JP2023068009A - Map information creation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a map information creation method and a current position estimation method designed to improve the accuracy of estimating a position in the movement direction of a mobile object.

SOLUTION: A first onboard unit 5 mounted in a measurement vehicle 5 includes a LiDAR 23B. The first onboard unit 5 receives a reflected wave of a laser having been transmitted to a road surface and reflected by the road surface and recognizes edges of a demarcation line formed on the road surface on the basis of the strength of the reflected wave. The first onboard unit 5 transmits road surface information including the recognized edges of the demarcation line to a server device 3 via a network N. The server device 3 adds, on the basis of the road surface information, the edge information indicating the edges of the demarcation line and continuation information indicating that the section exists continuously to the information about the demarcation line included in the map information. A second onboard unit mounted in a vehicle 6 acquires information about the demarcation line from the server device 3, compares the edges of the demarcation line recognized by a LiDAR 41B with information indicating the edges of the demarcation line acquired from the server device 3, and estimates the current position.

SELECTED DRAWING: Figure 1

COPYRIGHT: (C)2023,JPO&INPIT

Description

The present invention relates to a map information creation method, a map information structure, and a current position estimation method.

In self-driving vehicles, it is necessary to estimate the current position with high accuracy by matching the feature positions measured by sensors such as LiDAR (Light Detection and Ranging) and the feature positions of the map information for self-driving. . Patent Literature 1 describes an example of a method of estimating the current position using the position of a feature as a landmark detected using LiDAR and the feature of map information.

Further, Patent Document 2 describes that LiDAR is used to detect a white line, and the relative position of the white line in the lateral direction with respect to the vehicle or the direction in which the vehicle is facing with respect to the white line is detected with high accuracy.

JP 2017-72422 A JP 2017-215199 A

One example of the conventional technology described above is the problem that the position of a vehicle (moving body) in the moving direction cannot be estimated using white lines (division lines).

An object of the present invention is to address such problems. That is, it is an object of the present invention to provide, for example, a map information creation method and a current position estimation method that improve the position estimation accuracy in the moving direction of a mobile object.

According to a first aspect of the present invention, there is provided a lane marking information creation method for solving the above-mentioned problems. At least one of edge information indicating an edge of the marking line and non-edge information included in the marking line and indicating that the marking is not the edge is added.

A current position estimation method according to claim 2 is a current position estimation method for estimating the current position of a moving body, and obtains a plurality of pieces of edge information of lane markings to which latitudes and longitudes are assigned, which are included in surrounding map information. acquiring a plurality of pieces of information indicating end portions of the lane markings on the road surface using a sensor arranged on the moving object; a positional relationship between the acquired pieces of edge information; and a step of estimating the current position based on a positional relationship between pieces of information indicating end portions.

In the lane marking information creation device according to claim 3, a latitude and longitude on the lane marking is given to each of the point information indicating the lane marking composed of a plurality of pieces of point information, and the end portion of the lane marking and non-end information that is included in the marking line and indicates that it is not the edge.

In the lane marking information creation program according to claim 4, a latitude and a longitude on the lane marking are given to each of the point information indicating the lane marking composed of a plurality of point information, and the end portion of the lane marking is added to the point information. and non-end information that is included in the marking line and indicates that it is not the edge.

A current position estimating device according to claim 5 is a current position estimating device for estimating the current position of a moving object, and acquires a plurality of end information of lane markings to which latitudes and longitudes are assigned, which are included in surrounding map information. an acquisition unit, a sensor arranged on the moving object to acquire a plurality of pieces of information indicating end portions of the lane markings on the road surface, a positional relationship between the acquired pieces of edge information, and the end portions of the lane markings on the road surface. and an estimating unit for estimating the current position based on the positional relationship between pieces of information indicating .

A recording medium according to claim 8 is characterized in that the marking line information creating program according to claim 4 is recorded.

According to claim 9, a current position estimation program causes a computer to execute the current position estimation method according to claim 2.

The road surface information acquisition device according to claim 10 is a road surface information acquisition device that acquires road surface information, and includes a shape recognition unit that recognizes a marking line on the road surface, and an end portion that recognizes both ends of the recognized marking line. and an interpolation unit for interpolating a sequence of points between the two ends.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing one embodiment of a driving support system that implements a road surface information acquisition method, a map information creation method, and a current position estimation method of the present invention; 2 is a functional configuration diagram of a first vehicle-mounted device shown in FIG. 1; FIG. 2 is a functional configuration diagram of the server device shown in FIG. 1; FIG. FIG. 10 is an explanatory diagram for explaining map information before the end of the lane marking is recorded; FIG. 10 is an explanatory diagram for explaining map information after the end of the lane marking is recorded; 2 is a functional configuration diagram of a second vehicle-mounted device shown in FIG. 1; FIG. 2 is a flowchart showing a procedure of road surface information acquisition processing executed by the first vehicle-mounted device shown in FIG. 1; 4 is a graph showing reflection intensity when a laser beam is applied to the division line when there is no fading at the edge. FIG. 10 is a graph showing the reflection intensity when laser is irradiated to the division line when there is blurring at the end. FIG. 4 is a graph showing the intensity distribution of reflected waves when a laser beam is applied to the division line when there is no fading at the edge. FIG. 10 is a graph showing intensity distribution of reflected waves when a laser beam is applied to the division line when there is a faint edge. FIG. 2 is a flowchart showing a driving support processing procedure executed by a second vehicle-mounted device shown in FIG. 1; FIG. 11 is an explanatory diagram for explaining map information in which end portions of lane markings are recorded in another embodiment; FIG. 9 is a functional configuration diagram of a first vehicle-mounted device in another embodiment;

Hereinafter, in the lane marking information creation method according to an embodiment of the present invention, information indicating a lane marking formed on a road surface, edge information indicating the end of the lane marking, and the section continuously exist. and at least one of continuous information indicating that the As a result, the edge of the lane marking can be used as a landmark and can be used to estimate the current position, thereby improving the position estimation accuracy in the moving direction of the moving object.

