WO2021104180A1 - 地图生成方法、定位方法、装置、设备、存储介质及计算机程序 - Google Patents

地图生成方法、定位方法、装置、设备、存储介质及计算机程序 Download PDF

Info

Publication number
WO2021104180A1
WO2021104180A1 PCT/CN2020/130627 CN2020130627W WO2021104180A1 WO 2021104180 A1 WO2021104180 A1 WO 2021104180A1 CN 2020130627 W CN2020130627 W CN 2020130627W WO 2021104180 A1 WO2021104180 A1 WO 2021104180A1
Authority
WO
WIPO (PCT)
Prior art keywords
road
traffic light
road element
image
map
Prior art date
Application number
PCT/CN2020/130627
Other languages
English (en)
French (fr)
Inventor
付万增
王哲
石建萍
Original Assignee
上海商汤临港智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 上海商汤临港智能科技有限公司 filed Critical 上海商汤临港智能科技有限公司
Priority to JP2022506803A priority Critical patent/JP2022542712A/ja
Priority to KR1020227000811A priority patent/KR20220018594A/ko
Publication of WO2021104180A1 publication Critical patent/WO2021104180A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
    • G01C21/30—Map- or contour-matching
    • G01C21/32—Structuring or formatting of map data
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38—Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3804—Creation or updating of map data
    • G01C21/3807—Creation or updating of map data characterised by the type of data
    • G01C21/3815—Road data
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38—Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3804—Creation or updating of map data
    • G01C21/3833—Creation or updating of map data characterised by the source of data
    • G01C21/3841—Data obtained from two or more sources, e.g. probe vehicles
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38—Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3863—Structures of map data
    • G01C21/3867—Geometry of map features, e.g. shape points, polygons or for simplified maps
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38—Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3863—Structures of map data
    • G01C21/387—Organisation of map data, e.g. version management or database structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/05—Geographic models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle

Definitions

  • the embodiment of the present invention relates to a map generation method, positioning method, device, equipment, storage medium and computer program.
  • Maps are the core technology of autonomous driving systems. Compared with user-oriented static electronic maps, high-precision maps not only have higher position accuracy, but also contain more road and traffic elements, and even include real-time dynamic traffic and road condition information. For example, high-precision maps are one of the foundations for autonomous vehicles to perform automatic positioning and driving.
  • HD Map High Definition Map
  • the embodiments of the present invention provide a map generation method, positioning method, device, equipment, storage medium, and computer program.
  • an embodiment of the present invention provides a map generation method, including: acquiring position information of at least one road element on a road collected by a sensor, and each road element corresponds to an identifier; and for each of the at least one road element The road element generates a point set corresponding to the road element according to the location information of the road element, wherein the point set includes a plurality of location points representing the location of the road element; and the identification of each road element on the road is combined with The corresponding point set is stored in association with the road identifier to generate a map.
  • an embodiment of the present invention provides a positioning method, including: obtaining coordinates of a location point corresponding to at least one road element within a first preset range around the vehicle from a map, wherein each road element corresponds to a point
  • the point set includes a plurality of location points that characterize the location of the road element, and the map stores the coordinates of the location points included in the point set corresponding to the road element; obtains the image collected by the vehicle;
  • the matching result between the at least one road element and the image within the first preset range determines the position of the vehicle.
  • an embodiment of the present invention provides a map generation device, including: an acquisition module for acquiring position information of at least one road element on the road collected by a sensor, and each road element corresponds to an identifier; a generation module for Each road element in the at least one road element generates a point set corresponding to the road element according to the location information of the road element, where the point set includes a plurality of location points that characterize the location of the road element; storing; The module is used for associating and storing the identification of each road element on the road and the corresponding point set with the identification of the road to generate a map.
