscope can be used, for example.

[0052] The map information storage device 11 stores wide-area map information covering a relatively wide area (e.g., a 10- to 30-km square) that includes the current location of the vehicle 10. This map information includes high-precision map information, including three-dimensional road surface information, information indicating the types and locations of road features such as lane markings and structures, and legal speed limits. Based on the current location of the vehicle 10, the map information storage device 11 receives wide-area map information from an external server via a base station via wireless communication via a wireless communication device (not shown) mounted on the vehicle 10, and stores the information in the storage device. Whenever positioning information is input from the positioning information receiver 3, the map information storage device 11 refers to the stored wide-area map information and outputs map information for a relatively narrow area (e.g., a 100- to 10-km square) that includes the current location indicated by the positioning information to the position estimation device 12, the object detection device 13, the lane planning device 14, the driving planning device 15, and the vehicle control device 16 via the in-vehicle network 17.

[0053] The position estimation device 12 performs position estimation processing, calculation processing, and determination processing. To this end, the position estimation device 12 includes a communication interface (IF) 21, a memory 22, and a processor 23. The communication interface 21, the memory 22, and the processor 23 are connected via a signal line 24. The communication interface 21 includes an interface circuit for connecting the position estimation device 12 to the in-vehicle network 17.

[0054] All or part of the functions of the position estimation device 12 are, for example, functional modules implemented by a computer program running on the processor 23. The processor 23 has a position estimation unit 230, a calculation unit 231, and a determination unit 232. Alternatively, the functional modules of the processor 23 may be dedicated arithmetic circuits provided in the processor 23. The processor 23 has one or more CPUs (Central Processing Units) and their peripheral circuits. The processor 23 may also have other arithmetic circuits such as a logical operation unit, a numerical operation unit, or a graphics processing unit. The memory 22 is an example of a storage unit, such as a volatile semiconductor memory and a non-volatile semiconductor memory. In addition, the memory 22 stores computer programs and various data of applications used in the information processing performed by the processor 23.

[0055] Each time the position estimation device 12 determines the current position of the vehicle 10 at the time the camera image is captured, it outputs the current position to the object detection device 13, the lane planning device 14, the driving planning device 15, and the vehicle control device 16. The operation of the position estimation device 12 will be described in detail later.

[0056] The object detection device 13 detects other objects and their types (e.g., vehicles) around the vehicle 10 based on camera images, etc. These other objects include other vehicles traveling around the vehicle 10. The object detection device 13 tracks the detected other objects and determines their trajectories. The object detection device 13 determines the lanes in which the other objects are traveling based on lane markings and the positions of other objects indicated in map information. The object detection device 13 outputs object detection information, including information indicating the types and positions of the detected other objects, as well as information indicating the lanes in which the objects are traveling, to the lane planning device 14 and the driving planning device 15, etc.

[0057] At a predetermined periodic driving lane plan generation time, the driving lane planning device 14 selects a lane within the road where the vehicle 10 is traveling based on map information, the navigation route, surrounding environment information, and the current location of the vehicle 10 within the most recent driving section (e.g., 10 km) selected from the navigation route, and generates a driving lane plan indicating the planned driving lane for the vehicle 10. For example, the driving lane planning device 14 generates the driving lane plan so that the vehicle 10 travels in a lane other than a passing lane. Each time the driving lane plan is generated, the driving lane planning device 14 outputs the generated driving lane plan to the driving planning device 15.

[0058] Furthermore, the driving lane planning device 14 determines whether a lane change is necessary in the nearest driving section selected from the navigation route based on the driving lane plan, map information, the navigation route, and the current position of the vehicle 10, and generates a lane change plan based on the determination result. Specifically, the driving lane planning device 14 determines whether a lane change is necessary to move to the lane leading to the vehicle 10's destination based on the navigation route and the current position of the vehicle 10. It also determines whether the vehicle 10 has entered another road at a merging point (merging) from the current driving road and whether the vehicle 10 has exited another road at a diverging point (branching) from the driving road. During merging and branching, the vehicle moves from the driving road's lane to the lane of another road, so a lane change is performed. The driving lane planning device 14 may also utilize surrounding environment information or vehicle status information to determine whether a lane change is necessary. Surrounding environment information includes the positions and speeds of other vehicles traveling around the vehicle 10. Vehicle status information includes the vehicle 10's current position, vehicle speed, acceleration, and direction of travel. When the lane change plan is generated, the driving lane planning device 14 outputs the driving lane plan to which the lane change plan is added to the driving planning device 15 .

[0059] The driving planning device 15 executes a driving planning process at a driving plan generation time set at a predetermined period. The process generates a driving plan representing the planned driving trajectory of the vehicle 10 for a predetermined time in the future (e.g., 5 seconds) based on the driving lane plan, map information, the current position of the vehicle 10, surrounding environment information, and vehicle status information. The driving plan is represented as a set of target positions of the vehicle 10 and target vehicle speeds at those target positions at each time from the current time to the predetermined time in the future. The driving plan generation cycle is preferably shorter than the driving lane plan generation cycle. The driving planning device 15 generates the driving plan so that a gap of at least a predetermined distance can be maintained between the vehicle 10 and other vehicles. Even if the driving lane plan includes lane changes, in which the vehicle 10 moves between lanes, if a gap of at least a predetermined distance cannot be maintained between the vehicle 10 and other vehicles, the driving planning device 15 generates the driving plan so that the vehicle 10 stops. Each time the driving plan is generated, the driving planning device 15 outputs the generated driving plan to the vehicle control device 16.