Further, in the lane marking information creation method according to an embodiment of the present invention, the information indicating the lane marking formed on the road surface includes the edge information indicating the edge of the lane marking, and the information included in the lane marking. and at least one of non-edge information indicating that it is not an edge. As a result, the edge of the lane marking can be used as a landmark and can be used to estimate the current position, thereby improving the position estimation accuracy in the moving direction of the moving object.

Further, the information indicating the partition line may include point information indicating latitude and longitude on the partition line, and at least one of the edge information and the non-edge information may be added to the point information. This makes it possible to easily assign edge information and non-edge information.

Further, as the non-end portion information, information indicating that there is no defect on the marking line and information indicating that there is a defect on the marking line may be identifiably assigned. . As a result, it is possible to avoid using defective portions on the lane markings as landmarks, thereby further improving the position estimation accuracy in the moving direction of the moving object.

A lane marking data structure according to an embodiment of the present invention is a data structure of lane marking data indicating a lane marking formed on a road surface, wherein the lane marking data indicates an end portion of the lane marking. It is characterized in that end information and continuation information indicating that the sections exist continuously can be added. As a result, the edge of the lane marking can be used as a landmark and can be used to estimate the current position, thereby improving the position estimation accuracy in the moving direction of the moving body.

A lane marking data structure according to an embodiment of the present invention is a data structure of lane marking data indicating a lane marking formed on a road surface, wherein the lane marking data indicates an end portion of the lane marking. It is characterized in that edge information and non-edge information that is included in the marking line and indicates that it is not the edge can be added. As a result, the edge of the lane marking can be used as a landmark and can be used to estimate the current position, thereby improving the position estimation accuracy in the moving direction of the moving body.

Further, a storage device according to an embodiment of the present invention is a storage device for storing lane marking information indicating lane markings formed on a road surface, wherein the lane marking information includes end portions of the lane markings. and non-end information included in the marking line and indicating that it is not the end. As a result, the edge of the lane marking can be used as a landmark and can be used to estimate the current position, thereby improving the position estimation accuracy in the moving direction of the moving object.

Further, a current position estimation method according to an embodiment of the present invention is a current position estimation method for estimating the current position of a mobile object based on information related to lane markings, wherein an end indicating the end of the lane marking is received from an external device. an acquisition step of acquiring section information; information indicating an end of a lane marking on a road surface recognized by a sensor arranged on the moving body; and the acquired end information. and an estimating step of estimating the current position of the . As a result, it is possible to improve the position estimation accuracy in the moving direction of the moving body.

Further, in the lane marking information creation device according to the embodiment of the present invention, the information indicating the lane marking formed on the road surface includes the end information indicating the end of the lane marking and the section continuously present. The present invention is characterized by comprising an imparting unit that imparts at least one of continuous information indicating that the
As a result, the edge of the lane marking can be used as a landmark and can be used to estimate the current position, thereby improving the position estimation accuracy in the moving direction of the moving object.

Further, a marking line information creating program that causes a computer to execute the above-described marking line information creating method may be used. Since the program is executed by a computer in this way, dedicated hardware or the like is not required, and it can be installed and functioned in a general-purpose information processing apparatus.

Also, the information processing program described above may be stored in a computer-readable recording medium. By doing so, the program can be distributed as a single unit in addition to being incorporated into the device, and version upgrades and the like can be easily performed.

A driving support system that implements the road surface information acquisition method, lane marking information creation method, and current position estimation method of the present invention will be described below with reference to FIGS. 1 to 6. FIG.

The driving support system 1 includes a first vehicle-mounted device 2 as a road surface information acquisition device, an external device, a server device 3 as a lane marking information creation device, and a second vehicle-mounted device 4 . The first vehicle-mounted device 2 is a device that acquires road surface information and transmits it to the server device 3. For example, the first vehicle-mounted device 2 is mounted on a measurement vehicle 5 that is a moving object that travels on the road for the purpose of creating map information. It is The server device 3 acquires road surface information from the first vehicle-mounted device 2 and creates map information. The server device 3 can communicate with the first vehicle-mounted device 2 via a network N such as the Internet, for example, and acquires road surface information from the first vehicle-mounted device 2 using the network N. Acquisition of the road surface information of the server device 3 is not limited to the form described above. You can let it run. In the following description, it is assumed that information is exchanged between the first vehicle-mounted device 2 and the second vehicle-mounted device 4 and the server device 3 via the network N. This is not a limitation, and information may be exchanged manually by an operator.

The second vehicle-mounted device 4 can communicate with the server device 3 via the network N. As shown in FIG. The second vehicle-mounted device 4 is a device that receives map information from the server device 3 and performs driving assistance, and is mounted on a vehicle 6 as a moving body that receives driving assistance, for example. Also, in this embodiment, the first and second vehicle-mounted devices 2 and 4 mounted on the vehicles 5 and 6 are described as examples of terminals capable of communicating with the server device 3. It may be a portable terminal capable of The reception of the map information by the second vehicle-mounted device 4 is not limited to the form described above. may be moved.