  • an embodiment of the present invention provides a positioning device, including: an acquisition module, configured to acquire from a map the coordinates of a location point corresponding to at least one road element within a first preset range around the vehicle, where each road The element corresponds to a point set, the point set includes a plurality of location points that represent the location of the road element, and the map stores the coordinates of the location points included in the point set corresponding to the road element; the acquisition module, It is also used to obtain an image collected by the vehicle; a positioning module is used to determine the location of the vehicle according to a matching result between the at least one road element and the image within the first preset range.
  • an embodiment of the present invention provides a positioning device, including: at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the At least one processor executes the map generation method described in the above first aspect and various possible implementation manners of the first aspect.
  • an embodiment of the present invention provides a positioning device, including: at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the At least one processor executes the positioning method described in the above second aspect and various possible implementation manners of the second aspect.
  • an embodiment of the present invention provides a computer-readable storage medium having computer-executable instructions stored in the computer-readable storage medium.
  • the processor executes the computer-executable instructions, the first aspect and the first aspect described above are implemented.
  • an embodiment of the present invention provides a computer program, including computer-readable code, which when the computer-readable code runs on a device, causes a processor in the device to execute the first aspect and the first aspect described above.
  • the map generation method described in various possible implementation manners, or the positioning method described in the above second aspect and various possible implementation manners of the second aspect are executed.
  • the map generation method, positioning method, device, equipment, storage medium, and computer program provided in this embodiment obtain the position information of at least one road element on the road collected by the sensor, and each road element corresponds to an identifier; for at least one road element For each road element in, generate a point set corresponding to the road element according to the location information of the road element, where the point set contains multiple location points that characterize the location of the road element; the identification of each road element on the road and the corresponding
  • the point set is stored in association with the road sign to generate a map, so that the road elements can be stored in the form of a point set and the map can be generated according to the point set corresponding to each road element, avoiding storage in the map in the form of complex curve equations Road elements, thereby reducing the amount of data for map analysis and increasing the speed of analysis.
  • FIG. 1 is a schematic flowchart of a map generation method provided by an embodiment of the present invention
  • FIG. 2 is a schematic diagram of a reference line of a road in a map provided by an embodiment of the present invention
  • Fig. 3 is a schematic diagram of an intersection in a map provided by an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of a road connection relationship provided by an embodiment of the present invention.
  • FIG. 5 is a schematic flowchart of a positioning method provided by an embodiment of the present invention.
  • FIG. 6 is a schematic flowchart of a positioning method provided by another embodiment of the present invention.
  • FIG. 7 is a schematic flowchart of a vehicle navigation in a positioning method provided by another embodiment of the present invention.
  • FIG. 8 is a schematic diagram of a flow of traffic light recognition in a positioning method provided by still another embodiment of the present invention.
  • Figure 9 is a schematic diagram of a traffic light provided by an embodiment of the present invention.
  • FIG. 10 is a schematic structural diagram of a map generating apparatus provided by an embodiment of the present invention.
  • FIG. 11 is a schematic structural diagram of a positioning device provided by an embodiment of the present invention.
  • FIG. 12 is a schematic structural diagram of a positioning device provided by another embodiment of the present invention.
  • FIG. 13 is a schematic diagram of the hardware structure of a map generating device provided by an embodiment of the present invention.
  • FIG. 14 is a schematic diagram of the hardware structure of a positioning device provided by an embodiment of the present invention.
  • High-precision maps usually use the OpenDrive format, which is a set of open source format specifications describing road networks.
  • High-precision maps in OpenDrive format are mostly vectorized maps, in which road elements such as lane lines are fitted and stored by curve equations.
  • the autonomous vehicle extracts the curve equation from the high-precision map for analysis, and locates its location through the analyzed road data and the collected information of the surrounding environment.
  • the road element is stored in the form of a point set and the map is generated, avoiding storing the road element in the form of a complicated curve equation in the map, thereby reducing the amount of data for map analysis and improving Resolution speed.
  • the autonomous vehicle when the autonomous vehicle is positioning or navigating, it directly extracts the point set of the road elements from the map and combines the image for positioning, which can effectively reduce or even avoid the analysis process of the curve equation of the road elements, thereby reducing the amount of data required for calculation. Improve positioning speed.
