JP2019144066A - Method and device for estimating road structure - Google Patents

Abstract

To accurately estimate the position of a road structure without being affected by the weather even when it is snowing at the time of the estimation.SOLUTION: The method for estimating a road structure according to the present invention includes the steps of: preparing a road structure model formed by learning a road structure around a vehicle, on a road structure-by-road structure basis; detecting the weather in a road on which the vehicle is running; adjusting a road structure model of prepared road structure models which is used to estimate the road structure according to a detected weather; and estimating the position of the road structure around the vehicle, using the adjusted road structure model.SELECTED DRAWING: Figure 5

Description

  The present invention creates a road structure model obtained by learning and modeling a road structure around the vehicle for each different road structure, and estimates the position of the road structure around the vehicle using the plurality of created road structure models The present invention relates to a road structure estimation method and a road structure estimation apparatus.

  Conventionally, a method for correcting map data stored in a car navigation device or the like is known. The invention described in Patent Document 1 compares travel locus data using GPS with map data, and corrects a shift in map data with respect to the travel locus data. Thereby, the position of the target on the map is corrected, and the self position of the vehicle is calculated using the corrected position of the target.

JP 2007-310198 A

  However, in the above-described conventional method, when the weather becomes snow, it is difficult to detect a road structure on the road such as a white line, and therefore the position of the road structure cannot be accurately estimated depending on the weather. There was a problem.

  Therefore, the present invention has been made in view of the above problems, and provides a road structure estimation method and a road structure estimation apparatus that can accurately estimate the position of the road structure without being influenced by the weather. Objective.

  A road structure estimation method and apparatus according to an aspect of the present invention detect the weather of a road on which a vehicle travels, and are used to estimate the road structure among a plurality of road structure models according to the detected weather. To adjust the road structure model. Then, the position of the road structure around the vehicle is estimated using the adjusted road structure model.

  According to the present invention, the position of the road structure can be accurately estimated regardless of the weather.

FIG. 1 is a block diagram showing the configuration of the road structure estimation apparatus according to the first embodiment of the present invention. FIG. 2 is a diagram for explaining a travel locus model used in the road structure estimation device according to the first embodiment of the present invention. FIG. 3 is a diagram for explaining a road structure model used in the road structure estimation apparatus according to the first embodiment of the present invention. FIG. 4 is a diagram for explaining a road structure model adjustment method by the road structure estimation device according to the first embodiment of the present invention. FIG. 5 is a flowchart showing a processing procedure of road structure estimation processing by the road structure estimation device according to the first embodiment of the present invention. FIG. 6 is a diagram for explaining a modification of the road structure model adjustment method by the road structure estimation device according to the first embodiment of the present invention. FIG. 7 is a diagram for explaining a road structure model adjustment method by the road structure estimation device according to the second embodiment of the present invention. FIG. 8 is a diagram for explaining a method of adjusting the estimation sensitivity of the road structure by the road structure estimation device according to the second embodiment of the present invention. FIG. 9 is a diagram for explaining a method of adjusting the estimation sensitivity of the road structure by the road structure estimation device according to the second embodiment of the present invention. FIG. 10 is a diagram for explaining a method for adjusting a road structure model by the road structure estimation device according to the third embodiment of the present invention. FIG. 11 is a diagram for explaining a method of adjusting the error range of the road structure position by the road structure estimation device according to the third embodiment of the present invention. FIG. 12 is a diagram for explaining a method of adjusting the error range of the road structure position by the road structure estimation device according to the third embodiment of the present invention.

[First Embodiment]
Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. In the description of the drawings, the same portions are denoted by the same reference numerals, and description thereof is omitted.

[Configuration of road structure estimation device]
With reference to FIG. 1, the structure of the road structure estimation apparatus which concerns on this embodiment is demonstrated. As shown in FIG. 1, the road structure estimation apparatus 1 includes a camera 2, a laser range finder 3, a GPS receiver 4, a gyro sensor 5, a map database 6, a controller 10, and a communication device 7. Moreover, the road structure estimation apparatus 1 may be mounted on a vehicle, and the mounted vehicle may be a vehicle having an automatic driving function.

  The camera 2 includes an image sensor such as a charge-coupled device (CCD) or a complementary metal oxide semiconductor (CMOS). The camera 2 is mounted on the host vehicle and photographs the surroundings of the host vehicle. The camera 2 has an image processing function and detects a road structure from the photographed image. The road structure is a constituent of the road, for example, a road edge or curb, a target on the road, a median strip, etc., in addition to a road lane marking or a stop line such as a white line or a yellow line. Targets are traffic lights, utility poles, traffic signs, etc. provided on roads and sidewalks. The camera 2 outputs the detected data to the controller 10.

  The laser range finder 3 is a sensor that detects a road structure around the host vehicle. Specifically, the laser range finder 3 scans the laser beam within a certain angle range, receives the reflected light at that time, and detects the time difference between the time of laser emission and the time of receiving the reflected light. The laser range finder 3 detects the relative distance and direction of the road structure with respect to the host vehicle, and outputs the detected data to the controller 10. The laser range finder 3 is installed around a bonnet, a bumper, a license plate, a headlight, a side mirror, and the like.

  The GPS receiver 4 detects the current location of the vehicle on the ground by receiving radio waves from an artificial satellite. The GPS receiver 4 outputs the detected data to the controller 10.

  The gyro sensor 5 detects the yaw rate (rotational angular velocity) around the vertical axis of the center of gravity of the host vehicle, and outputs the detected data to the controller 10.

  The map database 6 is a database stored in a car navigation device or the like, and stores various data necessary for route guidance such as road structure information and facility information. The map database 6 is a highly accurate map, and stores information including road structure information such as the number of road lanes, road boundaries such as white lines, curbs, road edges, and targets. The map database 6 outputs map information to the controller 10 in response to a request from the controller 10.

