JP2017111566A - Lane information estimation apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology of estimating information on a lane of a vehicle when lane markings on both sides of the vehicle cannot be detected.SOLUTION: A lane information estimation apparatus acquires an adjacent parameter including at least one of adjacent shape which represents a shape of a lane for an adjacent lane which is located adjacent to a lane where the vehicle is traveling, out of lanes representing areas where vehicles travel on a road, and a parameter representing a traveling state of the vehicle with respect to the adjacent shape. (S225) In the lane information estimation apparatus, a lane parameter estimation unit estimates a lane parameter including at least one of a lane shape representing a shape of a lane where the vehicle is traveling, and a parameter representing a traveling state of the vehicle with respect to the lane shape, by use of a situation where the adjacent lane is adjacent to the lane where the vehicle is traveling (S230).SELECTED DRAWING: Figure 6

Description

  The present invention relates to a technique for estimating information related to a lane in which a host vehicle travels.

  Conventionally, a technology for detecting a white line drawn on a road and acquiring information on a lane (hereinafter referred to as the own lane) on which the vehicle travels among lanes sandwiched between the white lines based on the detection result of the white line is known. It has been. Patent Document 1 proposes a technique for detecting a white line drawn on a road from an image showing the front of the host vehicle acquired by an in-vehicle camera, and acquiring information about the host lane based on the detection result of the white line.

JP 2009-181310 A

  However, when the front of the host vehicle is blocked by a large vehicle such as a truck, white lines on both sides of the host lane cannot be detected, and therefore information on the host lane such as the lane width and curvature cannot be acquired. The problem can arise. In addition, as a line showing a lane, the line represented by the arrangement | sequence of solid objects, such as not only a white line but the yellow line drawn on the road and the curb installed on the road surface, is mentioned. A line representing such a lane is called a lane marking.

  The present invention has been made in view of these problems, and an object of the present invention is to provide a technique for estimating information related to the own lane when the lane markings on both sides of the own lane are not detected.

  The own lane information estimation device (50) of the present invention includes an adjacent parameter acquisition unit (S225) and an own parameter estimation unit (S230, S275). The adjacent parameter acquisition unit is an adjacent shape that represents the shape of the lane for the adjacent lane that is adjacent to the own lane in which the host vehicle is traveling among the lane that represents the region in which the vehicle is traveling on the road, and the host vehicle for the adjacent shape. An adjacent parameter including at least one of the parameters representing the running state is acquired. The own parameter estimator uses the adjacent parameter to indicate that the adjacent lane and the own lane are adjacent to each other, and the own shape representing the shape of the lane with respect to the own lane, and the parameter representing the running state of the own vehicle with respect to the own shape The own lane parameter including at least one of them is estimated.

  According to such a configuration, even if a situation where the white lines on both sides of the own lane cannot be detected occurs, the adjacent lane and the own lane are adjacent if the adjacent parameter that is information on the adjacent lane can be acquired. Based on this, it is possible to detect the own parameter which is information on the own lane.

  In addition, the code | symbol in the parenthesis described in this column and a claim shows the correspondence with the specific means as described in embodiment mentioned later as one aspect, Comprising: The technical scope of this invention is shown. It is not limited.

The block diagram which shows the structure of the vehicle control system of the 1st Embodiment, and the own lane information estimation apparatus. The flowchart of a traveling control process. The figure which shows an example of the image which imaged the front of the own vehicle. The figure which shows the example of a detection of the white line in the image of FIG. The figure which shows an example of the road where a traveling control process is applied. The flowchart of a road parameter estimation process. The figure which shows the example of a detection of a ranging point. The figure which shows the detection point which consists of a new measurement base and a past point. The figure which shows the road parameter in an adjacent road. The flowchart which shows the estimation range setting process of 1st Embodiment. The flowchart which shows the estimation range setting process of 2nd Embodiment. The figure which shows the other example of the road where a traveling control process is applied.

Hereinafter, embodiments for carrying out the invention will be described with reference to the drawings.
[1. First Embodiment]
[1-1. Constitution]
A vehicle control system 1 shown in FIG. 1 is a system mounted on a vehicle. Hereinafter, a vehicle equipped with the vehicle control system 1 is referred to as a host vehicle. The vehicle control system 1 includes a laser radar 11, a GPS receiver 12, a direction indicator sensor 13, a map database (hereinafter referred to as map DB) 14, a vehicle speed sensor 15, a steering angle sensor 16, and a vehicle control unit 17. And an own lane information estimation device (hereinafter referred to as an estimation device) 50.

  The laser radar 11 is provided in the front part of the host vehicle. The laser radar 11 irradiates a laser beam on the road surface ahead of the traveling direction of the host vehicle, and receives a laser beam (hereinafter, reflected light) reflected by an object on the road surface. The object here includes at least a white line drawn on the road surface in addition to a three-dimensional object such as a guardrail or a curb stone installed on the road surface. The curbstone refers to, for example, a rod-shaped stone made of concrete or the like, which is installed as a boundary line that divides an area where the vehicle travels and a retreat area or a sidewalk for retreating the vehicle.

  Based on the received reflected light, the laser radar 11 measures the distance from the host vehicle to a point where the laser beam is reflected (hereinafter referred to as a ranging point) and the direction of the ranging point based on the traveling direction of the host vehicle. And the relative speed of the distance measuring point. Further, the laser radar 11 detects the received light intensity of the reflected light from the distance measuring point. The laser radar 11 outputs detection signals representing these detection results.