A functional configuration of the first vehicle-mounted device 2 is shown in FIG. The first vehicle-mounted device 2 includes a control section 21 , an input/output section 22 and a sensor section 23 .

A processor such as a CPU (Central Processing Unit) of the first vehicle-mounted device 2 functions as the control unit 21 , and controls the entire first vehicle-mounted device 2 . The control unit 21 uses the LiDAR 23B, which will be described later, to recognize the ends of lane markings such as white lines and yellow lines formed on the road surface, and transmits them to the server device 3 as road surface information. The control unit 21 may acquire peripheral information other than the road surface information and transmit it to the server device 3 .

The input/output unit 22 functions as a network interface or the like of the first vehicle-mounted device 2 and transmits road surface information.

The sensor unit 23 includes a GPS (Global Positioning System) receiver 23A, a transmitter, a LiDAR 23B as a receiver, and the like. In addition, in the present embodiment, the sensor unit 23 includes the LiDAR 23B as an example of a sensor capable of transmitting electromagnetic waves. The GPS receiver 23A detects current position information of the measurement vehicle 5 . As is well known, the GPS receiver 23A periodically receives radio waves oscillated from a plurality of GPS satellites, obtains current position information and time, and outputs the information to the control unit 21 .

The LiDAR 23B outputs a pulsed laser while changing the output direction in a predetermined detection area, receives the reflected wave of the laser, and generates point cloud information. The LiDAR 23B outputs multiple pulses of laser within the detection area, and generates point group information based on the reflected waves of the multiple pulses of the laser. Each piece of information that constitutes the point group information is information that indicates the output direction of the laser, the distance to the object that reflected the laser, and the intensity of the reflected wave. In this embodiment, the LiDAR 23B irradiates a laser toward the road surface, and uses the road surface as the detection area. Therefore, the point group information becomes information indicating the distance to the road surface as the object. Note that the LiDAR 23B may acquire peripheral information other than the road surface information by emitting a laser to areas other than the road surface.

The server device 3 is installed in an office or the like that provides map information.

A functional configuration of the server device 3 is shown in FIG. The server device 3 includes a storage unit 31 as a storage device, a control unit 32 and an input/output unit 33 .

The storage unit 31 functions as a storage device such as a hard disk of the server device 3 and stores map information. In this embodiment, it is assumed that the map information already includes information about lane markings. Information about the lane markings will be described with reference to FIG. In _the example shown in FIG. 4, it is composed of point information P ₁ , . Position information (latitude, longitude) is assigned to each of the point information P ₁ , . . . P ₁₄ .

A processor such as a CPU of the server device 3 functions as the control unit 32 and controls the entire server device 3 . Based on the road surface information such as the end of the lane marking transmitted from the first vehicle-mounted device 2, the control unit 32 determines that it is an end point among the point information _P1 , ..., _P14 as shown in FIG. Edge information indicating the edge of the division line is added to the recognized one (indicated by the white circle in the drawing). By adding the above-described end information to the information about the lane markings included in the map information in this way, the second vehicle-mounted device 4 that has received the map information can determine the end of the lane markings based on the information about the lane markings. It becomes possible to recognize the part.

The input/output unit 33 functions as a network interface or the like of the server device 3 , receives road surface information from the first vehicle-mounted device 2 , and transmits map information to the second vehicle-mounted device 4 .

A functional configuration of the second vehicle-mounted device 4 is shown in FIG. The second vehicle-mounted device 4 includes a sensor section 41 , a control section 42 and an input/output section 43 .

The sensor unit 41 includes a GPS receiver 41A, LiDAR 41B, and the like. Since the GPS receiver 41A and the LiDAR 41B have the same functions as the GPS receiver 23A of the first vehicle-mounted device 2 and the LiDAR 23B of the first vehicle-mounted device 2, detailed description thereof will be omitted here.

A processor such as a CPU of the second vehicle-mounted device 4 functions as the control unit 42 , and controls the entire second vehicle-mounted device 4 . The control unit 42 uses the information obtained from the sensor unit 41 and the map information obtained from the server device 3 to provide driving assistance (control of the steering wheel, accelerator, brake, etc., presentation of information on driving, etc.). In addition, below, in order to simplify description, it demonstrates that automatic driving|operation control is performed as driving|running|working assistance. The control unit 42 needs to estimate the current position of the host vehicle in order to perform automatic driving control. In this embodiment, the control unit 42 detects the current position based on the information indicating the end of the lane marking recognized using the LiDAR 41B and the end information of the lane marking included in the map information obtained from the server device 3. is estimated.

The input/output unit 43 functions as a network interface or the like of the second vehicle-mounted device 4 and receives map information.

Next, the operation of the driving support system 1 in this embodiment will be described. First, the road surface information acquisition process executed by the control unit 21 of the first vehicle-mounted device 2 (hereinafter simply referred to as the first vehicle-mounted device 2) will be described below with reference to FIG. By using the flowchart shown in FIG. 7 as a computer program, the first vehicle-mounted device 2 becomes a road surface information acquisition program that causes the computer to execute the road surface information acquisition method.

The first vehicle-mounted device 2 executes the road surface information acquisition process while the vehicle is running. In the road surface information acquisition process, the first vehicle-mounted device 2 controls the LiDAR 23B to acquire the above-described point group information regarding the road surface on which the vehicle is traveling (step S1). Next, the first vehicle-mounted device 2 extracts lane marking segments based on the point group information. Specifically, an orthoimage of the point cloud is generated based on the acquired point cloud information. Then, the orthorectified image is subjected to image processing to detect, for example, line segments (straight lines). Then, the detected line segments (straight lines) and the like are grouped, and marking line segments forming the outline of one marking line are extracted (step S2). Note that KS in FIG. 8 is an example of a marking line segment. Next, as shown in FIGS. 8 and 9, the first vehicle-mounted device 2 recognizes the end and non-end portions of the lane markings formed on the road surface on which the vehicle is traveling from the extracted lane marking segments (step S3). ). Next, the first vehicle-mounted device 2 interpolates the sequence of points between the recognized end portion and the non-end portion (that is, interpolates the sequence of points in the continuous portion between the end portions) (step S4), and return to step S1.

