JP2015069289A - Lane recognition device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To determine a branch lane quickly and accurately.SOLUTION: A lane recognition device 100 includes a control apparatus 10 which executes: an image acquisition function of acquiring a captured image obtained by a camera mounted on a vehicle which captures the surroundings of the vehicle; a lane marker detection function which detects a lane marker indicated on a road, on the basis of the captured image; and a lane recognition function which calculates relative positions of the detected lane markers with respect to the vehicle, and recognizes that there is a branch lane in a lane where the vehicle is traveling when a V-shaped lane marker having two line-shaped parts opened upward in the captured image at an angle of a predetermined range is detected in a side area located within a predetermined distance from the detected lane markers.

Description

  The present invention relates to a lane recognition device that recognizes the mode of a traveling lane in which a host vehicle travels.

  Regarding this type of device, when the road curvature or pitch angle calculated based on the lane marker of the road is larger than a predetermined value, the device recognizes that the road is a branch road (branch lane) for making a right or left turn Is known (Patent Document 1).

JP 2005-346382 A

  However, since the shape of the traveling road is not uniform, if it is determined based on a uniform threshold whether or not the traveling road of the host vehicle is a branch lane for turning right or left, erroneous determination is likely to occur. Become.

  The problem to be solved by the present invention is to quickly and accurately recognize a lane in which a branched lane exists even on traveling roads having different shapes.

  The present invention recognizes the presence of a branch lane in the traveling lane when a V-shaped lane marker having two linear portions that expand at an angle of a predetermined range in the upward direction of the captured image is detected. This solves the above problem.

  According to the present invention, since the presence of a branch road is determined based on the shape characteristics of the lane marker, for example, even if the road has a different shape with varying curvature and pitch angle, Can be recognized quickly and accurately.

1 is a block configuration diagram of a driving support system 1 including a lane recognition device 100 according to the present embodiment. It is the 1st figure for explaining the example of installation of the camera of this embodiment, and a road parameter. It is the 2nd figure for explaining the example of installation of the camera of this embodiment, and a road parameter. It is a figure for demonstrating a lane marker model. It is a flowchart which shows the control procedure of the lane recognition process of this embodiment. It is a flowchart which shows the control procedure of the lane marker detection process of this embodiment. It is a figure for demonstrating the setting method of the detection area (side area | region) of a V-shaped lane marker candidate point. It is a figure for demonstrating the setting method of the detection area (1st detection area) of a lane marker candidate point. It is a flowchart which shows the control procedure of the lane recognition process of this embodiment. It is a flowchart which shows the 1st control procedure of the detection process of the V-shaped lane marker shown in FIG. It is a figure for demonstrating the setting method of the detection area (2nd detection area) of a V-shaped lane marker candidate point. It is a figure for demonstrating an example of the detection method of a V-shaped lane marker. It is a flowchart which shows the 2nd control procedure of the detection process of the V-shaped lane marker shown in FIG. It is a figure for demonstrating the setting method of the detection area (3rd detection area) of a linear lane marker candidate point. It is a flowchart which shows the control procedure of the driving assistance process of this embodiment. It is a figure which shows an example of the aspect of the branch lane of this embodiment. It is a figure for demonstrating the lane recognition process in the branch lane of this embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the present embodiment, a case where the lane recognition device according to the present invention is applied to a driving support system for a vehicle will be described as an example.

  FIG. 1 is a diagram illustrating a block configuration of a vehicle driving support system 1 including a lane recognition device 100 according to the present embodiment. The lane recognition device 100 and the vehicle driving support system 1 including the lane recognition device 100 are mounted on a vehicle. The driving support system 1 for a vehicle executes part or all of a steering operation and acceleration / deceleration operation of the host vehicle, or outputs information necessary for driving the host vehicle to support the driving of the host vehicle. The lane recognition device 100 recognizes the lane in which the host vehicle travels. In the present embodiment, the recognized travel lane information is used for driving support processing. In addition, the branch lane of this embodiment includes the aspect in which the number of lanes increases.

  As shown in FIG. 1, the vehicle driving support system 1 includes a lane recognition device 100 and an in-vehicle device 200. The in-vehicle device 200 includes a vehicle controller 50, a drive system 60, a steering device 70, a driving support device 80, and a navigation device 90. The lane recognition device 100 and each in-vehicle device 200 are connected by a CAN (Controller Area Network) or other in-vehicle LAN in order to exchange information with each other.

  The vehicle controller 50 is an ECU (Electronic Control Unit) that controls the operation of the entire vehicle. The vehicle controller 50 acquires detection signals from various sensors 51 mounted on the vehicle. The acquired signal is used for vehicle control and sent to the driving support device 80 and the lane recognition device 100. The various sensors 51 detect the state of the vehicle, and specifically include a vehicle speed sensor, a steering angle sensor, and a current sensor for a steering actuator. A yaw rate sensor may also be included. The vehicle speed is detected by measuring the rotational speed of the output side of the transmission and the rotational speed of the wheels. The steering angle is detected as a steering angle detection signal by an angle detection mechanism such as a rotary encoder or a potentiometer after the rotational displacement of the steering shaft is amplified directly or by a gear mechanism or the like.

  The drive system 60 is a drive mechanism for the vehicle V, and drives the vehicle based on input signals from the driver's accelerator operation and brake operation, and control signals acquired from the vehicle controller 50 or the driving support device 80. The steering control device 70 includes a steering actuator. The steering actuator includes a motor and the like attached to the column shaft of the steering. The vehicle is moved based on an input signal obtained by a driver's steering operation and a control signal acquired from the vehicle controller 50 or the driving support device 80.

  The driving support device 80 uses the lane marker model that expresses the relative positional relationship between the traveling lane calculated by the lane recognition device 100 and the host vehicle so that the host vehicle travels while maintaining the center position of the lane. Alternatively, the steering device 70 is controlled so that the host vehicle travels while maintaining the lateral position set with respect to the lane. Based on information on the vehicle speed, steering angle, and current of the steering actuator obtained from the various sensors 51 of the vehicle controller 50, the steering control amount is calculated, and a current command is sent to the steering actuator, thereby traveling at the target travel position. Control as follows.

  The navigation device 90 includes a GPS (Global Positioning System) 91 that detects a current position, and map information that associates point information, road information, facility information, and the like with position information.

  Further, as shown in the figure, the lane recognition device 100 of this embodiment includes a control device 10, a camera 20 that captures the front of the vehicle and the surroundings of the vehicle including the side, and an output that outputs a lane recognition result. Device 30.