  Note that various data such as road structure information is not necessarily limited to that acquired from the map database 6, and may be acquired by a sensor included in the host vehicle. Moreover, you may acquire using vehicle-to-vehicle communication and road-to-vehicle communication. For example, when various data such as road structure information is stored in the server 20, the controller 10 can acquire these data at any time by communication. In addition, the controller 10 can periodically obtain the latest map information from the server 20 and update the map information it holds. In the present embodiment, the map database 6 includes a first map (a high-precision map, a map used for vehicle route guidance), and a second map may be included to supplement the first map. . This second map is created by the vehicle or the external server 20 when the first map alone is incomplete or when more detailed map information is required.

  The communication device 7 is a device for communicating with the server 20.

  The controller 10 is a circuit that processes data acquired from the camera 2, the laser range finder 3, the GPS receiver 4, the gyro sensor 5, and the map database 6, and is configured by, for example, an IC or an LSI. When the controller 10 grasps this functionally, the travel locus model creation unit 11, the road structure model creation unit 12, the road structure model adjustment unit 13, the road structure estimation unit 15, and the self-position estimation unit 17 Can be classified.

  The controller 10 includes a general-purpose electronic circuit including a microcomputer, a microprocessor, and a CPU, and peripheral devices such as a memory. And it operates as the road structure estimation apparatus 1 by executing a specific program. Each function of the controller 10 can be implemented by one or a plurality of processing circuits. The processing circuit includes a programmed processing device such as, for example, a processing device including an electrical circuit, and an application specific integrated circuit (ASIC) or conventional circuit arranged to perform the functions described in the embodiments. It also includes devices such as parts.

  The travel locus model creation unit 11 creates a travel locus model using the travel data of the host vehicle. As shown in FIG. 2, the travel locus model is composed of a plurality of points Xi (i = 1 to n) on the travel locus in a predetermined section (section surrounded by a one-dot chain line shown in FIG. 2). Each point Xi is composed of coordinates (x, y) and a vehicle direction θ (traveling direction).

  The traveling locus model creation unit 11 acquires a large number of coordinates (x, y) and vehicle orientation θ from the traveling locus of the host vehicle, learns and models the traveling locus of each point Xi using the acquired traveling data, A travel locus model is created so that the travel locus can be reproduced. The predetermined section is a section divided by a predetermined traveling time or traveling scene. The travel time is, for example, 3 to 4 seconds. The traveling scene is, for example, a scene from the entrance to the exit of the intersection, a right turn or a left turn scene. The reason for dividing by the travel time or the travel scene in this way is that when the travel time is long or the travel scene is complicated, various data enters and the accuracy of the travel locus model may be lowered.

  The travel trajectory model creation unit 11 can create a travel trajectory model with high accuracy by creating a travel trajectory model of a section divided by travel time and travel scene. The travel locus model creation unit 11 does not necessarily need to use the travel data of the host vehicle, and may use the travel data of other vehicles other than the host vehicle collected in the server 20 without using the travel data in advance. A created model may be used.

  Specifically, the travel locus model creation unit 11 obtains an average position for each point Xi in a predetermined section from the acquired data. Next, the travel locus model creation unit 11 calculates a variance-covariance matrix from the acquired travel data, and obtains a principal component vector. At this time, the traveling locus model creation unit 11 obtains the number that can sufficiently represent the traveling locus as the number of principal component vectors. Next, the traveling locus model creation unit 11 arranges the principal component vectors in a column and creates a matrix A. The matrix A is a trajectory control coefficient. The travel locus model creation unit 11 creates a travel locus model X by adding the average position of each point of the model and the matrix A and the control parameter q of the matrix A. The travel locus model X is expressed as follows using the equations (1) and (2).

  The traveling locus model creation unit 11 may store the created traveling locus model X in the server 20 or may be stored in the host vehicle. Note that only one traveling locus model X or a plurality of traveling locus models X may be generated. The predetermined section may be divided by a time such as 5 seconds or 10 seconds, or may be divided by an operation. As an example in the case of dividing by operation, a period from entering the intersection to exiting is defined as one section. The traveling locus model creation unit 11 may store the traveling locus model X in another vehicle using inter-vehicle communication.

  The road structure model creation unit 12 creates a road structure model using the road structure data obtained by detecting the road structure around the host vehicle. As shown in FIG. 3, the road structure model is composed of a plurality of points Li (i = 1 to n) on a white line in a predetermined section (section surrounded by a one-dot chain line shown in FIG. 3). In FIG. 3, white lines are used as an example of the road structure. Each point Li is composed of coordinates (x, y) and a white line direction φ. The road structure model creation unit 12 acquires a large number of coordinates (x, y) and white line direction φ from the road structure data, learns and models the white line of each point Li using the acquired data, and can reproduce the white line. A road structure model is created so that Note that the point Li = (x, y) is the position of the i-th white line obtained by dividing the white line in a certain section at intervals of 10 cm, for example. Further, the predetermined section may be divided by a time such as 5 seconds or 10 seconds.

  Specifically, the road structure model creation unit 12 obtains an average position for each point Li in the predetermined section from the acquired data. Next, the road structure model creation unit 12 calculates a variance-covariance matrix from the acquired data and obtains a principal component vector. At this time, the road structure model creation unit 12 obtains the number that can sufficiently express the road structure as the number of principal component vectors. Next, the road structure model creation unit 12 arranges the principal component vectors in columns and creates a matrix B. The matrix B is a trajectory control coefficient. The road structure model creation unit 12 creates a road structure model L by adding the average position of each point of the model and the control parameter p of the matrix B and the matrix B. The road structure model L is expressed as follows using the equations (3) and (4), similarly to the traveling locus models of the equations (1) and (2).

  The road structure model creation unit 12 creates a road structure model for each different road structure, not just the white line. For example, road structure models are created for curbs, targets, road edges, and the like. The plurality of created road structure models may be stored in the server 20 or in the host vehicle. Further, the road structure model may be stored in another vehicle using inter-vehicle communication.