Details of step S3 will be described with reference to FIGS. 8 and 9. FIG. FIG. 8 shows a case where there is no problem such as blurring of the lane markings, and FIG. 9 shows a case where there is a problem such as blurring of the lane markings. As an example of step S3, as shown in FIGS. 8 and 9, the intensity of the reflected wave (hereinafter referred to as "reflection intensity") on each of a plurality of lines L1 to L4 along the longitudinal direction of the partition line There is a method of recognizing the end and non-end of the marking line based on the change in the direction along the longitudinal direction of the line. The non-end portion is a portion where the end portion cannot be accurately detected due to defects such as fading on the marking line. In other words, the non-end portion is a portion where the reflection intensity of the marking line changes along the longitudinal direction, but the change is not due to the end of the marking line but to a defect such as blurring. is. In addition to faintness, thinning, staining, overlapping of lines, and the like may also be considered as defects.

The laser reflectance is high on the road surface on which the marking lines are formed, and the laser reflectance on the road surface on which the marking lines are not formed is low. Therefore, the LiDAR 23B receives a reflected wave from a road surface on which lane markings are formed with a higher intensity than a road surface on which lane markings are not formed. Therefore, the first vehicle-mounted device 2 estimates the position of the marking line from the reflected intensity of the laser (that is, the intensity of the reflected wave received by the LiDAR 23B), and sets lines L1 to L4 along the longitudinal direction on the marking line. .

In addition, as shown in FIG. 8, at the end of the demarcation line where there is no defect such as blurring, the reflection intensity abruptly changes on each of the lines L1 to L4 along the longitudinal direction of the demarcation line. Therefore, the edge of the lane marking without defects such as blurring can be used as a landmark because there is little variation when the LiDAR 41B mounted on the vehicle 6 detects the edge of the lane marking.

On the other hand, as shown in FIG. 9, at the end of the faint marking line, the reflection intensity gradually changes on all or part of the lines L1 to L4 along the longitudinal direction of the marking line. For this reason, when the LiDAR 41B mounted on the vehicle 6 detects the end of the lane marking, the edge of the lane marking with a blur tends to cause variations in the detection position, and the direction of travel of the vehicle 6 (longitudinal direction of the lane marking) is likely to occur. ) as landmarks for self-localization.

Therefore, in this embodiment, the first vehicle-mounted device 2 recognizes positions on the lines L1 to L4 where the reflection intensity changes abruptly as the end T1 of the partition line, and recognizes positions on the lines L1 to L4 where the reflection intensity changes gradually. The upper position is recognized as the non-end T2 of the marking line. Further, in FIG. 9, blurring occurs uniformly in the horizontal direction. However, actual blurring may not occur uniformly in the horizontal direction, and may occur only above, below, or only in the center of the marking line. Therefore, in the present embodiment, if the reflection intensity changes abruptly on all the lines L1 to L4, the first vehicle-mounted device 2 recognizes the position where the reflection intensity changes as the end T1 of the marking line. If even one of the lines has a gradual change in reflection intensity, the position where the reflection intensity changes is recognized as the non-end portion T2 of the marking line.

As a specific process, the first vehicle-mounted device 2 changes from a high state in which the reflection intensity on all the lines L1 to L4 is reflected on the lane markings to a low state in which the reflection intensity is reflected on the road surface other than the lane markings, and , and if the rates of change on all the lines L1 to L4 are equal to or greater than the first threshold, it is recognized as the end T1 of the marking line. In addition, the first on-vehicle device 2 changes the reflection intensity on all the lines L1 to L4 from a low state where the reflection intensity is reflected on the road surface other than the lane marking to a high state where the reflection intensity is reflected on the lane marking, and all the lines L1 to L4 If the rate of change on ˜L4 is equal to or greater than the first threshold, it is recognized as the end T1 of the marking line.

In addition, the first vehicle-mounted device 2 changes from a high state in which the reflection intensity on all the lines L1 to L4 is reflected on the lane markings to a low state in which the reflection intensity is reflected on the road surface other than the lane markings, and all the lines L1 to L1 ˜L4, if the rate of change is less than the first threshold, it is recognized as the non-end T2 of the marking line. In addition, the first vehicle-mounted device 2 changes the reflection intensity of even one of all the lines L1 to L4 from a low state of being reflected on the road surface other than the lane marking to a high state of being reflected on the lane marking, and If the rate of change is less than the first threshold, it is recognized as the non-end T2 of the marking line.

In addition, the first vehicle-mounted device 2 associates the recognized edge T1 or non-edge T2 with the edge T1 or non-edge T2 on the same demarcation line (the same continuous line in the case of broken lines). Specifically, if the reflectance between two ends T1 adjacent in the longitudinal direction, between two non-ends T2, or between the ends T1 and non-ends T2 is high, the first on-vehicle device 2 Two ends T1, two non-ends T2, or adjacent ends T1-non-ends T2 are stored in association with each other as being on the same division line.