  The camera 20 of the present embodiment is provided at a substantially central position in the vehicle width direction of the vehicle. In this embodiment, it attaches to the upper part of the front window in a vehicle interior. The camera 20 is a camera using an image sensor such as a CCD, and captures an image of the traveling direction of the vehicle V. The camera 20 is attached to the vehicle with a lower pitch angle along the direction of gravity. The pitch angle can be appropriately set according to the height of the vehicle and the performance of the camera 20. For this reason, the camera 20 of the present embodiment images a road surface several meters to several tens meters ahead of the vehicle. Further, in the present embodiment, not only the travel lane in which the host vehicle travels, but also the adjacent lane and the road in the adjacent lane can be imaged. In this case, it is preferable to use a wide-angle camera 20.

  The display 31 and the speaker 32 as the output device 30 of the present embodiment are installed around a dashboard that can be visually recognized by the driver of the vehicle, and provide information on the lane recognition result, information on driving assistance according to the lane recognition result, and the like. Present to the driver. Further, the output device 30 of the present embodiment has a wireless or wired communication function, and sends a vehicle recognition result to the driving support device 80 outside the lane recognition device 100 and other in-vehicle devices 200 in accordance with instructions from the control device 10. .

  The lane recognition device 100 of this embodiment recognizes the relative positional relationship between the host vehicle and the lane in which the host vehicle travels based on the captured image. The control device 10 of the lane recognition device 100 will be described. The control device 10 according to the present embodiment executes a ROM (Read Only Memory) 12 in which a program for recognizing a lane is stored, and a program stored in the ROM, so that the lane recognition device 100 and driving support for a vehicle are performed. The computer includes a CPU (Central Processing Unit) 11 as an operation circuit that functions as the system 1 and a RAM (Random Access Memory) 13 that functions as an accessible storage device.

  The control device 10 of the lane recognition device 100 according to the present embodiment has an image acquisition function, a lane marker detection function, a lane recognition function, and an obstacle information acquisition function. The control apparatus 10 of this embodiment performs each function by cooperation of the software for implement | achieving the said function, and the hardware mentioned above.

  Hereinafter, each function of the lane recognition device 100 according to the present embodiment will be described.

  First, the image acquisition function of the control apparatus 10 of this embodiment is demonstrated. The control device 10 acquires a captured image around the vehicle captured by the camera 20. The captured image includes an image of the road surface in front of and in front of the vehicle.

  The lane marker detection function of the control apparatus 10 of this embodiment is demonstrated. Based on the captured image, the control device 10 detects a lane marker that divides the travel lane in which the host vehicle travels. The detection method of the lane marker is not particularly limited, and techniques such as a white line detection function, a driving support function, and a key plane support function known at the time of filing can be used as appropriate. The control device 10 according to the present embodiment extracts an edge corresponding to a boundary between a road and a lane marker (white line) included in the captured image, and detects the lane marker by analyzing the extracted edge. The control device 10 acquires the position of the detected lane marker.

  The obstacle information acquisition function of the control device 10 of this embodiment will be described. The information regarding the obstacle which exists ahead of the said own vehicle of this embodiment is acquired. The information regarding the obstacle includes information regarding the presence / absence of the obstacle, the type of the obstacle, and the relative position with the obstacle. In the present embodiment, the information related to the obstacle may be detected by the control device 10 or may be detected by the driving assistance device 80 mounted on the vehicle.

  The control device 10 determines the presence or absence of an obstacle by determining whether or not the image captured by the camera 20 includes an image of another vehicle ahead. The obstacle detection method based on the captured image is not particularly limited. For example, a predefined image feature of an obstacle to be detected is compared with an image feature extracted from a captured image, and the captured image includes an image feature whose correlation with the image feature of the obstacle is equal to or greater than a predetermined value. In this case, it can be determined that there is an obstacle ahead of the host vehicle. The control device 10 defines the image characteristics of the obstacle to be detected in accordance with the type of obstacle (ordinary vehicle, four-wheel other vehicle such as a truck, two-wheel other vehicle such as a motorcycle, pedestrian, stationary object, etc.). By doing so, the type of the detected obstacle can be identified. Furthermore, the control device 10 can acquire the position information of the obstacle with respect to the host vehicle based on the relative position of the image determined to be an obstacle with respect to the captured image.

  When the external driving support device 80 has an obstacle detection function, the control device 10 can acquire information on the obstacle from the driving support device 80. The driving support device 80 may detect an obstacle based on a captured image of the camera 20 or may detect an obstacle based on radar information acquired by the radar device 81. The method for determining the presence / absence of an obstacle, the position of the obstacle, and the type of the obstacle based on the radar information is not particularly limited, and a technique known at the time of filing can be used.

  Information on the presence / absence of the obstacle, the position of the obstacle, and the obstacle related to the type of the obstacle acquired by the control device 10 is used for setting a detection area of a lane marker, which will be described later, and lane recognition processing.

  Next, the lane recognition function of the control device 10 of the present embodiment will be described. The control device 10 according to the present embodiment calculates the relative position of the lane marker based on the detection result of the lane marker. The relative position of the lane marker is a position relative to the host vehicle.

  The control device 10 according to the present embodiment uses a lane marker model based on the position of the host vehicle, and uses any of the road curvature of the traveling lane corresponding to the lane marker, the relative pitch angle with respect to the host vehicle, or the relative yaw angle with respect to the host vehicle. The road parameter including one or more is calculated as the relative position of the lane marker with respect to the host vehicle. Thereby, the relative position of the lane marker with respect to the host vehicle can be specifically calculated. In addition, driving assistance using these values can be performed.

  The control device 10 detects a lane marker on the road by processing an image captured by the camera 20, and then uses a plurality of parameters (hereinafter also referred to as "road parameters") representing the road shape and vehicle behavior to determine the road lane. Lane parameters are recognized by updating road parameters with time so that the lane marker model that mathematically represents the shape matches the detection result of the lane marker. And the control apparatus 10 estimates the road shape and vehicle behavior of the recognized lane. Moreover, the control apparatus 10 performs the recognition process of the lane which the own vehicle drive | works using the road parameter obtained in this process.

Here, the road parameters of this embodiment will be described. 2A and 2B are diagrams for explaining a plurality of parameters expressing the road shape and the vehicle behavior. As shown in FIGS. 2A and 2B, the road parameters include the lateral displacement y _r of the center of gravity of the vehicle relative to the reference lane marker, the yaw angle φ _{r of} the vehicle relative to the lane, the pitch angle η of the vehicle, and the height from the road surface of the camera. H, road curvature (reciprocal of curvature radius) ρ, travel lane width W, and the like.

