JP7310313B2 - POSITION CORRECTION SERVER, POSITION MANAGEMENT DEVICE, MOBILE POSITION MANAGEMENT SYSTEM AND METHOD, POSITION INFORMATION CORRECTION METHOD, COMPUTER PROGRAM, IN-VEHICLE DEVICE, AND VEHICLE - Google Patents

Description

This disclosure relates to a position correction server, a position management device, a position management system and method for a moving object, a position information correction method, a computer program, an in-vehicle device, and a vehicle.

With the development of automatic driving technology for vehicles, etc., the importance of positioning technology is increasing. A global navigation satellite system (GNSS) using artificial satellites for positioning (positioning satellites) is mainly used for positioning of mobile objects such as vehicles. There are several types of global positioning satellite systems, but they all have in common that a receiver is used to measure the distance to a plurality of satellites by communication, and the position of a vehicle or the like is specified by the principle of triangulation.

From its principle, GNSS needs to be able to communicate with at least three positioning satellites in order to accurately measure its position. However, a situation may occur in which the receiver cannot communicate with the three positioning satellites due to the influence of topography or the influence of buildings on the side of the road. In that case, a technique of estimating the position of a moving object by so-called dead reckoning is used.

However, the position itself using GNSS originally contains an error, and the accumulation of the error due to dead reckoning may increase the error of the position information managed by the receiver. On the other hand, recently, we are developing a system that creates a bird's-eye view map of traffic conditions on a server by combining the positions of vehicles in a certain area with map information, and distributes the results to each vehicle for traffic support. Research is being conducted towards In order for the server to manage the traffic situation bird's-eye view map accurately, it is necessary to notify the server of position information as accurate as possible from each vehicle or the like. Therefore, if the error in the positioning result obtained by the GNSS receiver provided in the vehicle becomes large, the traffic situation bird's-eye view map becomes inaccurate, which may affect traffic assistance.

One solution for solving these problems is disclosed in Patent Document 1 listed below. Referring to FIG. 1, a position correction system 50 described in Patent Document 1 includes a position detection device mounted on a vehicle 64 for detecting the vehicle position by GPS (Global Positioning System), which is a kind of GNSS, and a road position. An in-vehicle communication device 70 receives beacon data 66, 68 including position information from an optical beacon 60 installed at a high position such as on an information pole 62 of the vehicle. The position occupied by the optical beacon head image in the image of the camera 72 when the camera 72 and the vehicle 64 are positioned directly below the optical beacon 60 is set in advance as a reference position, and the optical beacon extracted from the image of the camera 72. and a directly below determination device for determining whether or not the head image is positioned at the reference position.

The position correction system 50 further responds to the fact that the directly below determination device determines that the optical beacon head image exists at the reference position, and the optical beacon data indicates the vehicle position at the time when the optical beacon head image is captured. a position correction device for correcting the vehicle position maintained by the position detection device to match the position.

According to Patent Literature 1, this configuration enables accurate measurement of the vehicle position.

On the other hand, Japanese Patent Application Laid-Open No. 2002-200002 proposes a technique of recognizing roadside landmarks by image analysis of an in-vehicle camera without using the optical beacon 60, and utilizing them for position correction. Referring to FIG. 2, a position correction system 100 according to the technique described in Patent Document 2 includes a landmark position information storage device 112 that stores in advance images and positions of landmarks 114 that are features of the roadside; Connected to a mark position information storage device 112 for storing landmark information about one or more landmarks in the route that can be used for position correction in response to receiving a request from the vehicle 116 for information about landmarks in the route. and a server 110 that transmits to the Landmark information includes position information of landmarks that can be recognized from a camera image, landmark images for recognition, recognition evaluation values that indicate the likelihood of successful recognition, and the like. The in-vehicle control device mounted on the landmark 114 selects a landmark to be recognized based on the recognition result evaluation value included in the received landmark information. The in-vehicle control device evaluates the recognition result of the landmark to be recognized, and transmits the evaluation result value together with the image of the landmark recognized by the camera to the server. The server updates the recognition evaluation value and the landmark image for recognition of each piece of landmark information based on the information received from the in-vehicle control device. Landmarks include objects such as signs and signals on the road, paint on the road, and characteristic buildings near the road (paragraph 0049).

According to this position correction system 100, it is possible to improve the accuracy of estimating the position of the vehicle while reducing the load in landmark recognition processing.

JP 2010-276583 A JP 2015-108604 A

However, the invention disclosed in Patent Document 1 has the following problems. That is, the technique described in Patent Document 1 is based on the premise that a roadside device having the optical beacon 60 is provided on the roadside, so there is a problem that the scope of application is limited. In addition, since it is assumed that the vehicle passes directly under the optical beacon 60, it is necessary to install the optical beacon 60 for each lane, which also poses a problem of high installation costs.

Further, in the invention described in Patent Document 2, it is assumed that the in-vehicle control device recognizes the landmark image in the image from the in-vehicle camera. Therefore, the position correction capability depends on the performance of the on-vehicle camera and the recognition capability of the on-vehicle control device. On the other hand, since the in-vehicle camera moves with the vehicle, the image to be recognized is constantly changing. Therefore, teacher data used for learning for recognition becomes sparse, and there is a problem that it is difficult to improve recognition accuracy. Furthermore, image recognition requires a lot of computational resources. In general, there is a limit to the performance of an in-vehicle control device, and it is difficult to perform image recognition in a short time with high accuracy. Therefore, it must be said that the system disclosed in Patent Document 2 has a limit in its position correction capability. Therefore, it is desirable to be able to correct the position of the moving object without using the roadside unit equipped with the optical beacon 60 and the image of the vehicle-mounted camera.

Therefore, an object of the present disclosure is to provide a position correction server, a position management device, a position management system and method for a moving object, a method for correcting position information, a computer program, an in-vehicle device, and a vehicle that can accurately correct the position of a vehicle. That is.

A position correction server according to a first aspect of the present disclosure includes a mobile body information receiver that receives mobile body information including position information and time information of the mobile body from a mobile body; An image receiver for receiving a captured image from a device, and receiving a captured image including an arbitrary moving object image and moving object information based on an arbitrary moving object image and moving object information existing in the captured image a matching unit that matches a moving object that transmits moving object information to a machine; location information of the moving object information received from the moving object matched by the matching unit; captured images matched by the matching unit; and a correction processing unit that performs a process of correcting the position information maintained by the moving body using .

A position management device for a mobile object according to a second aspect of the present disclosure includes an information maintenance device that maintains mobile object information including its own position information and time information, and an information maintenance device that maintains mobile object information including its own position information and time information. In response to externally receiving correction information indicating that the location information maintained by the mobile information transmission device for transmission to the correction server and the information maintenance device should be corrected, using the correction information. and a correction processing unit that corrects the position information.

A position management system for a mobile object according to a third aspect of the present disclosure includes any of the position correction servers described above and any of the position management devices described above.

A position information correcting method according to a fourth aspect of the present disclosure includes steps of: a computer receiving moving body information including position information and time information of the moving body from a moving body; a step of receiving a captured image from an identifiable imaging device; and a captured image including an image of an arbitrary moving object based on an image of an arbitrary moving object existing in the captured image and information on the moving object. and a step of matching the mobile object transmitting the mobile object information in the step of receiving the mobile object information; location information of the mobile object information received from the mobile object matched in the matching step; and performing a process of correcting the position information maintained by the moving body using the captured image and its capturing time.

A computer program according to a fifth aspect of the present disclosure provides a computer with a mobile object information receiver that receives mobile object information including position information and time information of the mobile object from a mobile object, and an imaging range that can be specified. an image receiver for receiving a captured image from an imaging device; and a captured image including an image of an arbitrary moving object and a moving object based on an image of an arbitrary moving object existing in the captured image and information on the moving object. A matching unit that matches a moving object that transmits moving object information to an information receiver, location information of the moving object information received from the moving object matched by the matching unit, and a captured image that is matched by the matching unit and its image. It functions as a correction processing unit that performs a process of correcting the position information maintained by the moving body using the time.

A position management method of a mobile object according to a sixth aspect of the present disclosure includes steps of: a step of a computer maintaining mobile object information including its own position information and time information; and sending the correction information to the position correction server in response to externally receiving correction information indicating that the position information maintained by the step of maintaining the moving body information should be corrected. and using the position information to correct.

A computer program according to a seventh aspect of the present disclosure provides a computer with an information maintenance device that maintains mobile body information including its own position information and time information, and a mobile body information that is sent to a predetermined position correction server at a predetermined timing. and in response to externally receiving correction information indicating that the position information maintained by the information maintenance device should be corrected, using the correction information to update the position information. It functions as a correction processing unit for correction.

An in-vehicle device according to an eighth aspect of the present disclosure is an in-vehicle device equipped with the position management device for a moving body described above.

A vehicle according to a ninth aspect of the present disclosure is a vehicle equipped with the vehicle-mounted device described above.

It should be noted that this disclosure can be realized not only as a device having such a characteristic processing unit, but also as a method having steps of the characteristic processing as described above, or as a method that causes a computer to execute the steps. It can be realized as a program for Moreover, it can be realized as a semiconductor integrated circuit that realizes part or all of the device, or as a system including these devices. Other problems, objects, configurations, actions and effects of this disclosure will become apparent from the following description of the disclosure together with the accompanying drawings.

According to this disclosure, it is possible to provide a position correction server, a position management device, a position management system and method for a moving object, a method for correcting position information, a computer program, an in-vehicle device, and a vehicle that can perform position correction of a vehicle with high accuracy.