Further, as another example of step S3, as shown in FIGS. 10 and 11, the partitioning is performed based on the intensity distribution of the reflected waves reflected by the reflection areas A1 to A4 scanned along the longitudinal direction of the partitioning line. A method of recognizing the end of a line can be mentioned. The first vehicle-mounted device 2 estimates the position of the marking line from the reflection intensity of the laser, and sets reflection areas A1 to A4 along the longitudinal direction on the marking line.

As shown in FIG. 10, in the intensity distribution around the end of the marking line, where there is no problem such as blurring, the reflection intensity changes abruptly while the dispersion remains small. That is, in the example shown in FIG. 10, the intensity distribution in the reflection area A1 formed at the end of the partition line has a small dispersion and a large reflection intensity. A reflective area A2 adjacent to the reflective area A1, a reflective area A3 adjacent to the reflective area A2, and a reflective area A4 adjacent to the reflective area A3 are reflected on a road surface on which no marking line is formed. Therefore, the intensity distribution of the reflection areas A2 to A3 sharply decreases in intensity distribution as compared to the reflection area A1 while maintaining a small dispersion.

On the other hand, as shown in FIG. 11, the intensity distribution around the end of the demarcation line with a defect such as blurring has a large dispersion and the reflection intensity gradually changes. That is, in the example shown in FIG. 11, the intensity distribution in the reflection area A1 formed on the partition line has a small dispersion and a large reflection intensity. A reflective area A2 adjacent to the reflective area A1 and a reflective area A3 adjacent to the reflective area A2 are formed in a portion where a defect such as blurring occurs on the division line. Therefore, the intensity distribution in the reflection areas A2 and A3 has a large dispersion and the reflection intensity is smaller than that in the reflection area A1. Furthermore, since the reflection area A4 adjacent to the reflection area A3 is reflected on the road surface on which the division lines are not formed, its intensity distribution has a small dispersion and the reflection intensity is smaller than those of the reflection areas A2 and A3.

Therefore, in this embodiment, the first vehicle-mounted device 2 recognizes the position where the reflection intensity changes while the dispersion of the intensity distribution of the reflection areas A1 to A4 is kept small as the edge T1 of the marking line, and The position where the intensity distribution of A1 to A4 cannot be kept small and the reflection intensity changes is recognized as the non-end T2 of the marking line.

As a specific process, the first vehicle-mounted device 2 changes the intensity distribution of the reflection area from a state where the intensity distribution of the reflection area is less than the second threshold to a state where the intensity of the reflection area is reflected on the lane marking, and changes the intensity of the reflection area to a road surface other than the lane marking. The position where the reflected state changes to the low state is recognized as the end T1 of the marking line. In addition, the first vehicle-mounted device 2 shifts the intensity of the reflection area from a low state where the intensity is reflected by the road surface other than the lane marking to a high state where the intensity is reflected by the lane while the dispersion of the intensity distribution of the reflection area is less than the second threshold value. , is recognized as the end T2 of the marking line.

In the first onboard device 2, the dispersion of the intensity distribution of the reflection area becomes equal to or greater than the second threshold, and the intensity of the reflection area changes from a high state reflected on the lane marking to a low state reflected on the road surface other than the lane marking. This position is recognized as the non-end T2 of the marking line. In the first vehicle-mounted device 2, the dispersion of the intensity distribution of the reflection area becomes equal to or greater than the second threshold value, and the intensity of the reflection area changes from a low state reflected by road surfaces other than the lane markings to a high state reflected by the lane markings. This position is recognized as the non-end T2 of the marking line.

At a predetermined timing, the first vehicle-mounted device 2 detects the positions of the end T1 and the non-end T2 of the lane marking, the end T1 or the non-end on the same lane line associated with the recognized ends T1 and T2. Information about T2 is transmitted to the server device 3.

Next, the lane marking information creation process performed by the control unit 32 of the server device 3 (hereinafter simply referred to as the server device 3) will be described with reference to FIGS. 5 and 6. FIG. By making the lane marking information creation process into a computer program, the server device 3 becomes a road surface information acquisition program that causes a computer to execute the lane marking information creation method.

When the server device 3 receives the road surface information including the positions of the edge and non-edge of the lane marking, the information about the lane marking shown in FIG. Either non-end portion information (defective) or non-end portion information (no defect) indicating that it is not an included end portion is added. More specifically, as shown in FIG. 5, the server device 3 stores points P ₁ to P ₁₄ stored as information about the lane markings that correspond to the positions of the received end and non-end portions of the lane marking. If there is, the edge information and non-edge information (defective) are added to the point. In the example shown in FIG. 5, edge information is assigned to points P ₁ , P ₈ , and P ₁₃ indicated by white circles, and non-edge information (defective) is assigned to point P ₇ indicated by hatching. As described above, any of the edge information indicating the edge of the lane marking, the non-edge information (defective) indicating that it is not an edge included in the lane marking, or the non-edge information (no defect) Continuity information indicating that the sections are continuously present may be added to the corresponding point information instead of adding the . That is, continuous information is added to point information that is not considered to be an edge (non-edge information (defective) and non-edge information (no defect)), and continuity information is added to point information that is determined to be an edge. may not be given.

If there is no point corresponding to the position of the end or non-end of the received lane marking among the points P ₁ to P ₁₄ stored as the information on the lane marking, the server device 3 stores the information on the lane marking. , points corresponding to the positions of the end and non-end portions of the division line are added to the points, and end information and non-end information (defective) are added to the points. In the example shown in FIG. 5, points P ₁₅ , P _{16 , and P 18 to which end information indicated by white circles are added, and point P 17 to} which non-end information (defective) indicated by diagonal lines is added. be done.

