Road marking recognition device

Abstract

The present application provides a kind of as road marking recognition device external recognition device (50), with: laser radar (5), as vehicle-mounted detector, electromagnetic wave is irradiated to the surroundings of vehicle, and the external condition of surroundings is detected in time sequence;And operation part (11), as road surface information acquisition unit, based on the information detected by laser radar (5), the road surface information of the road traveled by vehicle is acquired.The operation part (11) includes identification part (111), which respectively identifies the information representing road marking on the first area of the road at a specified distance from the vehicle at different positions on the road;And mapping part (114), which maps the information representing road marking identified by identification part (111) according to the direction of travel and speed of vehicle.

Description

Technical Field

[0001] This invention relates to an external identification device for recognizing the external conditions of a vehicle. Background Technology

[0002] As such a device, there is a known device that scans by varying the irradiation angle of the laser emitted from the lidar around a first axis parallel to the height direction and a second axis parallel to the horizontal direction, and detects the curb stones at the end of the road based on the position information of each detection point (see, for example, Patent Document 1).

[0003] In the aforementioned device, the number of detection points obtained through scanning is large, resulting in a significant processing burden for obtaining location information based on each detection point.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Application Publication No. 2020-149079. Summary of the Invention

[0007] One technical solution of the present invention is a road marking recognition device comprising an on-board detector that irradiates electromagnetic waves around the vehicle to detect the surrounding external conditions in a time sequence, and a road information acquisition unit that obtains road surface information of the road on which the vehicle is traveling based on the information detected by the on-board detector. The road information acquisition unit includes: an identification unit that identifies information representing road markings in a first area of ​​the road at a predetermined distance from the vehicle at different positions on the road; and a mapping unit that maps the information representing road markings identified by the identification unit based on the vehicle's direction of travel and speed. Attached Figure Description

[0008] The objectives, features, and advantages of the present invention are further illustrated by the following description of embodiments in conjunction with the accompanying drawings.

[0009] Figure 1A It is a diagram showing the conditions of vehicles driving on the road.

[0010] Figure 1B Show Figure 1A An example of point cloud data for a given location.

[0011] Figure 2 It is a block diagram that roughly shows the main structural components of the vehicle control unit.

[0012] Figure 3A This is a diagram illustrating the detection area of ​​the lidar.

[0013] Figure 3BThis is a diagram illustrating the detection area of ​​the lidar.

[0014] Figure 4A It was extracted from Figure 3B The image shows the detection area as observed from the perspective of [the viewpoint].

[0015] Figure 4B It is shown on a two-dimensional diagram Figure 4A The first region of the diagram.

[0016] Figure 5A This is a schematic diagram illustrating the projection angle of a lidar.

[0017] Figure 5B This is a diagram illustrating the second and third regions.

[0018] Figure 6 It is by Figure 2 A flowchart illustrating an example of the processing performed by the CPU of the controller.

[0019] Figure 7 This is an explanation Figure 6 The flowchart of the S10 process.

[0020] Figure 8 This is an explanation Figure 6 The flowchart for the S20 process.

[0021] Figure 9 This is an explanation Figure 6 The flowchart for the S30 process. Detailed Implementation

[0022] The embodiments of the invention will now be described with reference to the accompanying drawings.

[0023] The road marking recognition device of the present invention can be applied to vehicles with autonomous driving capabilities, i.e., autonomous vehicles. It should be noted that sometimes the vehicle using the road marking recognition device of this embodiment is referred to as "this vehicle" to distinguish it from other vehicles. This vehicle can be any one of the following: an engine vehicle with an internal combustion engine as the driving source, an electric vehicle with a drive motor as the driving source, or a hybrid vehicle with both an engine and a drive motor as driving sources. This vehicle can operate not only in an autonomous driving mode that does not require driver intervention, but also in a manual driving mode based on driver intervention.

[0024] When an autonomous vehicle is driving in autonomous driving mode (hereinafter referred to as automatic driving or autonomous driving), it identifies the surrounding external conditions based on detection data from onboard detectors such as cameras and LiDAR (Light Detection and Ranging). Based on the identification results, the autonomous vehicle generates a driving trajectory (target trajectory) from the current point in time to a predetermined time elapsed, and controls the driving actuators to make the vehicle travel along the target trajectory.

[0025] Figure 1A This is a diagram showing the vehicle 101, which is an autonomous driving vehicle, driving on road RD. Figure 1B This diagram illustrates an example of detection data (measurement points) obtained by a lidar mounted on vehicle 101. The lidar's measurement point is the point information reflected back from a specific point on the surface of an object by the emitted laser light. Specific point information includes the distance from the laser source to the point, the intensity of the reflected laser light, and the relative velocity of the object at that point. Furthermore, information will be obtained from... Figure 1B The data formed by multiple measurement points shown is called point cloud data. Figure 1B The middle shows with Figure 1A The point cloud data corresponding to the location. This vehicle 101 is based on... Figure 1B The point cloud data shown identifies the external conditions around the vehicle, more specifically identifying road structures and objects (including road markings), and generates a target trajectory based on the identification results.

[0026] However, as a method to fully identify the external conditions around the vehicle, increasing the number of illumination points of electromagnetic waves emitted from vehicle-mounted detectors such as lidar can be considered (in other words, increasing the density of electromagnetic wave illumination points, thereby increasing the density of point cloud data). On the other hand, increasing the number of illumination points may increase the processing load on the vehicle-mounted detector, increase the volume of detection data (point cloud data) obtained by the vehicle-mounted detector, and thus increase the processing load on the point cloud data. In particular, such as Figure 1B As shown, when there are many objects (people, buildings, etc. in addition to trees) beside the road, the capacity of point cloud data increases further. Furthermore, when dealing with such problems, it may be necessary to increase the size of the device, such as by increasing the number of lasers. Therefore, taking these points into consideration, the road marking recognition device is configured as follows in the embodiment.

[0027] <Outline>

[0028] The road marking recognition device of this embodiment intermittently illuminates the vehicle 101 traveling on road RD with illumination light (a type of electromagnetic wave) from the lidar 5 of the vehicle 101 in the direction of travel, and discretely acquires point cloud data at different locations on road RD. The illumination range of the illumination light emitted from lidar 5 forms a long strip in the road width direction intersecting with the direction of travel, so that the point cloud data acquired by lidar 5 based on the previous illumination light and the point cloud data acquired by lidar 5 based on the current illumination light are uninterruptedly connected in the direction of travel of road RD.

[0029] More specifically, the illumination range is defined as a first strip-shaped area along the road RD at a predetermined distance (corresponding to the depth distance described later) from the vehicle 101 based on the vehicle 101's speed (driving speed). A predetermined illumination point density is set for this first area. Then, the total number of illumination points emitted by the lidar 5 is controlled by setting the illumination point density for areas other than the first area to be lower than the illumination point density for the first area, or by stopping illumination for other areas. In other words, the number of illumination points of the lidar 5 can be reduced without reducing the accuracy of recognizing the position (distance from the vehicle 101) and size of objects (including road markings) based on point cloud data.