[0060] The vehicle control device 16 controls various components of the vehicle 10 based on the vehicle 10's current position, speed, and yaw rate, as well as the driving plan generated by the driving planning device 15. For example, the vehicle control device 16 calculates the vehicle 10's steering angle, acceleration, and angular acceleration based on the driving plan, the vehicle 10's speed, and yaw rate, and sets the steering amount, accelerator pedal position, or braking amount to achieve the desired steering angle, acceleration, and angular acceleration. Furthermore, the vehicle control device 16 outputs a control signal corresponding to the set steering amount via the in-vehicle network 17 to an actuator (not shown) that controls the vehicle 10's steering wheels. Furthermore, the vehicle control device 16 calculates the fuel injection amount based on the set accelerator pedal position, and outputs a control signal corresponding to the fuel injection amount via the in-vehicle network 17 to a drive device (not shown) such as the vehicle 10's engine. Alternatively, the vehicle control device 16 outputs a control signal corresponding to the set braking amount via the in-vehicle network 17 to the vehicle 10's brakes (not shown).

[0061] The vehicle control device 16 has an automatic control driving mode in which the automatic control system 1 drives the vehicle 10 and a manual control driving mode in which a driver drives the vehicle 10. The vehicle control device 16 is capable of automatically controlling all of the actions of the vehicle 10, including driving, braking, and steering, when the automatic control driving mode is applied. The vehicle control device 16 in the automatic control driving mode changes the driving of the vehicle 10 from the automatic control driving mode to the manual control driving mode in the case where a control change notification is given by the driver to agree to a request to change from automatic control to manual control. Thus, in the case where the vehicle 10 cannot be safely driven by automatic control, the driver is able to drive the vehicle 10 by manually controlling the vehicle 10 using a steering wheel, an accelerator pedal, a brake pedal (not shown), and the like. In the manual control driving mode, the vehicle 10 manually controls at least one of the actions of driving, braking, and steering. The vehicle control device 16 generates a control signal in accordance with the amount of operation of the steering wheel, the accelerator pedal, and the brake pedal. Furthermore, it is also possible to change from automatic control to manual control in accordance with a request by the driver.

[0062] In Figure 2 , the map information storage device 11, the position estimation device 12, the object detection device 13, the travel lane planning device 14, the driving plan device 15, and the vehicle control device 16 are described as different devices (for example, Electronic Control Units: ECUs), but all or a part of these devices can be configured as one device.

[0063] Figure 3 is one example of an action flow related to the position accuracy determination processing of the position estimation device 12 of the present embodiment. Referring to Figure 3 , the position accuracy determination processing of the position estimation device 12 will be described below. The position estimation device 12 performs the position accuracy determination processing in accordance with the action flow shown in Figure 3 each time a camera image is input.

[0064] First, the position estimation section 230 estimates a first estimated position of the vehicle 10 at the time of the present camera image capturing based on the camera image representing the road features on the road surface around the vehicle 10 at the time of the present camera image capturing and the position information of the road features on the road surface (step S101). The position estimation section 230 is one example of a first position estimation section. For example, the position estimation section 230 compares the travel lane division lines identified within the camera image and the travel lane division lines represented in the map information input from the map information storage device 11 to find the first estimated position and the first estimated azimuth angle of the vehicle 10 at the time of the camera image capturing. The travel lane division lines are one example of road features.

[0065] Next, the position estimation unit 230 estimates the second estimated position (position estimated by dead reckoning) and second estimated heading angle of the vehicle 10 at the time of the current camera image capture based on the position of the mobile object at the time of the previous camera image capture, the amount of movement of the vehicle 10 from the previous camera image capture time to the current camera image capture time, and the amount of change in heading angle of the vehicle 10 from the previous camera image capture time to the current camera image capture time (step S102). The position estimation unit 230 is an example of a second position estimation unit. The position estimation unit 230 integrates the vehicle speed obtained from the vehicle speed information of the vehicle 10 to determine the amount of movement of the vehicle 10 from the previous camera image capture time to the current camera image capture time. Furthermore, the position estimation unit 230 integrates the yaw rate obtained from the yaw rate information of the vehicle 10 to determine the amount of change in heading angle of the vehicle 10 from the previous camera image capture time to the current camera image capture time. The position estimation unit 230 obtains a second estimated position and a second estimated azimuth estimated by dead reckoning using the position and azimuth of the vehicle 10 at the last camera image capturing time, the movement amount, and the azimuth change amount.

[0066] Next, the calculation unit 231 Figure 4 As shown, the first estimated position and the first estimated azimuth and the second estimated position and the second estimated azimuth are input to the prediction filter, and the current position and the current azimuth of the vehicle 10 at the time of this camera image shooting are calculated (step S103). Figure 4 : is a diagram illustrating a prediction filter. As the prediction filter, a Kalman filter, for example, can be used. In addition, a filter other than the Kalman filter can also be used as the prediction filter.