  Here, since the vehicle travels along the white line on the road, there is a correlation between the travel locus and the shape of the white line, which is an approximation. Therefore, there is also a correlation between the travel locus model and the road structure model, which is an approximation. Therefore, there is a correlation between the control coefficient A of the travel locus model and the control coefficient B of the road structure model, and there is also a correlation between the control parameter q of the travel locus model and the control parameter p of the road structure model.

  The road structure model adjustment unit 13 detects the weather of the road on which the vehicle is traveling, and adjusts the road structure model used for estimating the road structure among the plurality of road structure models according to the detected weather. I do. Specifically, in the road structure model adjustment processing, processing for selecting a road structure model to be used for estimation of the road structure from a plurality of road structure models is performed. Details of the road structure model adjustment processing will be described later.

  The road structure estimation unit 15 estimates the position of the road structure around the vehicle using the road structure model adjusted by the road structure model adjustment unit 13. Specifically, the road structure estimation unit 15 calculates the position of the road structure around the vehicle using the adjusted road structure model, and the calculated position of the road structure and the position of the road structure recorded in the map information The position of the calculated road structure is corrected so as to match. For example, the position of the road structure is corrected by using an optimization method such as the Gauss-Newton method and a method of gradually approaching the true value in the repeated calculation while adjusting the control parameter.

  When the road structure estimation unit 15 corrects the position of the road structure and determines that the calculated position of the road structure matches the position of the road structure recorded in the map information, the position of the road structure End the estimation of. The optimization method such as Gauss-Newton method is used to determine whether or not they match, and when it is determined that it has converged and approached the true value, the calculated road structure position and the road structure recorded in the map information Is determined to match.

  Further, the road structure estimation unit 15 may calculate a trajectory from the travel trajectory model created by the travel trajectory model creation unit 11 and estimate the position of the road structure based on the calculated trajectory. In this case, the road structure estimation unit 15 acquires the distance from each point on the trajectory to the road structure and the direction θ of the host vehicle when the distance is acquired. Then, the position of the road structure is estimated using the acquired distance to the road structure and the vehicle orientation θ.

  When the estimation of the position of the road structure is thus completed, the road structure estimation unit 15 links to the map database 6 and records the estimated position of the road structure.

  The self-position estimating unit 17 estimates the self-position using the position of the road structure recorded in the map database 6 by the road structure estimating unit 15. In this embodiment, since the position of the road structure on the map is estimated with high accuracy, the self-position estimation unit 17 can estimate the self-position with high accuracy by using the distance and direction from the host vehicle to the road structure. it can. When the vehicle's own position is estimated using the position of the road structure, when the vehicle travels around the road structure, the distance and direction to the road structure are detected by a sensor, and the detected distance and direction, and This means that the position of the host vehicle is estimated from the acquired position of the road structure.

[Adjustment processing of road structure model]
As shown in FIG. 4, the road structure model adjustment unit 13 detects the weather of the road on which the vehicle is traveling, and uses the road structure to estimate the road structure from a plurality of road structure models according to the detected weather. Select a structural model.

  The road structure model creation unit 12 creates road structure models for different road structures, and FIG. 4 illustrates a case where four road structure models are created as an example. The road structure model illustrated in FIG. 4 includes a road structure model created for a white line, a road structure model created for a curb, a road structure model created for a target, and a road structure created for a road edge. It is a model.

  The road structure model with white lines is a road structure model created for road lane markings such as white lines and yellow lines. Therefore, the road structure model with white lines is a two-dimensional model representing the texture of the road surface, and has a property that is weak against three-dimensional occlusion.

  The curb road structure model is a road structure model created for a step set at the boundary between a roadway and a sidewalk. Therefore, it is a model representing a three-dimensional three-dimensional object having a substantially fixed shape with a height of about 20 cm, and has a weak property against two-dimensional occlusion.

  The target road structure model is a road structure model created for road signs and traffic lights installed around the road. Therefore, it is a model representing a three-dimensional solid object having various shapes, and has a strong property against two-dimensional and three-dimensional occlusion. However, it is vulnerable to moving objects such as animals.

  The road structure model at the road edge is a road structure model created for the boundary between the paved portion of the road and the soil portion outside it. Therefore, it is a model representing a three-dimensional solid object with a small difference in height, and has a property more resistant to three-dimensional occlusion than a road structure model with white lines.

  As shown in FIG. 4, when the road structure model adjustment unit 13 determines the weather and determines that it is not snowing, the road structure model adjustment unit 13 performs adjustment so that the road structure is estimated using all road structure models. At this time, the road structure model adjustment unit 13 may perform adjustment so as to weight each road structure model. For example, the white line model is set to 0.5, the curb model is set to 0.1, the target model is set to 0.2, and the road end model is set to 0.2, so that weighting is performed to be 1.0.

  Further, the road structure model adjusting unit 13 may adjust the number of points on the road structure acquired from each road structure model. For example, 50 points on the white line are acquired from the white line model, 30 points on the curb from the curb model, 10 points on the target from the target model, and on the road end from the road end model 40 points may be acquired.

  Next, a case where the weather is determined and it is determined that there is snow will be described. As shown in FIG. 4, when the detected weather is snowfall, the road structure model adjustment unit 13 adjusts so as to estimate the position of the road structure around the vehicle using a road structure model other than the white line model. To do. That is, when it is determined that there is snow due to the weather, adjustment is performed so that the position of the road structure is estimated without using the white line model.

  Similarly, when the detected weather is snowfall, the road structure model adjustment unit 13 performs adjustment so as to estimate the position of the road structure around the vehicle using a road structure model other than the road edge model. That is, when it is determined that there is snow due to the weather, adjustment is performed so that the position of the road structure is estimated without using the road end model.

  When the weather is snowing, there is a high possibility that the white line and the road edge on the road will be hidden, and the accuracy of the data of the white line and the road edge obtained in that situation is low and the number is also reduced. Therefore, when the road structure is estimated using such data, the estimation accuracy of the road structure may be lowered. Therefore, when the weather is snowing, the road structure is estimated without using the white line model or the road edge model so that the estimation accuracy of the road structure can be prevented from decreasing, and the self-position estimation accuracy can be reduced. I try to improve.