[0030] Further details are provided regarding this road marking recognition device.

[0031] <Structure of Vehicle Control Device>

[0032] Figure 2 This is a block diagram showing the main structural components of a vehicle control device 100, including a road marking recognition device. The vehicle control device 100 includes a controller 10, a communication unit 1, a positioning unit 2, an internal sensor group 3, a camera 4, a lidar 5, and a driving actuator AC. The communication unit 1, positioning unit 2, internal sensor group 3, camera 4, lidar 5, and actuator AC are communicatively connected to the controller 10 via CAN communication lines, etc. Additionally, the vehicle control device 100 includes an external recognition device 50, which is part of the road marking recognition device. The external recognition device 50 identifies the external conditions (including road markings) around the vehicle based on detection data from onboard detectors such as the camera 4 and lidar 5.

[0033] The communication unit 1 communicates with various servers (not shown) via a network including wireless communication networks such as the Internet and mobile phone networks, and periodically or at any time obtains map information, driving record information, and traffic information from the servers. The network includes not only public wireless communication networks but also closed communication networks set up for each designated management area, such as wireless LAN, Wi-Fi, and Bluetooth. The obtained map information is output to the storage unit 12, and the map information is updated. The positioning unit (GNSS unit) 2 has a positioning sensor that receives positioning signals transmitted from positioning satellites. Positioning satellites are artificial satellites such as GPS satellites and quasi-zenith satellites. The positioning unit 2 uses the positioning information received by the positioning sensor to determine the current position (latitude, longitude, and altitude) of the vehicle 101.

[0034] The internal sensor group 3 is a collective term for multiple sensors (internal sensors) that detect the driving status of the vehicle 101. For example, the internal sensor group 3 includes a vehicle speed sensor to detect the vehicle speed, an acceleration sensor to detect the longitudinal and lateral acceleration (lateral acceleration) of the vehicle 101, a speed sensor to detect the rotational speed of the driving source, and a yaw rate sensor to detect the rotational angular velocity of the vehicle 101's center of gravity around its vertical axis. Sensors that detect driver operations in manual driving mode, such as operations on the accelerator pedal, brake pedal, and steering wheel, are also included in the internal sensor group 3.

[0035] Camera 4 has imaging elements such as CCD (charge-coupled device) and CMOS (complementary metal-oxide-semiconductor) to capture images of the surroundings (front, rear, and sides) of the vehicle 101. LiDAR 5 receives scattered light from the vehicle 101 in all directions in response to the illumination light, thereby determining the distance, position, shape, etc. of objects from the vehicle 101 to the surrounding area.

[0036] The actuator AC is a driving actuator used to control the movement of the vehicle 101. When the driving source is an engine, the actuator AC includes a throttle actuator for adjusting the opening of the engine's throttle valve (throttle opening). When the driving source is a drive motor, the drive motor is included in the actuator AC. The brake actuator that operates the braking system of the vehicle 101 and the steering actuator that drives the steering system are also included in the actuator AC.

[0037] The controller 10 is composed of an electronic control unit (ECU). More specifically, the controller 10 is configured as a computer having an arithmetic unit 11 such as a CPU (microprocessor), a storage unit 12 such as ROM (read-only memory) and RAM (random access memory), and other peripheral circuits (not shown) such as I / O interfaces. It should be noted that multiple ECUs with different functions, such as an engine control ECU, a drive motor control ECU, and a braking device ECU, can be set separately, but... Figure 2 For convenience, controller 10 is shown as a collection of these ECUs.

[0038] The storage unit 12 stores high-precision, detailed map information (referred to as high-precision map information). This high-precision map information includes road location information, road shape (curvature, etc.), road slope information, location information of intersections and forks in the road, information on the number of lanes (driving lanes), lane width, and the location information of each lane (the center position of the lane, information on lane boundaries), location information of landmarks (traffic lights, signs, buildings, etc.) that serve as markers on the map, and information on the road surface contours, such as road surface undulations. In addition to the two-dimensional map information described later, the storage unit 12 also stores information about various control programs, thresholds used by the programs, and setting information for vehicle-mounted detectors such as the LiDAR 5 (illumination point information described later).

[0039] The arithmetic unit 11 has a functional structure comprising an identification unit 111, a delimitation unit 112, a setting unit 113, a mapping unit 114, and a driving control unit 115. It should be noted that, as... Figure 2 As shown, the identification unit 111, the delineation unit 112, the setting unit 113, and the mapping unit 114 are included in the external identification device 50, which is a road marking identification device. As described above, the external identification device 50 identifies the external conditions (including road markings) around the vehicle based on detection data from vehicle-mounted detectors such as the camera 4 and the lidar 5. The identification unit 111, the delineation unit 112, the setting unit 113, and the mapping unit 114 included in the external identification device 50 will be described in detail later.

[0040] In autonomous driving mode, the driving control unit 115 generates a target trajectory based on the external conditions around the vehicle identified by the external recognition device 50, and controls the actuator AC to make the vehicle 101 drive along the target trajectory. It should be noted that in manual driving mode, the driving control unit 115 controls the actuator AC based on driving commands (steering operations, etc.) from the driver obtained by the internal sensor group 3.

[0041] <Detection area of ​​lidar>

[0042] The external recognition device 50, which serves as a road marking recognition device, sets a detection area for the lidar 5. Figure 3A and Figure 3B This diagram illustrates the detection area of ​​the lidar 5, which detects objects and other features on the road RD in the direction of travel. In this embodiment, the objects include road surface shapes such as unevenness, steps, and undulations; objects on the road RD (such as road-related equipment like traffic signals, signs, ditches, walls, fences, and guardrails); objects on the road RD (including other vehicles and road obstacles); and road markings on the road surface. Road markings include white lines (including lines of different colors such as yellow), curb lines, road studs, and may also be called lane marks. Furthermore, objects among the objects and other features pre-defined as detection targets are called detection targets. The lidar 5 receives scattered light from the illumination light; therefore, the first area defined as the illumination range of the illumination light is the detection area. In other words, the first area in this embodiment is the illumination range, which is the detection area.

[0043] Figure 3A The area RS shown is where vehicle 101 is located. Figure 1A The road segment identified by the identification unit 111 during the travel of the road RD. In the embodiment, the area corresponding to the road RD enclosed by the dividing lines RL and RB of the road RD is called the road segment.

[0044] Figure 3A The detection area BX shown corresponds to the first area set as the illumination range at a location on the road RD at a specified depth distance of 101 from this vehicle. Figure 3A The length of the arrow line represents the depth distance L. As shown by the arrow line, in this embodiment, the distance from the installation position of the lidar 5 to the lower center of the detection area BX is called the depth distance. The depth distance is set to a predetermined length based on the vehicle speed of the vehicle 101.

[0045] The detection area BX is set to include the area that should be observed when driving in autonomous driving mode. As an example, it can also be set so that the center position of the detection area BX in the road width direction overlaps with the center position of the road segment in the road width direction. Furthermore, the width (length in the road width direction) of the detection area BX is set to be longer than the road width in the depth distance L. In addition, the depth width of the detection area BX (the vertical width of the strip area constituting the detection area BX) is set to a predetermined length based on the vehicle speed of the vehicle 101.