[0067] Furthermore, the position estimation unit 230 may estimate the lane on the road where the vehicle 10 is located based on the lane markings indicated in the map information and the current position and current heading of the vehicle 10. Whenever the position estimation unit 230 determines the current position, current heading, and lane of the vehicle 10 at the time the camera image is captured, it outputs this information to the object detection device 13, the lane planning device 14, the driving planning device 15, the vehicle control device 16, and the like.

[0068] Next, the position estimation unit 230 estimates a third estimated position of the vehicle 10 at the current camera image capturing time based on the position and azimuth of the vehicle 10 at the camera image capturing time before the current camera image capturing time, and the movement amount and azimuth change of the vehicle 10 from the previous camera image capturing time to the current camera image capturing time (step S104). The previous camera image capturing time may be used as the camera image capturing time before the current camera image capturing time. In this case, the third estimated position is the same as the second estimated position, so step S104 is omitted. Alternatively, the camera image capturing time before the current camera image capturing time, such as the previous camera image capturing time, may be used.

[0069] Next, the determination unit 232 calculates the first difference between the first estimated position and the second estimated position of the vehicle 10 at the time of the current camera image capture (step S105). The first difference represents the distance between the first estimated position and the second estimated position. The first difference can also be divided into a longitudinal difference in the direction consistent with the direction of travel of the vehicle 10 and a transverse difference in the direction perpendicular to the direction of travel of the vehicle 10. In addition, the first difference can also be the distance between two positions on the map. In addition, the first difference is an example of the second difference in the claims.

[0070] Figure 5 This figure illustrates an example of a first difference. Vehicle 10 moves from its position at the previous camera image capturing time t1 to a first estimated position P1 at the current camera image capturing time t2. Determination unit 232 calculates the lateral difference between the lateral positions at first estimated position P1 and second estimated position P2 at the current camera image capturing time t2 as first difference D1. If the first difference is large, possible causes include a malfunction in camera 2, a mismatch between map information and the current terrain, or a malfunction in vehicle speed sensor 6 and / or yaw rate sensor 7.

[0071] A first difference D1 between a first estimated position P1 of the vehicle 10 estimated based on an image representing a road feature and position information of the road feature and a second estimated position P2 of the vehicle 10 estimated by dead reckoning indicates the relative validity of the second estimated position P2 with respect to the first estimated position P1.

[0072] Next, the determination unit 232 calculates the second difference between the third estimated position of the vehicle 10 at the time of this camera image capture and the current position (step S106). The second difference represents the distance between the third estimated position and the current position. The second difference can also be divided into a longitudinal difference in a direction consistent with the direction of travel of the vehicle 10 and a transverse difference in a direction orthogonal to the direction of travel of the vehicle 10. In addition, the second difference can also be the distance between two positions on the map. When the second difference is large and the first difference is small, the cause is considered to be a sideways slip of the vehicle 10. In addition, the second difference is an example of the first difference in the claims.

[0073] Figure 6 : is a diagram illustrating an example of the second difference amount. The vehicle 10 moves from the position at the previous camera image capturing time t3 to the third estimated position P3 at the current camera image capturing time t2. The determination unit 232 determines the difference between the lateral position at the current position P0 of the vehicle 10 and the lateral position at the third estimated position P3 at the current camera image capturing time t2 as the second difference amount D2. Figure 6 In the example shown, when the vehicle 10 is traveling on a curved road, the vehicle 10 may slide sideways depending on the state of the road surface.

[0074] The second difference D2 between the third estimated position P3 of the vehicle 10 estimated by dead reckoning and the current position P0 of the vehicle 10 calculated using the prediction filter indicates the validity of the current position P0 of the vehicle 10 with respect to the third estimated position P3.

[0075] Next, the determination unit 232 determines the accuracy of the current position of the vehicle 10 based on the first difference amount and / or the second difference amount, and ends the series of processes (step S107). Next, the determination process performed by the determination unit 232 will be described in detail.

[0076] Figure 7 FIG is a diagram illustrating an example of determination processing. Figure 7 As shown, the states of the accuracy of the current position of the vehicle 10 are determined as normal 1 (normal), normal 2 (Hands-off), abnormal 1 (Hands-on), and abnormal 2 (Transition Demand: TD).

[0077] When the accuracy of the current position of the vehicle 10 is determined to be normal 1 (normal), the driving of the vehicle 10 is safely and automatically controlled based on the current position of the vehicle 10 estimated by the position estimation device 12 .

[0078] When the accuracy of the current position of the vehicle 10 is determined to be normal 2 (Hands-off), the accuracy of the current position of the vehicle 10, although lower than normal 1 (normal), is within the permissible range of control of the automatic control system 1. Therefore, the driving of the vehicle 10 is safely and automatically controlled based on the current position of the vehicle 10 estimated by the position estimation device 12.