FIG. 1 is a schematic diagram showing a schematic configuration of a conventional position correction system described in Patent Document 1. As shown in FIG. FIG. 2 is a schematic diagram showing a schematic configuration of a conventional position correction system described in Patent Document 2. As shown in FIG. FIG. 3 is a diagram showing a schematic configuration of a position correction system according to this disclosure. FIG. 4 is a schematic diagram illustrating the principle of vehicle position correction according to this disclosure. FIG. 5 is a schematic diagram showing a schematic configuration of a vehicle related to this disclosure. FIG. 6 is a block diagram showing a schematic hardware configuration of an in-vehicle device used in this disclosure. FIG. 7 is a block diagram showing the hardware configuration of the HMI controller used in this disclosure. FIG. 8 is a flow chart showing a control structure of a computer program (hereinafter simply referred to as "program") that implements data upload processing in the in-vehicle device in this disclosure. FIG. 9 is a flow chart showing a control structure of a program that implements position correction processing in the in-vehicle device according to this disclosure. FIG. 10 is a diagram showing the appearance of an edge server according to this disclosure. FIG. 11 is a block diagram showing a schematic hardware configuration of the edge server shown in FIG. 12 is a block diagram showing a functional configuration of the edge server shown in FIG. 10; FIG. FIG. 13 is a flow chart showing the control structure of a program for realizing each function shown in FIG. FIG. 14 is a schematic diagram for explaining a problem related to vehicle identification when two vehicles of the same model are close to each other. FIG. 15 is a schematic diagram showing a method for solving the problem of vehicle identification when two vehicles of the same model are close to each other. FIG. 16 is a schematic diagram for explaining a problem related to vehicle identification when three vehicles of the same model are close to each other. FIG. 17 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the first disclosure. FIG. 18 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the second disclosure. FIG. 19 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the third disclosure. FIG. 20 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the fourth disclosure.

[Description of Embodiments of this Disclosure]
In the following description and drawings, identical parts are provided with identical reference numerals. Therefore, detailed description thereof will not be repeated. It should be noted that at least part of the following disclosures may be combined arbitrarily.

(1) A position correction server according to a first aspect of the present disclosure includes a mobile object information receiver that receives mobile object information including position information and time information of the mobile object from the mobile object, and an imaging range that specifies an image receiver for receiving a captured image from a possible imaging device; and a captured image including an image of an arbitrary moving object and movement based on an image of an arbitrary moving object existing in the captured image and information on the moving object. a matching unit that matches a moving object that transmits moving object information to the object information receiver; location information of the moving object information received from the moving object matched by the matching unit; a correction processing unit that performs a process of correcting the position information maintained by the moving body using the imaging time.

Based on the image from the imaging device and the position information of the moving object, the image at a specific imaging time and the moving object imaged by the imaging device at that time are matched. An error between the position information from the moving body at the imaging time and the position information of the moving body obtained from the image is calculated, and the position coordinates of the moving body are corrected by the error. It is possible to correct the position of a moving object using an existing imaging device without using an image from an in-vehicle camera.

(2) Preferably, the position correction server further includes a mobile object tracking device for tracking a mobile object whose position information is to be corrected based on the mobile object information, In response to the mobile information from a first mobile being tracked by the device, a captured image including an image of the first mobile is extracted from captured images received by the image receiver. The correction processing unit is calculated from the captured image and the position information transmitted from the first moving object at the imaging time of the captured image extracted by the image extracting unit and received by the moving object information receiver and a position correction processing unit that performs a process of correcting the maintained position information of the first moving body from the position of the first moving body that is determined.

Only the moving body subject to position correction is subject to position correction. Position correction is not performed for a moving object that maintains highly accurate position information. As a result, it is possible to efficiently correct the position information of the moving object while saving computational resources.

(3) More preferably, the position correction processing unit includes a position storage device that stores the position of the feature in the real world, a feature recognition unit that recognizes the feature that exists in the captured image, and a feature recognition unit. a reading device for reading from the position storage device the real-world position of the feature recognized by the first moving object, and for reading the position information maintained by the first moving object based on the position read from the position storage device; A calculation device for calculating a correction amount for correction is included.

The position of the feature is always present in the background of the image as it is imaged by a fixed imaging device. Based on the image and the moving object image in the image, the position of the moving object can be calculated from the positional relationship between the two. Since the image of the image pickup device whose position is fixed is used instead of the image of the on-board camera, the position of the moving object can be calculated with high accuracy, and the position can be corrected accurately.

(4) More preferably, the calculation device, based on the relationship between the image of the feature present in the captured image and the image of the moving object, the geographical position of the feature, and the characteristics of the imaging device, A relative position calculation device for calculating a relative position between a mobile object and a feature, and a geographical position calculation device for calculating the geographical position of the mobile object by adding the geographical position of the feature to the relative position and a correction amount calculation device for calculating, as a correction amount, an error between the position information maintained by the first moving body and the geographical position calculated by the geographical position calculation device.

When the imaging range of the imaging device is fixed, the background of the image is basically always the same. Recognition of features therein can therefore be performed with high accuracy. Also, the positions of the features are known in advance. Therefore, by calculating the relative position of the moving body with respect to the feature from the image and adding it to the position of the feature, the position of the moving body can be easily specified.

(5) Preferably, the matching unit responds to the fact that a second moving body of the same type as the first moving body exists within a predetermined distance from the first moving body, the first A forward image acquiring device for acquiring respective forward images from a moving object and a second moving object; An image inspection unit that inspects whether or not an image exists, and if there is only one forward image in which an image of a moving object of the same type as the first moving object does not exist in the forward image, , and a device for matching the captured image with the moving object that transmitted the forward image.

When a plurality of mobile objects of the same type exist close to each other, it is not possible to identify the mobile object that has transmitted the position information using only the image. In such a case, forward images are acquired from both sides, and location information is transmitted among a plurality of moving objects based on whether or not an image of the same type of moving object as that moving object exists in the images. Identifies moving objects that have traveled

(6) More preferably, the position correction server further includes a moving object tracking device that tracks the moving object whose position information is to be corrected based on the moving object information, and the moving object information is an appearance related to the appearance of the moving object. The matching unit identifies, for each moving object to be tracked by the moving object tracking device, an imaging device capable of specifying an imaging area existing near the moving object, and an image receiver. identifies an image of the moving object in the captured image extracted by the image extraction unit using the image extraction unit for extracting an image captured by the specified image capturing device from the captured image received by the specified image capturing device, and determines the image; and a moving body identification device for identifying the moving body in the captured image based on the appearance information.

If the position of the imaging device is fixed, the background of the image it captures is always the same. By using information about the appearance of a moving object, it is possible to easily identify the image of the moving object having that appearance in such an image and identify its position with high accuracy, so that the position information of the moving object can also be corrected with high accuracy. .

(7) More preferably, the correction processing unit, based on the position of the feature within the imaging region of the captured image extracted by the image extracting unit, and the position of the image of the moving object in the captured image, a moving body position identifying unit that identifies the geographical position of the moving body in the mobile body, the geographical position of the moving body identified by the moving body position identifying part, and the captured image, which is transmitted from the moving body at the time when the captured image is captured; and a device for correcting location information maintained by the mobile based on the location information received by the information receiver.

The position information of the moving object calculated with reference to the features in the received image is considered to be accurate position information. The position information transmitted from the moving object at the time the image was captured may contain an error, but the correct position calculated at the same time can be used to correct the error in the position information.

(8) Preferably, the device that performs correction processing is transmitted from the mobile object at the geographical position of the mobile object specified by the mobile object position specifying unit and the time when the captured image was captured, and the mobile object information receiver An error calculation device for calculating an error between the geographical position determined by the received position information and an instruction to add the error calculated by the error calculation device to the position information maintained by the mobile object is transmitted to the mobile object. and a transmitting device.

The error between the position of the moving object specified from the image and the position information transmitted from the moving object at the same time when the image was captured can be considered as an error of the position information maintained by the moving object. can. By adding this error to the position information maintained by the moving body, the moving body can correct the position information maintained by itself and calculate the correct position.

(9) More preferably, the location information server further includes a moving body tracking device that tracks a moving body whose position information is to be corrected based on the moving body information, and the moving body information is an appearance related to the appearance of the moving body. The matching unit, in response to the image receiver receiving the captured image, recognizes an image of the mobile being tracked by the mobile tracking device that is present in the captured image. Identification for identifying a moving object in the captured image among moving objects tracked by a moving object tracking device using a body image recognition unit, an image recognized by the moving object image recognition unit, and appearance information and a correction processing unit, for the moving object recognized by the moving object image recognition unit, the position of the image of the feature in the captured image, the position of the image of the moving object, and the geography of the feature. a mobile body position calculation device for calculating the position of the mobile body at the time when the captured image was captured based on the position of the target, and the position calculated by the mobile body position calculation device and the image of the mobile body are recognized. and a correction execution unit for executing a process of correcting the position information maintained by the moving body based on the position information received by the moving body information receiver when the captured image is captured.

Tracking for position information correction is performed only for the moving object whose position information is to be corrected. In response to receiving the image from the imaging device, based on the image, an image of a mobile object present in the image is identified and its geographic location is identified. The moving body and the moving body tracked by the moving body tracking device are matched using the appearance information of the moving body to identify the moving body in the image. The result is used to correct the position information of the moving object. Since the processing for position information correction can be performed only for the mobile object to be corrected, computational resources can be saved. When the image is received from the imaging device, the detection of the target moving object in the image and the position correction process are started, so the correction process can be performed efficiently.

(10) More preferably, in the position correction server, the mobile body information further includes accuracy-related information related to the accuracy of the position information maintained by the mobile body, and the mobile body tracking device, based on the accuracy-related information, performs a predetermined A tracking device that tracks a moving object having a position information acquisition device that is highly likely to operate with an accuracy lower than the accuracy of position information to be corrected.

This configuration saves computational resources for position correction.

(11) Preferably, the accuracy-related information includes the number of positioning satellites that can be communicated in a satellite positioning system used by the mobile, and the tracking device detects that the number of positioning satellites that the mobile can communicate with is equal to or less than a threshold value. Includes equipment for tracking certain moving objects.

A mobile object to be tracked can be determined by a clear criterion of the number of positioning satellites that can be communicated with.

(12) More preferably, the satellite positioning system includes GPS.

Since it targets vehicles equipped with GPS, the most popular satellite positioning system, many vehicles can benefit from this system.

(13) Preferably, the position correction server further confirms that the accuracy-related information related to the accuracy of the location information maintained by the moving body has changed to a value indicating that the accuracy of the location information has become higher than a predetermined accuracy. and an excluding device for executing a process of excluding the moving object from a target of position information correction by correction in response to the correction.