[0046] Figure 3BThis diagram schematically illustrates detection areas BX, BX1, and BX2 located at different depth distances from the vehicle 101 on road RD. The example detection areas BX, BX1, and BX2 are used to illustrate their variability. Additionally, Figure 4A It was extracted from Figure 3B The diagram shows detection areas BX, BX1, and BX2 as observed from the perspective of [the device / entity]. Detection area BX corresponds to a first area set on road RD at a depth distance of, for example, L = 180m from the vehicle 101. Detection area BX1 corresponds to a first area set on road RD at a depth distance of, for example, L1 = 100m from the vehicle 101. Detection area BX2 corresponds to a first area set on road RD at a depth distance of, for example, L2 = 10m from the vehicle 101.

[0047] In this embodiment, the external detection device 50 sets a first region based on the driving speed of the vehicle 101. The faster the vehicle 101 travels, the longer the braking distance; therefore, to avoid collisions, a longer depth distance is required to detect distant objects. Thus, it has the function of changing the position of the detection region BX according to the driving speed of the vehicle 101. For example, when the vehicle 101 is stationary, a detection region BX2 is set as the first region on road RD at a depth distance of L2 = 10m. Furthermore, when the vehicle 101 is traveling at a speed of 100 km / h, a detection region BX1 is set as the first region on road RD at a depth distance of L1 = 100m. Moreover, when the vehicle 101 is traveling at a speed of 180 km / h, a detection region BX is set as the first region on road RD at a depth distance of L = 180m.

[0048] like Figure 3B and Figure 4A As shown, when viewed from the perspective of LiDAR 5, the detection area BX is smaller the further away from the depth distance of the vehicle 101. However, when the road width increases with the increase in depth distance, the first area sometimes increases with the increase in depth distance.

[0049] <Size of the detection area>

[0050] Figure 4B It is shown on a two-dimensional diagram Figure 4A The image shows the detection areas. Figure 4BFrom the perspective of the LiDAR 5, a smaller detection area appears larger on a two-dimensional map. This is because the detection area BX, which is farther from the depth distance L of the vehicle 101, has a larger depth width D. More specifically, in this embodiment, the detection area BX, as the first region, is set at a position where the depth distance increases with the vehicle 101's speed, as described above. Because the vehicle 101 travels a longer distance within the measurement interval of the LiDAR 5 when the vehicle speed is high, in order to avoid creating blank areas in the point cloud data (in other words, to avoid gaps in the direction of travel), the depth width of the detection area BX, which is the first region, is increased, thereby expanding the acquisition range of the point cloud data for each measurement.

[0051] <Angle of projection of the illuminating light>

[0052] Figure 5A This is a schematic diagram illustrating the projection angle θ0 (the angle relative to the horizontal direction of the illumination light) of the lidar 5. The external identification device 50 adjusts the depth distance on the path RD of the illuminated light by changing the projection angle θ0, thereby changing the illumination direction of the illumination light vertically.

[0053] exist Figure 5A In the case where the illumination light shines on road RD at a depth distance L2 of 10m, the road surface is illuminated at an incident angle θ2. Furthermore, when the illumination light shines on road RD at a depth distance L1 of 100m, the road surface is illuminated at an incident angle θ1. Additionally, when the illumination light shines on road RD at a depth distance L1 of 180m, the road surface is illuminated at an incident angle θ.

[0054] Typically, when the road gradient in the direction of travel is zero, the projection angle θ0 coincides with the incident angle of the irradiated light on the road surface. The depth distance L is then calculated using the following formula (1).

[0055] L=1 / tanθ0×H (1)

[0056] It should be noted that the symbol H represents the height of the road surface, or more specifically, the vertical distance from the location of the lidar 5 to the road surface.

[0057] The external identification device 50 decreases the projection angle θ0 when it wants to increase the depth distance, and increases the projection angle θ0 when it wants to decrease the depth distance. For example, if the depth distance changes from 70m to 100m, the external identification device 50 decreases the projection angle θ0 to ensure that the light shines on the 100m depth distance location. Conversely, if the light does not shine on the road RD when it has a downhill slope, the external identification device 50 increases the projection angle θ0 to ensure that the light shines on the road RD.

[0058] <Number of points illuminated by the illumination light>

[0059] The external identification device 50 calculates the illumination points of the laser radar 5 within the detection area BX1 set above. More specifically, the external identification device 50 calculates the illumination points based on the angular resolution calculated according to the minimum size of the detection object (e.g., 15cm both horizontally and vertically) and the depth distance (e.g., L1 (100m)). In this case, the angular resolution requires 0.05 degrees in both the vertical direction (also known as the vertical direction) and the horizontal direction (also known as the road width direction) of the grid points described later. It should be noted that when detecting detection objects smaller than 15cm or when setting the detection area at a depth distance longer than L1 (100m), the number of illumination points within the detection area can be increased to improve the angular resolution.

[0060] The external recognition device 50 calculates, for example, illumination points arranged in a grid within the detection area BX1, and aligns the grid spacing in the vertical and horizontal directions with the angular resolution. To increase the vertical angular resolution, the detection area BX1 is divided vertically according to the number of points based on the angular resolution, narrowing the grid spacing and thus increasing the number of illumination points. Conversely, to decrease the vertical angular resolution, the detection area BX1 is divided vertically according to the number of points based on the angular resolution, widening the grid spacing and thus reducing the number of illumination points. The same applies to the horizontal direction.

[0061] The external identification device 50 generates information indicating the position of the illumination point calculated based on the angular resolution (hereinafter referred to as illumination point information), and stores it in the storage unit 12 in a corresponding relationship with the position information indicating the current driving position of the vehicle 101.

[0062] When the vehicle 101 is driving in autonomous driving mode, the external identification device 50 reads the illumination point information corresponding to the current driving position of the vehicle 101 from the storage unit 12, and sets the illumination point of the lidar 5 within the detection area according to the illumination point information. As a result, the illumination light from the lidar 5 illuminates the set illumination point.

[0063] It should be noted that the illumination light of the lidar 5 can be emitted by grating scanning, or by emitting the illumination light intermittently only on the illumination points arranged in a grid pattern within the detection area, or by other methods.

[0064] <Structure of External Identification Device>

[0065] Detailed description of the external recognition device 50 as a road marking recognition device.

[0066] As described above, the external identification device 50 includes an identification unit 111, a delineation unit 112, a setting unit 113, a mapping unit 114, and a lidar 5.

[0067] <Identification Department>

[0068] The identification unit 111 identifies the road structure and the objects to be detected on the road RD in the direction of travel of the vehicle 101, based on image information captured by the camera 4 or detection data measured by the lidar 5. Road structure refers to, for example, straight roads, curves, forks, tunnel entrances and exits.

[0069] Furthermore, the recognition unit 111 detects road markings, for example, by performing brightness filtering on data representing a flat road surface. In this case, the recognition unit 111 determines that a road marking is present when the height of a road surface with brightness exceeding a predetermined threshold is approximately the same as the height of a road surface that does not exceed the threshold.