[0079] If the accuracy of the current position of vehicle 10 is determined to be abnormal 1 (Hands-on), there is a possibility that the accuracy of the current position of vehicle 10 exceeds the permissible range of control of automatic control system 1. Therefore, if the determination unit 232 determines that the accuracy of the current position of vehicle 10 is abnormal 1 (Hands-on), it notifies the driver via UI 5 of a request for driving intervention for vehicle 10. Specifically, the determination unit 232 notifies the driver of a grip request to grip the steering wheel. Automatic control system 1 automatically controls the driving of vehicle 10 while the driver grips the steering wheel. If the driver determines that driving vehicle 10 under automatic control is unsafe, the driver can operate the steering wheel.

[0080] If the accuracy of the current position of vehicle 10 is determined to be abnormal 2 (TD), it is estimated that the accuracy of the current position of vehicle 10 exceeds the permissible range of control of automatic control system 1. If determination unit 232 determines that the accuracy of the current position of vehicle 10 is abnormal 2 (TD), it notifies the driver via UI 5 of a request for further intervention in the driving of vehicle 10. Specifically, determination unit 232 notifies the driver via UI 5 of a control change request to change the driving of vehicle 10 from automatic control to manual control.

[0081] Figure 8 This is an example of an operation flow related to the determination process of the position estimation device 12 of this embodiment. Figure 8 According to the action flow shown, the determination process is executed.

[0082] First, the following describes a determination process when the current state of the accuracy of the current position of the vehicle 10 is normal 1 (normal).

[0083] First, the determination unit 232 determines whether the first difference quantity satisfies the state deterioration condition (step S201). The state deterioration condition is a condition for determining whether the state of the accuracy of the current position of the vehicle 10 has deteriorated from the current state. As the state deterioration condition, it can be cited that the ratio of the first difference quantity exceeding the state deterioration threshold value among the multiple first difference quantities obtained between the current camera shooting time and the time before a predetermined time (for example, 1 second) is greater than the reference value. For example, the condition th1 for changing from normal 1 (normal) to normal 2 (Hands-off) can be that the state deterioration threshold value is set to 0.1m and the reference value is set to 0.5. In addition, the condition th2 for changing from normal 1 (normal) to abnormal 1 (Hands-on) can be that the state deterioration threshold value is set to 0.3m and the reference value is set to 0.6. In addition, the condition th3 for changing from normal 1 (normal) to abnormal 2 (TD) can be that the state deterioration threshold value is set to 0.5m and the reference value is set to 0.8.

[0084] In this specification, determining that condition 1 is satisfied means that the plurality of first difference amounts obtained between the current camera shooting time and the time before a predetermined time (e.g., 1 second) satisfy condition 1 but do not satisfy condition 2. In this specification, determining that condition 2 is satisfied means that the plurality of first difference amounts obtained between the current camera shooting time and the time before a predetermined time (e.g., 1 second) satisfy condition 2 but do not satisfy condition 3. This also applies to other determination processes.

[0085] If the first difference quantity satisfies the state deterioration condition (step S201 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processing ends. If the first difference quantity satisfies the condition th1, the state of the accuracy of the current position of the vehicle 10 is changed from normal 1 (normal) to normal 2 (hands-off). If the first difference quantity satisfies the condition th2, the state of the accuracy of the current position of the vehicle 10 is changed from normal 1 (normal) to abnormal 1 (hands-on). If the first difference quantity satisfies the condition th3, the state of the accuracy of the current position of the vehicle 10 is changed from normal 1 (normal) to abnormal 2 (TD).

[0086] On the other hand, if the first difference does not satisfy the state deterioration condition (step S201 - "No"), the determination unit 232 determines whether the second difference satisfies the state deterioration condition (step S202). The state deterioration condition can be exemplified by at least one of the multiple second difference quantities calculated between the current camera capture time and a predetermined time (e.g., 1 second) prior, exceeding a state deterioration threshold. For example, condition th1 for changing from Normal 1 (Normal) to Normal 2 (Hands-off) can be setting the state deterioration threshold to 0.2 m. Condition th2 for changing from Normal 1 (Normal) to Disturbance 1 (Hands-on) can be setting the state deterioration threshold to 0.4 m. Condition th3 for changing from Normal 1 (Normal) to Disturbance 2 (TD) can be setting the state deterioration threshold to 0.6 m. Furthermore, if the curvature of the road on which the vehicle 10 is traveling is determined to be greater than a reference curvature, the state deterioration threshold can be increased compared to when the curvature is below the reference curvature. This is because the change in position tends to be greater when the road is curved. Furthermore, regarding the second difference amount, the determination of the change from normal 1 (normal) to normal 2 (Hands-off) may not be performed.

[0087] If the second difference quantity satisfies the state deterioration condition (step S202 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processing ends. If the second difference quantity satisfies the condition th1, the state of the accuracy of the current position of the vehicle 10 is changed from normal 1 (normal) to normal 2 (hands-off). If the second difference quantity satisfies the condition th2, the state of the accuracy of the current position of the vehicle 10 is changed from normal 1 (normal) to abnormal 1 (hands-on). If the second difference quantity satisfies the condition th3, the state of the accuracy of the current position of the vehicle 10 is changed from normal 1 (normal) to abnormal 2 (TD).