  The GPS receiver 12 receives a GPS signal transmitted from a GPS satellite and detects the current position and traveling direction of the vehicle. The current position is represented using an absolute position represented by latitude and longitude coordinates.

The direction indicator sensor 13 outputs a detection signal indicating the operation state of the direction indicator.
The map DB 14 records map data representing roads on which vehicles can travel, which is stored in a database according to latitude and longitude coordinates.

The vehicle speed sensor 15 detects the speed of the host vehicle and outputs a detection signal indicating the detected speed.
The steering angle sensor 16 detects the steering angle of the host vehicle and outputs a detection signal indicating the detected speed.

  The vehicle control unit 17 executes various controls for causing the host vehicle to travel based on the road parameters estimated by the estimation device 50. The vehicle control unit 17 includes, for example, an engine ECU that controls the engine, a brake ECU that controls the braking force, and a steering ECU that controls the steering angle.

  The estimation device 50 is mainly configured by a known microcomputer having a CPU 51 and a semiconductor memory (hereinafter, memory 52) such as a RAM, a ROM, and a flash memory. Various functions of the estimation device 50 are realized by the CPU 51 executing a program stored in a non-transitional physical recording medium. In this example, the memory 52 corresponds to a non-transitional tangible recording medium that stores a program. Further, by executing this program, a method corresponding to the program is executed. Note that the number of microcomputers constituting the estimation device 50 may be one or more.

  The estimation device 50 executes a travel control process to be described later according to a program stored in the memory 52. The estimation device 50 may realize some or all elements of the functions realized by executing the traveling control process using hardware that combines a logic circuit, an analog circuit, and the like.

[1-2. processing]
[1-2-1. Travel control processing]
The travel control process executed by the CPU 51 of the estimation device 50 will be described with reference to the flowchart of FIG. The travel control process is a process for acquiring information representing the travel environment ahead of the host vehicle and controlling the travel of the host vehicle based on the information representing the travel environment.

  The information representing the driving environment includes information representing the shape of the lane such as the lane width that is the width of the lane and the curvature of the lane, that is, the road. The information representing the driving environment includes an offset representing the distance from the center position in the width direction of the lane to the position of the host vehicle, and a yaw angle representing an angle between the lane, that is, the tangential direction of the road and the traveling direction of the host vehicle, It includes information representing the traveling state of the host vehicle with respect to the lane shape. Information representing these driving environments such as lane width, lane curvature, offset, and yaw angle is referred to as a road parameter.

  That is, the shape of the lane refers to the state of the lane in the lane extending direction and the width direction, such as a curvature representing the degree of bending of the lane in the lane extending direction and a lane width. The traveling state of the vehicle means, for example, in which position in the lane width direction the vehicle is traveling or in which direction the vehicle is traveling in the lane. The state of running. The curvature is the reciprocal of the radius of curvature. The yaw angle is an angle formed by the tangent line at the intersection of a straight line passing through the vehicle from the center of the radius of curvature and the white line representing the lane and the traveling direction of the vehicle.

  The travel control process is a process for controlling the travel of the host vehicle based on the road parameters. The traveling control process is repeatedly executed while the ACC switch is on. In addition, when the subject is abbreviate | omitted in the following description, CPU51 is made into a subject.

  In S100, the CPU 51 acquires a distance measuring point. Specifically, on the basis of the detection signal from the laser radar 11, the relative position represented by the distance and azimuth with the vehicle as the base point and the received light intensity are obtained for each ranging point.

  In S110, CPU51 detects the white line drawn on the road about the own lane. Specifically, the current position detected by the GPS receiver 12 is acquired, and based on the current position and the relative position acquired from the laser radar 11, an arbitrary point is set as the origin for each distance measuring point. Specify a position in the absolute coordinate system.

  Then, based on the fact that the received light intensity of the laser light reflected at the white line is larger than the received light intensity of the laser light reflected around the white line, a distance measuring point having a received light intensity equal to or greater than a predetermined threshold is extracted.

  For example, as shown in FIG. 3, when white lines 102, 102, 111, and 112 are drawn on the road ahead of the host vehicle, the distribution of the received light intensity of the laser beam reflected at the distance measuring points on the road ahead of the host vehicle is assumed. Is represented as shown in FIG. A point surrounded by a dotted line in FIG. 4 corresponds to a distance measuring point having a received light intensity of a predetermined threshold value or more.

  The CPU 51 detects, as a white line, a line estimated based on the arrangement of distance measuring points whose extracted light reception intensity is equal to or greater than a predetermined threshold. Returning to FIG. 3, in the following, the white line on the right side of the host vehicle 100 is referred to as a first right lane line 101 and a second right lane line 102 in order of increasing proximity to the host vehicle 100. Similarly, a white line on the left side of the host vehicle 100 is referred to as a first left lane line 111 and a second left lane line 112 in the order closer to the host vehicle 100. Of the adjacent lanes adjacent to the own lane 90, the adjacent lane on the right side of the own lane 90 is referred to as the right adjacent lane 91, and the adjacent lane on the left side is referred to as the left adjacent lane 92.

  In the example of FIG. 3, the CPU 51 detects the first left lane line 111 as the left white line for the own lane 90 and detects the first right lane line 101 as the right white line for the own lane 90. Since such a white line detection method is well known, further description thereof is omitted here. The CPU 51 records the position of the distance measuring point representing the white line in the memory 52 for each white line.