[0070] When the second and third regions are defined by the delineation unit 112 (described later), and the setting unit 113 (described later) sets the lidar 5 to a detection region BY1 as the second region and a detection region BZ1 as the third region in addition to the detection region BX1 as the first region, the lidar 5 also irradiates illumination light onto the detection regions BY1 and BZ1 in addition to the detection region BX1.

[0071] The recognition unit 111 uses the temporal detection data (measurement points) detected by the lidar 5 in the detection area BX1 to generate three-dimensional point cloud data, and uses the temporal detection data (measurement points) detected by the lidar 5 in the detection areas BY1 and BZ1 respectively to generate three-dimensional point cloud data.

[0072] It should be noted that when obtaining detection data by illuminating the detection area BX1 with illumination light from the lidar 5, it is not necessary to generate 3D point cloud data in the detection areas BY1 and BZ1.

[0073] <Identification of Road Structure>

[0074] The identification unit 111 further details its identification of road structures. The identification unit 111 identifies curbs, walls, ditches, guardrails, or road markings of the road RD (road RD) in the direction of travel (i.e., ahead) contained in the point cloud data generated based on the detection area BX1 as the road RD's boundary lines RL and RB. Based on this, it identifies the road structure in the direction of travel indicated by the boundary lines RL and RB. As described above, road markings include white lines (including lines of different colors), curb lines, road studs, etc., and the markings indicated by these road markings define the driving lanes of the road RD. In this embodiment, the lines corresponding to the road RD's boundary lines RL and RB defined by the above-mentioned markings are referred to as road markings.

[0075] The identification unit 111 identifies the area enclosed by the dividing lines RL and RB as the area corresponding to road RD (the aforementioned road segment). It should be noted that the method for identifying road segments is not limited to this, and other methods can also be used for identification.

[0076] In addition, the recognition unit 111 identifies road surface shapes such as bumps, steps, and undulations exceeding 15cm and objects exceeding 15cm in length and width, among the three-dimensional objects such as road RD in the road segment in the direction of travel (i.e., ahead) contained in the point cloud data generated based on the detection area BX1 as detection objects.

[0077] <Identification of Detected Objects>

[0078] The identification unit 111 identifies the detection objects in a further detailed manner. When the second region and the third region are delineated by the delineation unit 112 (described later), and in addition to the detection region BX1, illumination light is also shone from the lidar 5 onto the detection region BY1 corresponding to the second region and the detection region BZ1 corresponding to the third region, the identification unit 111 identifies objects such as road surface shapes of approximately 100cm and objects of approximately 100cm horizontally and vertically in the road segment in the direction of travel (i.e., the road RD) contained in the point cloud data generated based on the detection regions BY1 and BZ1 as detection objects.

[0079] It should be noted that, when the detection data is obtained by the lidar 5 illuminating the detection area BX1 with illumination light, the recognition unit 111 can also identify the road surface shape and the detection object in the road segment ahead of the direction of travel based on the point cloud data generated based on the detection area BX1.

[0080] <Delineation Department>

[0081] The delineation unit 112 delineates the aforementioned first area as the illumination range of the laser radar 5. More specifically, as... Figure 4B As shown, the delineation unit 112 delineates a first strip-shaped region with a predetermined depth and width D1 at a predetermined distance (e.g., depth distance L1) from the vehicle 101 in the direction of travel on the road RD.

[0082] As described above, the delineation unit 112 calculates the depth distance L1 from the current position of the vehicle 101 to the first region and the depth width of the first region (the width of the strip area constituting the first region) D1 based on the vehicle speed of the vehicle 101 detected by the vehicle speed sensor of the internal sensor group 3. The same applies to the depth distance L2 from the current position of the vehicle 101 to the first region and the depth width D2 of the first region, as well as the depth distance L from the current position of the vehicle 101 to the first region and the depth width D of the first region.

[0083] Because of this configuration, when the LiDAR 5 intermittently illuminates the detection area BX1 corresponding to the first area from the vehicle 101 moving on the road RD, the first area can be delineated in a way that the point cloud data detected by the LiDAR 5 based on the previous illumination light and the point cloud data detected by the LiDAR 5 based on the current illumination light are uninterruptedly connected in the direction of travel. Thus, continuous detection data from the LiDAR 5 in the direction of travel can be obtained.

[0084] The delineation unit 112 calculates the length W1 in the road width direction of the strip region constituting the first region (corresponding to the detection region BX1) based on the road structure identified by the identification unit 111. The delineation unit 112 calculates the length W1 such that the length W1 is sufficiently longer than the road width of the road RD at the depth distance L1. The length W2 at the depth distance L2 and the length W at the depth distance L are calculated in the same way.

[0085] With this configuration, the detection data of LiDAR 5 includes the left and right ends (boundary lines RL and RB) of the road RD.

[0086] In addition, such as Figure 5B As shown, the delineation unit 112 can also delineate a second region on the road RD that is closer to the vehicle 101 than the first region corresponding to the detection region BX1 (in other words, the distance from the vehicle 101 is shorter than the depth distance L1) and a third region on the road RD that is farther from the vehicle 101 than the first region (in other words, the distance from the vehicle 101 is longer than the depth distance L1 + depth width D1).

[0087] <Settings Department>

[0088] Based on the first region defined by the delineation unit 112, the setting unit 113 sets a detection area BX1 corresponding to the first region for the lidar 5. As described above, in this embodiment, the detection area BX1 corresponds to the illumination range of the illumination light.

[0089] Furthermore, when the second and third regions are defined by the delineation unit 112, the setting unit 113 sets a detection region BY1 corresponding to the second region and a detection region BZ1 corresponding to the third region for the lidar 5 based on the second and third regions. In the embodiment, the detection regions BY1 and BZ1 also correspond to the illumination range of the illumination light.

[0090] <Different Angled Resolutions>

[0091] Based on the required angular resolution for measuring the lidar 5, the setting unit 113 calculates and sets the number of illumination points (in other words, the illumination point density) within the detection area BX1 corresponding to the first area. As described above, the illumination points arranged in a grid within the detection area BX1 are calculated, and the vertical and horizontal spacing of the grid points corresponds to the angular resolution.

[0092] Furthermore, when the setting unit 113 sets a detection area BY1 corresponding to the second region and a detection area BZ1 corresponding to the third region for the lidar 5, it calculates and sets the number of illumination points for the first resolution for the detection area BX1, and calculates and sets the number of illumination points for the second resolution, which is lower than the density of the first resolution, for both the detection areas BY1 and BZ1. The first resolution, for example, corresponds to the number of illumination points required for detecting three-dimensional objects exceeding 15cm on a road RD at a location with a detection depth distance L1 of 100m. The second resolution, for example, corresponds to the number of illumination points required for detecting three-dimensional objects exceeding 100cm on a road RD at a location with a detection depth distance L1 of 100m.

[0093] <Mapping Department>

[0094] Mapping unit 114 displays a data mapping showing the location of the detected object (including the location of road markings) based on time-series point cloud data measured in real time by lidar 5, for example, with... Figure 4B On the same two-dimensional graph, a series of continuous position data are generated.