  • FIG. 1 is a schematic flowchart of a map generation method provided by an embodiment of the present invention. As shown in Figure 1, the method includes the following steps.
  • the sensor may be an image sensor, a laser sensor, etc., which is not limited herein.
  • Road elements include at least one of the following: left/right boundary line of the road (left/right boundary line), drivable boundary line, dashed lane line, start line (start line), stop line (stop line), pedestrian crossing ( crosswalk), parking spaces (park), obstacles (obstacle), intersection boundaries, traffic lights (traffic light), street lights and traffic signs.
  • the location information of the road element may be the location coordinates of the point cloud data corresponding to the road element.
  • a data collection vehicle can be used to drive on a designated road.
  • the data collection vehicle is equipped with image sensors and laser sensors.
  • the image sensor collects the image containing the road elements on the road, and the three-dimensional point cloud data of the road elements on the road is scanned by laser sensors such as laser radar.
  • the map generating device obtains the data collected by the data collection vehicle, can identify the road elements on the road by performing target recognition on the image, and generate a corresponding mark for each road element. Then, the map generation device combines the positioning information of the data collection vehicle during the driving process, and the installation parameters of the image sensor and the laser sensor on the data collection vehicle to perform coordinate system conversion to determine the location information of the road elements on the road.
  • S102 Generate a point set (PointSet) corresponding to each road element according to the location information of each road element, where the point set corresponding to one road element includes multiple location points that characterize the location of the road element.
  • PointSet point set
  • the map generating device can generate the corresponding point set of each road element according to the location information of each road element.
  • the location information of a certain road element includes the three-dimensional point cloud data corresponding to the road element, and the map generating device can select or calculate multiple location points representing the location of the road element from the three-dimensional point cloud data of the road element, and then Generate the point set corresponding to the road element.
  • the location point that characterizes the location of a road element may be the location point of the edge contour of the road element, or may be a point on the smallest circumscribed polygon or polyhedron of the road element, which is not limited here.
  • the point set corresponding to the left/right boundary line is composed of multiple position points on the left/right boundary line.
  • the point set corresponding to the drivable boundary line is composed of multiple position points on the drivable boundary line.
  • the point set corresponding to the dashed lane line is composed of two end points on the dashed lane line.
  • the point set corresponding to the starting line is composed of two end points on the starting line.
  • the point set corresponding to the stop line is composed of two end points on the stop line.
  • the point set corresponding to the crosswalk is composed of multiple vertices of the polygon area to which the crosswalk belongs.
  • the point set corresponding to the parking space is composed of four vertices of the rectangular area to which the parking space belongs.
  • the point set corresponding to the obstacle is composed of multiple vertices of the polyhedron to which the obstacle belongs.
  • two or more points on the road element can be selected to form a point set
  • a point set can be selected from the plane area or space area where the road element is located.
  • the point set in addition to the above-mentioned method of generating the point set, there may also be other methods of generating the point set, which are not limited here.
  • two endpoints can be directly selected and added to the point set.
  • the density of the location points included in the corresponding point set is determined by its curvature.
  • the density of the point concentration location corresponding to the left/right boundary line is determined according to the curvature of the left/right boundary line.
  • the selected density and number of the point concentrated location points can be determined, so that as few location points as possible can be used to accurately characterize the road elements of different curvatures, so as to avoid the inability to accurately represent the road due to too few location points.
  • the position of the element is determined, so that as few location points as possible can be used to accurately characterize the road elements of different curvatures, so as to avoid the inability to accurately represent the road due to too few location points.
  • S103 Store the identifiers of the road elements on the road and the corresponding point sets in association with the identifiers of the roads to generate a map.
  • the map generating device will associate and store A with A1, A2, and A3 during the map generation process, and A1, A2, A3 are associated and stored with their corresponding point sets, so that when the autonomous vehicle locates on the map, the corresponding point set can be quickly queried from the map according to the road element identification.