[0088] On the other hand, if the second difference amount does not satisfy the state deterioration condition (step S202 - "No"), the determination unit 232 determines that the state of the accuracy of the current position of the vehicle 10 has not changed (step S205), and the series of processes ends. If the current state of the accuracy of the current position of the vehicle 10 is normal 1 (normal), the processes of steps S203 and S204 are not performed.

[0089] Next, a determination process when the current state of the accuracy of the current position of the vehicle 10 is normal 2 (Hands-off) will be described below.

[0090] First, the determination unit 232 determines whether the first difference quantity satisfies a state deterioration condition (step S201). The state deterioration condition can be exemplified by the ratio of the first difference quantities exceeding the state deterioration threshold value among a plurality of first difference quantities calculated between the current camera capture time and a predetermined time (e.g., 1 second ago) being greater than or equal to a reference value. For example, condition th2 for changing from Hands-off to Hands-on can be a state deterioration threshold value of 0.3 m and a reference value of 0.6. Alternatively, condition th3 for changing from Hands-off to Hands-on can be a state deterioration threshold value of 0.5 m and a reference value of 0.8.

[0091] If the first difference value satisfies the state deterioration condition (step S201 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the first difference value satisfies the condition th2, the state of the accuracy of the current position of the vehicle 10 changes from normal 2 (hands-off) to abnormal 1 (hands-on). If the first difference value satisfies the condition th3, the state of the accuracy of the current position of the vehicle 10 changes from normal 2 (hands-off) to abnormal 2 (TD).

[0092] On the other hand, when the first difference does not satisfy the state deterioration condition (step S201-"No"), the determination unit 232 determines whether the second difference satisfies the state deterioration condition (step S202). As the state deterioration condition, it can be cited that at least one of the plurality of second difference values ​​obtained between the current camera shooting time and the time before a predetermined time (for example, 1 second) is above the state deterioration threshold. For example, the condition th2 for changing from normal 2 (Hands-off) to abnormal 1 (Hands-on) may be that the state deterioration threshold is set to 0.4m. In addition, the condition th3 for changing from normal 2 (Hands-off) to normal 2 (Hands-off) may be that the state deterioration threshold is set to 0.6m. When the curvature of the road on which the vehicle 10 is traveling is determined to be greater than the reference curvature, the state deterioration threshold may be increased compared to the case where the curvature is below the reference curvature.

[0093] If the second difference value satisfies the state deterioration condition (step S202 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the second difference value satisfies the condition th2, the state of the accuracy of the current position of the vehicle 10 changes from normal 2 (hands-off) to abnormal 1 (hands-on). If the second difference value satisfies the condition th3, the state of the accuracy of the current position of the vehicle 10 changes from normal 2 (hands-off) to abnormal 2 (TD).

[0094] On the other hand, when the second difference does not satisfy the state deterioration condition (step S202-"No"), the determination unit 232 determines whether the first difference satisfies the state improvement condition (step S203). The state improvement condition is a condition for determining whether the state of the accuracy of the current position of the vehicle 10 is better than the current state. As a state improvement condition, it can be cited that the ratio of the first difference below the state improvement threshold among the multiple first differences obtained between the time of the camera shooting this time and the time before the predetermined time (for example, 1 second) is above the reference value. For example, the condition th4 for changing from normal 2 (Hands-off) to normal 1 (normal) can be to set the state improvement threshold to 0.1m and the reference value to 0.8.

[0095] If the first difference value satisfies the state improvement condition (step S203 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the first difference value satisfies the condition th4, the state of the accuracy of the current position of the vehicle 10 is changed from normal 2 (hands-off) to normal 1 (normal).

[0096] On the other hand, if the first difference does not satisfy the state improvement condition (step S203 - "No"), the determination unit 232 determines whether the second difference satisfies the state improvement condition (step S204). The state improvement condition can be exemplified by the requirement that at least one of the plurality of second difference values ​​calculated between the current camera capture time and a predetermined time (e.g., 1 second ago) is below a state improvement threshold. For example, condition th4 for changing from Normal 2 (Hands-off) to Normal 1 (Normal) can be setting the state improvement threshold to 0.2 m.

[0097] In a case where the second difference amount satisfies the state improvement condition (step S204 - "Yes"), the determination section 232 changes the state of the precision of the current position of the vehicle 10 (step S206), and ends the series of processing. In a case where the second difference amount satisfies the condition th4, the state of the precision of the current position of the vehicle 10 is changed from Normal 2 (Hands-off) to Normal 1 (Normal).

[0098] On the other hand, in a case where the second difference amount does not satisfy the state improvement condition (step S203 - "No"), the determination section 232 determines that the state of the precision of the current position of the vehicle 10 is not changed (step S205), and ends the series of processing.

[0099] Next, the determination processing in a case where the current state of the precision of the current position of the vehicle 10 is Abnormal 1 (Hands-on) will be described below.