[0095] More specifically, the mapping unit 114 acquires the position information of all three-dimensional objects (excluding road markings) on the two-dimensional map stored in the storage unit 12, and performs coordinate transformation and calculation of the relative positions of the three-dimensional objects, etc., with the position of the vehicle 101 as the center, based on the moving speed and moving direction (e.g., azimuth angle) of the vehicle 101. Whenever the mapping unit 114 acquires point cloud data from the lidar 5 using the measurement described later (step S310), it performs coordinate transformation of the relative positions of the three-dimensional objects, etc., based on the acquired point cloud data with the position of the vehicle 101 as the center and records them on the two-dimensional map.

[0096] Furthermore, the mapping unit 114 acquires the position information of all road markings on the two-dimensional map stored in the storage unit 12, and performs coordinate transformation and calculation of the relative positions of the road markings with the position of the vehicle 101 as the center, based on the moving speed and direction of the vehicle 101. Whenever the mapping unit 114 acquires point cloud data using the measurement described later (step S310), it performs coordinate transformation of the relative positions of the road markings based on the acquired point cloud data with the position of the vehicle 101 as the center and records them on the two-dimensional map.

[0097] <Explanation of the flowchart>

[0098] Figure 6 It is shown Figure 2 A flowchart illustrating an example of processing performed by the arithmetic unit 11 of the controller 10 according to a predetermined program. Figure 6 The process shown in the flowchart is repeated, for example, in a predetermined cycle, during the operation of this vehicle 101 in autonomous driving mode.

[0099] First, in step S10, the arithmetic unit 11 sets the vertical illumination range of the laser radar 5's illumination light. More specifically, by changing the projection angle θ0, the long strip of illumination light in the road width direction illuminates a location at a predetermined depth distance L1 (e.g., 100m) on the road RD, thereby adjusting the vertical illumination range of the illumination light, and proceeding to step S20.

[0100] By adjusting the illumination range of the irradiation light in the vertical direction in step S10, even when the depth of the irradiation light varies due to the height H and slope of the road surface RD, the illumination range of the irradiation light can be corrected in the vertical direction before the measurement in step S30 described later. (See below for further details.) Figure 7 The processing of step S10 is explained in detail.

[0101] In step S20, the arithmetic unit 11 sets the horizontal (road width direction) illumination range of the laser radar 5. More specifically, the length W1 of the strip-shaped illumination range in the road width direction is adjusted to be longer than the road width at a location on the road RD at a predetermined depth distance L1 (e.g., 100m), so that the illumination light illuminates the left and right ends of the road, and then proceeds to step S30.

[0102] By adjusting the illumination range of the illumination light in the road width direction in step S20, even if the left and right ends (boundary lines RL and RB) of the road RD are outside the illumination range of the illumination light due to the curvature of the road RD, the illumination range of the illumination light can be corrected in the road width direction before the measurement in step S30 described later. (See below for further details.) Figure 8 The process of step S20 is explained in detail.

[0103] In step S30, the arithmetic unit 11 outputs an illumination instruction to the lidar 5 via a CAN communication line, etc., and determines road surface information showing the road markings based on the point cloud data obtained by the lidar 5 according to the illumination instruction. More specifically, it detects white lines (including lines of different colors such as yellow), curb lines, road studs, etc., which are predetermined detection objects, and proceeds to step S40. (See below for further details.) Figure 9 The processing of step S30 is explained in detail.

[0104] In step S40, the arithmetic unit 11 maps the relative position of the road markings obtained based on the point cloud data determined in step S30 onto a two-dimensional map, generates a series of continuous two-dimensional position data, and proceeds to step S50.

[0105] More specifically, whenever point cloud data is acquired by the lidar 5 through the measurement (step S310) described later, the mapping unit 114 of the calculation unit 11 performs coordinate transformation on the relative position of the road markings obtained based on the acquired point cloud data, with the position of the vehicle 101 as the center, and records it on a two-dimensional map.

[0106] In step S50, the arithmetic unit 11 determines whether to end the process. If the vehicle 101 continues to drive in autonomous driving mode, the arithmetic unit 11 determines that step S50 is negative (S50: No), returns to step S10, and repeats the above process. By returning to step S10, the measurement of three-dimensional objects based on point cloud data is periodically and repeatedly performed during the driving of the vehicle 101. On the other hand, if the vehicle 101 ends its autonomous driving mode, the arithmetic unit 11 determines that step S50 is positive (S50: Yes), and ends the process. Figure 6 The processing.

[0107] Figure 7 This is a detailed explanation of step S10 performed by the arithmetic unit 11. Figure 6 The flowchart for processing ).

[0108] In step S110, the arithmetic unit 11 uses the detection data detected by the lidar 5 to obtain three-dimensional point cloud data and proceeds to step S120. The measurement of the lidar 5 performed in step S110 to obtain the point cloud data can also be referred to as the first prediction determination. The first prediction determination is used to determine whether the illumination light of the lidar 5 illuminates the detection area BX1 (corresponding to the first area) on the road RD at a depth distance L1 (e.g., 100m).

[0109] In step S120, the arithmetic unit 11 performs separation processing and proceeds to step S130. More specifically, by using known methods, data representing three-dimensional objects on the road RD are detected and separated based on the point cloud data generated in step S110, resulting in point cloud data representing a flat road surface. Three-dimensional objects include, in addition to those located at the left and right ends of the road RD such as curbs, walls, ditches, and guardrails, other vehicles such as moving two-wheeled vehicles.

[0110] In step S130, the calculation unit 11 determines whether road surface data exists in the point cloud data obtained in step S120. For example, the calculation unit 11 calculates a histogram of height data based on the point cloud data, and determines that road surface data exists if the total number of points on a mountain with a peak is above a predetermined threshold. Furthermore, the calculation unit 11 obtains the road surface height H based on the location of the mountain with the peak. Other methods can also be used to determine the presence or absence of road surface data and to obtain the road surface height H.

[0111] If the calculation unit 11 has road surface data (in other words, flat data with a value higher than the noise level in the road width direction), it determines step S130 as affirmative (S130: Yes) and proceeds to step S140. If the calculation unit 11 does not have road surface data (in other words, only noise level data exists), it determines step S130 as negative (S130: No) and proceeds to step S170.

[0112] The situation in step S170 is that the illumination light does not illuminate the road RD due to the aforementioned downhill slope, etc. In step S170, the arithmetic unit 11 sends an instruction to the lidar 5 to increase the projection angle θ0 compared to the current angle, so that the illumination light illuminates the road RD, and then returns to step S110, and repeats the above process.

[0113] The situation in step S140 is that the illumination light is shining on the road RD, so the lidar 5 receives the scattered light. In step S140, the calculation unit 11 calculates the road surface height H based on the point cloud data obtained in step S120, as described above, and proceeds to step S150.

[0114] In step S150, the calculation unit 11 substitutes the road surface height H calculated in step S140 and the projection angle θ0 set in the lidar 5 into the above formula (1) to calculate the depth distance, and proceeds to step S160.