  In S115, the CPU 51 determines whether or not white lines on both the left and right sides of the own lane 90, that is, both the first left lane line 111 and the first right lane line 101 are detected in S110. For example, when one of the white lines of the own lane 90 is rubbing or hidden by a large vehicle traveling in front of the own vehicle, both the first left lane line 111 and the first right lane line 101 are An undetected situation can arise. Further, when the road ahead of the host vehicle is branched, a situation may occur in which the white line on the branched side is not detected. Furthermore, when a retreat area is provided on the road ahead of the host vehicle, a situation may occur in which the white line on the side where the retreat area is provided is not detected.

  The CPU 51 shifts the process to S120 when the left and right white lines for the own lane 90 are not detected as described above, and shifts the process to S125 when the left and right white lines are detected. However, the case where the white lines on the left and right sides of the own lane 90 are not detected here refers to the case where one of the white lines on the left and right sides of the own lane 90 is detected. That is, it means a case where one of the first left lane line 111 or the first right lane line 101 is detected.

  In S120, the CPU 51 determines whether or not the GPS signal is accurate. Specifically, it is determined that the GPS signal is accurate when the value of the horizontal accuracy decrease rate included in the GPS signal is equal to or less than a predetermined threshold value. If the GPS signal is accurate, the process proceeds to S125, and if the GPS signal is not accurate, the process proceeds to S130.

  When the white line on both the left and right sides is detected for the own lane 90 or when the GPS signal is correct, the CPU 51 maintains the road parameter of the own lane 90 recorded in the memory 52 as it is and performs the processing. The process proceeds to S160. The road parameter of the own lane 90 is an offset representing the lane width, the curvature of the lane, and the distance from the center position in the width direction of the own lane 90 to the own vehicle 100. The yaw angle that represents the angle formed by the straight direction of the host vehicle 100 and the tangential direction of the lane. Note that the curvature of the lane and the tangential direction of the lane are equal to the curvature of the white line and the tangential direction of the white line.

  In S130, where the white line on both the left and right sides of the own lane 90 is not detected and the GPS signal is incorrect, the CPU 51 executes a road parameter estimation process. When only one of the white lines on the left and right sides of the own lane 90 is detected, the road parameter estimation process is performed on the lane adjacent to the side where the one white line is located with respect to the own vehicle 100. This is a process for estimating the road parameters of the own lane 90 based on the road parameters of a certain adjacent lane.

  In the following, as shown in FIG. 5 as an example, the host vehicle 100 is traveling on a road bifurcated in front, the right adjacent lane 91 is detected, and the white line on the left side of the host lane 90 is shown. The process will be described using an example in which the first left dividing line 111 is not detected. The right adjacent lane 91 being detected means that both the first right lane line 101 and the second right lane line 102 that are white lines representing the right adjacent lane 91 are detected.

In S135, the detection signal of the direction indicator sensor 13 is acquired.
In S140, based on the detection signal of the direction indicator sensor 13, it is determined whether or not a lane change is performed. If the lane change is performed, the process proceeds to S145, and if not, the process proceeds to S150.

In S145, the first road parameter estimated in the road parameter estimation process is recorded in the memory 52 as the road parameter of the own lane.
In S150, the second road parameter estimated in the road parameter estimation process is recorded in the memory 52 as the road parameter of the own lane.

  In S160, based on the road parameter of the own lane recorded in the memory 52, a control instruction for causing the own vehicle 100 to travel in the center of the lane selected by the driver is output to the vehicle control unit 17. Then, the traveling control process ends.

[1-2-2. Road parameter estimation process]
Next, the road parameter estimation process executed in S130 of the travel control process will be described using the flowchart of FIG.

  In step S205, information on past points is acquired from the memory 52. The past point refers to a plurality of distance measuring points whose received light intensity is equal to or higher than a predetermined threshold among the detection points extracted in S100 in the past.

  In S210, a detection signal is acquired from the vehicle speed sensor 15 and the steering angle sensor 16, and based on the speed and the steering angle of the host vehicle, in which direction the host vehicle is in which direction from when the past point is detected to the present time. Estimate the momentum representing how much has moved. Based on the estimated momentum, the current position of the past point in the absolute coordinate system is estimated. Hereinafter, the distance measuring point and the past point at the present time are collectively referred to as a detection point.

  The CPU 51 records, in the memory 52, the positions of a plurality of ranging points whose received light intensity is equal to or greater than a predetermined threshold among the detection points extracted in S100, and the current positions of past points as detection point positions. To do.

  As an example, a distance measuring point represented by a black circle in FIG. 7 is represented by a white circle in FIG. 8 representing t seconds after the distance measuring point is acquired. In FIG. 8, the points represented by white circles correspond to past points at the time of FIG. 8, and the points represented by black circles correspond to new distance measuring points at the time of FIG. As described above, since a method of reusing a distance measuring point detected in the past at the present time after t seconds is well known, further explanation is omitted.

  In S215, the CPU 51 acquires a detection point representing an adjacent lane. The adjacent lane here refers to a case where white lines on both the left and right sides of the own lane 90 are not detected in S115, in other words, only one of the white lines on both the left and right sides of the own lane 90 is detected. In addition, the lane adjacent to the side where the one white line is located when viewed from the host vehicle 100 is referred to.

  In the example shown in FIG. 5, the right adjacent lane 91 sandwiched between the first right lane line 101 and the second right lane line 102 corresponds to the adjacent lane. In the example shown in FIG. 5, the first right lane marking 101 corresponds to the adjacent lane marking in the claims. In this step, the CPU 51 acquires a detection point representing the first right lane line 101 and a detection point representing the second right lane line 102 as detection points representing the right adjacent lane 91.