[0100] First, the determination section 232 determines whether the first difference amount satisfies the state deterioration condition (step S201). As the state deterioration condition, it can be cited that the ratio of the first difference amount that exceeds the state deterioration threshold value among a plurality of first difference amounts calculated between the present camera shooting time and the time a predetermined time (for example, 1 second) before is a reference value or more. For example, the condition th3 that changes from Abnormal 1 (Hands-on) to Abnormal 2 (TD) can be such that the state deterioration threshold value is set to 0.5 m and the reference value is set to 0.8.

[0101] In a case where the first difference amount satisfies the state deterioration condition (step S201 - "Yes"), the determination section 232 changes the state of the precision of the current position of the vehicle 10 (step S206), and ends the series of processing. In a case where the first difference amount satisfies the condition th3, the state of the precision of the current position of the vehicle 10 is changed from Abnormal 1 (Hands-on) to Abnormal 2 (TD).

[0102] On the other hand, in a case where the first difference amount does not satisfy the state deterioration condition (step S201 - "No"), the determination section 232 determines whether the second difference amount satisfies the state deterioration condition (step S202). As the state deterioration condition, it can be cited that at least one of the second difference amounts calculated between the present camera shooting time and the time a predetermined time (for example, 1 second) before is the state deterioration threshold value or less among a plurality of second difference amounts. For example, the condition th3 that changes from Abnormal 1 (Hands-on) to Abnormal 2 (TD) can be such that the state deterioration threshold value is set to 0.6 m. In a case where the curvature determination of the road on which the vehicle 10 is traveling is greater than a reference curvature, the state deterioration threshold value can also be increased more than in a case where it is below the reference curvature.

[0103] If the second difference value satisfies the state deterioration condition (step S202 - "YES"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the second difference value satisfies the condition th3, the state of the accuracy of the current position of the vehicle 10 changes from "hands-on" 1 to "hands-on" 2 (TD).

[0104] On the other hand, if the second difference does not satisfy the state deterioration condition (step S202 - "No"), the determination unit 232 determines whether the first difference satisfies the state improvement condition (step S203). The state improvement condition can be exemplified by the ratio of the first difference values ​​obtained between the current camera capture time and a predetermined time (e.g., 1 second ago) in which the first difference values ​​are below the state improvement threshold being greater than a reference value. For example, condition th5 for changing from abnormal 1 (hands-on) to normal 2 (hands-off) can be a state improvement threshold of 0.3 m and a reference value of 1.0.

[0105] If the first difference value satisfies the state improvement condition (step S203 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the first difference value satisfies the condition th5, the state of the accuracy of the current position of the vehicle 10 is changed from "unsound 1" (hands-on) to "normal 2" (hands-off).

[0106] On the other hand, if the first difference does not satisfy the state improvement condition (step S203 - "No"), the determination unit 232 determines whether the second difference satisfies the state improvement condition (step S204). The state improvement condition can be exemplified by the requirement that at least one of the plurality of second difference values ​​calculated between the current camera capture time and a predetermined time (e.g., 1 second ago) is below a state improvement threshold. For example, condition th5 for changing from abnormal 1 (hands-on) to normal 2 (hands-off) can be setting the state improvement threshold to 0.4 m.

[0107] If the second difference value satisfies the state improvement condition (step S204 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the second difference value satisfies the condition th5, the state of the accuracy of the current position of the vehicle 10 is changed from "unsound 1" (hands-on) to "normal 2" (hands-off).

[0108] On the other hand, when the second difference amount does not satisfy the state improvement condition (step S204 —No), the determination unit 232 determines that the state of the accuracy of the current position of the vehicle 10 has not changed (step S205 ), and ends the series of processes.

[0109] Next, the following describes the determination process for when the current state of the accuracy of the current position of vehicle 10 is Defective 2 (TD). When the current state of the accuracy of the current position of vehicle 10 is Defective 2 (TD), the determination process is executed until the driver agrees to the control transfer request.

[0110] First, the determination unit 232 determines whether the first difference quantity satisfies the state improvement condition (step S203). If the current state of the accuracy of the current position of the vehicle 10 is abnormal 2 (TD), the processing of steps S201 and S202 is not performed. As a state improvement condition, it can be cited that the ratio of the first difference quantities below the state improvement threshold among the multiple first difference quantities calculated between the current camera shooting time and the time before a predetermined time (for example, 1 second) is greater than the reference value. For example, the condition th5 for changing from abnormal 2 (TD) to hand-off 2 can be to set the state improvement threshold to 0.3m and the reference value to 1.0.

[0111] If the first difference value satisfies the state improvement condition (step S203 - "Yes"), the determination unit 232 changes the state of the accuracy of the current position of the vehicle 10 (step S206), and the series of processes ends. If the first difference value satisfies the condition th5, the state of the accuracy of the current position of the vehicle 10 is changed from Uncorrected 2 (TD) to Hands-off 2 (Hands-off).

[0112] On the other hand, if the first difference does not satisfy the state improvement condition (step S203 - "No"), the determination unit 232 determines whether the second difference satisfies the state improvement condition (step S204). The state improvement condition can be exemplified by the requirement that at least one of the plurality of second difference values ​​calculated between the current camera capture time and a predetermined time (e.g., 1 second ago) is below a state improvement threshold. For example, condition th5 for changing from "Dead" 2 (TD) to "Hands-off" 2 can be a state improvement threshold of 0.4 m.