  When there is snow on the road, it is difficult to detect white lines and road edges. Therefore, the presence or absence of snow is determined from images taken on the road around the vehicle taken by the camera 2. If it is determined that there is snow, the white line model and the road end model are not used regardless of whether it is snowing or not. You may adjust as follows.

  The road structure model adjustment unit 13 measures the amount of snow on the road when the detected weather is snowfall, and if the measured amount of snow on the road is equal to or greater than a predetermined value, the curbstone. Adjustment is performed so that the position of the road structure around the vehicle is estimated using a road structure model other than the model. In other words, when it is determined that there is snowfall by judging the weather, and the amount of snow on the road is greater than or equal to a predetermined value, adjustment is made so that the position of the road structure is estimated without using the curb model.

  If the snow on the road becomes as high as the curb, the curb along the road is likely to be hidden, and the accuracy of the curb data obtained in that situation is low and the number is low. Therefore, when the road structure is estimated using such data, the estimation accuracy of the road structure may be lowered. Therefore, when it is determined that the weather is snowing and the amount of snow on the road is greater than or equal to a predetermined value, the road structure is estimated without using the curb model to prevent a decrease in the estimation accuracy of the road structure. This improves the accuracy of self-position estimation.

  In addition, what is necessary is just to set the predetermined value for determining the amount of snow accumulation based on the height of a curb. Further, as a method for detecting the amount of snow on the road, it may be detected from an image on the road around the vehicle taken by the camera 2.

  Furthermore, the road structure model adjustment unit 13 measures the amount of snowfall when the detected weather is snowfall, and the road structure other than the target model when the measured snowfall amount is equal to or greater than a predetermined value. Adjust to estimate the position of the road structure around the vehicle using the model. That is, when it is determined that there is snowing by determining the weather and the snowfall amount is a predetermined value or more, adjustment is performed so as to estimate the position of the road structure without using the target model.

  If the amount of snowfall increases and the visibility becomes worse, there is a high possibility that signs and traffic lights along the road will not be visible, and the accuracy of the target data obtained in that situation will be poor and the number will be reduced. Therefore, when the road structure is estimated using such data, the estimation accuracy of the road structure may be lowered. Therefore, when it is determined that the weather is snowing and the amount of snowfall is greater than or equal to a predetermined value, the estimation of the road structure can be prevented from being reduced by estimating the road structure without using the target model. As a result, the accuracy of self-position estimation is improved.

  Note that the predetermined value for determining the snowfall amount may be set to a snowfall amount value such that visibility is deteriorated. Moreover, what is necessary is just to detect from the image around the vehicle image | photographed with the camera 2 as a snowfall detection method.

  As described above, when the weather is snowing, the road structure model used for estimating the road structure is selected by the method shown in FIG. The road structure model adjustment unit 13 performs adjustment so that the road structure is estimated using the selected road structure model. At this time, the road structure model adjustment unit 13 may weight the selected road structure model, or may adjust the number of points on the road structure acquired from each road structure model.

[Road structure estimation process]
Next, with reference to the flowchart of FIG. 5, the processing procedure of the road structure estimation process by the road structure estimation apparatus 1 according to the present embodiment will be described.

  In step S101, the road structure model adjustment unit 13 detects the weather of the road on which the vehicle is traveling. The road structure model adjustment unit 13 detects the weather based on the image captured by the camera 2 and information from the outside. Specifically, the road structure model adjustment unit 13 determines whether or not the weather is snowing, whether or not the amount of snow on the road is greater than or equal to a predetermined value, whether or not the amount of snowfall is greater than or equal to a predetermined value, etc. I do.

  In step S102, the road structure model adjustment unit 13 performs a process of adjusting a road structure model used for estimating the road structure among a plurality of road structure models according to the weather detected in step S101. In the present embodiment, as a road structure model adjustment process, a process for selecting a road structure model to be used for estimating a road structure from a plurality of road structure models is performed. The specific adjustment process is as described with reference to FIG.

  In step S103, the road structure estimation unit 15 estimates the position of the road structure around the vehicle using the road structure model adjusted in step S102. In the present embodiment, the position of the road structure around the vehicle is estimated using a road structure model selected according to the presence / absence of snowfall, the amount of snowfall, and the amount of snow on the road. Then, when the position of the road structure is estimated, the road structure estimation unit 15 links to the map database 6 and records the estimated position of the road structure.

  In step S104, when the self-position estimating unit 17 estimates the self-position using the position of the road structure recorded in the map database 6 in step S103, the road structure estimating process according to the present embodiment is terminated. In the road structure estimation process of the present embodiment, the position of the road structure on the map is accurately estimated, so that the self-position estimation unit 17 can accurately estimate the self-position.

[Modification]
Next, a modification of the first embodiment will be described. In the above-described embodiment, the road structure model used for estimating the road structure is selected according to the weather. However, the road structure model may be selected according to the place where the vehicle is traveling in addition to the weather. .

  For example, as shown in FIG. 6, there are cases where there are places where a snow removal part and a snow removal part are mixed in an intersection and a sufficient number of accurate detection values cannot be obtained. . In FIG. 6, there are a point 51 where a sufficient number of accurate detection values cannot be obtained and a point 53 where a sufficient number of accurate detection values can be obtained. Moreover, there is no white line in the intersection, and the road edge does not exist unless it is outside the intersection. Therefore, when it is determined that the vehicle is in the intersection based on the position information from the GPS receiver 4 or the image taken by the camera 2, the white line model and the road end model can be adjusted not to be used.

  If it is determined from the image taken by the camera 2 that the vehicle is traveling near the construction site, the white line may be redrawn or the road edge may be changed. You may adjust so that a road edge model may not be used.

  In addition, when it is determined that the road on which the vehicle is traveling is an agricultural road or a mountain road, the white line and the road edge model may be deteriorated due to poor maintenance. You may adjust it.