  In S220, the road parameters for the right adjacent lane 91 are estimated. Specifically, the road parameter for the right adjacent lane 91 is, as shown in FIG. 9, the lane width W, which is the width of the lane sandwiched between the first right lane line 101 and the second right lane line 102, The curvature R of the lane, the offset e indicating the distance from the center O in the width direction of the right adjacent lane 91 to the position M of the host vehicle 100, the straight traveling direction P of the host vehicle 100 and the tangential direction Q of the right adjacent lane 91 are formed. This means the yaw angle θ representing the angle. The position M of the host vehicle 100 may be any position as long as it is a position in the host vehicle 100, and may be an installation position of the laser radar 11, for example.

  In this step, the CPU 51 uses the axis extending through the center position O in the width direction of the right adjacent lane 91 and extending in the tangential direction of the right adjacent lane 91 as the y axis, and is an axis perpendicular to the y axis and passing through the position M of the host vehicle 100. Set the coordinate system with x as the x-axis. Then, the CPU 51 calculates the position (X, Y) of each detection point representing the right adjacent lane 91 in the coordinate system. A point represented by a black circle in FIG. 9 is an example of each detection point representing the right adjacent lane 91.

  The CPU 51 represents the relationship between the position (X, Y) of the detection point representing the right adjacent lane 91 and the lane width W, the curvature R, the offset e, and the yaw angle θ that are the road parameters of the right adjacent lane 91 (1). Based on the equation, each road parameter is estimated by using a Kalman filter or the like. Then, each estimated road parameter is recorded in the memory 52.

  In the formula (1), when the adjacent lane is the right lane of the own lane, the value of the lane width W is used as a positive value, and when the adjacent lane is the left lane of the own lane, the lane The value of the width W is used as a negative value. In the example of FIG. 9, the right adjacent lane 91 is the lane on the right side of the own lane 90, so the value of the lane width W is used as a positive value.

  In S225, CPU51 acquires yaw angle (theta) and curvature R as an adjacent parameter among the road parameters about the right adjacent lane 91 estimated in S220. The adjacent parameter refers to a road parameter for the right adjacent lane 91 acquired in this step.

  In S230, the yaw angle θ and the curvature R, which are adjacent parameters acquired in 220, are recorded in the memory 52 as the yaw angle and the curvature of the own lane 90. Here, using the fact that the own lane 90 and the right adjacent lane 91 are adjacent, in other words, it is estimated that the own lane 90 and the right adjacent lane 91 are parallel due to the adjacent. The yaw angle θ and the curvature R of the right adjacent lane 91 are used as they are as the yaw angle and the curvature of the own lane 90.

  The term “parallel” as used herein refers to a state that is not completely parallel but falls within the error range from the parallel state, and indicates a state in which substantially the same effect as in the case of the parallel state can be obtained.

  In S235, an estimated range setting process is performed. In the example shown in FIG. 5, the estimated range setting process is performed by setting an estimated range I that is a range in which it is estimated that the first left lane line 111 that is the other white line that is not detected among the white lines of the own lane 90 exists. It is a process to set.

  In S240 to S260, the CPU 51 determines the detection points located on the opposite lane side that is opposite to the side on which the right adjacent lane 91 is located with respect to the own vehicle 100 among the detection points acquired in S100. A classification is made as to whether or not it is within the estimated range I.

  That is, in S240, the CPU 51 acquires a detection point located on the side adjacent to the own vehicle 100 on the side adjacent to the adjacent lane. In S245, it is determined whether or not the detection point is located within the estimated range I based on the position of the detection point. In S250, a detection point located outside the estimated range I is recorded in the memory 52 as an out-of-range detection point. In S255, the detection point located within the estimated range I is recorded in the memory 52 as the in-range detection point. And the process of S240-S260 is repeated until classification | category is complete | finished about all the detection points acquired in S240 (S260; YES).

In S <b> 265, a detection point representing the first right lane marking 101 that is the adjacent lane marking is acquired from the memory 52.
In S270, the in-range detection point is acquired from the memory 52.

  In S275, the detection points representing the in-range detection points and the first right dividing line 101, specifically, the detection points represented by white circles and black circles in FIG. Based on the positions of these detection points, the road parameters of the own lane 90 are estimated in the same manner as the road parameters for the right adjacent lane 91 described above.

  That is, in this step, although not shown, the CPU 51 uses the axis extending through the center position in the width direction of the own lane 90 and extending in the tangential direction of the own lane as the y axis, and is perpendicular to the y axis and passes through the position of the own vehicle 100. Set the coordinate system with the x axis as the axis. And CPU51 calculates the position (Xj, Yj) of each detection point showing the own lane 90 in this coordinate system.

  The CPU 51 represents the relationship between the position (Xj, Yj) of the detection point representing the own lane 90 and the lane width Wj, the curvature Rj, the offset ej, and the yaw angle θj, which are the road parameters of the own lane 90, as shown in equation (2). Based on this, each road parameter is estimated by using a Kalman filter or the like.

  However, the curvature Rj and the yaw angle θj among the road parameters for the own lane 90 have already been estimated in S230 as using the values of the curvature R of the right adjacent lane 91 and the curvature R of the adjacent lane as they are. is there.

In this step, the CPU 51 estimates only the lane width Wj and the offset ej of the own lane 90 by using a Kalman filter or the like based on the equation (2).
In S280, the CPU 51 stores the yaw angle θj and curvature Rj for the own lane 90 estimated in S230, and the lane width Wj and the offset ej for the own lane 90 estimated in S270 as the first road parameters in the memory 52. To record.