If the accuracy of the positional information of the moving body has increased, it is meaningless to continue the correction. By excluding such moving bodies from correction targets, computational resources can be saved, and the saved computational resources can be allocated to correction of position information regarding other vehicles.

(14) More preferably, the mobile object information further includes appearance information relating to the appearance of the mobile object, and the matching unit recognizes the image of the mobile object included in the captured image based on the appearance information included in the mobile object information. an image recognizing unit that recognizes the image of the moving object, and a device that associates the captured image in which the image of the moving object is recognized by the image recognizing unit with the moving object that transmits the moving object information to the moving object information receiver.

Since the position of the imaging device is fixed, the background is always constant. Therefore, the position of the moving body within the image can be easily specified using the appearance information of the moving body.

(15) A position management device for a mobile object according to the second aspect of the present disclosure includes an information maintenance device that maintains mobile object information including its own position information and time information, and an information maintenance device that maintains mobile object information including its own position information and time information. In response to externally receiving correction information indicating that the position information maintained by the mobile information transmitting device for transmission to a predetermined position correction server and the information maintaining device should be corrected, the correction information is transmitted. and a correction processing unit that is used to correct the position information.

The mobile can use the location information received from the server to correct the location information it maintains. As a result, mobile units can use services that use location information in a more appropriate state.

(16) Preferably, the information maintenance device has a function of maintaining position information by a satellite positioning system, and the mobile object information transmitting device preferably includes the satellite positioning information in the mobile object information when transmitting the mobile object information to the position correction server. Transmit including the number of positioning satellites that the system can communicate with.

By transmitting the number of positioning satellites to the server, the server can determine whether or not the position information of the mobile object is highly accurate. When the number of positioning satellites is less than a predetermined number and the mobile body corrects its own position by dead reckoning, it becomes possible to receive a position information correction service from the server.

(17) More preferably, the correction information includes an error between the position information maintained by the information maintenance device and the correct position information, and the correction processing unit adds the error to the position information maintained by the information maintenance device. contains an adder that adds

With this configuration, it is possible to easily correct the position information of the moving body.

(18) More preferably, the position management device further includes a stop device for stopping the operation of the correction processing unit in response to the number of positioning satellites with which the satellite positioning system can communicate reaches or exceeds a predetermined number. include.

If the number of communicable positioning satellites reaches a predetermined number or more, the accuracy of position information by the satellite positioning system will be sufficiently high. By stopping the position information correction process, computational resources can be saved and allocated to other processes.

(19) A mobile body position management system according to a third aspect of the present disclosure includes any of the position correction servers described above and any of the position management devices described above.

Using an image from an imaging device capable of specifying an imaging range, the server can correct the position information of the moving object. There is no need to use constantly changing information from on-board cameras. As a result, it is possible to easily obtain the accurate position of the mobile object, and efficiently maintain the accuracy of the position information managed by the mobile object at a high level of accuracy. (20) Position information according to the fourth aspect of this disclosure The correction method comprises the steps of: a computer receiving moving body information including position information and time information of the moving body from the moving body; and the computer receiving the captured image including the image of the arbitrary moving object and the information on the moving object based on the image of the arbitrary moving object and the information on the moving object existing in the captured image. A step of matching a moving body to be transmitted, position information of the moving body information received from the matching moving body in the matching step, and a captured image matched in the matching step and its capturing time to move. and performing processing to correct body-maintained position information.

Using an image from an imaging device capable of specifying an imaging range, the server can correct the position information of the moving object. There is no need to use constantly changing information from on-board cameras. As a result, the accurate position of the mobile object can be easily obtained, and the accuracy of the position information managed by the mobile object can be efficiently maintained at a high level of accuracy.

(21) A computer program according to a fifth aspect of the present disclosure provides a computer with a mobile object information receiver that receives mobile object information including position information and time information of the mobile object from a mobile object; an image receiver that receives a captured image from an imaging device that can identify a captured image that includes an image of an arbitrary moving object based on an image of an arbitrary moving object that exists in the captured image and information on the moving object; A matching unit that matches a moving object that transmits moving object information to a moving object information receiver, location information of the moving object information received from the moving object that has been matched by the matching unit, and a captured image that has been matched by the matching unit. It functions as a correction processing unit that performs a process of correcting the position information maintained by the moving body using the imaging time.

Based on the image from the imaging device and the positional information from the moving body, the image captured at a specific imaging time and the moving body captured by the imaging device at that time are matched. An error between the position information from the moving body at the imaging time and the position information of the moving body obtained from the image is calculated, and the position coordinates of the moving body are corrected by the error. It is possible to correct the position of a moving object using an existing imaging device without using an image from an in-vehicle camera.

(22) A mobile object location management method according to a sixth aspect of the present disclosure includes a step in which a computer maintains mobile object information including its own location information and time information; sending the information to a predetermined location correction server; and in response to the computer externally receiving correction information indicating that the location information maintained by the step of maintaining the mobile information should be corrected; and correcting the position information using the correction information.

By transmitting the mobile object information including the position information and time information of the mobile object to the position correction server, the correction information for correcting the position information of the mobile object is received from the server, and the position information can be maintained at a more accurate value.

(23) A computer program according to a seventh aspect of the present disclosure provides a computer with an information maintenance device that maintains mobile body information including its own location information and time information; In response to externally receiving correction information indicating that the location information maintained by the mobile information transmission device for transmission to the correction server and the information maintenance device should be corrected, using the correction information. It functions as a correction processing unit that corrects the position information.

By transmitting the mobile object information including the position information and time information of the mobile object to the position correction server, the correction information for correcting the position information of the mobile object is received from the server, and the position information can be maintained at a more accurate value.

(24) An in-vehicle device according to an eighth aspect of the present disclosure is an in-vehicle device equipped with the position management device for a moving body described above.

By transmitting mobile information including location information and time information of a mobile device equipped with an in-vehicle device to a location correction server, correction information for correcting the location information of the mobile device is received from the server and maintained by this in-vehicle device Position information can be maintained at a higher accuracy value.

(25) A vehicle according to a ninth aspect of the present disclosure is a vehicle equipped with the in-vehicle device described above.

The moving body information including the position information and time information of the vehicle equipped with the above-described in-vehicle device is transmitted to the position correction server. As a result, the in-vehicle device receives correction information for correcting the position information of the moving object from the server, and the position information maintained by the in-vehicle device can be maintained at a value with higher accuracy.

[Detailed description of the embodiment]
<First Disclosure>
<composition>
FIG. 3 is a diagram showing a schematic configuration of the position correction system 150 according to the first disclosure. Referring to FIG. 3, this position correction system 150 is installed, for example, on the roadside of a road 174 passing through skyscrapers 162, 164, 166, 168, 170 and 172, or on the wall of skyscraper 162, etc. Infrastructure camera 160 whose position and imaging range are known, vehicles equipped with the in-vehicle device according to this disclosure, such as vehicles 176 and 178, and vehicle type, location, and vehicle received wirelessly from these vehicles Based on vehicle information such as time information maintained, image data received from infrastructure sensor equipment such as infrastructure camera 160, position data of moving objects by LiDAR, and image capturing time, create a traffic situation bird's-eye view map, an edge server 180 for maintaining, managing and distributing traffic assistance information to the on-board devices of each vehicle. From the infrastructure sensor equipment such as the infrastructure camera 160, the image of the camera or sensor data such as point cloud data detected by the sensor such as LiDAR is periodically uploaded to the edge server 180 together with the information indicating the imaging time.

In the situation shown in FIG. 3, for example, when a vehicle such as a vehicle 176 exists in a relatively open space in the front, rear, left, and right, communication with three positioning satellites is possible. When a vehicle is in a place where buildings 162, 164, 168 and 170 are present, it may happen that it cannot communicate with three positioning satellites and can only communicate with two or one positioning satellite. . In such a case, for example, the in-vehicle device of the vehicle 176 performs a process of estimating the position information by adding the movement information accumulated by the subsequent dead reckoning to the position information finally acquired by GNSS. The edge server 180 according to this disclosure, which will be described below, tracks vehicles in such a situation, and based on the image from the infrastructure camera 160 and its imaging time, as well as the vehicle position information received from each vehicle. It has a function to create correction information that corrects position information and send it to the vehicle in question.

As will be described later, position information estimated by dead reckoning includes an error that is considerably larger than that of position information measured by GNSS. This error is represented, for example, by circles drawn around vehicles 176 and 178 shown in FIG. If this error is large, multiple vehicles may be present within this circle, as shown in FIG. As in this disclosure, when the vehicles included in the image captured by the infrastructure camera 160 are of the same type, it may not be possible to determine which vehicle is the target vehicle. In this disclosure, the position of the target vehicle is not corrected in such a case. can also identify the target vehicle.

FIG. 4 is a schematic diagram for explaining one method of vehicle position correction in this disclosure. Referring to FIG. 4, in one approach of this disclosure, for some landmarks (eg, road paint 202), the edge server 180 pre-identifies and stores the exact location (Rx, Ry). The image of the vehicle 200 whose own position is estimated by dead reckoning as the vehicle position (Cx, Cy) exists together with the paint 202 in the image captured by the infrastructure equipment device such as the infrastructure camera 160. When the edge server 180 recognizes, it calculates the difference (ΔPx, ΔPy) based on the position (Rx, Ry) of the paint 202 and the distance 204 between the paint 202 and the vehicle 200 in the image, and uses the value to 200 exact positions are calculated. Further, edge server 180 calculates the error between the thus calculated accurate position of vehicle 200 and the position (Cx, Cy) received from vehicle 200 and estimated by vehicle 200, and calculates the error. Transmit to vehicle 200 . After this, the vehicle 200 obtains the vehicle position information by repeating the process of adding the error received from the edge server 180 to the position of the vehicle estimated at predetermined time intervals by the vehicle 200 itself through dead reckoning.

Of course, it may be a specific building or the like in the image instead of the paint of the road. It is also possible to calculate the position of the vehicle in relation to multiple features.