  In addition to the weather, a road structure model may be selected according to the situation in which the vehicle is traveling. For example, if it is determined from the image taken by the camera 2 that the western sun around the vehicle is strong, the white line detection accuracy is lowered due to a decrease in the color contrast of the image due to the western sun, so the white line model is not used. You may adjust it.

[Effect of the first embodiment]
As described above in detail, the road structure estimation device 1 according to the present embodiment creates road structure models for different road structures. Then, according to the weather of the road on which the vehicle travels, a process for adjusting the road structure model used for estimating the road structure among a plurality of road structure models is performed, and the vehicle surroundings using the adjusted road structure model The position of the road structure is estimated. Thereby, since the road structure model can be adjusted according to the weather, the position of the road structure can be accurately estimated without being influenced by the weather.

  Moreover, in the road structure estimation apparatus 1 according to the present embodiment, as a process for adjusting the road structure model, a process for selecting a road structure model to be used for estimating the road structure from a plurality of road structure models is performed. Thereby, since the optimal road structure model can be selected according to the weather, the position of the road structure can be accurately estimated without being influenced by the weather.

  Furthermore, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall, the vehicle surroundings using a road structure model other than the road structure model created for the white line among the plurality of road structure models The position of the road structure is estimated. As a result, it is possible to estimate the road structure without using a white line road structure model that is highly likely to be hidden in the case of snowfall, so it is possible to prevent a reduction in the estimation accuracy of the road structure. .

  Further, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall, a vehicle using a road structure model other than the road structure model created for the road edge among a plurality of road structure models. Estimate the position of the surrounding road structure. As a result, it is possible to estimate the road structure without using a road structure model at the edge of the road that is highly likely to be hidden in the case of snowfall, so it is possible to prevent a decrease in the estimation accuracy of the road structure. Become.

  Furthermore, in the road structure estimation apparatus 1 according to the present embodiment, when the detected weather is snowfall and the measured amount of snow on the road is a predetermined value or more, the curb road structure among the plurality of road structure models. The position of the road structure around the vehicle is estimated using a road structure model other than the model. As a result, it is possible to estimate the road structure without using a curb road structure model that is likely to be hidden if there is a lot of snow on the road. It becomes possible.

  Moreover, in the road structure estimation apparatus 1 according to the present embodiment, when the detected weather is snowfall and the measured snowfall amount is equal to or greater than a predetermined value, the road structure model other than the target road structure model among the plurality of road structure models The position of the road structure around the vehicle is estimated using this road structure model. As a result, it is possible to estimate the road structure without using a target road structure model that is highly likely to disappear when the amount of snowfall increases, so it is possible to prevent a decrease in the estimation accuracy of the road structure. Become.

[Second Embodiment]
Hereinafter, a second embodiment of the present invention will be described with reference to the drawings. In the description of the drawings, the same portions are denoted by the same reference numerals, and description thereof is omitted. In the present embodiment, the road structure model adjustment processing by the road structure model adjustment unit 13 is different from the first embodiment. Therefore, detailed description of other configurations and processes is omitted.

[Adjustment processing of road structure model]
As shown in FIG. 7, the road structure model adjustment unit 13 detects the weather of the road on which the vehicle travels, and performs a process of adjusting the estimated sensitivity of the road structure in the plurality of road structure models according to the detected weather. Do. FIG. 7 illustrates a case where four road structure models are created as an example. The road structure model illustrated in FIG. 7 includes a road structure model created for white lines, a road structure model created for curbs, a road structure model created for targets, and a road structure created for road edges. It is a model.

  As shown in FIG. 7, when the road structure model adjustment unit 13 determines the weather and determines that it is not snowing, the road structure model adjusting unit 13 makes the estimated sensitivity of all road structure models lower than when the weather is snowing. Adjust to. Then, the road structure model adjustment unit 13 performs adjustment so that the road structure is estimated using all road structure models whose estimation sensitivity is lowered. At this time, the road structure model adjustment unit 13 may weight each road structure model, or may adjust the number of points on the road structure acquired from each road structure model.

  Here, the estimated sensitivity of the road structure will be described with reference to FIGS. FIG. 8 is a diagram illustrating a case where the number of detected points on the white line is large when the position of the white line is estimated as the road structure. FIG. 9 is a diagram illustrating a case where the number of detected points on the white line is small when the position of the white line is estimated as the road structure.

  As shown in FIG. 8, when the weather is clear rather than snowfall, many points on the white line as the road structure can be detected. In FIG. 8, five points P1 to P5 are detected. Can be detected. A curve W in FIG. 8 represents the actual position of the white line.

  In this case, if the estimation sensitivity of the road structure is increased, the white line estimation result H is strongly influenced by the positions of the points P1 to P5, and thus becomes a curved meandering left and right. On the other hand, when the estimation sensitivity of the road structure is lowered, the white line estimation result L becomes a smooth curve because the influence of the positions of the points P1 to P5 is weakened. Therefore, when the weather is clear rather than snowfall, many points on the road structure can be detected. Therefore, the road structure estimation accuracy can be improved by lowering the road structure estimation sensitivity.

  On the other hand, as shown in FIG. 9, when the weather is snowing, the number of points on the white line that can be detected decreases, and in FIG. 9, only three points P1, P3, and P5 can be detected. Points P1, P3, and P5 in FIG. 9 represent the same points as the points P1, P3, and P5 in FIG. The curve W also represents the actual position of the white line as in FIG.

  In this case, if the road structure estimation sensitivity is lowered, the number of points is small, and the white line estimation result L becomes a curve connecting intermediate positions of the points P1, P3, and P5. On the other hand, if the estimation sensitivity of the road structure is increased, the white line estimation result H becomes a curve connecting the positions of the points P1, P3, and P5, so that the position of the white line can be accurately estimated even if the number of points is small. can do. Therefore, when the weather is snowing, the number of points on the road structure that can be detected is reduced. Therefore, the road structure estimation accuracy can be improved by increasing the road structure estimation sensitivity.