In S285, the CPU 51 acquires out-of-range detection points from the memory 52.
In S290, similarly to S220, the road parameter is estimated using the position of the out-of-range detection point. That is, assuming that the out-of-range detection point is a detection point representing the left lane of the own lane 90, the road parameter represented by the out-of-range detection point is calculated using a Kalman filter or the like based on the equation (1). presume.

In S295, the road parameter estimated in S290 is recorded in the memory 52 as the second road parameter. And this road parameter estimation process is complete | finished.
[1-2-3. Estimated range setting process]
Next, the estimation range setting process executed in S235 of the road parameter estimation process will be described using the flowchart of FIG.

In S300, the CPU 51 acquires the curvature of the adjacent lane from the memory 52, specifically, the curvature R of the first right lane line 101 shown in FIG.
In S310, the CPU 51 is located on the opposite side to the first right lane line 101 in the own lane 90 when viewed from the own vehicle 100 among the detection points acquired in S210, and the first left lane line 111 as a candidate. A plurality of detection points constituting a line to be obtained are acquired from the memory 52. The plurality of detection points constituting a candidate line for the first left partition line 111 correspond to points represented by white circles and white squares in FIG.

  In S330, the CPU 51 extracts a plurality of candidate lines represented by the arrangement of a plurality of detection points. For example, in FIG. 5, the candidate line refers to a line 121 represented by a sequence of detection points represented by white circles, lines 122 and 123 represented by a sequence of detection points represented by white circles and white squares, and the like. .

  In S350, the CPU 51 sets a candidate line whose curvature is closest to the adjacent curvature R among the plurality of candidate lines as a virtual dividing line, and calculates an estimated range I that is a band-shaped range having a predetermined width including the virtual dividing line. Set. For example, in the example of FIG. 5, the candidate line 121 represented by the arrangement of detection points represented by white circles most closely matches the adjacent curvature R, so that the candidate line represented by the arrangement of detection points represented by white circles. 121 is a virtual lane marking 121.

  The width of the estimation range may be the same width as the white line, for example. In this case, for example, the width of the white line may be set to A (cm), and a band-shaped range of ± A / 2 (cm) centered on the virtual dividing line 121 may be set as the estimated range I. The CPU 51 records information representing the estimated range I in the memory 52. And this estimation range setting process is complete | finished.

[1-3. effect]
According to the first embodiment described in detail above, the following effects can be obtained.
(1a) The estimation device 50 is an adjacent that represents the shape of a lane for an adjacent lane (right adjacent lane 91) that is a lane adjacent to the own lane 90 on which the host vehicle travels out of the lane that represents an area on which the vehicle travels on the road. An adjacent parameter including at least one of the shape and the parameter representing the running state of the host vehicle with respect to the adjacent shape is acquired (S225). Moreover, the estimation apparatus 50 uses the adjacent parameter and the adjacent lane and the own lane 90 to be adjacent to each other, and the own shape representing the shape of the lane for the own lane 90 and the traveling state of the own vehicle with respect to the own shape. The own lane parameter including at least one of the parameters representing the position is estimated (S230, S275).

  The adjacent parameter refers to the curvature R and the yaw angle θ among the road parameters of the adjacent lane. The own parameter refers to the curvature Rj, the lane width Wj, the yaw angle θj, and the offset ej, which are road parameters of the own lane 90.

  Here, in this embodiment, the adjacent lane refers to the right adjacent lane 91, and the parameter representing the adjacent shape refers to the curvature R of the right adjacent lane 91. However, the parameter representing the adjacent shape may be at least one of the curvature R and the lane width W of the right adjacent lane 91. Further, the parameter representing the traveling state of the host vehicle 100 with respect to the adjacent shape refers to the yaw angle θ with respect to the right adjacent lane 91.

  On the other hand, the parameter representing the self-shape refers to the lane width Wj and the curvature Rj in the present embodiment, but may be at least one of these. Further, in the present embodiment, the parameter representing the traveling state of the host vehicle with respect to the host vehicle shape refers to the yaw angle θj and the offset ej with respect to the host vehicle lane 90, but may be at least one of these.

  In addition, using the fact that the adjacent lane and the own lane 90 are adjacent to each other means that the right adjacent lane 91 and the own lane 90 are estimated to be parallel due to being adjacent as described above. It means using.

  According to this, even if a situation occurs in which the white lines on both the left and right sides of the own lane 90 cannot be detected in front of the host vehicle, at least one white line (the first right lane line 101) about the own lane 90 can be detected. For example, the road parameter of the own lane 90 can be estimated based on the adjacent parameter of the adjacent lane (right adjacent lane 91) located on the one white line side. And various control for making the own vehicle 100 drive | work can be performed using the estimated road parameter of the own lane 90. FIG.

  (1b) The estimation device 50 may acquire the curvature R of the adjacent lane (right adjacent lane 91), and use the acquired curvature R of the right adjacent lane 91 as an estimated value of the curvature Rj in the own lane 90 (S230). ). According to this, since the process which estimates curvature Rj in the own lane 90 using a Kalman filter etc. based on (2) Formula can be abbreviate | omitted, the processing load in CPU51 can be reduced.

  (1c) The estimation device 50 acquires the yaw angle θ representing the traveling direction of the host vehicle 100 with respect to the direction in which the adjacent lane (right adjacent lane 91) extends, and the acquired yaw angle θ with respect to the right adjacent lane 91 is determined by the host lane 90. It may be used as an estimated value of the yaw angle θj representing the traveling direction of the host vehicle 100 with respect to the direction in which the vehicle extends (S230). In the present embodiment, the direction in which the adjacent lane (right adjacent lane 91) extends refers to the tangential direction of the adjacent lane (right adjacent lane 91), and the direction in which the own lane extends refers to the tangential direction of the own lane 90. According to this, since the process which estimates yaw angle (theta) j in the own lane 90 using a Kalman filter etc. based on (2) Formula can be abbreviate | omitted, the processing load in CPU51 can be reduced.