[0113] In a case where the second divergence amount satisfies the state improvement condition (step S204 - "Yes"), the determination section 232 changes the state of the precision of the current position of the vehicle 10 (step S206), and ends the series of processing. In a case where the second divergence amount satisfies the condition th5, the state of the precision of the current position of the vehicle 10 is changed from the abnormality 2 (TD) to the normality 2 (Hands-off).

[0114] On the other hand, in a case where the second divergence amount does not satisfy the state improvement condition (step S204 - "No"), the determination section 232 determines that the state of the precision of the current position of the vehicle 10 is not changed (step S205), and ends the series of processing.

[0115] In the above-described determination processing, the determination 233 determines the state of the precision of the current position of the vehicle 10 on the basis of the first divergence amount or the second divergence amount. The determination 233 can also determine the state of the precision of the current position of the vehicle 10 on the basis of the first divergence amount and the second divergence amount. In this case, the determination 233 determines that the state of the precision of the current position is deteriorated in a case where the first divergence amount satisfies the state deterioration condition (step S201 - "Yes") and the second divergence amount satisfies the state deterioration condition (step S202 - "Yes"). In addition, the determination 233 determines that the state of the precision of the current position is improved in a case where the first divergence amount satisfies the state improvement condition (step S203 - "Yes") and the second divergence amount satisfies the state improvement condition (step S204 - "Yes").

[0116] Next, with reference to Figure 9 and Figure 10 , the method of considering for determining the above-described state deterioration condition and the state improvement condition will be described below. Figure 9 and Figure 10 are diagrams for explaining the method of considering for determining the state deterioration condition and the state improvement condition.

[0117] In the example shown in Figure 9 , the vehicle 900 shows the first estimated position (estimated position using road features) at times t1 to t4, the vehicle 901 shows the second estimated position (estimated position based on dead reckoning) at times t1 to t4, and the vehicle 902 shows the current position (estimated position using a prediction filter) at times t1 to t4. It is assumed that an abnormality of a camera occurs at time t1, and the first estimated position shown by the vehicle 900 moves by δ (1 m) in the lateral direction from the second estimated position shown by the vehicle 901. The abnormality of the camera continues thereafter as well. It is assumed that the vehicle travels at a constant speed, and the first estimated position, the second estimated position, and the current position thereafter are calculated by simulation. The current position shown by the vehicle 902 approaches the first estimated position shown by the vehicle 900 as time elapses.

[0118] In Figure 10 In FIG. 1 , the amount of lateral positional deviation of the current position of the vehicle 902 is shown over time for the cases where the vehicle speed is 40 km / h, 80 km / h, and 136 km / h. Calculation of each position is performed every 100 ms.

[0119] From time t1 (20000 ms) to approximately one second later (21000 ms), the current position of vehicle 902 moves to 90% of the deviation δ (1 m) due to an abnormality in the first estimated position of vehicle 900. The current position of vehicle 902 moves laterally by approximately 0.1 m in 100 ms, approximately 0.3 m in 150 ms, and approximately 0.5 m in 300 ms. The lateral deviation of vehicle 902's current position is approximately the same for vehicle speeds of 40 km / h, 80 km / h, and 136 km / h. The first and second difference amounts are calculated approximately every 32 ms.

[0120] Regarding the first difference quantity state deterioration condition, according to Figure 10 The results shown consider condition th1, which sets the state deterioration threshold to 0.1 m and the reference value to 0.5; condition th2, which sets the state deterioration threshold to 0.3 m and the reference value to 0.6; and condition th3, which sets the state deterioration threshold to 0.5 m and the reference value to 0.8.

[0121] In addition, regarding the state deterioration condition of the second difference quantity, according to Figure 10 The results shown consider the condition th1 where the state deterioration threshold is set to 0.2 m, the condition th2 where the state deterioration threshold is set to 0.4 m, and the condition th3 where the state deterioration threshold is set to 0.6 m.

[0122] In addition, regarding the state improvement condition of the first difference quantity, according to Figure 10 The results shown consider the condition th4, where the state improvement threshold is set to 0.1 m and the reference value is set to 0.8, and the condition th5, where the state improvement threshold is set to 0.3 m and the reference value is set to 1.0.

[0123] In addition, regarding the state improvement condition of the second difference quantity, according to Figure 10 The results shown consider that the condition improvement threshold is set to 0.2m under condition th4 and to 0.4m under condition th5.

[0124] According to the position accuracy determination device of the present embodiment described above, it is possible to determine the accuracy of the estimated current position of the vehicle.

[0125] In the present disclosure, the position accuracy determination device, position accuracy determination computer program, and position accuracy determination method of the above-described embodiments may be modified as appropriate without departing from the gist of the present disclosure. Furthermore, the technical scope of the present disclosure is not limited to these embodiments but also includes the inventions described in the claims and their equivalents.

[0126] For example, in the above embodiment, the determination unit determines the accuracy of the vehicle's current position based on the first or second difference amount. However, the determination unit may determine the accuracy of the vehicle's current position based on the second difference amount.