  Next, a case where the weather is determined and it is determined that it is snowing will be described. As shown in FIG. 7, when the detected weather is snowfall, the road structure model adjustment unit 13 makes the estimated sensitivity of the white line model higher than when the weather is not snowy. Similarly, the road structure model adjustment unit 13 increases the estimated sensitivity of the road end model when the detected weather is snowing compared to when the weather is not snowing.

  When the weather is snowing, there is a high possibility that the white line and the road edge on the road will be hidden, and the number of white line and road edge data obtained in that situation will be reduced. Therefore, when the estimation sensitivity is lowered, the estimation accuracy of the road structure may be lowered as described with reference to FIG. Therefore, when the weather is snowing, the estimation accuracy of the white line model and the road edge model is increased to prevent a decrease in the estimation accuracy of the road structure, and as a result, the estimation accuracy of the self-location is improved. ing.

  The road structure model adjustment unit 13 measures the amount of snow on the road when the detected weather is snowfall, and if the measured amount of snow on the road is equal to or greater than a predetermined value, the curbstone. The estimated sensitivity of the model is set higher than when the amount of snow is less than a predetermined value.

  If the snow on the road rises to the height of the curb, the curb along the road is more likely to be hidden, and the number of curb data obtained in that situation is reduced. Therefore, when the estimation sensitivity is lowered, the estimation accuracy of the road structure may be lowered as described with reference to FIG. Therefore, when it is determined that the weather is snowing and the amount of snow on the road is greater than or equal to a predetermined value, by reducing the estimation sensitivity of the curb model, it is possible to prevent a decrease in the estimation accuracy of the road structure, As a result, the self-position estimation accuracy is improved.

  Further, the road structure model adjustment unit 13 measures the amount of snowfall when the detected weather is snowing, and if the measured amount of snowfall is equal to or greater than a predetermined value, the road structure model adjustment unit 13 increases the target model estimation sensitivity. The amount of snowfall is set higher than when it is less than a predetermined value.

  If the amount of snowfall increases and visibility becomes worse, there is a high possibility that signs and traffic lights along the road will not be visible, and the number of target data obtained in that situation will be reduced. Therefore, when the estimation sensitivity is lowered, the estimation accuracy of the road structure may be lowered as described with reference to FIG. Therefore, if it is determined that the weather is snowing and the amount of snowfall is greater than or equal to a predetermined value, the estimation sensitivity of the target model can be increased to prevent a reduction in the estimation accuracy of the road structure, and in turn The position estimation accuracy is improved.

  Note that when the number of target data that can be detected is small regardless of the weather, the target model estimation sensitivity may be increased regardless of whether the weather is snowing or not.

  As described above, when the weather is snowing, the estimation sensitivity of the road structure in the road structure model is adjusted by the method shown in FIG. The road structure model adjustment unit 13 performs adjustment so that the road structure is estimated using the road structure model with the estimated sensitivity adjusted. At this time, the road structure model adjustment unit 13 may weight the adjusted road structure model, or may adjust the number of points on the road structure acquired from each road structure model.

[Effects of Second Embodiment]
As described above in detail, the road structure estimation device 1 according to the present embodiment performs a process of adjusting the estimation sensitivity of the road structure in a plurality of road structure models as the process of adjusting the road structure model. Thereby, since it can adjust to the optimal estimation sensitivity according to the weather, a road structure can be estimated correctly regardless of the weather.

  Further, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall, the estimated sensitivity of the road structure model created for the white line among a plurality of road structure models is used, when the weather is not snowy. Higher than. Thereby, since the estimation sensitivity of the road structure model of the white line with high possibility of being hidden in the case of snowfall can be made high, it becomes possible to prevent the estimation accuracy of the road structure from being lowered.

  Furthermore, in the road structure estimation apparatus 1 according to the present embodiment, when the detected weather is snowfall, the estimated sensitivity of the road structure model created for the road edge among the plurality of road structure models is used. Higher than if not. Thereby, since the estimation sensitivity of the road structure model of the road edge which is highly likely to be hidden in the case of snowfall can be increased, it is possible to prevent the estimation accuracy of the road structure from being lowered.

  Further, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall and the measured amount of snow on the road is equal to or greater than a predetermined value, the curb road structure among a plurality of road structure models. The estimated sensitivity of the model is set higher than when the amount of snow is less than a predetermined value. As a result, it is possible to increase the estimation sensitivity of the curb road structure model, which is likely to be hidden when there is a lot of snow on the road, and thus it is possible to prevent a decrease in the estimation accuracy of the road structure.

  Furthermore, in the road structure estimation apparatus 1 according to the present embodiment, when the detected weather is snowfall and the measured snowfall amount is equal to or greater than a predetermined value, the road structure model of the target among the plurality of road structure models. The estimated sensitivity is set higher than when the amount of snowfall is less than a predetermined value. As a result, it is possible to increase the estimation sensitivity of the road structure model of a target that is highly likely to disappear when the amount of snowfall increases, and thus it is possible to prevent a decrease in the estimation accuracy of the road structure.

[Third Embodiment]
Hereinafter, a third embodiment of the present invention will be described with reference to the drawings. In the description of the drawings, the same portions are denoted by the same reference numerals, and description thereof is omitted. In the present embodiment, the road structure model adjustment processing by the road structure model adjustment unit 13 is different from the first and second embodiments. Therefore, detailed description of other configurations and processes is omitted.

[Adjustment processing of road structure model]
As shown in FIG. 10, the road structure model adjustment unit 13 detects the weather of the road on which the vehicle travels, and adjusts the error range of the position of the road structure in the plurality of road structure models according to the detected weather. Process. In FIG. 10, a case where four road structure models are created will be described as an example. The road structure model illustrated in FIG. 10 includes a road structure model created for white lines, a road structure model created for curbstones, a road structure model created for targets, and a road structure created for road edges. It is a model.