  Then, on the screen coordinate system x, y as shown in FIG. 3, the lane marker model is expressed by the equation (1) using the road parameters. The expression method of the lane marker model is not particularly limited, and a model known at the time of filing can be used.

In the formula (1), a to e are road parameters. When the height h of the camera 1 from the road surface is constant, each road parameter represents the following road and white line shape or vehicle behavior. That, a is the vehicle lateral displacement y _r in the lane, b is the ρ curvature of the road, c is the yaw angle phi _r with respect to the road of the vehicle (the optical axis of the camera 1), d is the vehicle ( The pitch angle η with respect to the road of the optical axis of the camera 1 and e corresponds to the lane width W of the road. Lateral displacement y _r indicates the position lateral displacement of the lane marker as a reference for the vehicle. In this embodiment, the lateral displacement y _r denotes a lateral displacement from the center of the lane.

  When an arbitrary point on the real coordinate system x (vehicle left-right direction), y (vehicle vertical direction), z (vehicle front-back direction) fixed to the vehicle is projected onto the screen coordinate system x, y, the following formula ( 2).

  However, f is a lens parameter and is a coefficient corresponding to the focal length of the lens. Assuming that the road curvature ρ is not so large and the road surface is a plane, the coordinates of the road white line with respect to the vehicle center line (camera center line) ahead of Z [m] are given by the following equation (3). However, this assumption is an assumption set for the simplification of the model. By increasing the order of the model, the general condition can be established.

  The following formula (4) is obtained by erasing X, Y, and Z from the above formulas (2) to (3).

  The above equation (1) is obtained by normalizing each parameter using the equation (5) with the camera height h having the smallest fluctuation among the variables being constant.

  In the initial state, since the shapes of the road and lane markers and the vehicle behavior are unknown, for example, a value corresponding to the median value is set as an initial value for each road parameter. That is, for example, the lane center is set for the lateral displacement amount correspondence parameter a of the host vehicle in the lane, the straight line is set for the road curvature correspondence parameter b, the yaw angle correspondence parameter c for the lane is zero degrees, and the pitch for the lane The angle corresponding parameter d is set to η degrees in a stopped state, and the lane width corresponding parameter e is set to a general road lane width.

  When detecting the lane marker, the control device 10 detects a lane marker that divides a traveling lane in which the host vehicle travels at a predetermined position on the right side and a predetermined position on the left side with respect to the center of the captured image. And a second detection area is set inside the two first detection areas set on the right and left sides. The first detection area is an area for accurately detecting the lane marker of the traveling lane in which the host vehicle travels, and the second detection area is a V-shaped lane marker (V-shaped lane marker characteristically seen in a branch lane described later). This is an area for accurately detecting a linear lane marker, which is regarded as a lane marker, and so on.

  Further, when the control device 10 acquires the information that the obstacle described above has been detected, the third detection area excluding the obstacle detection area in which the obstacle is predicted to exist from within the second detection area. Set. When there is an obstacle, the image of the lane marker on the road hidden by the obstacle is not included in the captured image, so the lane marker cannot be detected. Moreover, there is a possibility that a diagram based on the edge of the obstacle is erroneously detected as a lane marker. For this reason, the control apparatus 10 of this embodiment sets a 3rd detection area | region as an area | region for detecting the V-shaped lane marker characteristically seen in a branch lane in the scene where an obstruction exists.

  The control apparatus 10 detects the lane marker described on the road from the captured image. The method for detecting the lane marker expressed on the road based on the captured image of the in-vehicle camera 20 is not particularly limited, and a method known at the time of filing can be used.

  The control device 10 detects various lane markers, and determines whether or not the detected lane marker includes a V-shaped lane marker. In the present embodiment, the V-shaped lane marker is defined as a lane marker having two linear portions that expand at an angle of a predetermined value range upward in the captured image. The upward direction of the captured image is a direction in which an image corresponding to the sky exists. In other words, this is the opposite direction when the direction in which the image corresponding to the road surface on which the host vehicle is traveling is defined as the downward direction. Lane markers on both sides of the road show a shape that expands downward in the captured image near the vanishing point. In the present embodiment, since it is determined whether or not it is a V-shaped lane marker on the condition that the shape expands upward, a lane marker indicating a branch road can be identified from the lane marker near the vanishing point.

  In the V-shaped lane marker of the present embodiment, the upper end points of the two linear portions are expanded from each other, and the lower end points approach or intersect. The lower end points may be separated by a predetermined distance or may intersect each other. That is, the two linear portions may be in contact with each other in the valley portion of the V-shaped lane marker of the present embodiment, or may be separated from each other. In other words, in the V-shaped lane marker of the present embodiment, the distance between the end points that intersect when two linear portions are extended is less than a predetermined value, and the intersection angle of the extended linear portions is within a predetermined value range. It can be defined as a lane marker with a shape that is an angle. Less than a predetermined value includes zero. That is, the two lane markers may touch at the end points.

  When a lane marker having the above-described morphological features is detected in a lateral region within a predetermined distance from the lane marker detected in the previous process, the control device 10 according to the present embodiment treats this as a V-shaped lane marker. Recognize as The side region may be defined as a region within a predetermined distance from the left and right lane markers detected in the past (previous processing), or the left side region may be defined by the distance from the right lane marker, The right side region may be defined by the distance from the lane marker.

  In this embodiment, when the branch road is viewed from the upstream side of the branch point, the detection area of the V-shaped lane marker is determined based on the knowledge that the V-shaped lane marker by the branch road is seen in the right or left end area of the road. Limited to the side area of Road signs displayed on road surfaces such as rhombus figures indicating the presence of pedestrian crossings or bicycle crossings in Japan are written near the center between lane markers, so when you see the road sign from the upstream side of the branch point, The image corresponding to the road sign is generally found in the central area of the road. In this embodiment, a V-shaped lane marker indicating a branch road is determined from a road sign on the road surface by determining whether or not it is a V-shaped lane marker on the condition that it is present in the right or left end region of the road. Can be identified.

Although it does not specifically limit, the control apparatus 10 of this embodiment sets a side area | region according to the width of the lane divided by the lane marker detected in the past process.
For example, the control device 10 acquires the lane width (distance between lane markers) of the traveling lane on which the host vehicle travels from the detection result of the lane marker before Δt seconds in a scene where the host vehicle gradually approaches the branch point. The position within a predetermined distance according to the lane width from the right (or left) lane car is set as the left (or right) side area. This predetermined distance is obtained by adding or subtracting the margin obtained experimentally to the detected lane width.