Another approach is to use a neural network model that takes an image from an infrastructure camera as input and estimates the position of the vehicle in that image. In this case, the camera that captures the image is an infrastructure camera, and the background of the captured image is basically always the same. It has the advantage that it can be done with much higher precision than This disclosure described below uses the latter approach.

FIG. 5 is a schematic diagram showing a schematic configuration of vehicle 200 related to this disclosure. Referring to FIG. 5, a vehicle 200 related to this disclosure includes a group of sensors such as a millimeter wave radar 212, an in-vehicle camera 214, and a LiDAR 216, and an edge server that periodically outputs sensor data as detection data from these sensors. 180 and receives traffic assistance information from the edge server 180 to provide traffic assistance for the driver.

FIG. 6 is a block diagram showing a schematic hardware configuration of the in-vehicle device 210. As shown in FIG. Referring to FIG. 6, in-vehicle device 210 includes an HMI (Human-Machine Interface) controller 232 connected to an in-vehicle LAN; 230 as antennas for fifth-generation mobile communication systems (so-called "5G"), intelligent transport systems (so-called "ITS"), GPS, which is a type of GNSS, and Wi-Fi. It includes a functioning integrated antenna 240, an automatic driving controller 234 connected to the HMI controller 232 and the external communication controller 230 by an in-vehicle LAN, and a driving system controller 236 connected to the in-vehicle LAN.

A monitor 242 and a plurality of ECUs (Electronic Control Units) 244 and 246 are connected to the HMI controller 232 .

An automatic driving ECU 248 is connected to the automatic driving controller 234 in addition to the millimeter wave radar 212 , the in-vehicle camera 214 and the LiDAR 216 .

A plurality of ECUs for vehicle travel control are connected to the travel system controller 236 .

FIG. 7 is a block diagram showing the hardware configuration of HMI controller 232 shown in FIG. 6 and used in this disclosure. 7, the HMI controller 232 includes a computer 350, a touch panel 352 connected to the computer 350, a monitor 242 such as an LCD connected to the computer 350 and providing a user I/F together with the touch panel 352, It also includes a speaker/microphone 360 for providing the traffic assistance information by voice to the user and the user I/F of the computer 350 by voice.

The computer 350 includes a CPU (Central Processing Unit), a bus 372 connected to the CPU 370 and providing a communication path for data and control signals between the CPU 370 and each unit in the HMI controller 232, and a boot-up controller for the computer 350. Wireless communication with an edge server 180 via a ROM 374 that stores programs, a RAM 376 that can be written and read at any time, an auxiliary storage device 378 such as a hard disk or SSD, and an external communication controller 230 and an integrated antenna 240 shown in FIG. and a USB memory port 384 connected to the bus 372 to which the USB memory 362 can be attached/detached.

The computer 350 is further connected to a bus 372 to provide an input/output I/F 382 serving as an interface between various ECUs 356 and various sensors 358 and the computer 350, and an interface between the bus 372 and the speaker/microphone 360. and a network I/F 392 connected to the bus 372 and the in-vehicle LAN 394 .

FIG. 8 is a flowchart showing a control structure of a computer program (hereinafter simply referred to as "program") that implements data upload processing in the in-vehicle device 210 in this disclosure. This program is executed by the CPU 370 of the HMI controller 232 shown in FIG.

Referring to FIG. 8, this program is repeatedly executed at predetermined intervals. It includes step 400 of reading the data received and stored in RAM 376 or auxiliary storage device 378 from those storage devices, and step 402 of uploading the read information to edge server 180 and completing the process.

The information uploaded to the edge server 180 in step 400 includes location information, vehicle type, time information, GPS performance, number of satellites with which GPS can communicate, vehicle speed, acceleration, etc. of the vehicle equipped with the in-vehicle device 210 including the HMI controller 232. include.

FIG. 9 is a flow chart showing a control structure of a program that implements position correction processing in the in-vehicle device according to this disclosure. Referring to FIG. 9, this program is activated each time data is received. This program consists of step 500 for judging whether or not the received data is sensor data and branching the flow of control according to the result, and storing the received sensor data in the RAM 376 or auxiliary storage when the judgment in step 500 is affirmative. Step 502 for saving in the device 378 and ending the process; and step 504 for branching the control flow.

The program further includes a step 508 in which, in response to an affirmative determination in step 504, the received position correction information is stored in RAM 376 or auxiliary storage device 378 and execution of the program ends; is negative, the data is processed according to the type of the data, and the execution of the program is terminated. The position correction information saved in the process of step 508 is then added to the position information calculated by dead reckoning. Although not shown, this position correction information is cleared when the number of communicable satellites returns to three or more.

FIG. 10 is a diagram showing the appearance of the edge server 180 used in this disclosure. 10, this edge server 180 includes a computer 640 having a USB memory port 652 and a DVD (Digital Versatile Disc) drive 650, a keyboard 646, a mouse 648, and a monitor 642.

FIG. 11 is a block diagram showing a schematic hardware configuration of the edge server shown in FIG. Referring to FIG. 11, computer 640 includes, in addition to USB memory port 652 and DVD drive 650, CPU 656, bus 666 connected to CPU 656, USB memory port 652 and DVD drive 650, numerical calculations, etc., in parallel. A GPU (Graphics Processing Unit) 657 that can be executed on a computer, a read-only memory (ROM) 658 that stores a boot-up program, etc., a RAM 660 that is connected to a bus 666 and stores program instructions, system programs, work data, etc. Includes hard disk 654 . Edge server 180 further includes a network interface 644 that provides connection to network 668 that allows communication with other terminals. The GPU 657 can execute a large amount of numerical calculations generated in image processing, neural networks, and the like concurrently and at high speed.

A computer program for causing the edge server 180 to function as each functional unit of the edge server 180 according to each disclosure of the present invention is stored and distributed on the DVD 662 or USB memory 664 attached to the DVD drive 650 or USB memory port 652. , are further transferred to the hard disk 654 from these. Alternatively, the program may be transmitted to computer 640 over network 668 and stored on hard disk 654 . Programs are loaded into RAM 660 during execution. Programs may be loaded directly into RAM 660 from DVD 662 , from USB memory 664 , or via network 668 .

This program includes an instruction string consisting of a plurality of instructions for causing the computer 640 to function as the edge server 180 and its respective functional units according to the above disclosures. Some of the basic functions necessary to cause computer 640 to perform this operation are the operating system or third party programs running on computer 640 or various programming toolkits or programs installed on computer 640 that can be dynamically linked. provided by the library. Accordingly, the program itself need not necessarily include all the functionality necessary to implement the system, apparatus and method of this disclosure. This program, in instructions, dynamically invokes the appropriate functions or programs in a programming toolkit or program library at run time in a controlled manner to achieve the desired result. , apparatus or method. Of course, the program alone may provide all necessary functions.

12 is a block diagram showing a functional configuration of the edge server shown in FIG. 10; FIG. Referring to FIG. 12, edge server 180 has the following configuration to enable such operations. The edge server 180 receives signals from a plurality of infrastructure sensor facilities (including either or both of cameras and LiDAR) and signals from vehicle-mounted sensors (including either or both of cameras and LiDAR) as described above. is received via wireless communication or wired communication, and the received data is sorted according to its contents.

The edge server 180 further analyzes the sensor data received by the reception processing unit 700, and outputs analysis results such as the position, speed, attributes, etc. of each moving body including vehicles and pedestrians existing within the managed range. a moving body tracking unit 702 for tracking, an analysis result storage unit 704 for storing the analysis results output by the moving body tracking unit 702, and a moving object such as a vehicle and a pedestrian based on the signal received by the reception processing unit 700 Attribute detection unit 706 for detecting attributes including vehicle type, type, time information, GPS performance, number of satellites with which GPS can communicate, etc.; Integrating processing for integrally managing the analysis results stored in the analysis result storage unit 704 and the attributes stored in the attribute storage unit 708, integrating them with the high-precision map information, and generating and outputting a traffic situation bird's-eye view map. 710 and an analysis result storage unit 712 that stores the traffic situation bird's-eye view map output by the integrated processing unit 710 . As for the method of analyzing sensor data performed by the moving object tracking unit 702, any known method may be employed.

The edge server 180 further obtains information such as position information, vehicle speed, acceleration, etc. of moving bodies such as vehicles existing in the area to be managed based on the signals received by the reception processing unit 700 from each vehicle and terminal. , a vehicle/terminal tracking unit 714 for tracking a moving object based on the information, and a vehicle/terminal information storage for storing moving object information such as each vehicle and terminal obtained by the vehicle/terminal tracking unit 714. unit 716, based on the number of GPS communicable satellites stored in the analysis result storage unit 712, the number of GPS communicable satellites is less than 3, or the GPS accuracy (performance) is equal to or less than a predetermined threshold value. A tracked object determination unit 718 that identifies a certain vehicle and marks the vehicle/terminal information of the vehicle stored in the vehicle/terminal information storage unit 716 as a vehicle requiring correction of the position information, and the integration processing unit 710 generates and analyzes and a driving support information generation unit 734 for generating driving support information based on the analysis results stored in the result storage unit 712 .

The edge server 180 further receives images from the infrastructure cameras to be managed via the reception processing unit 700 as a function unit for matching the vehicle to be tracked with the vehicle in the image, and converts the image data and the imaging time. an infrastructure image receiving unit 720 separated into an infrastructure image receiving unit 720, an infrastructure image storage unit 721 for storing the image of the vehicle in the image received by the infrastructure image receiving unit 720 and its imaging time, a vehicle/terminal tracking unit 714, and a vehicle/terminal tracking unit 714; In response to the vehicle/terminal tracking unit 714 connected to the terminal information storage unit 716 receiving the position information from the vehicle whose position is to be corrected, whether or not a vehicle of the same model exists within a predetermined range from the vehicle. and an image specifying unit 721 for specifying, in the infrastructure image storage unit 721, the infrastructure image at the time when the vehicle recently passed through the imaging area point of the infrastructure camera in the management area. 726, a vehicle position estimation unit 724 for estimating the position of the vehicle whose position information is received by the vehicle/terminal tracking unit 714 from the infrastructure image specified by the image specifying unit 726, and a vehicle/terminal information storage unit 716. An error calculation unit 728 for calculating an error between the stored position of the vehicle specified by the image specifying unit 726 at the time when the image was taken and the position of the vehicle estimated by the vehicle position estimation unit 724. including.