  As shown in FIG. 10, when the road structure model adjustment unit 13 determines that the weather is not snowing by determining the weather, the error range of the position of the road structure in all the road structure models is represented by the case where the weather is snowing. Adjust to make it larger. Then, the road structure model adjustment unit 13 performs adjustment so that the road structure is estimated using all road structure models having a large error range. At this time, the road structure model adjustment unit 13 may weight each road structure model, or may adjust the number of points on the road structure acquired from each road structure model.

  Here, the error range of the position of the road structure will be described with reference to FIGS. FIG. 11 is a diagram illustrating a case where the number of detected points on the white line is large when the position of the white line is estimated as the road structure. FIG. 12 is a diagram illustrating a case where the number of detected points on the white line is small when the position of the white line is estimated as the road structure.

  As shown in FIG. 11, when the weather is clear rather than snowfall, many points on the white line that is a road structure can be detected. In FIG. 11, five points P1 to P5 are detected. Can be detected. A curve W in FIG. 11 represents the actual position of the white line.

  In this case, if the error range of the position of the road structure is reduced, the white line estimation result S passes through the vicinity of the center of the points P1 to P5, and thus becomes a curved meandering left and right. On the other hand, when the error range of the position of the road structure is increased, the white line estimation result B passes through the vicinity of the positions of the points P1 to P5 and thus becomes a smooth curve. Therefore, when the weather is clear rather than snowfall, many points on the road structure can be detected. Therefore, the road structure estimation accuracy can be improved by increasing the error range of the road structure position.

  On the other hand, as shown in FIG. 12, when the weather is snowing, the number of points on the white line that can be detected decreases, and in FIG. 12, only three points P1, P3, and P5 can be detected. Points P1, P3, and P5 in FIG. 12 represent the same points as the points P1, P3, and P5 in FIG. A curve W also represents the actual position of the white line, as in FIG.

  In this case, if the error range of the position of the road structure is increased, the number of points is small, and the white line estimation result B becomes a curve connecting the intermediate positions of the points P1, P3, and P5. On the other hand, when the error range of the position of the road structure is reduced, the white line estimation result S becomes a curve connecting the positions of the points P1, P3, and P5, so that the position of the white line can be accurately determined even if the number of points is small. Can be estimated. Therefore, when the weather is snowing, the number of points on the road structure that can be detected is reduced. Therefore, the road structure estimation accuracy can be improved by reducing the error range of the road structure position.

  Next, a case where the weather is determined and it is determined that it is snowing will be described. As shown in FIG. 10, when the detected weather is snowfall, the road structure model adjustment unit 13 makes the error range of the white line model smaller than when the weather is not snowy. Similarly, the road structure model adjustment unit 13 makes the error range of the road end model smaller when the detected weather is snowing than when the weather is not snowing.

  When the weather is snowing, there is a high possibility that the white line and the road edge on the road will be hidden, and the number of white line and road edge data obtained in that situation will be reduced. Therefore, if the error range is increased, the estimation accuracy of the road structure may be lowered as described with reference to FIG. Therefore, when the weather is snowing, the error range of the white line model and the road edge model can be reduced to prevent a decrease in the estimation accuracy of the road structure, which in turn improves the estimation accuracy of the self-location. ing.

  The road structure model adjustment unit 13 measures the amount of snow on the road when the detected weather is snowfall, and if the measured amount of snow on the road is equal to or greater than a predetermined value, the curbstone. The error range of the model is made smaller than when the amount of snow is less than a predetermined value.

  If the snow on the road rises to the height of the curb, the curb along the road is more likely to be hidden, and the number of curb data obtained in that situation is reduced. Therefore, if the error range is increased, the estimation accuracy of the road structure may be lowered as described with reference to FIG. Therefore, when it is determined that the weather is snowing and the amount of snow on the road is greater than or equal to a predetermined value, by reducing the error range of the curb model, it is possible to prevent a reduction in the estimation accuracy of the road structure, As a result, the self-position estimation accuracy is improved.

  Further, the road structure model adjusting unit 13 measures the amount of snowfall when the detected weather is snowing, and if the measured amount of snowfall is equal to or greater than a predetermined value, the road structure model adjusting unit 13 sets the error range of the target model. The amount of snowfall is made smaller than when it is less than a predetermined value.

  If the amount of snowfall increases and visibility becomes worse, there is a high possibility that signs and traffic lights along the road will not be visible, and the number of target data obtained in that situation will be reduced. Therefore, if the error range is increased, the estimation accuracy of the road structure may be lowered as described with reference to FIG. Therefore, if it is determined that the weather is snowing and the amount of snowfall is greater than or equal to a predetermined value, the error range of the target model can be reduced to prevent a decrease in the estimation accuracy of the road structure, and in turn The position estimation accuracy is improved.

  If the number of target data that can be detected is small irrespective of the weather, the error range of the target model may be reduced regardless of whether the weather is snowing or not.

  As described above, when the weather is snowing, the error range of the position of the road structure in the road structure model is adjusted by the method shown in FIG. The road structure model adjustment unit 13 performs adjustment so that the road structure is estimated using the road structure model in which the error range is adjusted. At this time, the road structure model adjustment unit 13 may weight the adjusted road structure model, or may adjust the number of points on the road structure acquired from each road structure model.

[Effect of the third embodiment]
As described above in detail, the road structure estimation device 1 according to the present embodiment performs a process of adjusting the error range of the position of the road structure in the plurality of road structure models as the process of adjusting the road structure model. Thereby, since it can adjust to the optimal error range according to the weather, the position of the road structure can be accurately estimated without being influenced by the weather.

  Further, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall, the error range of the road structure model created for the white line among the plurality of road structure models is displayed, when the weather is not snowy. Smaller than. As a result, the error range of the road structure model of the white line that is highly likely to be hidden in the case of snowfall can be reduced, so that it is possible to prevent a decrease in the estimation accuracy of the road structure.

  Furthermore, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall, the error falls within the error range of the road structure model created for the road edge among the plurality of road structure models. Make it smaller than not. As a result, the error range of the road structure model at the road edge that is highly likely to be hidden in the case of snowfall can be reduced, so that it is possible to prevent a decrease in the estimation accuracy of the road structure.