  (1d) As described above, the curvature Rj and the yaw angle θj are used by using the fact that the own lane 90 and the adjacent lane (the right adjacent lane 91) are adjacent to each other, that is, by being estimated to be parallel. As for, the road parameter of the adjacent lane may be used as it is as the road parameter of the own lane. However, for example, a parameter such as the offset ej cannot be used as it is as the road parameter of the own lane 90.

  Therefore, the estimation device 50 is a lane marking as a boundary that divides the own lane 90 and another area in the own lane 90, and the adjacent lane-side lane marking is defined as an adjacent lane marking (first right lane marking 101). That is, from the position of the adjacent lane line and the adjacent parameter, the adjacent lane line and the own lane line 90 are adjacent to each other, and the lane line on the opposite side of the adjacent lane line in the own lane line 90 is used. The position of the non-adjacent lane marking (first left lane marking 111) may be estimated (S270). And the estimation apparatus 50 is the road parameter of the own vehicle 100 based on the position of the adjacent lane line (first right lane line 101) and the position of the non-adjacent side lane line (first left lane line 111). The own parameter may be estimated (S275).

  In the present embodiment, a lane marking means a white line drawn on a road. The anti-adjacent dividing line refers to the first left dividing line 111 in the present embodiment. The position of the adjacent lane marking means the position of a plurality of detection points representing the first right lane marking 101 that is the adjacent lane marking. Further, the position of the anti-adjacent dividing line means the position of a plurality of detection points representing the first left dividing line 111 that is the anti-adjacent dividing line.

  According to this, it is possible to estimate a parameter that cannot use the road parameter of the adjacent lane as it is among the road parameters of the own lane 90 based on the road parameter of the adjacent lane.

  (1e) For example, as shown in FIG. 5, in a road having a branch road 140, detection points representing a plurality of lane markings can be acquired on the branch road 140 side. It is desirable to estimate whether it corresponds to the line 111 and then to estimate its own parameters such as the lane width Wj and the offset ej based on the position of the detection point.

  Therefore, the estimation device 50 acquires the adjacent curvature R, which is the curvature of the adjacent lane (right adjacent lane 91), and the adjacent lane line (first right lane line 101) in the own lane 90 when viewed from the own vehicle 100. You may acquire the position of the some detection point which comprises the line which is located in the other side and becomes a candidate of an anti-adjacent side division line (1st left division line 111) (S310).

  The estimation device 50 extracts a plurality of candidate lines 121 to 123 represented by the arrangement of a plurality of detection points (S330), and the candidate line whose curvature is closest to the adjacent curvature R among the plurality of candidate lines 121 to 123. 121 may be a virtual lane marking 121, and an estimation range I that is a band-shaped range having a predetermined width including the virtual lane marking 121 may be set (S350). The predetermined width may be, for example, a width of a dividing line such as a white line width or a curb width, or may be a width arbitrarily determined.

  Then, the estimation device 50 obtains at least the position of the detection point existing within the estimation range I (S270), and sets the line represented by the detection point located within the estimation range I as the anti-adjacent division line (first left You may estimate that it is a lane marking 111).

  According to this, since it is estimated that there is a detection point that represents a line that is a candidate for an anti-adjacent lane marking line in the estimation range I including the virtual lane marking 121 whose curvature matches the adjacent curvature R, the branch path 140 is Even if it is provided, the anti-adjacent side lane marking can be estimated. As a result, the lane width Wj and the offset ej of the own lane 90 can be estimated.

  In addition, since the white line and curb representing the dividing line have a predetermined width, the detection point located within the estimation range I does not necessarily represent the center position of the white line or curb. For this reason, if the virtual lane marking 121 is estimated as the first left lane marking 111, there may be an error with respect to a line obtained by arranging the white lines and the central positions of the curbs.

  According to the present embodiment, since the anti-adjacent lane marking (first left lane line 111) is estimated based on the detection points within the estimation range I including the virtual lane marking 121, the virtual lane marking 121 is assumed to be the first left The position of the first left marking line 111 can be estimated with higher accuracy than when the marking line 111 is used.

  In the first embodiment, the estimation device 50 corresponds to an example of the own lane information estimation device. The estimation device 50 corresponds to an example of an adjacent parameter acquisition unit, an adjacent parameter acquisition unit, an adjacent position acquisition unit, an anti-adjacent side estimation unit, a detection point position acquisition unit, a detection point acquisition unit, and a range setting unit. S225 corresponds to an example of processing as an adjacent parameter acquisition unit, S230 and S275 correspond to an example of processing as an adjacent parameter acquisition unit, and S265 corresponds to an example of processing as an adjacent position acquisition unit. S270 corresponds to an example of processing as an anti-adjacent side estimation unit and a detection point position acquisition unit. S310 corresponds to an example of processing as a detection point acquisition unit, and S350 corresponds to an example of processing as a range setting unit.

[2. Second Embodiment]
[2-1. Difference from the first embodiment]
Since the basic configuration of the second embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences. Note that the same reference numerals as those in the first embodiment indicate the same configuration, and the preceding description is referred to.

  In the first embodiment described above, the estimated range I is set based on the adjacent curvature R in the estimated range setting process. On the other hand, the second embodiment is different from the first embodiment in that the estimation range I is set based on the lane width W of the adjacent lane 91.