  For example, the control device 10 acquires the lane width (distance between lane markers) of the traveling lane on which the host vehicle travels from the detection result of the lane marker before Δt seconds in a scene where the host vehicle gradually approaches the branch point. The position separated from the right (or left) lane car by a predetermined distance or more according to the lane width is set as the left (or right) side region. This predetermined distance is obtained by adding or subtracting an experimentally obtained margin to the lane width.

  Although not particularly limited, the lateral region of the present embodiment is 5 to 20% of the lane width, preferably about 8% to 12%, specifically 20 cm to 50 cm, preferably about 25 cm to 30 cm. The predetermined distance that defines the side distance can be set as appropriate based on the vehicle speed of the host vehicle, the road type of the traveling lane of the host vehicle, and the like.

  The control device 10 according to the present embodiment determines whether or not the V-shaped lane marker is detected from the second detection area after the V-shaped lane marker is detected in the side area. When the distance from the branch point is larger than the branch point, that is, when the distance from the branch point is large, there is a high possibility that the V-shaped lane marker corresponding to the image of the branch road is visible on the side (both sides) of the lane.

  For this reason, when a V-shaped lane marker is detected in the side area, it can be determined that a branch lane exists in the travel lane, but the lane marker that separates the lane and images of other objects are displayed because the distance is long. The possibility of overlapping and being detected as a V-shaped lane marker cannot be denied. That is, when the distance from the branch point is large, a plurality of objects in front of the host vehicle may overlap and form an image similar to a V-shaped lane marker.

  In the present embodiment, after a V-shaped lane marker is detected in the side area, a process for confirming whether or not a branch lane actually exists is performed. Specifically, when the V-shaped lane marker is detected in the second detection area set at the center side after the V-shaped lane marker is detected in the side area, the control device 10 Recognize that there is a branch lane in the travel lane. That is, a V-shaped lane marker is detected in a lateral region of the traveling lane in the host vehicle that travels at a first point relatively separated from the branch point, and then approaches the branch point relatively than the first point. When a V-shaped lane marker is detected in the second detection area in the host vehicle traveling at the second point, the detected V-shaped lane marker is an image of a branch lane, and the branch lane is actually a running lane. Then recognize. Thereby, it can prevent misidentifying the V-shaped image based on another object as the V-shaped lane according to the image of a branch lane.

  In this embodiment, a V-shaped lane marker can appear in a scene approaching a branching point by using the fact that the location where the V-shaped lane marker appears over time changes according to the positional relationship with the branching point. A highly sensitive second detection region is set, and the features of the branch lane can be detected with high accuracy. The control device 10 of the present embodiment can detect the V-shaped lane marker with high accuracy by changing the position of the detection region according to the degree of proximity to the branch point.

  Although not particularly limited, when the control device 10 acquires information indicating that the above-described obstacle has been detected, whether or not a linear lane marker that is regarded as a V-shaped lane marker is detected from the third detection region. You may make it judge. In the present embodiment, the linear lane marker is defined as a linear lane marker having a predetermined length or more and having a predetermined angle with respect to the xy axis of the captured image. If there is an obstacle ahead, a part of the V-shaped lane marker may be hidden behind the obstacle. In addition, when a part is hidden by an obstacle, the characteristic shape of the V-shaped lane marker changes. In the present embodiment, when a linear lane marker that is a feature of a branch lane when an obstacle is present appears in the third detection region, it is determined that the traveling lane of the host vehicle includes the branch lane. By narrowing down the detection area, the influence of the obstacle can be eliminated, and the features of the branch lane hidden by the obstacle can be detected with high accuracy.

  Further, the control device 10 calculates the road curvature of the traveling lane corresponding to the lane marker, the relative pitch angle with respect to the own vehicle, or the relative yaw angle with respect to the own vehicle, which is calculated based on the relative position of the lane marker of the traveling lane in which the own vehicle travels. When the time-dependent deviation of the amount of road parameters including any one or more of the above becomes a predetermined threshold value or more, it may be determined whether or not a V-shaped lane marker has been detected. If the road parameter value calculated during travel varies and the deviation over time increases, the travel lane may include a branch lane, and the width of the lane marker may change. In the present embodiment, the branch lane can be detected with high accuracy by examining the characteristics of the shape of the lane marker in the branch lane in a scene where there is a high possibility that the branch lane exists.

  When a V-shaped lane marker is detected, that is, when the detected lane marker includes a V-shaped lane marker, the control device 10 of the present embodiment has a branch lane in the travel lane on which the host vehicle travels. Recognize that exists. From the viewpoint that the V-shaped lane marker is a lane marker having a characteristic shape detected on a branch road, when the V-shaped lane marker is detected, a branch in which the traveling lane of the host vehicle branches in front of the traveling direction. Recognize that there is a lane. When the second detection area is set, the control device 10 recognizes that a branch lane exists in the travel lane of the host vehicle when a V-shaped lane marker is detected from the second detection area. Only when the V-shaped lane marker exists in the second detection area, the presence or absence of the branch lane can be recognized with high accuracy by recognizing whether or not the branch lane exists in the travel lane.

  In addition, when the control device 10 acquires information indicating that an obstacle exists in front of the host vehicle and a linear lane marker is detected from the third detection area, a branch lane exists on the traveling road. recognize. When there is an obstacle ahead, the influence of the presence of the front obstacle is eliminated by recognizing that there is a branch lane in the traveling lane only when a linear lane marker is present in the third detection area. Thus, the presence or absence of the branch lane can be recognized with high accuracy.

  Further, the control device 10 according to the present embodiment sends the detected relative position of the lane marker to the driving support device 80 and other various in-vehicle devices 200. Specifically, the control device 10 obtains the road parameters calculated based on the relative positions of the detected lane markers with respect to the host vehicle, and the lane identification information for identifying the traveling lane corresponding to the road parameters, the driving support device 80 and others. To various in-vehicle devices 200. In the present embodiment, the control device 10 does not send road parameters to the driving support device 80 when it is recognized that a branch lane exists on the road. It is difficult to calculate accurate road parameters when there are branch lanes on the road. In the present embodiment, when there is a possibility that the accuracy of the relative position is lowered due to the presence of the branch lane, the transmission of the information on the relative position including the road parameter is prohibited to ensure the accuracy of the driving support. The driving assistance device 80 temporarily stops driving assistance or performs driving assistance using past road parameters.