Vehicle position estimator 724 includes a two-stage neural network. The first-stage neural network is for recognizing objects in an image, including attributes such as position, color, size, and object type in the image. Such techniques already exist and can estimate multiple objects, including their types, from a single image. The vehicle position estimator 724 further includes a second neural network that recognizes the position in the image of an object that is of a vehicle type among the objects recognized by the first-stage neural network. This second neural network replaces other objects in the image with images normally captured by the infra-camera, and the background image normally captured by the infra-camera contains only the image of the target vehicle. It takes the placed image as input and estimates the geographical position of the vehicle.

For training of this second neural network, an image in which a vehicle is arranged in a background image is used as an input, and training data is used in which the geographical position of the vehicle at that time is used as training data. The vehicle does not need to be actually placed at that position, but the background image is an image of the vehicle, with the size, shape, and pose calculated based on its assumed vehicle position. It can be used as an input. The same applies to the color of the vehicle, and by image processing, a plurality of images of the same image, differing only in the color of the vehicle, can be created and used as input data for training. It is assumed that the background of the image captured by the infrastructure camera is always the same. Therefore, if training data is prepared for each infrastructure camera, the training data will not become sparse, and recognition accuracy can be increased. Therefore, the vehicle position estimating unit 724 prepares a trained one for each infrastructure camera, and uses a neural network dedicated to the infrastructure camera that captures the image specified by the image specifying unit 726 as the vehicle position estimating unit 724 .

The edge server 180 further generates information to be distributed to each vehicle stored in the vehicle/terminal information storage unit 716 from the driving support information generated by the driving support information generation unit 734, or generates information to be distributed to each vehicle stored in the vehicle/terminal information storage unit 716. Based on the error, among the vehicles stored in the vehicle/terminal information storage unit 716, positional error correction information for transmission to the vehicle corresponding to the positional information received by the vehicle/terminal tracking unit 714 is generated. and a transmission processing unit 732 for transmitting the driving support information and the error correction information generated by the information generation unit 730 to respective destinations (in-vehicle devices). The information about the transmission destination is generated by the information generation unit 730 based on the vehicle address stored in the vehicle/terminal information storage unit 716 and attached to each piece of information.

FIG. 13 is a flow chart showing the control structure of a program for realizing each function shown in FIG. Referring to FIG. 13, this program is activated each time edge server 180 receives information. The program executes step 750 for determining whether the received information is from the vehicle and branching the control flow depending on the result, and in response to a positive determination at step 750, Step 754 of identifying the identification information of the vehicle; Step 756 of updating the tracking information, analysis information and attribute information of the identified target vehicle based on the received location information; Step 758 for judging whether or not the position information is from the target vehicle and branching the control flow; Based on the position information of each vehicle stored in the vehicle/terminal information storage unit 716 of FIG. and step 762 for determining whether it exists and branching control flow accordingly. When the determination result of step 762 is affirmative, the execution of this program is terminated without correcting the vehicle position information.

Further, when the determination in step 762 is negative, that is, when it is determined that there is no vehicle of the same type in the vicinity of the vehicle, the vehicle transmits the image. , Recognizing the image including the image of the vehicle and the geographical position of the vehicle in the image captured by the image from the infrastructure camera existing in the vicinity of the position information received from the vehicle, and the vehicle image in the image a step 764 of reading out from the vehicle/terminal information storage unit 716 the location information of the vehicle that was received by the edge server 180 at the time when the image identified in step 764 was captured; Step 768 calculates the error of the position information of the vehicle based on the difference between the position information received from the vehicle read out in step 766 and the position information obtained from the image. and a step 770 of transmitting the information as position correction information to the in-vehicle device of the next vehicle (the vehicle that transmitted the information when this program was started) and ending the execution of this program.

The program further processes the received information in response to a negative determination at step 750, i.e., that the received information is not from the vehicle, according to the type of information, and the program includes a step 752 that terminates execution of . The execution of this program is also ended when the determination at step 758 is negative, that is, when it is determined that the information is not from the vehicle being tracked for correcting the position information.

At step 764, a neural network is used to estimate the position of the vehicle based on the infrastructure camera image. This neural network is pre-trained for each infrastructure camera. For the training, training data is used in which the image taken from the camera is used as input data and the actually measured geographical position of the vehicle existing in the image is used as training data. As the vehicle, various images such as a large number of vehicle types, vehicle types, and colors are used. Vehicle positions used as training data should be determined in a consistent manner. A similar method is also used when estimating the vehicle position.

Note that in this disclosure, the neural network model described above is trained for each infrastructure camera. A convolutional neural network is suitable in this case as the neural network. The image captured by the infrastructure camera is basically the same and does not change at any time. Therefore, there is little possibility that the teacher data for specifying the position of the vehicle existing in the same background will be sparse, and there is no possibility that an impossibly large number of images will be prepared as teacher data. It is not necessary to place the actual vehicle within the imaging range of the infrastructure camera as the image of the training data. For example, only one image (or a plurality of images depending on the weather) that does not include the vehicle, which is captured by the infrastructure camera, may be prepared, and an image of the vehicle captured separately may be arranged there. In this case, the direction, the position, and the size of the vehicle image are determined in advance using a conversion formula, and the images are synthesized. However, it is desirable to specify vehicle positions with a certain degree of granularity in order to prevent teacher data from becoming sparse. For example, a map may be divided into 30 cm square rectangles, and information indicating in which of these rectangles the center of the vehicle is located may be used as teacher data. If abundant training data is available, this granularity can be increased to segment the map into smaller rectangles, such as 20 cm square, or 15 cm square, and the vehicle location can be specified at the location of the rectangle. . That is, the amount of training data for the model can be controlled by limiting the accuracy of the vehicle position measurement to the side length of the rectangle.

In principle, the neural network described above may be learned for all infrastructure cameras instead of for each infrastructure camera. However, since the background differs depending on the infrastructure camera, it is difficult to sufficiently prepare training data, and it takes a long time to learn the neural network.

<motion>
The position correction system 150 having the configuration described above operates as follows. First, the vehicle position estimation unit 724 of the edge server 180 shown in FIG. 12 is trained for each infrastructure camera by the method described above, and a neural network model of the vehicle position estimation unit 724 is prepared for each infrastructure camera.

Normally, HMI controller 232 of each vehicle shown in FIG. 7 receives sensor data from each sensor and stores it in RAM 376 or auxiliary storage device 378 . The HMI controller 232 periodically executes the program shown in FIG. It is transmitted wirelessly to the edge server 180 shown in FIG.

Referring to FIG. 12, upon receiving this information, reception processing unit 700 of edge server 180 uses the vehicle information to determine whether or not the vehicle has not been managed until now. A new vehicle entry is generated in the moving object tracking unit 702, the attribute detection unit 706, and the vehicle/terminal tracking unit 714, and the same identification number is given. The reception processing unit 700 gives the position information to the vehicle/terminal tracking unit 714 , and the vehicle/terminal tracking unit 714 stores them in the vehicle/terminal information storage unit 716 . Or, when new vehicle information for the same vehicle is received, the previous vehicle record is updated with the new record.

When the reception processing unit 700 receives sensor information from a vehicle or the like, the moving object tracking unit 702 gives the sensor information to the moving object tracking unit 702 and the attribute detection unit 706 . The moving object tracking unit 702 recognizes the existence of moving objects such as vehicles existing within the management range from the sensor information, and generates information regarding the position, moving direction, moving speed, etc. of each moving object. These pieces of information are stored in the analysis result storage unit 704 .

On the other hand, the attribute detection unit 706 detects attributes such as color, size, type (pedestrian/adult, pedestrian/child, vehicle (large vehicle, small vehicle, motorcycle, bicycle), etc. of a moving object based on the given sensor data. Detected and stored for each moving object in the attribute storage unit 708. An integration processing unit 710 integrates this information and the attributes of each moving object maintained in the analysis result storage unit 704 to create a traffic situation bird's-eye view map. , is stored in the analysis result storage unit 712 .

Further, when the reception processing unit 700 receives position information from a vehicle or the like, the reception processing unit 700 gives the information to the vehicle/terminal tracking unit 714 . The vehicle/terminal tracking unit 714 stores vehicle information of vehicles existing within the management range, and periodically grasps, maintains and manages the current position of the vehicle based on its update time, position information, moving speed, moving direction, and the like.

The driving support information generation unit 734 periodically reads the traffic situation overview map from the analysis result storage unit 712 , generates driving support information by a known method, and supplies the driving support information generation unit 734 with the driving support information. Upon receiving this driving assistance information, the driving assistance information generation unit 734 determines the address of the vehicle to which the driving assistance information is to be transmitted based on the position information stored in the vehicle/terminal information storage unit 716, and sends the information to the transmission processing unit 732. Send driving assistance information to that address via

Upon receiving an infrastructure image, the reception processing unit 700 gives this image to the moving object tracking unit 702 , the attribute detection unit 706 , and the infrastructure image reception unit 720 . The infrastructure image reception unit 720 stores this image in the infrastructure image storage unit 721 .

On the other hand, when the position information is received, the same vehicle model presence/absence determination unit 722 determines whether or not the vehicle that transmitted the information is a tracking target for correcting the position information (step 758 in FIG. 13). Furthermore, if it is a tracking object, whether or not a vehicle of the same model exists within a predetermined range from that position is checked by the same model presence/absence determination unit 722 based on the information in the vehicle/terminal information storage unit 716 . (step 760). If there is such a vehicle (YES at step 762), then this disclosure does not correct the position information. If no other such vehicle exists (NO at step 762), the processes of steps 764, 766, 768 and 770 of FIG. 13 are executed in this order, and position correction information is transmitted to the vehicle in question.

Referring to FIG. 9, the vehicle that has received this correction information executes the processing of step 508 through the route of step 500→step 504 in FIG. That is, the received position correction information is stored in the RAM 376 or the auxiliary storage device 378 (see FIG. 7). Hereinafter, this position correction information is added to the position information calculated by dead reckoning. When the number of positioning satellites with which GPS can communicate becomes three or more, the position correction information is cleared.