[2-2. processing]
Next, an estimation range setting process executed by the estimation apparatus 50 (CPU 51) of the second embodiment instead of the process shown in FIG. 10 of the first embodiment will be described using the flowchart of FIG. In the present embodiment, a case will be described in which the host vehicle 100 is traveling on a road having a retreat area 141 as shown in FIG.

In S <b> 320, the CPU 51 acquires the lane width W in the adjacent lane 91 from the memory 52.
In S <b> 340, the CPU 51 sets a temporary lane marking 131 on the side opposite to the first right lane marking 101 as viewed from the host vehicle 100, with a lane width W separated from the first right lane marking 101. That is, a line represented by a point where the position of the detection point representing the first right lane marking 101 is moved by the lane width W to the opposite side of the right adjacent lane 91 in the width direction is set as a temporary lane marking 131. To do.

In S360, as in S350 shown in FIG. 10, a band-shaped range having a predetermined width including the temporary marking line 131 is set as the estimated range I.
[2-3. effect]
According to the second embodiment described in detail above, in addition to the effects (1a) to (1d) of the first embodiment described above, the following effects can be obtained.

  (2a) The estimation device 50 acquires the adjacent lane width (lane width W) that is the lane width of the adjacent lane 91, and calculates the lane width W from the first right lane line 101 that is the adjacent lane line to the host vehicle 100 side. An estimated range I that is a band-shaped range having a predetermined width may be set at a distance (S360). Moreover, the estimation apparatus 50 may acquire the position of the lane marking that exists at least in the estimation range I (S270). And the estimation apparatus 50 may estimate that the lane marking represented by the detection point located in the estimation range I is an anti-adjacent side lane marking (first left lane marking 111).

  According to this, since the estimated range I is set based on the temporary lane line 131 obtained by translating the already detected first right lane line 101 to the opposite lane side by the lane width W, The processing load on the CPU 51 when setting the estimated range I can be reduced.

  In the second embodiment, the estimation device 50 corresponds to an example of a range setting unit and a lane line position acquisition unit. Further, S360 corresponds to an example of processing as a range setting unit, and S270 corresponds to an example of processing as a lane marking position acquisition unit.

[3. Other Embodiments]
As mentioned above, although the form for implementing this invention was demonstrated, this invention is not limited to the above-mentioned embodiment, It can implement in various deformation | transformation.

  (3a) In the above embodiment, based on the example shown in FIG. 5, the first right lane line 101 that is the white line on the right side of the own lane 90 is detected, and the first left lane line 111 that is the white line on the left side. If the right adjacent lane 91 is an adjacent lane, the first right lane line 101 is an adjacent lane line, and the first left lane line 111 is an anti-adjacent lane line, 50 operations have been described. However, it is not limited to this. For example, when the first left lane line 111 is detected and the first right lane line is not detected, the estimation device 50 sets the left adjacent lane 92 as the adjacent lane and the first left lane line 111 as the adjacent side lane. A similar process may be executed to estimate the own parameter with the first right lane marking 101 as the non-adjacent lane marking.

  (3b) In the above embodiment, the estimation device 50 uses the yaw angle θ and the curvature R in the road parameters of the adjacent lane 91 as the yaw angle θj and the curvature Rj of the own lane 90, but the present invention is not limited to this. Absent. For example, the estimation device 50 may use at least one of the yaw angle θ, the curvature R, and the lane width W among the road parameters for the adjacent lane 91 as its own parameters. That is, the adjacent parameter refers to at least one road parameter other than the offset e that can be used as it is as the own parameter among the road parameters for the adjacent lane 91.

  (3c) In the above embodiment, the offset ej and the lane width Wj are estimated based on the expression (2) as parameters that cannot use the road parameters of the adjacent lane 91 as they are, but the present invention is not limited to this. For example, any of the road parameters of the own lane 90 may be estimated based on the equation (2).

  (3d) In the above embodiment, the estimation device 50 recognizes an area sandwiched between white lines as a lane, but may recognize an area sandwiched between lane markings other than the white line as a lane. For example, the lane marking may be a yellow line drawn on a road, a line represented by a road survey object that is a three-dimensional object such as a curb or a guardrail, or the like.

  (3e) In the above-described embodiment, the estimation device 50 does not execute the road parameter estimation processing when white lines on both the left and right sides of the own lane 90 are detected. However, the present invention is not limited to this. For example, the estimation device 50 may execute the road parameter estimation process when a white line on both the left and right sides of the own lane 90 is detected, but when a part of the white line is rubbed. Good.

  (3f) In the above embodiment, the estimation device 50 estimates the anti-adjacent lane marking (first left lane marking 111) using the detection points within the estimation range I set based on the virtual lane marking 121. However, this is not a limitation. For example, the estimation device 50 may estimate the virtual lane marking 121 as the first left lane marking 111 when accuracy is not required.

  That is, the description will be made using the flowchart shown in FIG. 10. The estimation device 50 acquires the adjacent curvature R in S300, and in S310, the adjacent lane line (first right lane line) in the lane 90 when viewed from the host vehicle 100. 101), the positions of a plurality of detection points that constitute a line that is a candidate for an adjacent division line that is located on the opposite side may be acquired. In S330, the estimating apparatus 50 extracts a plurality of candidate lines 121 to 123 represented by the arrangement of detection points, and selects a candidate line 121 having a curvature closest to the adjacent curvature among the plurality of candidate lines 121 to 123. Unlike the first embodiment, the virtual lane marking 121 may be estimated as the non-adjacent side lane marking (first left lane marking 111).