[0115] In step S160, the arithmetic unit 11 adjusts the illumination range of the laser radar 5 in the vertical direction (also known as the depth direction), and then ends the process. Figure 7 The processing. More specifically, regarding setting the illumination range for the next illumination of the lidar 5, by changing the projection angle θ0, the detection area BX1 (corresponding to the first area) at the location where the illumination light is illuminating at a depth distance L1 (e.g., 100m) is set.

[0116] Figure 8 This is a detailed explanation of step S20 performed by the arithmetic unit 11. Figure 6 The flowchart for processing ).

[0117] In step S210, the arithmetic unit 11 uses the detection data detected by the lidar 5 to obtain three-dimensional point cloud data and proceeds to step S220. The measurement performed by the lidar 5 in step S210 to obtain the point cloud data can also be called the second prediction. The second prediction is a measurement used to determine whether the road end on the road RD at a location with a depth distance L1 (e.g., 100m) is included in the illumination range of the lidar 5.

[0118] In step S220, the arithmetic unit 11 performs separation processing and proceeds to step S230. More specifically, by using known methods, data on three-dimensional objects such as those on the road RD are detected based on the point cloud data generated in step S210, and then separated from the road surface data to obtain point cloud data representing the three-dimensional objects. In addition to the aforementioned curb stones, walls, ditches, guardrails, etc., the three-dimensional objects also include other vehicles such as moving two-wheeled vehicles.

[0119] In step S230, the calculation unit 11 determines whether the separated point cloud data contains data representing three-dimensional objects, etc. If data representing three-dimensional objects, etc., is present, the calculation unit 11 determines step S230 to be affirmative (S230: Yes) and proceeds to step S240. If no data representing three-dimensional objects, etc., is present (in other words, only flat road surface data exists in the road width direction intersecting the direction of travel), the calculation unit 11 determines step S230 to be negative (S230: No) and proceeds to step S250.

[0120] In step S240, the computation unit 11 extracts data representing the road end based on point cloud data representing three-dimensional objects, and proceeds to step S250. For example, data representing curbs, walls, ditches, guardrails, etc., which are approximately parallel to the direction of travel of the vehicle 101, are extracted. On the other hand, data representing three-dimensional objects, etc., that intersect the direction of travel in the road width direction are not extracted.

[0121] In step S250, the arithmetic unit 11 adjusts the illumination range of the laser radar 5 in the horizontal direction (road width direction), and then ends the process. Figure 8 The processing involves setting the illumination range for the next illumination of the lidar 5. Specifically, the length W1 of the detection area BX1 (corresponding to the first area) in the road width direction is set to be longer than the road width of the road RD at the depth distance L1. Thus, the left and right ends (boundary lines RL and RB) of the road RD at the depth distance L1 are included in the next illumination range.

[0122] For example, if data representing one of the left and right ends is not extracted in step S240, the position of the end of the detection area BX1 (corresponding to the first area) in the road width direction is set to be increased by a predetermined amount based on the positions of the two ends and the length W1 of the two ends in the road width direction of the detection area BX1 for the next irradiation.

[0123] Furthermore, if step S230 is determined to be negative (S230: No) because there is no data representing three-dimensional objects (for example, there is only flat road surface data in the road width direction intersecting the direction of travel, and there is no data representing the left and right ends of the road RD), in step S250, the calculation unit 11 sets the length W1 of the detection area BX1 for the next irradiation to be expanded from the two ends of the currently set detection area BX1 (corresponding to the first area) to the left and right sides by a predetermined magnification, based on the positions of the two ends and the length W1 of the currently set detection area BX1 in the road width direction.

[0124] It should be noted that, in order to ensure temporal continuity, the length W1 of the detection area BX1 (corresponding to the first area) in the road width direction set in step S250 can also be filtered by state space model, Kalman filter, etc. The position of the detection area BX1 in the road width direction (the position of the left end and the right end) can also be set according to the position information located by the positioning unit 2 and the high-precision map information stored in the storage unit 12.

[0125] Figure 9 This is a detailed explanation of step S30 performed by the arithmetic unit 11. Figure 6 The flowchart for processing ).

[0126] In step S310, the arithmetic unit 11 uses the detection data detected by the lidar 5 to obtain three-dimensional point cloud data and proceeds to step S320. The measurement performed by the lidar 5 in step S310 to obtain the point cloud data can be referred to as this measurement. This measurement is used to simultaneously detect road markings and specified three-dimensional objects that are predetermined detection targets using the lidar 5.

[0127] In step S320, the arithmetic unit 11 performs separation processing and proceeds to step S330. More specifically, by using known methods, data on three-dimensional objects such as those on the road RD are detected and separated from the point cloud data generated in step S310, thereby leaving point cloud data representing a flat road surface. Three-dimensional objects include, for example, curbs, walls, ditches, guardrails, etc., exceeding 15cm, as well as other vehicles such as moving two-wheeled vehicles.

[0128] In step S330, the calculation unit 11 obtains the road surface height distribution based on the point cloud data separated in step S320, and proceeds to step S340. For example, the calculation unit 11 calculates a histogram of height data based on the point cloud data, and calculates the average number of points for all points on the mountain with the peak. Alternatively, it calculates the value of the peak's location (the value of the height data). This calculated value represents the road surface height H of the entire road surface, including the road markings. Other methods can also be used to obtain the road surface height distribution.

[0129] In step S340, the calculation unit 11 performs brightness filtering on the point cloud data separated in step S320, removing data with brightness below a predetermined threshold. This processing is based on the idea that the brightness of parts of the road surface that do not correspond to road markings falls below the aforementioned threshold. The aforementioned threshold can also be a fixed value. Alternatively, a threshold that varies depending on the depth distance of the detection area BX can be set; more specifically, a threshold that becomes smaller as the depth distance increases. This is because both the reflection intensity of the road surface and the reflection intensity of the white lines tend to decrease with increasing distance, so by making the threshold variable according to distance, the separation accuracy between the road surface and the white lines can be further improved. When the calculation unit 11 retains road surface data exceeding the aforementioned threshold (in other words, there is a possibility of road markings), it proceeds to step S350.

[0130] In step S350, the calculation unit 11 obtains the height distribution of the road surface based on the point cloud data left in step S340, and proceeds to step S360. For example, the calculation unit 11 calculates a histogram of height data based on the point cloud data, and calculates the average value of all points on the mountain with the peak. Alternatively, it calculates the value of the peak's location (the value of the height data).

[0131] In step S360, the arithmetic unit 11 compares the height of the road surface (the entire road surface) calculated in step S330 with the height of the road surface (the road surface whose brightness exceeds the threshold) calculated in step S350, and proceeds to step S370.

[0132] In step S370, the calculation unit 11 determines whether there are road markings based on the comparison result of step S360. For example, if the absolute value of the difference between the average value of the data of mountains on the entire road surface and the average value of the data of mountains corresponding to the portion with brightness exceeding the threshold is below a predetermined determination threshold, that is, if the height of the portion with brightness exceeding the threshold (first portion) is approximately the same as the height of the portion with brightness below the threshold (second portion), then step S370 is determined to be affirmative (S370: Yes), the processing in Figure 9 ends, and the process proceeds to step S40 in Figure 6. At this time, the calculation unit 11 (recognition unit 111) detects the portion with brightness exceeding the threshold as a road marking and recognizes the detected position as information indicating a road marking.