  Next, the procedure of the lane recognition process and the driving support process according to the present embodiment will be described with reference to FIGS.

  As shown in FIG. 4, the lane recognition device 100 of the driving support system 1 according to the present embodiment performs edge extraction processing S1, lane marker detection processing S2, road parameter calculation processing S3, lane recognition processing S4, road parameters as lane recognition processing. The output process S5 is performed. Furthermore, the driving assistance apparatus 80 performs driving assistance processing S6.

  In the edge extraction process S1, an edge of an image corresponding to a lane marker is extracted from a captured image including a road image. In general, since the lane marker is provided with a color that can be distinguished from a road, an edge can be extracted based on the contrast of the color.

  In the lane marker detection process S2, the continuity of the edges extracted in the edge extraction process S1 is examined, and the presence and position of the lane marker are detected based on edge information constituting the lane marker. Edge extraction is performed at least in a range including an image of the traveling lane of the host vehicle. If necessary, it can be performed in a range that includes an image of a traveling lane adjacent to the traveling lane of the host vehicle and an adjacent traveling lane adjacent to the traveling lane of the host vehicle. The control device 10 can calculate the position (relative position) of the lane marker with respect to the host vehicle based on the position of the lane marker in the captured image and the parameters (installation position, optical axis setting, etc.) of the camera 20. The control device 10 calculates the relative position of the host vehicle with respect to the lane marker of the lane in which the host vehicle travels.

  FIG. 5 shows a subroutine of the lane marker detection process S2. In step S101, since the shapes of the road and lane markers and the vehicle behavior are unknown in the initial state, for example, a value corresponding to the median value is set as an initial value for each road parameter. For example, the lane center is set as the lateral displacement amount correspondence parameter a of the host vehicle in the lane, the straight line is set as the road curvature correspondence parameter b, the yaw angle correspondence parameter c for the lane is zero degrees, and the pitch angle correspondence to the lane is set. The parameter d is set to η degrees in a stopped state, and the lane width corresponding parameter e is set to a general road lane width. The lane width may be read from the road information of the road to which the position information detected by the GPS 91 belongs by referring to the map information 92 including the road width and other road information stored in the navigation device 90.

  In step S102, as shown in FIGS. 6A and 6B, initial setting of a region for detecting lane marker candidate points is performed. As shown in FIG. 6A, the control device 10 sets the side areas SL and SR for detecting the V shape further upstream from the branch point. As shown in FIG. 6B, in the present embodiment, the control device 10 divides the traveling lane in which the host vehicle travels into a predetermined position on the right side and a predetermined position on the left side with respect to the center A of the captured image. First detection areas Q1R and Q1L for detecting lane markers are set. The area for detecting the lane marker candidate points for dividing the lane is set for the travel lane in which the host vehicle travels. If necessary, the first detection area may be set for another traveling lane adjacent to the traveling lane. Furthermore, you may set a 1st detection area | region about the other traveling lane adjacent to a traveling lane. In the initial state, it is expected that there is a large gap between the white line model in which the initial value is set as the road parameter and the road white line on the actual screen. Therefore, it is desirable to set a region as large as possible.

  In step S103, edge point information is input. In step S104, in the initial state, the detection area set in S102 is used. If the lane marker has already been detected by the previous processing, the detection area of the lane marker candidate point is set so that the previous detection position is at the center of the area. Furthermore, when the road parameter is calculated in the previous process, the lane marker candidate point detection area may be set at a position offset in the vehicle change direction from the lane marker model obtained in the previous process.

  In subsequent step S105, an edge point existing in the lane marker candidate point detection area is detected as a lane marker candidate point. In step S106, it is confirmed whether or not the number of lane marker candidate points is equal to or greater than a predetermined value. If the number is less than the predetermined value, it is determined that no lane marker is included in the lane marker candidate point detection area. If it is larger than the predetermined value, the process proceeds to step S107, and the detected lane marker candidate point information is output. The lane marker candidate point information is used in the next road parameter calculation process. Similar processing is performed not only on the lane marker of the traveling lane in which the host vehicle travels but also on other traveling lanes, and the lane marker of the traveling lane is also detected. In step S104, the lane marker candidate point detection area is set at a position offset by the road width W with respect to the previously detected lane marker based on the road parameter calculated in the previous process, so that the lane marker of the adjacent lane is An area that can exist can be set as a candidate point detection area.

  In the road parameter calculation process S3, any one of the road curvature of the traveling lane corresponding to each lane marker calculated based on the relative position of the own vehicle with respect to each lane marker, the relative pitch angle with respect to the own vehicle, or the relative yaw angle with respect to the own vehicle. Calculate road parameters including at least two. The control device 10 calculates road parameters for the travel lane on which the host vehicle travels.

  Specifically, two lane markers at the center of the image are selected, and road parameters based on the travel lane on which the host vehicle travels are calculated. This calculation method can be calculated by, for example, a method of fitting a plurality of detected lane marker candidate points to the lane marker model equation (1) by the least square method. Furthermore, it converts into an actual physical quantity using Formula (5). Based on the road parameters of the lane marker model detected from the captured image of the camera 20 in this manner, the relative positional relationship between the host vehicle and each traveling lane is recognized.

  Returning to FIG. 4, the lane recognition process S4 is performed in parallel with the road parameter calculation process S3 after the lane marker detection process S2. The control device 10 calculates a road parameter as a relative position of the host vehicle with respect to the lane marker.

FIG. 7 shows a subroutine of the lane recognition process S4. In step S201, road parameters of the host vehicle lane within a predetermined time are stored, and average values ρ _p , φ _rp , η _p , y _rp , W _p are calculated. Next, in step S202, whether or not changes in the current road parameters ρ, φ _r , η with respect to the past road parameters ρ _p , φ _rp , η _p are within predetermined values ρ _d , φ _rd , η _d . Determine. If it is within the predetermined value for all parameters, the process proceeds to step S207, it is determined that there is no branch lane in the currently traveling lane, and the lane recognition can be performed normally, and the process proceeds to step S208. In step S <b> 208, the control device 10 outputs the calculated road parameter to the driving support device 80. If any one of the parameters ρ, φ _r , η exceeds a predetermined value in step S202, it is determined that there is a possibility that a branch lane exists, and the process proceeds to step S203.