  That is, in this embodiment, the estimation apparatus 50 omits S140 to S170 shown in FIG. 6, and in S175, unlike the first embodiment, the detection point representing the virtual lane line 121 and the first right lane line 101 are displayed. The own parameter may be estimated using the detected detection point. Further, in S185, the estimation device 50, unlike the first embodiment, the detection points excluding the detection points representing the virtual partition lines 121 among the detection points acquired in S310 of the present embodiment are set as out-of-range detection points. You may get it.

  (3g) A plurality of functions of one constituent element in the above embodiment may be realized by a plurality of constituent elements, or a single function of one constituent element may be realized by a plurality of constituent elements. . Further, a plurality of functions possessed by a plurality of constituent elements may be realized by one constituent element, or one function realized by a plurality of constituent elements may be realized by one constituent element. Moreover, you may abbreviate | omit a part of structure of the said embodiment. In addition, at least a part of the configuration of the above embodiment may be added to or replaced with the configuration of the other embodiment. In addition, all the aspects included in the technical idea specified only by the wording described in the claim are embodiment of this invention.

  (3h) The above-described estimation device 50, the vehicle control system 1 including the estimation device 50 as a constituent element, a program for causing the estimation device 50 to function, and a non-transitory actual recording medium such as a semiconductor memory storing the program The present invention can also be realized in various forms such as a method for estimating lane information.

  50 ECU, 51 CPU.

Claims (7)

An adjacent shape that represents the shape of a lane for an adjacent lane that is adjacent to the own lane in which the host vehicle is traveling among lanes that represent a region in which the vehicle is traveling on a road, and a parameter that represents the traveling state of the host vehicle with respect to the adjacent shape An adjacent parameter acquisition unit (S225) that acquires an adjacent parameter including at least one of
From the adjacent parameter, by using that the adjacent lane and the own lane are adjacent to each other, at least of the own shape representing the shape of the lane for the own lane, and the parameter representing the traveling state of the own vehicle with respect to the own shape An own parameter estimating unit (S230, S275) for estimating an own lane parameter including one;
Self-lane information estimation device (50) comprising: The own lane information estimation device according to claim 1,
The adjacent parameter acquisition unit acquires the curvature of the adjacent lane,
The own parameter estimation unit (S230) uses the curvature of the adjacent lane as an estimated value of curvature in the own lane. It is the own lane information estimation device according to claim 1 or 2,
The adjacent parameter acquisition unit acquires a yaw angle representing a traveling direction of the host vehicle with respect to a direction in which the adjacent lane extends,
The own parameter estimation unit (S230) uses a yaw angle with respect to the adjacent lane as an estimated value of a yaw angle representing a traveling direction of the own vehicle with respect to a direction in which the own lane extends. It is the own lane information estimation device according to any one of claims 1 to 3,
An adjacent position acquisition unit (S265) that acquires a position of an adjacent lane line that is a lane line as a boundary that divides the own lane and another area in the own lane,
From the position of the adjacent lane marking and the adjacent parameter, the adjacent lane and the own lane are adjacent to each other, and the opposite lane is the lane marking opposite to the adjacent lane marking in the own lane. An anti-adjacent side estimation unit (S270) for estimating the position of the side marking line,
The own parameter estimating unit (S275) estimates the own parameter based on the position of the adjacent lane marking and the position of the non-adjacent lane marking. The own lane information estimation device according to claim 4,
The adjacent parameter acquisition unit acquires an adjacent lane width that is a lane width of the adjacent lane,
A range setting unit (S360) for setting an estimated range which is a band-shaped range having a predetermined width at a position where the adjacent lane width is separated from the adjacent lane marking toward the own vehicle;
A lane line position acquisition unit (S270) that acquires at least the position of a lane line existing within the estimated range;
With
The said anti-adjacent side estimation part estimates the lane line located in the said estimation range as the said anti-adjacent side lane line The own lane information estimation apparatus. The own lane information estimation device according to claim 4,
The adjacent parameter acquisition unit acquires an adjacent curvature that is a curvature of the adjacent lane,
A detection point acquisition unit (S310) that acquires positions of a plurality of detection points that are located on the opposite side of the adjacent lane line as viewed from the own vehicle and that constitute a line that is a candidate for the non-adjacent side lane line. )When,
A plurality of candidate lines represented by the arrangement of the plurality of detection points are extracted (S330), a candidate line having a curvature closest to the adjacent curvature among the plurality of candidate lines is defined as a virtual lane line, and the virtual lane line A range setting unit (S350) for setting an estimation range that is a band-shaped range having a predetermined width including
A detection point position acquisition unit (S270) that acquires positions of a plurality of detection points existing at least within the estimated range;
With
The said anti-adjacent side estimation part estimates the line represented by the some detection point located in the said estimation range as the said anti-adjacent side division line Self-lane information estimation apparatus. The own lane information estimation device according to claim 4,
A detection point acquisition unit (S310) that acquires positions of a plurality of detection points that are located on the opposite side of the adjacent lane line as viewed from the host vehicle and that constitute a line that is a candidate for the adjacent lane line.
With
The adjacent parameter acquisition unit acquires an adjacent curvature that is a curvature of the adjacent lane,
The anti-adjacent side estimation unit extracts a plurality of candidate lines represented by the arrangement of the plurality of detection points, and sets a candidate line having a curvature closest to the adjacent curvature among the plurality of candidate lines as a virtual partition line. The own lane information estimation device that estimates the virtual lane marking as the non-adjacent lane marking.