As described above, according to this disclosure, when position information is received from a vehicle, it is determined whether or not the vehicle is being tracked for position correction. If it is the object, by confirming that there is no vehicle of the same model in the vicinity, the infrastructure image including the image of the vehicle captured when the vehicle passed the specific point is specified. (step 764). From that image, the exact position of the vehicle is deduced. That is, the vehicle included in the image and the image of the tracked object are matched. The error between the position estimated in this way and the position information transmitted from the vehicle at the time the image was captured is transmitted to the vehicle as correction information for the position error. The vehicle stores this correction information, and thereafter, every time position information is output by dead reckoning, this error is added to the position information to correct the position information by dead reckoning. When the number of positioning satellites with which the vehicle's GPS can communicate becomes three or more, this correction information is cleared, and the GPS output is adopted as the position information.

Therefore, it is possible to correct the position of the vehicle whose positional information is degraded by using infrastructure equipment that is already in widespread use, such as an infrastructure camera. It is not necessary to arrange many facilities like the optical beacon of Patent Document 1, and it is not necessary to use the image of the vehicle-mounted camera as in Patent Document 2. The background of the image of the infrastructure camera is basically always the same as compared with the image of the vehicle-mounted camera, and the detection of the vehicle and the position can also be performed with high accuracy.

An image of the vehicle and its position in the image captured by the infrastructure camera are recognized. From the vehicle information about the vehicle managed by the vehicle/terminal tracking unit 714, the position information transmitted from the vehicle when the image was captured and stored in the vehicle/terminal tracking unit 714 is read. The two positions are compared and the error between them is calculated and transmitted as the vehicle position error information.

It should be noted that this disclosure is mainly intended to solve the problem when a vehicle travels in an area of high-rise buildings as shown in FIG. But these problems are not limited to high-rise buildings. For example, a similar problem can arise in a tunnel or the like. Infracameras are often installed in tunnels as well, so the position correction system according to this disclosure can be used. The approximate positional error in dead reckoning is a median of several meters in open skies, with a maximum value of about 10 meters. Dozens of meters, the maximum value is one hundred and several tens of meters. These errors are quite large.

According to experiments, the error in position detection by the system disclosed above using infrastructure cameras can be kept within several meters. The time required for analysis can be set to several tens of milliseconds for position detection and less than one second for vehicle type identification. Also, the image of the infrastructure camera can be zoomed in. By zooming in and narrowing down the imaging range, the accuracy can be further increased. Note that if the vehicle can be specified, the position can be corrected even after the vehicle has passed the imaging area of the infrastructure camera.

-Modified example-
In the system of the first disclosure, the same vehicle model presence/absence determination unit 722 shown in FIG. It is determined whether or not If such other vehicle exists, the position information correction process based on the received position information is not executed. Since it is rare for two or more vehicles of exactly the same model to be very close to each other at the same time, it is considered that such a method according to the disclosure does not pose any particular problem.

However, although it is rare, such a situation cannot be ruled out. For example, FIG. 14 shows a case where two vehicles 780 and 782 of exactly the same model are close to each other with a distance L between their centers. Further, if the error range of the position information of the vehicle 780 is a circle 784 with radius R centered on the vehicle 780, and the error range of the position information of the vehicle 782 is a circle 786 centered on the vehicle 782 and overlapping the circle 784, then the position received from the vehicle 780 is Even if the information is really the position of the vehicle 780, it cannot be determined whether the vehicle is the vehicle 780 from the image. If two vehicles exist close to each other in this way and both are of the same model, there are many cases where these two vehicles also exist in the infrastructure image. Even if only one of these vehicles can be recognized in the infrastructure image, it cannot be determined from the image or from the position information which of the two vehicles the vehicle is. Therefore, in this case, the position information is not corrected in the first disclosure.

However, even in such a case, depending on the conditions, it may be possible to identify the vehicle that has transmitted the position information. FIG. 15 is a schematic diagram showing a method for solving the problem of vehicle identification when two vehicles of exactly the same model are close to each other. Referring to FIG. 15, it is assumed that both vehicles 780 and 782 are equipped with in-vehicle cameras that capture forward images. Since both are the same model, it is highly likely that these conditions will be satisfied. In this case, since the vehicle 780 exists in the imaging area 790 of the vehicle-mounted camera of the vehicle 782, there should be an image of the vehicle 780 in the image. On the other hand, an image of a vehicle of the same type as the vehicles 780 and 782 should not be captured in the imaging area 788 of the vehicle-mounted camera of the vehicle 780 .

Therefore, as described above, when two vehicles of the same model exist in close proximity, and if these vehicles are equipped with an on-vehicle camera that captures an image of the front, the image is sent to the edge server 180. Instruct those vehicles to transmit. The vehicle in the received image is image-recognized, and if a vehicle of the same type as the vehicle that sent the position information is in the image, the vehicle that sent the position information is the vehicle 782 behind the two. I understand. On the other hand, if such a vehicle is not shown, it can be determined that the vehicle is the leading vehicle 780 . Therefore, the geographical position of the image of the leading vehicle in the infrastructure image is recognized in the same manner as in the first disclosure, and is compared with the positional information of the leading vehicle to generate correction information.

FIG. 16 is a schematic diagram for explaining a problem regarding vehicle identification when three vehicles of exactly the same model are close to each other. In this example, in addition to vehicles 780 and 782, a vehicle 792 of exactly the same type is running in a lane adjacent to the lane of vehicle 780 and the like. Even in such a case, the vehicle that has transmitted the position information can be identified under certain conditions.

For example, in this case as well, if these three vehicles have on-board cameras that capture forward images, these vehicles are instructed to transmit them. The onboard camera images received from these vehicles can be classified into those containing images of vehicles of the same type and those not containing images. For example, vehicles of the same model do not exist in the imaging area 796 of the vehicle-mounted camera of the vehicle 792 . Therefore, there should be no image of the same vehicle as these in the image. The same is true for vehicle 780 with imaging area 788 . However, in the case of a vehicle 782 having an imaging area 790, since the vehicle 780 exists in the imaging area 790, the image of the vehicle 780 should be included in the front captured image.

If the image from the vehicle that has transmitted the position information includes the image of the same vehicle type, and the images from the other two vehicles do not include the image of the same vehicle, this condition is satisfied in FIG. , only when the vehicle that has transmitted the position information is the vehicle 782 . Therefore, in this case, vehicle 782 can be identified as the vehicle that has transmitted the position information.

On the other hand, when the image from the vehicle that has sent the position information does not include the image of the same model of vehicle, and the image from one of the other two vehicles does not include the image of the same model of vehicle. , it cannot be determined whether the vehicle that has transmitted the position information is the vehicle 780 or the vehicle 792 . Therefore, in this case, the vehicle that has transmitted the position information cannot be identified.

Similarly, when the image from the vehicle that has transmitted the position information includes the image of the same type of vehicle, and the image from one of the other vehicles also includes the image of the same type of vehicle, Vehicles that have sent location information cannot be identified. In addition, if the image from the vehicle that has sent the position information does not include the image of the same model of vehicle, and the images from the other two vehicles both include the image of the same model of vehicle, It can be determined that a vehicle that has transmitted an image that does not include an image of the vehicle is the vehicle that has transmitted the position information.

According to this idea, the number of images of vehicles of the same model included in the image from the vehicle that has transmitted the position information is different from the number of images of the same model of vehicle included in the images from other vehicles. If it is, the vehicle that has sent the position information can be identified, otherwise it cannot be identified.

In this disclosure, according to this idea, even if a vehicle of the same type exists in the vicinity of the vehicle that has transmitted the position information, when a specific condition is satisfied, the vehicle that has transmitted the position information is identified, Correct the position information of the vehicle.

FIG. 17 is a flow chart showing the control structure of the program executed by the edge server according to this disclosure. Referring to FIG. 17, this program includes the same steps 750, 752, 754, 756, 758 and 760 as shown in FIG. Unlike FIG. 13, this program follows step 760 and determines whether or not there is only one vehicle of the same type in the vicinity. Step 832 of acquiring front camera images from all of the vehicles of the same type, Step 834 of recognizing the vehicles in each image including the type of vehicle acquired in Step 832, Step 834 followed by position information determines whether or not the number of vehicles of the same model in the front image from the vehicle that has transmitted the is different from the number of vehicles of the same model in the other images, and branches the control flow according to the determination. and step 836 . If the determination result of step 836 is negative, execution of this program ends. That is, the position information of the vehicle that has transmitted the position information is not corrected.

On the other hand, when the determination in step 836 is affirmative, step 763 is executed to specify the position of the vehicle in the group of the same model based on the number of vehicles of the same model included in the image from the vehicle that has transmitted the position information. , and the same steps 764, 766, 768, and 770 as shown in FIG. 13 are executed, and execution of this program ends. When the determination at step 830 is affirmative, the processing from step 764 onwards is similarly executed. That is, the vehicle in the image and the tracked object are matched. When the vehicle of the same model included in the front image from the vehicle that has transmitted the position information is unique, for example, depending on the number of the vehicles, it can be determined that the vehicle is at a specific position (often the head or the tail) within a group of the same model. I understand. As a result, it is possible to identify the location of a vehicle within the infrastructure image that meets these conditions, and use that information to correct the location information for that vehicle.

As described above, in this modified example, position information is transmitted only when the forward image from the vehicle that has transmitted the position information satisfies a specific condition (the number of vehicles of the same vehicle type in the image is unique). Vehicles are identified in infrastructure images and their position information is corrected. In other cases, correction of position information is not performed.

Even in such a rare case that a plurality of vehicles of the same type are close to each other, if a certain condition is satisfied, the vehicle to be tracked for correction can be identified and its position information can be corrected.