In addition, the server device 3 gives non-edge information (no problem) to points between the edge and non-edge on the same division line. In the example shown in FIG. 5, the points P ₂ to P ₆ , P ₉ , P ₁₁ , and P ₁₄ indicated by black circles are provided with non-end information (no defect).

Next, the driving assistance processing performed by the second vehicle-mounted device 4 will be described with reference to the flowchart of FIG. 12 . The server device 3 executes the current position estimation method by executing this driving support process. First, the second vehicle-mounted device 4 acquires point cloud information from the LiDAR 41B (step S10). Next, the second vehicle-mounted device 4 performs object detection from the point cloud information, and detects the end of the lane marking based on the point cloud information from the LiDAR 41B (step S11).

Specifically, the second vehicle-mounted device 4 detects an object and recognizes its type (building, pedestrian, other vehicle, etc.) by executing a so-called object recognition process based on the point cloud information. This makes it possible to recognize the type of object around the vehicle and the distance to the object. In addition, the second vehicle-mounted device 4 uses the same determination method as the first vehicle-mounted device 2 to detect the edge of the lane marking and recognize the distance to the edge.

Next, the second vehicle-mounted device 4 communicates with the server device 3 and acquires map information around the current position detected by the signal from the GPS receiver 41A (step S12). After that, the second vehicle-mounted device 4 estimates the current position using the recognized object or the end of the lane marking as a landmark (step S13). That is, in step S13, the second vehicle-mounted device 4 estimates the current position based on the positional relationship between the position information of the feature included in the map information and the position of the object recognized in step S10. Also, the current position is estimated by comparing the positional relationship between the information about the lane markings included in the map information and the positions of the end portions of the lane markings recognized in step S11.

After that, the second vehicle-mounted device 4 executes driving assistance based on the estimated current position (step S14), and returns to step S10 again.

According to the above-described embodiment, the laser transmitted to the road surface receives the reflected wave reflected by the road surface, and the end of the marking line formed on the road surface is recognized based on the intensity of the reflected wave. there is As a result, the end of the marking line can be recognized with high accuracy. In addition, since the end of the recognized lane marking can be used as a landmark to estimate the current position, it is possible to improve the accuracy of estimating the position of the vehicle 6 in the moving direction.

Further, according to the above-described embodiment, the laser is transmitted from the LiDAR 23B arranged on the measurement vehicle 5. FIG. In this way, by mounting the LiDAR 23B on the measurement vehicle 5, it is possible to easily recognize the ends of lane markings in a wide range.

Further, according to the embodiment described above, the end portions of the marking lines are recognized based on changes in reflection intensity on the lines L1 to L4 along the longitudinal direction of the marking lines. As a result, it is possible to improve the recognition accuracy of the end of the marking line.

Further, according to the above-described embodiment, the end of the demarcation line is recognized based on the change along the longitudinal direction of the reflection intensity on the plurality of lines L1 to L4 arranged in the direction perpendicular to the longitudinal direction. . As a result, it is possible to further improve the recognition accuracy of the end of the marking line.

Further, according to the above-described embodiment, the positions where the intensity of the reflected waves on all the lines L1 to L4 changes along the longitudinal direction and the rate of change is equal to or greater than the first threshold value are the ends of the demarcation lines. recognized as As a result, a portion where an end cannot be accurately recognized due to a problem such as blurring is not recognized as an end of a marking line.

Further, according to the above-described embodiment, the intensity of the reflected wave changes along the longitudinal direction on all the lines L1 to L4, and the position where the rate of change is less than the first threshold value is at the end of the demarcation line. It is recognized as a non-end. As a result, it is possible to recognize, as a non-end portion of the marking line, a portion where an end portion cannot be accurately recognized due to a defect such as blurring.

Further, according to the embodiment described above, the end of the marking line is recognized based on the intensity distribution of the reflected waves in the reflection areas A1 to A4 scanned along the longitudinal direction of the marking line. As a result, it is possible to improve the recognition accuracy of the end of the marking line.

Further, according to the embodiment described above, the position where the intensity of the reflected wave changes while the dispersion of the intensity distribution of the reflected wave is kept below the second threshold value is recognized as the edge of the marking line. As a result, a portion where an end cannot be accurately recognized due to a problem such as blurring is not recognized as an end of a marking line.

Further, according to the embodiment described above, the position where the dispersion of the intensity distribution of the reflected wave becomes equal to or greater than the second threshold value and the intensity of the reflected wave changes is recognized as the non-end portion of the marking line. As a result, it is possible to recognize, as a non-end portion of the marking line, a portion where an end portion cannot be accurately recognized due to a defect such as blurring.

Further, in the above-described embodiment, the information about the division line included in the map information includes end information indicating the end of the division line, continuation information indicating that the division exists continuously, at least one of As a result, the ends of the lane markings can be used as landmarks for estimating the current position, so that the position estimation accuracy in the moving direction of the vehicle 6 can be improved.

In the above-described embodiment, the information about the lane marking included in the map information includes edge information indicating the edge of the lane marking and non-edge information indicating that the lane marking is not the edge. and are given. As a result, the ends of the lane markings can be used as landmarks for estimating the current position, so that the position estimation accuracy in the moving direction of the vehicle 6 can be improved.

In the above-described embodiment, the information about the division lines included in the map information is point information indicating the latitude and longitude on the division lines, and the point information is added to the edge information and the non-edge information. This makes it possible to easily assign edge information and non-edge information.