[0133] If the absolute value of the difference between the average value of the mountain data across the entire road surface and the average value of the mountain data corresponding to the portion of brightness exceeding the threshold is greater than a predetermined judgment threshold, the calculation unit 11 determines step S370 as negative (S370: No) and returns to step S310. A negative determination in step S370 does not mean that the point cloud data of the high-brightness portion left in step S340 is data from the road surface (e.g., the high-brightness portion exists on a three-dimensional object). Furthermore, as an example of other judgment thresholds, a positive determination can be made if the absolute value of the difference between the peak position value of the height data across the entire road surface and the peak position value of the height data corresponding to the portion of brightness exceeding the threshold is less than a predetermined judgment threshold, and a negative determination is made if the absolute value is greater than the judgment threshold.

[0134] The implementation method described above achieves the following effects.

[0135] (1) The external identification device 50, which serves as a road marking recognition device, includes: a lidar 5, which acts as a vehicle-mounted detector, irradiating the area around the vehicle 101 with electromagnetic light and detecting the surrounding external conditions in a time sequence; and a calculation unit 11, which acts as a road information acquisition unit, acquiring road information of the road RD on which the vehicle 101 is traveling based on the information detected by the lidar 5. The calculation unit 11 includes: an identification unit 111, which identifies information representing road markings in a first area (corresponding to detection area BX1) on the road RD at different positions on the road RD at a predetermined distance from the vehicle 101; and a mapping unit 114, which maps the information representing road markings identified by the identification unit 111 based on the travel direction and speed of the vehicle 101.

[0136] Because the calculation unit 11 is configured to acquire road surface information for a first area on the road RD, the number of illumination points of the LiDAR 5's illumination light can be reduced compared to acquiring road surface information for the entire road surface area of ​​the road RD at once. Furthermore, the mapping unit 114 maps the information identified by the recognition unit 111 constituting the calculation unit 11 based on the vehicle 101's direction of travel and speed, thus allowing the vehicle 101's movement to be considered as actual scanning travel. Therefore, even if the first area does not cover the entire road surface area in the direction of travel of the road RD, road surface information for the entire road surface area of ​​the road RD can still be collected.

[0137] As described above, the number of illumination points of the lidar 5 can be reduced without decreasing the recognition accuracy of the position and width of the road markings detected by the external recognition device 50, which is a road marking recognition device. The reduction in the number of illumination points also reduces the processing load of the computing unit 11.

[0138] In situations where the shooting environment of camera 4 is deteriorated due to factors such as backlighting or weather, it is extremely beneficial to accurately identify road markings of road RD based on point cloud data obtained by lidar 5.

[0139] (2) In the road marking recognition device (external recognition device 50) described in (1) above, the recognition unit 111 also recognizes information indicating predetermined detection objects at different positions on the road RD, namely, the second area (corresponding to the detection area BY1) on the road RD that is closer to the vehicle 101 than the first area (corresponding to the detection area BX1) and the third area (corresponding to the detection area BZ1) on the road RD that is farther away from the vehicle 101 than the first area.

[0140] Because of this configuration, compared to the case where the arithmetic unit 11 only obtains road surface information in the first region (corresponding to the detection region BX1) of the road RD, the range of road surface information obtained in the direction of travel of the road RD can be expanded. Even in this case, compared to the case where road surface information of the entire road surface area of ​​the road RD is obtained at once, the number of illumination points of the LiDAR 5 illumination light can be reduced.

[0141] For example, in a scenario where an object suddenly lands on the path of the vehicle 101, deviating from the detection area BX1, the object can be identified if it lands in the detection area BY1 or BZ1.

[0142] (3) In the road marking recognition device (external recognition device 50) of (2) above, the recognition unit 111 recognizes the information representing the detection object in the first region (corresponding to the detection region BX1) with a first resolution, and recognizes the information representing the detection object in the second and third regions (corresponding to the detection regions BY1 and BZ1) with a second resolution lower than the first resolution.

[0143] Because of this configuration, it is possible to ensure high recognition accuracy in, for example, the first region, while reducing recognition accuracy in the second and third regions compared to the first region, thereby appropriately reducing the number of illumination points of the LiDAR 5 illumination light.

[0144] (4) In the road marking recognition device (external recognition device 50) of (1) above, the calculation unit 11 further includes a delineation unit 112. The delineation unit 112 calculates a predetermined distance based on the speed of the vehicle 101 and delineates a long strip-shaped area on the road RD that intersects with the road RD at a predetermined distance from the vehicle 101 in the direction of travel and has a depth and width based on the vehicle speed in the road width direction as the first area.

[0145] Because of this configuration, the delineation unit 112 can delineate the first region (corresponding to the detection region BX1) by illuminating the left and right ends of the road RD with illumination light. When the lidar 5 intermittently irradiates the detection region BX1 corresponding to the first region from the vehicle 101 moving on the road RD, the delineation unit 112 can also delineate the first region in a way that the road surface information obtained by the calculation unit 11 based on the previous illumination light and the road surface information obtained by the calculation unit 11 based on the current illumination light are continuously connected in the direction of travel.

[0146] (5) In the road marking recognition device (external recognition device 50) of (4) above, the calculation unit 11 further includes a setting unit 113, which sets the illumination direction (projection angle θ0) of the laser radar 5 in such a way as to detect the illumination light scattered by the road surface in at least the first region (corresponding to the detection region BX1).

[0147] Because of this configuration, the illumination range of the illumination light can be corrected when the depth of the illumination light changes due to factors such as the height H and slope of the road surface of road RD. Furthermore, the illumination range of the illumination light can also be corrected when the left and right ends (boundary lines RL and RB) of road RD are removed from the illumination range due to curvature of road RD.

[0148] (6) In the road marking recognition device (external recognition device 50) described in (5) above, the recognition unit 111 also recognizes the road surface data of the travel direction of the vehicle 101 based on the information detected in the prediction setting before the measurement of the lidar 5, and the setting unit 113 sets the vertical direction and the irradiation range in the road width direction for irradiating the first area on the road RD with irradiation light based on the specified distance and the road surface data.

[0149] Because of this configuration, the irradiation range of the irradiated light can be corrected in the vertical direction and the road width direction before this measurement.

[0150] (7) In the road marking recognition device (external recognition device 50) of (6) above, the setting unit 113 establishes a correspondence between the illumination position of the laser radar 5 and the position of each grid point obtained by dividing the first region (corresponding detection region BX1) in the road width direction based on the angular resolution as the first resolution.

[0151] Because of this configuration, the number of illumination points of the LiDAR 5 in the road width direction within the detection area BX1 can be appropriately reduced.

[0152] (8) In the road marking recognition device (external recognition device 50) of (7) above, the setting unit 113 establishes a correspondence between the illumination position of the laser radar 5 and the position of each grid point obtained by further dividing the first region (corresponding detection region BX1) in the vertical direction based on the angular resolution as the first resolution.