  If a V-shaped lane marker is detected in step S204, the process proceeds to step S205, where it is recognized that a branch lane exists in the traveling lane of the host vehicle. If it is recognized that a branch lane exists in the travel lane, the process proceeds to step S206, the output of road parameters is stopped, and the driving support function based on the currently calculated road parameters is interrupted. If no V-shaped lane marker is detected in S204, the process proceeds to step S207, where it is recognized that there is no branch lane in the traveling lane of the host vehicle. If it is recognized that there is no branch lane, the process proceeds to step S208, where road parameters are output in a normal cycle, and a driving support function based on the currently calculated road parameters is executed.

  An example of a V-shaped lane marker detection method in step S203 of FIG. 7 will be described with reference to FIG.

  In step S301, a road parameter is fetched, and a side area or a side area and a second detection area for detecting a V-shaped lane marker are set based on the road parameter fetched in S302. As shown in FIG. 6A, the control device 10 sets a lateral region within a predetermined distance from the detected lane marker. Further, as shown in FIG. 9A, the control device 10 is set to a predetermined position on the right side and a predetermined position on the left side with respect to the center of the captured image in order to detect a pair of left and right lane markers that divide the traveling lane. A second detection area Q2 for detecting a V-shaped lane marker is set inside the first detection areas Q1R and Q1L. In the present embodiment, from the knowledge that the V-shaped lane marker VL, which is a characteristic shape of the branch lane, is likely to appear in the second detection region Q2 when the host vehicle approaches the branch point to some extent. A second detection area is provided near the center of the travel lane between the first detection areas Q1R and Q1L.

  In S303, the edge image of the second detection area is captured, and in S304, probabilistic Hough transform is executed on the edge image to extract a linear portion. In this processing, the processing information in step S103 of FIG. 5 may be acquired in S303.

  In S305, the distance between the end points of the two linear portions of the extracted linear portions and the inclination are evaluated. In the present embodiment, a lane marker having two linear portions that expand at an angle of a predetermined value range in the upward direction of the captured image is detected as a V-shaped lane marker. Specifically, as shown in FIG. 9B, when the linear portion L1 and the linear portion L2 are detected, it is determined whether or not these are V-shaped lane markers.

  In the present embodiment, a lane marker having two linear portions that expand at an angle of a predetermined value range upward in the captured image is determined as a V-shaped lane marker. The control device 10 determines whether or not there are two linear portions, and if there are linear portions, the two linear portions are expanded upward in the captured image, and the angle of the expansion is predetermined. Determine whether it is in the range. Although this is an example of a determination method, in the present embodiment, whether or not the marker is a V-shaped lane marker is determined based on the following two points.

  The first condition is that the distance between the end points that intersect when the two linear portions L1 and L2 are extended is less than a predetermined value. Incidentally, the end points that intersect when extending the linear portions L1 and L2 shown in FIG. 9B are E11 and E21 (the end points E12 and E22 do not intersect even if they are extended). The threshold value (predetermined value of the distance) serving as an evaluation criterion for the distance between the end points E11 and E21 is set from the viewpoint of determining whether or not the two linear parts are continuous. You may define according to road width and road classification.

  The second condition is that the extended linear portions L1 and L2 expand to the upper side of the captured image, and the intersection angle θ is within a predetermined value range. This is because if the intersection angle is too small or the intersection angle is in the vicinity of 180 degrees, the lane marker is likely to be a simple linear object instead of a V-shape. Although not particularly limited, when the crossing angle is about 20 degrees to 160 degrees, it can be determined that the shape is V-shaped. Note that the V-shaped lane marker detection method is not limited to this, and one linear portion including a portion bent at an angle of a predetermined value range may be detected as a V-shaped lane marker. The lane marker that expands downward is a lane marker on both sides of the lane that intersects at the vanishing point, and is determined not to be a lane marker on a branch road.

  In this embodiment, in step S305, when one or more sets of linear portions satisfying the above two conditions exist, the process proceeds to step S306, and it is determined that a V-shaped lane marker exists. On the other hand, if there is no linear portion that satisfies the above two conditions, the process proceeds to step S307, and it is determined that there is no V-shaped lane marker.

  A method for detecting a V-shaped lane marker when an obstacle such as another vehicle exists in front of the host vehicle will be described with reference to FIGS. In step S401, road parameters are acquired, and in step S402, the obstacle information acquisition function acquires the presence of a front obstacle and its position information. The obstacle information may be calculated by the control device 10 itself, or may be calculated by the external driving support device 80.

  The control device 10 sets a third detection region for detecting a V-shaped lane marker based on the road parameter captured in S403. As illustrated in FIG. 11, the third detection region is set as a third detection region Q3 in the second detection region Q2 described with reference to FIG. 9A except the obstacle detection region QV where the obstacle exists. . In step S404, an edge image is captured. In step S405, probabilistic Hough transformation is performed on the edge image to extract a linear portion.

  When there are obstacles such as other vehicles ahead, the lane marker displayed on the road is hidden and the characteristic V-shape of the branch road cannot be detected. For this reason, in this detection process, a V-shaped lane marker is detected by a different method. In S406, the control device 10 evaluates the length and direction of the extracted linear portion. Specifically, the control device 10 determines whether or not a linear lane marker having a predetermined length or more and having a predetermined angle with respect to the xy axis of the captured image is detected from the third detection region. When a part of the left and right sides of the V-shaped lane marker characteristically seen in the branch lane is concealed by a front obstacle, there is a high possibility that the right or left part of the V-shape will be recognized. For this reason, on the premise that there is a forward obstacle, a linear shape having a predetermined length or more corresponding to the feature of the right or left portion of the V-shape and having a predetermined angle with respect to the xy axis of the captured image If a lane marker is detected, it is assumed that a V-shaped lane marker exists.

  In step S406, if there is a linear portion whose length is equal to or greater than the predetermined value and whose inclination with respect to the xy axis of the captured image is within the predetermined value range, a part of the left and right sides of the V-shaped lane marker is hidden. It is determined that the characteristic is a lane marker, and the process proceeds to S407 to determine that a V-shaped lane marker exists. This is because if the intersection angle with the xy axis is too small, or if the intersection angle with the xy axis is in the vicinity of 90 degrees or 180 degrees, the lane marker is likely to be a simple linear object instead of a V-shape. Although not particularly limited, when the angle of intersection with the xy axis is about 10 degrees to 80 degrees, it can be determined that the shape is V-shaped. On the other hand, if there is no linear part that satisfies the condition of step S406, the process proceeds to S408, and it is determined that there is no V-shaped lane marker.