In the above-described embodiment, the non-edge information includes non-edge information (no defect) indicating that there is no defect on the lane marking and non-edge information indicating that there is a defect on the lane marking. Part information (defective) and , are given so as to be identifiable. As a result, it is possible to avoid using the defective portion on the lane marking as a landmark.
It is possible to improve the accuracy of position estimation in the direction of movement.

Further, in the above-described embodiment, the second vehicle-mounted device 2 acquires the edge information indicating the edge of the lane marking from the server device 3, which is an external device, and the road surface recognized by the LiDAR 41B arranged in the vehicle 6. The current position of the vehicle 6 is estimated based on the information indicating the edge of the upper lane marking and the obtained edge information. As a result, it is possible to improve the position estimation accuracy of the moving direction of the vehicle 6 .

In the above-described embodiment, edge information and non-edge information are given to the points constituting the information about the lane markings already stored in the storage unit 31 of the server device 3. not a thing The server device 3 may create new information about lane markings based on the road surface information received from the first vehicle-mounted device 2 . In this case, for example, when the server device 3 receives the road surface information including the positions of the edge and non-edge of the lane marking, it creates information about the lane marking as shown in FIG.

The server device 3 also adds edge information to the points P ₂₀ , P ₂₇ , P ₃₀ , P ₃₁ , P ₃₄ , and P ₃₇ indicating the positions of the edges received from the first vehicle-mounted device 2 . In addition, the server device 3 adds non-end information (defective) to the points P ₂₆ and P ₃₃ indicating the positions of the non-ends received from the first vehicle-mounted device 2 . Next, the server device 3 determines points P ₂₀ -P ₂₆ , points P ₂₇ -P ₃₀ , points P ₃₁ -P ₃₃ , point P Points P _{21 to} P ₂₅ , points P ₂₈ to P ₂₉ , points P ₃₂ , and points P ₃₅ to P ₃₆ are arranged at equal intervals along the division line between ₃₄ and point P 37 , and the points P ₂₁ to Non-end information (no defect) is given to P ₂₅ , points P ₂₈ to P ₂₉ , point P ₃₂ , and points P ₃₅ to P ₃₆ .

In addition, the server device 3 receives from the first vehicle-mounted device 2 the points other than the points P ₂₀ , P ₂₇ , P ₃₀ , P 31 , P 34 , and P 37 indicating the positions of the ends, and the partitions exist continuously at points other than the points P 20 , P 27 , P 30 , P ₃₁ , P ₃₄ , and P ₃₇ . Continuity information indicating that the edge is not (in other words, information indicating that it is not the edge) may be added.

Further, in the example shown in FIG. 13 described above, the server device 3 assigns points P ₂₈ to P ₂₉ , point P ₃₂ , and points P 35 to P ₃₆ to the information about the dashed lines, and points P ₂₈ to P 29 to the points P ₂₈ to P ₂₉ , point P ₃₂ , and points P ₃₅ to P ₃₆ are provided with non-end information (no defect), but the present invention is not limited to this.

In the present embodiment, the first vehicle-mounted device 2 is mounted on the measurement vehicle 6 dedicated to measurement, and the second vehicle-mounted device 4 is mounted on the vehicle 6 receiving driving assistance, but the present invention is not limited to this. The vehicle-mounted device 4 mounted on the vehicle 6 receiving driving assistance may have the functions of both the first vehicle-mounted device 2 and the second vehicle-mounted device 4 .

Also, in the above-described embodiment, the first vehicle-mounted device 2 recognizes the end portion and the non-end portion, but the present invention is not limited to this. The first in-vehicle device 2 may transmit only the point cloud information to the server device 3, and the server device 3 may recognize the end portion and the non-end portion.

In the above-described embodiment, the non-end information (defective) and the non-end information (no defect) were identifiable as the non-end information, but the present invention is not limited to this. It is not necessary to assign non-end information to points other than points corresponding to ends among the points on the division line so as to distinguish defective portions from non-defective portions.

Also, in the above-described embodiment, the end portion where a problem such as blurring has occurred is recognized as a non-end portion, but the present invention is not limited to this. Even if a problem such as blurring occurs, it may be recognized as an edge.

In the above-described embodiment, the server device 3 adds edge information and non-edge information to the information about the lane markings, but the present invention is not limited to this. For example, an operator of a map maker may manually add edge information and non-edge information by looking at the road surface information transmitted from the first vehicle-mounted device 2 . In the above-described embodiment, the map information including the information about the lane markings is stored and held in the server 3 (storage unit 31). The vehicle-mounted device 2 and the second vehicle-mounted device 4 can also store/hold at least part of the map information. Further, the process of adding edge information and non-edge information described above may be performed on the measurement vehicle side (first vehicle-mounted device 2). That is, the map information generation processing including the processing for recognizing the end portion and non-end portion of the division line described above (step S3 in FIG. 7) and the processing for adding end portion information or non-end portion information is performed by the server device 3. It may be implemented, or may be implemented by the measurement vehicle side (first vehicle-mounted device 2).

Further, in the above-described embodiment, the sensor unit 23 mounted on the first vehicle-mounted device 2 includes the LiDAR 23B as an example. Instead of or in addition to this, the sensor section 23 mounted on the first vehicle-mounted device may include a photographing section 23C that photographs the road surface on which the measurement vehicle 5 travels, as shown in FIG. In this case, the control unit 21 of the first vehicle-mounted device 2 acquires the image of the road surface captured by the image capturing unit 23C from the image capturing unit 23C, and based on the luminance information of the acquired image, the Recognize the end of the parcel line. The photographing unit 23C is composed of, for example, a stereo camera capable of detecting the distance to an object to be photographed.