<Second Disclosure>
In the position correction system 150 according to the first disclosure and its modification described above, the correction of the position information of the vehicle is started when the position information is received from the vehicle. In correcting the positional information, an image is specified when the vehicle transmitting the positional information is traveling at a specific position, and the position of the vehicle is estimated from the image by a neural network. Correction information is generated based on the error between the estimated position and the position transmitted by the vehicle when the image was taken. Image recognition using a neural network is used to identify the target vehicle in the image, and if multiple vehicles are recognized, processing is performed to select one that satisfies specific conditions. However, the invention is not so limited. Even with the same image recognition, if the license plate of the vehicle can be image-recognized, even if there are a plurality of vehicles of the same model in the image, the vehicle that has transmitted the position information can be easily identified.

FIG. 18 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the second disclosure. In the flowchart shown in FIG. 18, instead of step 760 in FIG. and step 902 of identifying an infrastructure image at the time when the vehicle having the license plate of the vehicle passes through the specific point, among the infrastructure images from the infrastructure cameras. After step 902, steps 766, 768 and 770 of FIG. 13 in the infrastructure image are executed and execution of this program ends.

If the license plate of the vehicle can be recognized from the image, it is easy to identify the target vehicle. Therefore, it is not necessary to perform complex matching for vehicle identification as in the variant of the first disclosure.

<Third Disclosure>
FIG. 19 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the third disclosure. The systems disclosed in the first and second disclosures start the process of correcting the position information when the position information is received from the target vehicle. However, the trigger for starting correction processing is not limited to reception of position information. For example, correction processing may be performed periodically for each of the target vehicles. This third disclosure is such an example. As a hardware configuration, a server similar to that disclosed in the first disclosure shown in FIGS. 10 and 11 can be used.

Referring to FIG. 19, the program for realizing the third disclosure is for each vehicle to be tracked for correction stored in vehicle/terminal information storage unit 716 shown in FIG. and includes a step 940 of performing the following process 942 .

A process 942 determines whether or not an infrastructure camera exists near the current vehicle based on the position information of each vehicle stored in the vehicle/terminal information storage unit 716 for the target vehicle being processed. includes a step 950 for ending the process 942 and starting the process for the next target vehicle if the infrastructure camera does not exist.

The program further includes step 952 of acquiring an image of an infrastructure camera present near the target vehicle when the determination in step 950 is affirmative, step 954 of recognizing the vehicle in the acquired image, and A step 956 of determining whether or not the target vehicle being processed exists among the vehicles, and ending the execution of the processing 942 if it does not exist; Step 958 of specifying the position of the target vehicle using a neural network similar to the vehicle position estimation unit 724 shown in FIG. and a step 960 of calculating the position of the vehicle based on the history of the position of the vehicle stored in the vehicle/terminal information storage unit 716 .

The program further includes a step 962 for calculating an error between the position of the target vehicle identified in step 958 and the position of the target vehicle calculated in step 960, and a step 962 for processing the calculated error as correction information. and a step 964 of transmitting to the target vehicle and terminating execution of process 942 .

By repeatedly executing the processing shown in FIG. 19, the position of the target vehicle whose position information is low in accuracy can be corrected using the infrastructure image, as in the first and second disclosures.

<Fourth Disclosure>
FIG. 20 is a flow chart showing the control structure of a program that implements the functions of the edge server according to the fourth disclosure. The system according to the fourth disclosure is triggered by receiving the image from the infrastructure image, and corrects the position information of the vehicle based on the image of the vehicle in the image.

Referring to FIG. 20, this program executes the following processing when infrastructure image reception unit 720 shown in FIG. 12 receives an infrastructure image from the infrastructure camera.

The program is activated in response to receiving an infrastructure image and performs step 980 to recognize vehicle images in the infrastructure image and, following step 980, perform the following process 984 for each recognized vehicle. including step 982;

A process 948 determines whether or not the vehicle being processed is a target vehicle for position correction based on the information stored in the vehicle/terminal information storage unit 716 (FIG. 12). It includes a step 990 to end and the same steps 958, 960, 962 and 964 as shown in FIG. 19, which are performed when the determination of step 990 is affirmative.

The vehicle recognition performed at step 990 uses a different method than the first disclosure. When the infrastructure image includes images of a plurality of vehicles, the background image of the infrastructure camera replaces all images other than the image of the vehicle to be determined here, as in the first disclosure. However, in this case, an image to be judged and all position information and attribute information of the vehicle to be judged are input, and a correct/incorrect judgment is made to indicate whether or not the image is the image of the vehicle to be judged. A convolutional neural network trained with two labels as teacher data is used for judgment. By adopting the SoftMax function in the final layer of this neural network, this neural network receives an image of a vehicle included in the image for an input including an image and target vehicle information. It indicates the probability that the image is a vehicle image. All target vehicles stored in the vehicle/terminal information storage unit 716 are judged by this neural network. It is determined that the image of the vehicle at is that of the input target vehicle.

In this fourth disclosure, in response to receiving an infrastructure image, images of all vehicles included in the image are recognized, and for each vehicle image, an object stored in the vehicle/terminal information storage unit 716 is recognized. It is determined whether or not it is an image. By performing such a determination, it is possible to determine whether or not the vehicle in the infrastructure image is the target vehicle. The position can be compared to calculate the error.

In any of the above disclosures, the images used are those captured by infrastructure cameras. The background of the image captured by the infrastructure camera may always be regarded as the same in the short term. Therefore, compared to recognizing landmarks using an in-vehicle camera, when recognizing a vehicle in an image using a neural network, for example, the training data does not become sparse, and image recognition using a neural network does not become sparse. Accuracy can be improved. In addition, there is an effect that there is no need to newly provide a special facility such as an optical beacon because an infrastructure camera, which is already in widespread use, is used. 15 and 16 show an example of using a front image for vehicle matching, but this disclosure is not limited to using a front image. For example, a rear image may be used.

The present disclosure is merely illustrative, and this disclosure is not limited to the disclosures set forth above. The scope of this disclosure is indicated by each claim after taking into consideration the description of the detailed description of the invention, and all changes within the meaning and range of equivalents to the wording described therein include.

50, 100, 150 position correction system 60 optical beacon 62 pole 64, 116, 176, 178, 200, 780, 782, 792 vehicle 66, 68 beacon data 70 in-vehicle communication device 72 camera 110 server 112 landmark position information storage device 114 Landmark 160 Infrastructure camera 162, 164, 166, 168, 170, 172 Skyscraper 174 Road 180 Edge server 202 Paint 204 Distance 210 In-vehicle device 212 Millimeter wave radar 214 In-vehicle camera 216 LiDAR
230 External communication controller 232 HMI controller 234 Automatic operation controller 236 Driving system controller 240 Integrated antennas 242, 642 Monitors 244, 246, 250, 252, 254, 256 ECU
248 Autonomous driving ECU
350, 640 Computer 352 Touch panel 356 Various ECUs
358 Various sensors 360 Speaker/microphone 362, 664 USB memory 370, 656 CPU
372, 666 Bus 374, 658 ROM
376, 660 RAM
378 auxiliary storage device 380 wireless communication unit 382 input/output I/F
384, 652 USB memory port 390 Audio processing I/F
392, 644 network I/F
394 In-vehicle LAN
400, 402, 500, 502, 504, 508, 512, 750, 752, 754, 756, 758, 760, 762, 763, 764, 766, 768, 770, 830, 832, 834, 836, 900, 902, 940, 950, 952, 954, 956, 958, 960, 962, 964, 980, 982, 990 Step 646 Keyboard 648 Mouse 650 DVD drive 654 HDD
657 GPUs
662 DVDs
668 network 700 reception processing unit 702 moving body tracking units 704, 712 analysis result storage unit 706 attribute detection unit 708 attribute storage unit 710 integration processing unit 714 vehicle/terminal tracking unit 716 vehicle/terminal information storage unit 718 tracking target determination unit 720 infrastructure image Reception unit 721 Infrastructure image storage unit 722 Same vehicle model presence/absence determination unit 724 Vehicle position estimation unit 726 Image identification unit 728 Error calculation unit 730 Information generation unit 732 Transmission processing unit 734 Driving support information generation unit 784, 786 Circles 788, 790, 796 Imaging Areas 942, 984 Processing

Claims (23)