  Returning to FIG. 4, in the road parameter output process S <b> 5, the control device 10 outputs the calculated parameters and the lane identification information for identifying the lane targeted for the lane recognition process to the driving support device 80. In the driving support process S6, the driving support device 80 performs driving support for the host vehicle using the road parameters acquired from the lane recognition device 100.

  Hereinafter, the driving support process S6 will be described. In the present embodiment, the recognition result of whether or not a branch lane exists on the road is used in the driving support device 80.

  The driving support device 80 of the present embodiment includes a vehicle information acquisition function that acquires vehicle information related to travel from a vehicle, a road parameter calculated by the lane recognition function of the lane recognition device 100, and a travel lane corresponding to the road parameter. It has a lane identification information acquisition function for acquiring lane identification information to be identified. The driving support device 80 also has a target lateral position setting function for setting a target lateral position with respect to the traveling lane based on the acquired vehicle information, parameters, and lane identification information, a target lateral position of the traveling lane, and the host vehicle. And a driving support function for supporting the traveling of the vehicle so that the relative positional relationship is maintained.

Based on FIG. 12, the driving assistance process of this embodiment is demonstrated.
In step S <b> 301, the driving support device 80 acquires each sensor signal as vehicle information from the various sensors 51 of the vehicle controller 50. Specifically, the vehicle signal includes signals from a current sensor of a vehicle speed sensor, a steering angle sensor, and a steering actuator.

  In step S302, the driving support device 80 acquires the road parameters calculated by the lane recognition device 100 and lane identification information for identifying the lane to be subjected to the lane recognition process.

In step S303, the driving support device 80 sets a target lateral position for driving the host vehicle based on the road parameters acquired in step S302 and the lane identification information for identifying the lane to be subjected to the lane recognition process. For example, when the road parameter of the traveling lane is acquired, the center of the lane can be set as the target lateral position by setting y _r = −W / 2. When road parameters based on the right adjacent lane are acquired, y _r = W / 2 is the center of the lane. Of course, the position offset from the center of the lane can be set as the target lateral position according to the situation. In addition, when the lane to be recognized is switched, if the already acquired value is used, the target lateral position may change abruptly. For this reason, when the lane is switched, by setting the lateral displacement amount immediately after the switching as the target lateral position, it is possible to support driving while maintaining the previous lateral position.

  In step S304, the driving support device 80 calculates a steering control amount corresponding to the sensor signal, the road parameter, and the target lateral position. Based on the calculated steering control amount, a vehicle control command is sent to the vehicle controller 50 and / or the steering device 70.

  In step S305, the vehicle controller 50 and / or the steering device 70 drives the steering actuator 71 according to the steering control amount included in the control command. Thereby, the own vehicle can drive | work in the state which maintained the predetermined positional relationship with respect to the target lateral position of a predetermined travel road.

  As described above, when a V-shaped lane marker is detected, the lane recognition device 100 according to the present embodiment determines that a branch lane exists in the host vehicle travel lane, and stops outputting road parameters. This is because the distance between the lane markers is not constant in the branch lane, and the relative position between the lane marker and the host vehicle may not be accurately calculated. In the present embodiment, when it is determined that a branch lane exists, the driving of the host vehicle is controlled based on inaccurate road parameters by not sending road parameters to the driving support device 80. To prevent.

  Specifically, FIG. 13 shows a scene in a branch lane in which the driving support system 1 of the present embodiment functions effectively. As shown in FIG. 13, since there is a branch lane, the lane marker on the main line is interrupted and a part thereof is missing. For this reason, the lane markers that delimit the main line are not the same width. Focusing on the right lane, the width of the two lane markers that delimit the right lane is wide on the near side, but branches off on the far side, which is the traveling direction, and suddenly becomes narrow. Note that the “branch lane” in the present embodiment means a lane in which the number of lanes decreases.

  Furthermore, specifically, the operation of the lane recognition process of the present embodiment will be described based on FIG. 14 by taking a branch lane as an example. As shown in FIG. 14, since a branch lane exists in the travel lane of the road on which the host vehicle travels, the lane marker P1L on one side of the travel lane deviates from the main lane as the lane marker of the branch lane. Note that the travel lane on the right side of the travel lane in the figure does not include a branch lane.

  As shown in FIG. 14, the intervals Q11 and Q12 between the lane markers P1L and P1R of the travel lane change along the traveling direction E of the host vehicle V. The distances Q21 and Q22 between the lane markers P2L and P2R in the travel lane do not change along the traveling direction E of the host vehicle V. In the situation of FIG. 14, the lane marker of the travel lane on which the host vehicle is traveling is diminished due to branching.

  In such a case, the lane marker PL1 deviating in the direction of the branch lane may be misrecognized as a lane marker of the traveling lane of the host vehicle, and the driving assistance may be continued on the assumption that the lane marker is normally detected. There is. Such driving assistance leads the vehicle in the wrong direction. Moreover, since the lane marker model of a general lane and the lane marker model of a branch lane do not match in shape, a road parameter with an abnormal value may be calculated. In the present embodiment, in order to prevent such a phenomenon, the presence of a branch lane is detected, and erroneous lane recognition and driving assistance are prevented from being continuously performed. For this reason, in this embodiment, after quickly and accurately determining whether or not a branch lane exists in the travel lane on which the host vehicle travels, the output of road parameters is temporarily stopped, and the driving support process is performed. It is possible to prevent erroneous driving support control from being performed by interruption (or execution based on road parameters calculated in the past).

  As described above, even when there is an obstacle in front of the host vehicle, the V-shaped lane marker can be detected in consideration of the portion hidden by the presence of the obstacle, so there is another vehicle ahead. Even in such a situation, it is possible to stably determine whether a branch lane exists.

  Since the lane recognition device 100 of the present embodiment is configured and operates as described above, the following effects can be obtained.

  [1] According to the lane recognition device 100 of the present embodiment, the presence of a branch road is determined based on the characteristic of the shape of the lane marker that is characteristically found in the branch lane. For example, based on the characteristic of the shape of the lane marker. Because it recognizes the presence of branch roads, it can quickly and accurately recognize whether there are branch lanes on the road of the vehicle, even on roads with different shapes that vary in curvature and pitch angle. it can.

  [2] According to the lane recognition device 100 of the present embodiment, since the V-shaped lane marker is detected in the side region where the image of the branch point is likely to appear when viewed from the own vehicle heading for the branch point, it is high. The presence of a branch point can be recognized with accuracy.