Specifically, when the first vehicle-mounted device 2 (control unit 21) executes the road surface information acquisition process illustrated in FIG. Then, the image of the road surface photographed by the photographing unit 23C is obtained from the photographing unit 23C. Then, as in step S3, after recognizing the end and non-end portions of the lane markings formed on the road surface on which the vehicle is traveling (step S3), the process returns to step S1. More specifically, in step S3, the first vehicle-mounted device (control unit 21) converts the acquired photographed image into an orthoimage, and uses the luminance information and the like of the orthoimage to determine the end portion of the lane marking (white line). , to recognize non-ends. As another example, the photographing unit 23C is configured with a monocular camera, an image of the road surface photographed by the monocular camera is obtained, and the image is associated with the point cloud information obtained from the LiDAR 23B. good too. In this case, on the point cloud information acquired from the LiDAR 23B (in other words, as 3D information), using luminance information etc. in the captured image, it is possible to recognize the end and non-end of the marking line (white line). can.

Here, the brightness of the road surface where the marking lines are formed in the photographed image is high, and the brightness of the road surface where the marking lines are not formed is low. Also, at the end of the demarcation line where there is no defect such as blurring shown in FIG. 8, the luminance intensity changes abruptly on each of the lines L1 to L4 along the longitudinal direction of the demarcation line in the captured image. Also, at the end of the demarcation line with blurring shown in FIG. 9, the reflection intensity gradually changes on each of the lines L1 to L4 along the longitudinal direction of the demarcation line in the photographed image.

Therefore, the first vehicle-mounted device 2 (control unit 21) sets lines L1 to L4 along the longitudinal direction of the partition lines on the captured image, and based on the result of detecting and recognizing changes in luminance along these lines, As in the above-described embodiment, the end and non-end of the demarcation line (white line) can be recognized. In other words, the "brightness information" of the marking lines in the photographed image of this modified example can be handled in the same manner as the "reflection intensity" of the marking lines in the above-described embodiment. In this case, "reflection intensity" in FIGS. 8 to 11 can be appropriately read as "luminance" in the photographed image.

Further, the sensor unit 41 mounted on the second vehicle-mounted device 4 includes, as an example, the LiDAR 41B, but similarly to the first vehicle-mounted device 2, instead of or in addition to this, the second vehicle-mounted device 4 The sensor unit 41 mounted on the vehicle 6 may include a photographing unit 41C that photographs the road surface on which the vehicle 6 travels. That is, the second vehicle-mounted device 4 may recognize the end of the lane marking from the captured image by the method described above, and may execute the driving support process shown in FIG. 12 .

It should be noted that the present invention is not limited to the above embodiments. That is, various modifications can be made without departing from the gist of the present invention.

2 First in-vehicle device (road surface information acquisition device)
3 Server device (external device, lane marking information creation device)
5 Vehicle (moving body)
6 Measurement vehicle (moving object)
23B LiDAR (sensor, transmitter, receiver)
31 storage unit (storage device)
41B LiDAR (sensor)
L1~L4 line A1~A4 reflection area

Claims (10)

Latitude and longitude on the lane marking are given to each of the point information indicating the lane marking composed of a plurality of pieces of point information, and edge information indicating the edge of the lane marking and information included in the lane marking are added. A marking line information creating method, characterized in that at least one of non-end portion information indicating that the line is not the end portion is added. A current position estimation method for estimating the current position of a mobile body,
a step of obtaining a plurality of pieces of end information of division lines to which latitudes and longitudes are assigned, which are included in surrounding map information;
a step of obtaining a plurality of pieces of information indicating end portions of lane markings on the road surface using sensors arranged on the moving object;
and estimating the current position based on the positional relationship between the acquired edge information and the positional relationship between information indicating the edge of the lane marking on the road surface. Location estimation method. Latitude and longitude on the lane marking are given to each of the point information indicating the lane marking composed of a plurality of pieces of point information, and edge information indicating the edge of the lane marking and information included in the lane marking are added. A marking line information creating apparatus, comprising: a marking unit for adding at least one of non-end information indicating that the line is not the edge. Latitude and longitude on the lane marking are given to each of the point information indicating the lane marking composed of a plurality of pieces of point information, and edge information indicating the edge of the lane marking and information included in the lane marking are added. A marking line information creating program for causing a computer to function as an imparting unit that imparts at least one of non-end portion information indicating that it is not an edge portion. A current position estimation device for estimating the current position of a mobile object,
an acquisition unit that acquires a plurality of pieces of end information of the lane markings to which latitudes and longitudes are assigned, which are included in the peripheral map information;
a sensor that is arranged on the moving body and acquires a plurality of pieces of information indicating the end of the lane marking on the road surface;
an estimating unit for estimating the current position based on the obtained positional relationship between the edge information and the positional relationship between the information indicating the edge of the lane marking on the road surface. Current position estimator. As the non-end portion information, information indicating that there is no defect on the lane marking and information indicating that there is a defect on the lane marking are added in an identifiable manner. The marking line information creation method according to claim 1. 4. The point information, instead of adding the non-end portion information, is added to the point information in place of adding the non-end portion information. 2. The lane marking information creation method according to 1. 5. A recording medium on which the lane marking information creation program according to claim 4 is recorded. A current position estimating program that causes a computer to execute the current position estimating method according to claim 2 . A road surface information acquisition device for acquiring road surface information,
a shape recognition unit that recognizes lane markings on the road surface;
an edge recognition unit that recognizes both ends of the recognized lane marking;
and an interpolation unit that interpolates a sequence of points between the two ends.

Images