  Further, in the road structure estimation device 1 according to the present embodiment, when the detected weather is snowfall and the measured amount of snow on the road is equal to or greater than a predetermined value, the curb road structure among a plurality of road structure models. The error range of the model is made smaller than when the amount of snow is less than a predetermined value. As a result, the error range of the curb road structure model, which has a high possibility of being hidden when the amount of snow on the road increases, can be reduced, so that it is possible to prevent a reduction in the estimation accuracy of the road structure.

  Moreover, in the road structure estimation apparatus 1 according to the present embodiment, when the detected weather is snowfall and the measured snowfall amount is equal to or greater than a predetermined value, the road structure model of the target among the plurality of road structure models. The error range is made smaller than when the snowfall amount is less than a predetermined value. Thereby, since the error range of the road structure model of the target that is highly likely to disappear when the amount of snowfall increases can be reduced, it is possible to prevent the estimation accuracy of the road structure from being lowered.

  The above-described embodiment is an example of the present invention. For this reason, the present invention is not limited to the above-described embodiment, and even if it is a form other than this embodiment, as long as it does not depart from the technical idea of the present invention, it depends on the design and the like. Of course, various modifications are possible.

DESCRIPTION OF SYMBOLS 1 Road structure estimation apparatus 2 Camera 3 Laser range finder 4 GPS receiver 5 Gyro sensor 6 Map database 7 Communication apparatus 10 Controller 11 Traveling track model creation part 12 Road structure model creation part 13 Road structure model adjustment part 15 Road structure estimation part 17 Self-position estimation unit 20 server

Claims (17)

A road structure model that is obtained by learning and modeling the road structure around the vehicle for each different road structure, and using the created road structure models, the road structure estimation that estimates the position of the road structure around the vehicle A device road structure estimation method comprising:
Detect the weather on the road where the vehicle is running,
According to the detected weather, a process for adjusting the road structure model used for estimating the road structure among the plurality of road structure models is performed,
A road structure estimation method, wherein a position of a road structure around a vehicle is estimated using the adjusted road structure model.   The road structure estimation method according to claim 1, wherein the process of adjusting the road structure model is a process of selecting a road structure model to be used for estimating a road structure from the plurality of road structure models. .   When the detected weather is snowfall, the position of the road structure around the vehicle is estimated using a road structure model other than the road structure model created for the white line among the plurality of road structure models. The road structure estimation method according to claim 2.   When the detected weather is snowfall, the position of the road structure around the vehicle is estimated using a road structure model other than the road structure model created for the road edge among the plurality of road structure models. The road structure estimation method according to claim 3, wherein the road structure is estimated. If the detected weather is snowfall, measure the amount of snow on the road,
If the measured amount of snow on the road is greater than or equal to a predetermined value, the position of the road structure around the vehicle using a road structure model other than the road structure model created for the curb among the plurality of road structure models The road structure estimation method according to claim 3, wherein the road structure is estimated. If the detected weather is snowfall, measure the amount of snowfall,
When the measured amount of snowfall is equal to or greater than a predetermined value, the position of the road structure around the vehicle is estimated using a road structure model other than the road structure model created for the target among the plurality of road structure models. The road structure estimation method according to any one of claims 3 to 5, wherein:   The road structure estimation method according to claim 1, wherein the process of adjusting the road structure model is a process of adjusting an estimation sensitivity of the road structure in the plurality of road structure models.   8. When the detected weather is snowfall, the estimated sensitivity of the road structure model created for the white line among the plurality of road structure models is made higher than that when the weather is not snowing. The road structure estimation method described.   9. When the detected weather is snowfall, the estimated sensitivity of a road structure model created for a road edge among the plurality of road structure models is set higher than that when the weather is not snowy. The road structure estimation method described in 1. If the detected weather is snowfall, measure the amount of snow on the road,
If the measured amount of snow on the road is greater than or equal to a predetermined value, the estimated sensitivity of the road structure model created for the curbstone among the plurality of road structure models is less than the predetermined value. The road structure estimation method according to claim 8 or 9, wherein the road structure is higher than the case. If the detected weather is snowfall, measure the amount of snowfall,
When the measured amount of snowfall is greater than or equal to a predetermined value, the estimated sensitivity of the road structure model created for the target among the plurality of road structure models is greater than when the measured amount of snowfall is less than the predetermined value. The road structure estimation method according to any one of claims 8 to 10, wherein the road structure is also increased.   The road structure estimation method according to claim 1, wherein the process of adjusting the road structure model is a process of adjusting an error range of a road structure position in the plurality of road structure models.   13. When the detected weather is snowfall, an error range of a road structure model created for a white line among the plurality of road structure models is made smaller than that when the weather is not snowing. The road structure estimation method described.   14. When the detected weather is snowing, an error range of a road structure model created for a road edge among the plurality of road structure models is made smaller than that when the weather is not snowing. The road structure estimation method described in 1. If the detected weather is snowfall, measure the amount of snow on the road,
When the measured amount of snow on the road is greater than or equal to a predetermined value, the measured snow amount is less than the predetermined value within the error range of the road structure model created for the curb among the plurality of road structure models 15. The road structure estimation method according to claim 13 or 14, wherein the road structure estimation method is smaller than the case. If the detected weather is snowfall, measure the amount of snowfall,
When the measured amount of snowfall is greater than or equal to a predetermined value, the error range of the road structure model created for the target among the plurality of road structure models is compared with the case where the measured amount of snowfall is less than the predetermined value. The road structure estimation method according to claim 13, wherein the road structure is also made smaller. A road structure model that is obtained by learning and modeling the road structure around the vehicle for each different road structure, and using the created road structure models, the road structure estimation that estimates the position of the road structure around the vehicle A device,
Detect the weather on the road where the vehicle is running,
According to the detected weather, a process for adjusting the road structure model used for estimating the road structure among the plurality of road structure models is performed,
A road structure estimation apparatus characterized by estimating a position of a road structure around a vehicle using the adjusted road structure model.

Images