a mobile information receiver that receives mobile information including location information and time information of the mobile from the mobile;
an image receiver that receives a captured image from an imaging device whose imaging range can be specified;
Based on an image of an arbitrary moving object existing in the picked-up image and the information on the moving object, a captured image containing the image of the arbitrary moving object and the information on the moving object are transmitted to the receiver for information on the moving object. a matching unit that matches the moving object;
The position information maintained by the moving object is determined by using the position information of the moving object information received from the moving object matched by the matching unit and the captured image matched by the matching unit and its image capturing time. A position correction server including a correction processing unit that performs correction processing,
further comprising a mobile object tracking device for tracking a mobile object whose position information is to be corrected based on the mobile object information;
The matching unit responds to the fact that the moving body information is moving body information from a first moving body being tracked by the moving body tracking device, and outputs a captured image including an image of the first moving body. from the captured image received by the image receiver,
The correction processing unit is transmitted from the first moving object at the image capturing time of the captured image extracted by the image extracting unit and is calculated from the captured image and the position information received by the moving object information receiver. a position correction processing unit that performs a process of correcting the position information maintained by the first moving body based on the position of the first moving body. The position correction processing unit
a position storage device that stores the positions of real-world features;
a feature recognition unit that recognizes features present in the captured image;
a reading device for reading the real-world position of the feature recognized by the feature recognition unit from the position storage device;
2. The position according to claim 1 , further comprising a calculation device for calculating a correction amount for correcting said position information maintained by said first moving body based on said position read from said position storage device. compensation server. The calculation device is
Based on the relationship between the image of the characteristic object present in the captured image and the image of the moving object, the geographical position of the characteristic object, and the characteristics of the imaging device, the moving object and the characteristic a relative position calculation device that calculates a relative position with respect to an object;
a geographical position calculation device for calculating the geographical position of the moving object by adding the geographical position of the feature to the relative position;
a correction amount calculation device for calculating, as the correction amount, an error between the position information maintained by the first moving body and the geographical position calculated by the geographical position calculation device; The position correction server according to claim 2 . The matching unit
in response to the presence of a second mobile object of the same type as the first mobile object within a predetermined distance from the first mobile object, the first mobile object and the second mobile object; a front image acquisition device that acquires each front image from the moving body of
For each of the front images acquired by the front image acquisition device,
an image inspection unit that inspects whether or not an image of a moving object of the same type as the first moving object exists;
Among the front images, if there is only one front image in which an image of the same type of moving object as the first moving object does not exist, the moving object that transmitted the front image and the captured image are present. 2. The location correction server of claim 1 , comprising a device for matching a . a mobile information receiver that receives mobile information including location information and time information of the mobile from the mobile;
an image receiver that receives a captured image from an imaging device whose imaging range can be specified;
Based on an image of an arbitrary moving object existing in the picked-up image and the information on the moving object, a captured image containing the image of the arbitrary moving object and the information on the moving object are transmitted to the receiver for information on the moving object. a matching unit that matches the moving object;
The position information maintained by the moving object is determined by using the position information of the moving object information received from the moving object matched by the matching unit and the captured image matched by the matching unit and its image capturing time. A position correction server including a correction processing unit that performs correction processing,
further comprising a mobile object tracking device that tracks a mobile object whose position information is to be corrected based on the mobile object information;
the moving body information includes appearance information relating to the appearance of the moving body;
The matching unit
For each moving object to be tracked by the moving object tracking device,
an image extraction unit that identifies an imaging device that is present in the vicinity of the moving object and whose imaging area can be specified, and extracts an image captured by the identified imaging device from the captured image received by the image receiver;
identifying an image of the moving body in the captured image extracted by the image extracting unit using the appearance information, and identifying the moving body in the captured image based on the image and the appearance information; A location correction server, comprising: a mobile identification device for The correction processing unit is
Based on the position of the feature within the imaging area of the captured image extracted by the image extracting unit and the position of the image of the moving body in the captured image, the geography of the moving body in the captured image a moving body position specifying unit that specifies a target position;
Based on the geographical position of the mobile body identified by the mobile body position identifying unit and the position information transmitted from the mobile body at the time when the captured image was captured and received by the mobile body information receiver 6. The position correction server according to claim 5 , further comprising a device for correcting said position information maintained by said moving body. The apparatus for performing the correcting process includes:
Geographical position determined by the geographical position of the mobile body specified by the mobile body position specifying unit and the position information transmitted from the mobile body at the time when the captured image was captured and received by the mobile body information receiver an error calculator for calculating the error between the position and
7. The position correction server according to claim 6 , further comprising a transmitting device that transmits an instruction to add the error calculated by the error calculating device to the position information maintained by the mobile object to the mobile object. a mobile information receiver that receives mobile information including location information and time information of the mobile from the mobile;
an image receiver that receives a captured image from an imaging device whose imaging range can be specified;
Based on an image of an arbitrary moving object existing in the picked-up image and the information on the moving object, a captured image containing the image of the arbitrary moving object and the information on the moving object are transmitted to the receiver for information on the moving object. a matching unit that matches the moving object;
The position information maintained by the moving object is determined by using the position information of the moving object information received from the moving object matched by the matching unit and the captured image matched by the matching unit and its image capturing time. A position correction server including a correction processing unit that performs correction processing,
further comprising a mobile object tracking device that tracks a mobile object whose position information is to be corrected based on the mobile object information;
the moving body information includes appearance information relating to the appearance of the moving body;
The matching unit
A moving body image recognition unit that recognizes an image of a moving body being tracked by the moving body tracking device that exists in the captured image in response to the image receiver receiving the captured image. and,
An identification device for identifying the moving object in the captured image among the moving objects tracked by the moving object tracking device, using the image recognized by the moving object image recognition unit and the appearance information. and
The correction processing unit is
for the moving body recognized by the moving body image recognition unit, based on the position of the image of the feature in the captured image, the position of the image of the moving body, and the geographical position of the feature a moving body position calculation device for calculating the position of the moving body at the time when the captured image was captured;
the moving body based on the position calculated by the moving body position calculation device and the position information received by the moving body information receiver from the moving body when the captured image in which the image of the moving body is recognized is captured; a correction execution unit for executing a process of correcting the position information maintained by the position correction server. The mobile information further includes accuracy-related information related to the accuracy of location information maintained by the mobile,
Based on the accuracy-related information, the moving object tracking device tracks a moving object having a position information acquisition device that is likely to operate with an accuracy lower than a predetermined accuracy as a target for position information correction. A location correction server according to any one of claims 1 to 8 , comprising a tracking device. The accuracy-related information includes the number of communicable positioning satellites in the satellite positioning system used by the mobile,
10. The location correction server according to claim 9 , wherein said tracking device includes a device for tracking a mobile with said number of positioning satellites with which said mobile can communicate is equal to or less than a threshold. 11. The location correction server of Claim 10 , wherein the satellite positioning system includes GPS. Furthermore, in response to accuracy-related information related to the accuracy of location information maintained by a mobile object changing to a value indicating that the accuracy of the location information has become higher than the predetermined accuracy, the mobile object is 12. The position correction server according to any one of claims 9 to 11 , further comprising an exclusion device for executing a process of excluding from a target of position information correction by said correction. The moving body information further includes appearance information about the appearance of the moving body,
The matching unit includes an image recognition unit that recognizes an image of the moving body included in the captured image based on the appearance information included in the moving body information;
a device for associating the captured image in which the image of the moving body is recognized by the image recognition unit with the moving body that transmits the moving body information to the moving body information receiver;
The position correction server according to any one of claims 1 to 4 . an information maintenance device that maintains mobile information including its own location information and time information;
a mobile body information transmitting device for transmitting the mobile body information to a predetermined position correction server at a predetermined timing;
a correction processing unit that, in response to externally receiving correction information indicating that the position information maintained by the information maintenance device should be corrected, corrects the position information using the correction information; A location management device for a mobile object , comprising:
The information maintenance device has a function of maintaining the position information by a satellite positioning system,
The mobile object information transmission device is a position management device, wherein when the mobile object information is transmitted to the position correction server, the mobile object information includes the number of positioning satellites with which the satellite positioning system can communicate. the correction information includes an error between the position information maintained by the information maintenance device and correct position information;
15. The position management device according to claim 14 , wherein said correction processing unit includes an adder for adding said error to said position information maintained by said information maintenance device. Further, claim 14 or claim 1, further comprising a stop device for stopping the operation of the correction processing unit in response to the number of positioning satellites with which the satellite positioning system can communicate reaches or exceeds a predetermined number. 5. The position management device according to 5 . A position management system for a moving body, comprising the position correction server according to any one of claims 1 to 13 and the position management device according to claim 14 or claim 15 . a step in which the computer receives mobile information including location information and time information from the mobile;
a step in which a computer receives a captured image from an imaging device whose imaging range can be specified;
In the step of receiving a captured image including the image of the arbitrary moving body and the moving body information based on the moving body information and the image of the moving body existing in the captured image, the computer receives the moving body information. a step of matching with a mobile body that transmits body information;
The computer maintains the moving body using the position information of the moving body information received from the matching moving body in the matching step and the captured image matched in the matching step and its capturing time. A method for correcting position information, comprising:
Furthermore, the computer includes a moving body tracking step of tracking a moving body whose position information is to be corrected based on the moving body information,
In the matching step, the computer, in response to the mobile information being mobile information from a first mobile being tracked by the mobile tracking step, generates an image of the first mobile. An image extraction step of extracting a captured image containing from the captured image received in the image receiving step,
In the step of performing the correcting process, the computer transmits position information from the first mobile object at the imaging time of the captured image extracted in the image extraction step and received by the mobile object information receiver; and performing a process of correcting the position information maintained by the first moving body based on the position of the first moving body calculated from the captured image. the computer,
a mobile information receiver that receives mobile information including location information and time information of the mobile from the mobile;
an image receiver that receives a captured image from an imaging device whose imaging range can be specified;
Based on an image of an arbitrary moving object existing in the picked-up image and the information on the moving object, a captured image containing the image of the arbitrary moving object and the information on the moving object are transmitted to the receiver for information on the moving object. a matching unit that matches the moving object;
The position information maintained by the moving object is determined by using the position information of the moving object information received from the moving object matched by the matching unit and the captured image matched by the matching unit and its image capturing time. A computer program that functions as a correction processing unit that performs correction processing ,
Further, causing the computer to function as a mobile object tracking device for tracking a mobile object whose position information is to be corrected based on the mobile object information,
The matching unit responds to the fact that the moving body information is moving body information from a first moving body being tracked by the moving body tracking device, and outputs a captured image including an image of the first moving body. from the captured image received by the image receiver,
The correction processing unit is transmitted from the first moving object at the image capturing time of the captured image extracted by the image extracting unit and is calculated from the captured image and the position information received by the moving object information receiver. a position correction processing unit that performs a process of correcting the position information maintained by the first moving body based on the position of the first moving body. a step in which the computer maintains mobile information including its own location information and time information;
a step in which the computer transmits the moving body information to a predetermined position correction server at a predetermined timing;
In response to the computer externally receiving correction information indicating that the position information maintained by the step of maintaining the moving object information should be corrected, the position information is corrected using the correction information. A position management method for a moving object, comprising:
maintaining the mobile information includes, by a computer, maintaining the location information by a satellite positioning system;
The transmitting step includes transmitting the mobile information including the number of positioning satellites with which the satellite positioning system can communicate to the predetermined location correction server at the predetermined timing. . the computer,
an information maintenance device that maintains mobile information including its own location information and time information;
a mobile body information transmitting device for transmitting the mobile body information to a predetermined position correction server at a predetermined timing;
Functioning as a correction processing unit that corrects the position information using the correction information in response to externally receiving correction information indicating that the position information maintained by the information maintenance device should be corrected. A computer program that causes
The information maintenance device has a function of maintaining the position information by a satellite positioning system,
A computer program according to claim 1, wherein the mobile object information transmitting device transmits the mobile object information including the number of positioning satellites with which the satellite positioning system can communicate when transmitting the mobile object information to the position correction server. An in-vehicle device equipped with the position management device for a moving object according to claim 14 . A vehicle equipped with the in-vehicle device according to claim 22 .

Images