[0127] Furthermore, in the above embodiment, the first estimated position represents the vehicle's estimated position at each camera capture time. In this case, it is impossible to determine an estimated position using road features between the previous and current camera capture times. Therefore, dead reckoning can be performed using the vehicle's estimated position at the previous camera capture time as the starting point to determine the vehicle's position between the estimated position at the previous camera capture time and the estimated position at the current camera capture time, using this as the first estimated position.

[0128] In addition, the above-mentioned determination process is an example, and the determination conditions should not be limited to the above.

Claims

1. A position accuracy determination device, characterized in that: have: a first position estimating unit for estimating a first position of the moving object at a first moment based on an image of a road feature on a road surface surrounding the moving object at a first moment and position information of the road feature on the road surface; a second position estimating unit for estimating a second position of the mobile object at the first moment based on the position of the mobile object at a second moment before the first moment and the movement amount and azimuth angle change amount of the mobile object from the second moment to the first moment; a calculation unit that inputs the first position and the second position at the first moment into a prediction filter to calculate a current position of the moving object at the first moment; as well as a determination unit that determines a state of accuracy of the current position of the mobile object based on a first difference amount indicating a distance between the current position of the mobile object at the first moment and a position estimated based on the position of the mobile object at a moment before the first moment and the amount of movement and azimuth angle change of the mobile object from the previous moment to the first moment, The determination unit determines that the state of accuracy of the current position of the moving object has deteriorated when at least one of the multiple first difference quantities calculated between the first moment and the moment before a predetermined time is greater than a predetermined first threshold value, and determines that the state of accuracy of the current position of the moving object has not deteriorated when all of the first difference quantities are less than the first threshold value.

2. The position accuracy determination device according to claim 1, wherein: The determination unit obtains a second difference amount indicating a distance between the first position and the second position of the moving object at the first time. The determination unit determines the accuracy of the current position of the moving object based on the first difference amount and the second difference amount. The determination unit determines that the state of accuracy of the current position of the moving object has deteriorated when the ratio of the second difference amount exceeding the predetermined second threshold value among the multiple second difference amounts calculated between the first moment and the moment before the predetermined time is greater than the first reference value; and determines that the state of accuracy of the current position of the moving object has not deteriorated when the ratio of the second difference amount exceeding the second threshold value is less than the first reference value.

3. The position accuracy determination device according to claim 1 or 2, wherein: The determination unit uses different criteria when determining that the accuracy of the current position of the moving object has changed from a good state to a poor state and when determining that the accuracy of the current position of the moving object has changed from a poor state to a good state.

4. The position accuracy determination device according to claim 2, wherein: The determination unit compares the second difference amount with a predetermined reference threshold value to determine the accuracy of the current position of the moving object.

5. The position accuracy determination device according to claim 1 or 2, wherein: The determination unit notifies the driver of a request for driving intervention with respect to the moving object via the notification unit based on the accuracy of the current position of the moving object.

6. The position accuracy determination device according to claim 1 or 2, wherein: The determination unit uses the second time as the previous time.

7. A computer-readable non-transitory storage medium storing a computer program for determining position accuracy, the computer program causing a processor to execute: estimating a first position of the moving object at the first moment based on an image representing a road feature on a road surface surrounding the moving object at the first moment and position information of the road feature on the road surface; estimating a second position of the mobile object at the first moment based on the position of the mobile object at a second moment before the first moment and the movement amount and azimuth change amount of the mobile object from the second moment to the first moment; inputting the first position and the second position at the first moment into a prediction filter to calculate the current position of the moving object at the first moment; as well as determining a state of accuracy of the current position of the mobile object based on a first difference amount indicating a distance between the current position of the mobile object at the first moment and a position estimated based on the position of the mobile object at a moment before the first moment and the amount of movement and azimuth change of the mobile object from the previous moment to the first moment; When at least one of the multiple first difference quantities calculated between the first moment and the moment before a predetermined time is greater than a predetermined first threshold value, it is determined that the state of accuracy of the current position of the moving object has deteriorated; when all of the first difference quantities are less than the first threshold value, it is determined that the state of accuracy of the current position of the moving object has not deteriorated.

8. A method for determining position accuracy performed by a position accuracy determining device, wherein: inferring a first position of the moving object at the first moment based on an image of a road feature on a road surface surrounding the moving object at the first moment and position information of the road feature on the road surface; Estimate a second position of the mobile object at the first moment based on the position of the mobile object at a second moment before the first moment and the movement amount and azimuth change amount of the mobile object from the second moment to the first moment, The first position and the second position at the first moment are input to a prediction filter to calculate the current position of the moving object at the first moment. determining the accuracy of the current position of the mobile object based on a first difference amount indicating a distance between the current position of the mobile object at the first moment and a position estimated based on the position of the mobile object at a moment before the first moment and the movement amount and azimuth angle change amount of the mobile object from the previous moment to the first moment, When at least one of the multiple first difference quantities calculated between the first moment and the moment before a predetermined time is greater than a predetermined first threshold value, it is determined that the state of accuracy of the current position of the moving object has deteriorated; when all of the first difference quantities are less than the first threshold value, it is determined that the state of accuracy of the current position of the moving object has not deteriorated.

Images