  • the embodiment of the present invention can store road elements in the form of point sets and generate a map according to the point set corresponding to each road element, avoiding storing road elements in the form of complex curve equations in the map, thereby reducing the amount of map analysis data and improving Resolution speed.
  • the map generated in this embodiment is organized in an XML (eXtensible Markup Language, extensible markup language) format.
  • the map mainly uses point sets to describe various information.
  • the expression form of point sets can be the lines between adjacent points, which avoids a large number of vector calculations.
  • the components of the map generated in this embodiment may include, but are not limited to: Header, Road(s), and Junction(s).
  • Header is used to describe map information.
  • the Header may include but is not limited to at least one of the following information: version information, map name, map version number, generation date, coordinate system value range, map manufacturer information, and coordinate system conversion tool.
  • the header can also store an origin coordinate (for example, the center of the map), which is used to transform the coordinate system of the data in the map.
  • Road(s) is used to describe road information. Among them, it mainly includes descriptions of various lanes and trajectories. For example, through the description of the drivable trajectory, the intersection in the practical sense is simplified to form a collection of multiple roads. Road(s) is also used to describe road elements that rely on road information, such as stop lines, crosswalks, and traffic lights.
  • each road part Road in the map has one attribute information.
  • the attribute information may include but is not limited to at least one of the following information: name (name_), length (len_), identifier (id_), which intersection belongs to (junction_). Where id_ is the unique identifier of the road, name_ and len_ are optional.
  • the attribute information of each road part Road may include information indicating whether the road is a one-way street or a two-way street.
  • each road part Road in the map corresponds to a piece of link information
  • the link information of a road part Road records other roads that the road can reach and other roads that can enter the road.
  • Junction(s) is used to describe how multiple roads are connected, guide the trajectory of vehicles, and describe the boundary conditions of the entire intersection.
  • a road can be divided into multiple road sections.
  • a road is not static.
  • the increase or decrease of the number of lanes and the change of lane line attributes all indicate the complexity of the lane. Therefore, in the map of this embodiment, roads are segmented to enable complicated road information description. At the same time, the segmentation also distinguishes the drivable area from the road boundary, which can better describe some special situations, such as the existence of special areas or passages between the vehicle travel area and the road boundary.
  • the basis for dividing road segments may include at least one of the following: changes in the number of lanes, changes in lane attributes, stop lines, crosswalks, and traffic lights.
  • the road element includes a left/right boundary line; the above method may further include: determining a reference line of the road according to the position information of the left/right boundary line of the road And generate the identification of the reference line, wherein the reference line is located at the center of the road; a plurality of position points are selected from the reference line to form a point set corresponding to the reference line; The identification of the reference line and its corresponding point set are stored in the map in association with the identification of the road.
  • an additional virtual reference line is set for the road when the map is generated.
  • the reference line does not exist in the real road scene. It is created when the map generation device generates the map.
  • Each road corresponds to a reference line, the reference line is located at the center of the road, and each reference line corresponds to a point set.
  • the map generating device can determine the center position of the road as the position information of the reference line according to the position information of the left/right boundary line of the road, and generate the identification of the reference line, and then select multiple position points from the reference line to form the corresponding reference line Point set, the reference line identification and its corresponding point set are associated with the road identification and stored in the map.
  • Fig. 2 is a schematic diagram of a reference line provided by an embodiment of the present invention.
  • the dash-dotted line in the figure is the reference line of the road, and the forward direction of the reference line can be specified.
  • the lanes to the left of the reference line along the S direction can be numbered 1, 2 ,3, the lanes on the right side of the reference line can be numbered -1, -2, -3 sequentially, and the lane numbers are stored in the map.
  • the road includes at least one lane, each lane corresponds to an identifier, and the road element includes the lane line of each lane; the above method may further include: for each lane, according to the The position information of the lane line of the lane determines the position information of the center line of the lane, where the center line of a lane is located at the center of the lane; for each lane, select multiple from the center line of the lane The position points constitute a point set corresponding to the center line of the lane; the identification of each lane on the road and the point set corresponding to the center line are stored in the map in association with the identification of the road.