  [3] According to the lane recognition device 100 of the present embodiment, after a V-shaped lane marker is detected in the lateral region, the captured image is detected in order to detect the lane marker that separates the traveling lane in which the host vehicle is traveling. When a second detection area for detecting a V-shaped lane marker is set between the first detection areas set to the left and right, and the V-shaped lane marker is detected from the second detection area, It is determined that the travel lane includes a branch lane. After the V-shaped lane marker is detected in the side area, that is, in a scene approaching the branch lane, there is a high possibility that the V-shaped lane marker which is a feature of the branch lane appears in the second detection area. In the present embodiment, the presence or absence of a branch lane can be accurately detected by changing the detection area according to the positional relationship with the branch point.

  [4] According to the lane recognition device 100 of the present embodiment, when an obstacle is detected in front of the host vehicle, the third detection area excluding the obstacle detection area where the obstacle exists from the second detection area. When a linear lane marker is detected from within the third detection region, it is determined that the traveling lane of the host vehicle includes a branching lane. If there is an obstacle, a part of the V-shaped lane marker may be hidden behind the obstacle. In addition, when a part is hidden by an obstacle, the characteristic shape of the V-shaped lane marker changes. In the present embodiment, when a linear lane marker that is a feature of a branch lane when an obstacle is present appears in the third detection region, it is determined that the traveling lane of the host vehicle includes the branch lane. By narrowing down the detection area, the influence of the obstacle can be eliminated, and the features of the branch lane hidden by the obstacle can be detected with high accuracy.

  [5] According to the lane recognition device 100 of the present embodiment, the road curvature of the traveling lane corresponding to the lane marker, the relative pitch angle with respect to the own vehicle, calculated based on the relative position of the lane marker of the traveling lane on which the host vehicle is traveling. Whether or not a V-shaped lane marker or a linear lane marker is detected when a time-dependent deviation of a road parameter including at least one of the relative yaw angles with respect to the host vehicle exceeds a predetermined threshold value. Perform the decision. When road parameters calculated during travel vary and deviations with time increase, branch lanes are included in the travel lane, and the width of the lane marker may change. In the present embodiment, the branch lane can be detected with high accuracy by examining the characteristics of the shape of the lane marker in the branch lane in a scene where there is a high possibility that the branch lane exists.

  [6] The driving support system 1 of the present embodiment stops the output of road parameters when there is a branch lane in the traveling lane of the own vehicle. It is possible to prevent the support process from being executed.

  The embodiment described above is described for facilitating the understanding of the present invention, and is not described for limiting the present invention. Therefore, each element disclosed in the above embodiment is intended to include all design changes and equivalents belonging to the technical scope of the present invention.

  That is, in the present specification, the present invention will be described by taking the driving support system 1 for a vehicle including the lane recognition device 100 and the in-vehicle device 200 including the driving support device 80 as an example, but the present invention is not limited to this. It is not something.

  In this specification, as an example of the lane recognition device according to the present invention, a lane recognition device 100 including the control device 10, the camera 20, and the output device 30 will be described as an example. However, the present invention is not limited to this. Absent.

  In the present specification, as an example of a lane recognition device according to the present invention including an image acquisition unit, a lane marker detection unit, a lane recognition unit, and an obstacle information acquisition unit, an image acquisition function, a lane marker detection function, Although the lane recognition device 100 including the control device 10 that executes the lane recognition means and the obstacle information acquisition function will be described as an example, it is not limited thereto.

DESCRIPTION OF SYMBOLS 1 ... Driving assistance system 100 ... Lane recognition apparatus 10 ... Control apparatus 11 ... CPU
12 ... ROM
13 ... RAM
DESCRIPTION OF SYMBOLS 20 ... Camera 30 ... Output device 31 ... Display 31 ... Speaker 200 ... In-vehicle device 50 ... Vehicle controller 51 ... Various sensors 60 ... Drive system 70 ... Steering device 80 ... Driving support device 81 ... Radar device 90 ... Navigation device 91 ... GPS
92 ... Map information, road information, road width information

Claims (6)

Image acquisition means for acquiring a captured image around the vehicle by a camera mounted on the host vehicle;
Lane marker detection means for detecting a lane marker displayed on the road based on the captured image;
In the case where a V-shaped lane marker having two linear portions expanding at an angle of a predetermined value range in the upward direction of the captured image is detected in a lateral region within a predetermined distance from the detected lane marker, A lane recognition device comprising: lane recognition means for recognizing that a branch lane exists in a travel lane in which the host vehicle travels.   2. The lane recognition device according to claim 1, wherein the lane recognition unit sets the predetermined distance that defines the side area according to a width of a lane that is divided by the lane marker detected in a past process.   The lane recognition means detects a V-shaped lane marker inside a pair of left and right first detection areas set at a predetermined position on the right side and a predetermined position on the left side with respect to the center of the captured image. When a detection area is set and the V-shaped lane marker is detected in the second detection area after the V-shaped lane marker is detected in the side area, The lane recognition device according to claim 1 or 2, which recognizes that a branch lane exists. Obstacle information acquisition means for acquiring information on obstacles existing in front of the host vehicle,
The lane recognition unit sets a third detection area excluding the obstacle detection area where the obstacle exists from the second detection area when the information indicating that the obstacle is detected is acquired. In the third detection area, a linear portion having an angle within a predetermined value range with respect to the xy axis of the captured image and having a predetermined length or more is detected as the V-shaped lane marker, and the V-shaped lane marker is detected. The lane recognition device according to claim 3, wherein when a lane is detected, it is recognized that a branch lane exists in the travel lane of the host vehicle.   The lane recognition means calculates the road curvature of the traveling lane corresponding to the lane marker, the relative pitch angle with respect to the own vehicle, or the relative yaw with respect to the own vehicle, calculated based on the relative position of the lane marker of the traveling lane on which the own vehicle is traveling. Determining whether or not the V-shaped or linear lane marker is detected when a time-dependent deviation of a road parameter including at least one of the corners exceeds a predetermined threshold value; The lane recognition device according to claim 1, wherein the lane recognition device is a lane recognition device. A lane recognition device according to any one of claims 1 to 5, and a travel support device that supports travel of the host vehicle,
The lane recognition means of the lane recognition device recognizes the relative position of each lane marker detected by the lane marker detection means, and sends the relative position to the travel support device,
The travel support device includes a steering control unit that controls a steering operation based on a target lateral position set based on the relative position,
The lane recognition means does not send the relative position to the travel support device when it recognizes that a branch lane exists in the travel lane of the host vehicle.