[0153] Because of this configuration, the number of illumination points of the LiDAR 5 in the vertical direction within the detection area BX1 can be appropriately reduced.

[0154] (9) In the road marking recognition device (external recognition device 50) of (8) above, the setting unit 113 establishes a correspondence between the illumination position of the laser radar 5 and the position of each grid point obtained by dividing the second and third regions (corresponding detection regions BY1 and BZ1) in the road width direction based on the angular resolution as the second resolution.

[0155] Because of this configuration, the number of illumination points of the LiDAR 5 in the road width direction within the detection areas BY1 and BZ1 can be appropriately reduced.

[0156] (10) In the road marking recognition device (external recognition device 50) of (9) above, the setting unit 113 establishes a correspondence between the illumination position of the laser radar 5 and the position of each grid point obtained by further dividing the second and third regions (corresponding detection regions BY1 and BZ1) in the vertical direction based on the angular resolution as the second resolution.

[0157] Because of this configuration, the number of illumination points of the LiDAR 5 in the vertical direction within the detection areas BY1 and BZ1 can be appropriately reduced.

[0158] The above-described embodiments can be modified in various ways. The following describes some of these modifications.

[0159] (Variation Example 1)

[0160] The mapping unit 114 can also map the data representing the height H of the road surface detected by the point cloud data obtained from different locations on the road RD based on the lidar 5 onto a one-dimensional map to generate a series of continuous road slope data.

[0161] More specifically, the mapping unit 114 performs coordinate transformation and calculation based on the relative position of the location at depth distance L1 on the road RD at each measurement time, with the location of the vehicle 101 as the center, according to the vehicle 101's moving speed and moving direction (e.g., azimuth angle). Whenever point cloud data is acquired from the lidar 5 through this measurement (step S310), the mapping unit 114 records the road surface height H at the location at depth distance L1 based on the acquired point cloud data on the aforementioned one-dimensional map. The information from the one-dimensional map can also be stored in the storage unit 12.

[0162] (Variation Example 2)

[0163] The road surface farther from vehicle 101 reflects the illumination light from lidar 5 less strongly, sometimes failing to detect sufficiently strong reflected light. The depth distance at which the reflection intensity of this road surface is reduced to the limit level that allows for detection is called the maximum road surface detection distance L′.

[0164] In Modification 2, when the depth distance L up to the detection area BX calculated based on the vehicle speed 101 exceeds the maximum road surface detection distance L′ (for example, when the depth distance is L = 150m and the maximum road surface detection distance is L′ = 110m), the height of the location at the specified depth distance L is estimated based on the height at the location of the maximum road surface detection distance L′ obtained (calculated) based on information detected by the lidar 5 and the road slope, and the detection area BX at that time is set based on the estimated height. Additionally, the detection area BX′ at the location of the maximum road surface detection distance L′ is also set simultaneously.

[0165] It should be noted that the depth and width of the detection area BX′ can also be expanded, and the detection area BX′ and the detection area BX can be connected into a single detection area.

[0166] Using Modification Example 2, even when it is difficult to detect the road surface itself due to its distance, it is possible to detect road surface irregularities, three-dimensional objects, etc., where the level of reflected light is higher than that of the road surface reflected light.

[0167] The above description is essentially an example, and the above embodiments and modifications do not limit the invention as long as they do not destroy its features. One or more of the above embodiments and modifications can be combined in any way, and modifications can also be combined with each other.

[0168] Using this invention, the processing load can be reduced while accurately identifying road markings.

[0169] The present invention has been described above in conjunction with preferred embodiments, but those skilled in the art should understand that various modifications and alterations can be made without departing from the scope of the following claims.

Claims

1. A road marking recognition device, characterized in that, An external identification device (50) is provided for detecting the surrounding external conditions in a time sequence by irradiating electromagnetic waves around the vehicle and for obtaining road surface information information of the road on which the vehicle is traveling based on the information detected by the vehicle detector (5). The road information acquisition unit (11) has: The identification unit (111) identifies information about road markings representing a first area on the road at a predetermined distance from the vehicle at different locations on the road; and The mapping unit (114) maps the information representing the road markings identified by the identification unit according to the vehicle's direction of travel and speed. The delineation unit (112) calculates the predetermined distance based on the vehicle speed and delineates a long strip-shaped area that intersects the road at a distance from the vehicle in the direction of travel and has a depth and width based on the vehicle speed in the road width direction as the first area.

2. The road marking recognition device according to claim 1, characterized in that, The identification unit (111) also identifies information representing predetermined detection objects in a second area on the road that is closer to the vehicle than the first area and a third area on the road that is farther from the vehicle than the first area, respectively, at different locations on the road.

3. The road marking recognition device according to claim 2, characterized in that, The identification unit (111) identifies information representing the road markings in the first region at a first resolution, and identifies information representing the detected object in the second region and the third region at a second resolution lower than the first resolution.

4. The road marking recognition device according to claim 1, characterized in that, The road information acquisition unit (11) further includes a setting unit (113), which sets the irradiation direction of the electromagnetic wave for the vehicle detector (5) to detect electromagnetic waves scattered by the road surface at least in the first area.

5. The road marking recognition device according to claim 4, characterized in that, The identification unit (111) also identifies road surface data in the direction of travel of the vehicle based on information detected in the prediction before the measurement by the on-board detector. The setting unit (113) sets the irradiation range in the vertical direction and the road width direction for irradiating electromagnetic waves onto the first area on the road, based on the specified distance and the road surface data.

6. The road marking recognition device according to claim 3, characterized in that, The setting unit (113) establishes a correspondence between the illumination position of the vehicle-mounted detector (5) and the position of each grid point obtained by dividing the first region in the road width direction based on the first resolution.

7. The road marking recognition device according to claim 6, characterized in that, The setting unit (113) establishes a correspondence between the illumination position of the vehicle-mounted detector (5) and the position of each grid point obtained by further dividing the first region in the vertical direction based on the first resolution.

8. The road marking recognition device according to claim 7, characterized in that, The setting unit (113) establishes a correspondence between the illumination position of the vehicle-mounted detector (5) and the position of each grid point obtained by dividing the second region and the third region in the road width direction based on the second resolution.

9. The road marking recognition device according to claim 8, characterized in that, The setting unit (113) establishes a correspondence between the illumination position of the vehicle-mounted detector (5) and the position of each grid point obtained by further dividing the second region and the third region in the vertical direction based on the second resolution.

10. The road marking recognition device according to claim 1, characterized in that, The identification unit (111) calculates the height difference between the first part of the road surface in the first area whose brightness exceeds the specified brightness and the second part whose brightness is below the specified brightness, based on the information detected by the vehicle detector (5). When the absolute value of the calculated difference is below the judgment threshold, the first part is detected as the road marking, and the detection position of the road marking is identified as information representing the road marking.

11. The road marking recognition device according to any one of claims 1 to 10, characterized in that, The vehicle-mounted detector is a lidar.

Images