  • the road may include one or more lanes.
  • One lane includes two lane lines.
  • the two lane lines serve as the boundary of the lane, and the area between the two lane lines is the area where the vehicle drives in the lane.
  • a virtual centerline is set for each lane when the map is generated.
  • the centerline does not exist in real lanes, and is created by the map generating device when generating the map.
  • Each lane corresponds to a centerline
  • the centerline is located at the center of the lane
  • each centerline corresponds to a point set.
  • the map generating device can determine the center position of the lane as the position information of the center line of the lane according to the position information of the lane line of a certain lane, and then select multiple position points from the center line to form its corresponding point set, and divide the lane
  • the mark of and the point set corresponding to its centerline are stored in the map in association with the mark of the road. In this way, the autonomous vehicle can obtain the point set corresponding to the center line of the current lane from the map during the driving process, and then drive in the lane along the center line, which can keep the vehicle in the center of the lane and avoid accidents.
  • turning information may be stored for each lane in the map, and the turning information may be used to indicate whether the current lane can turn left, turn right, turn around, and other information.
  • the method further includes: obtaining the speed limit value corresponding to each lane; and storing the speed limit value corresponding to each lane and the identification of each lane in the map in association with each other.
  • the traffic rules stipulate the speed limit value of some lanes.
  • the speed limit value of each lane can be associated with the identification of the lane and stored in the map.
  • the automatic driving vehicle can obtain the speed limit value of the current driving lane or the speed limit value of the designated lane from the map during the driving process, and then adjust the speed according to the speed limit value .
  • the map further includes at least one intersection, and each intersection corresponds to an identifier
  • the above method may further include: for each intersection, acquiring information about multiple roads associated with the intersection, And according to the information of the multiple roads, every two roads in the multiple roads that have a connection relationship are determined, wherein the connection relationship includes the starting road and the target road of the connection relationship; and the connection relationship and the intersection are determined.
  • the identifier of is stored in the map in association with each other.
  • multiple roads associated with a certain intersection are all connected to other roads through the intersection.
  • the information of the road may be the identification and location of the road.
  • the map generating device When the map generating device generates a map, it can determine every two roads that have a connection relationship based on the information of multiple roads. Among them, the existence of a connection relationship between two roads means that the vehicle can drive from one of the two roads to the other road at the intersection.
  • the map generating device may store the connection relationship corresponding to the intersection and the identification of the intersection in the map in association with each other.
  • intersection S connects four roads, namely A, B, C, and D. Among them, when the vehicle travels along road A to intersection S, it will turn right to road B, go straight to road C, and turn left to road D. Then the connection relationship of the intersection S may include but is not limited to A->B, A->C, A->D, and the specific data storage form is not limited here.
  • connection relationship of intersection S and intersection S can be stored in the map in association with each other.
  • the autonomous driving vehicle can obtain the connection relationship corresponding to the intersection from the map during the driving process, and then control the driving of the vehicle according to the driving route and the corresponding connection relationship.
  • the junction of the intersection in the map may include but is not limited to at least one of the following information: name (name_), identifier (id_). Where id_ is the unique identifier of Junction, and name_ is optional.
  • Junction is used when a road can be connected to multiple roads. The typical use scenario is to describe the connection information (Connection) between the roads in the intersection and the shape of the intersection (Boundary, Hole).
  • Figure 3 shows a schematic diagram of an intersection in the map provided in this embodiment. Among them, Junction Boundary indicates the boundary of the intersection, and Junction Hole indicates obstacles such as flower beds in the intersection.
  • the connection relationship between each road is represented by a Connection.
  • FIG. 4 is a schematic diagram of the road connection relationship provided by an embodiment of the present invention.
  • Figures 3 and 4 show the visualization effects of some road elements, such as connecting the points of the road elements and neighboring points to form a continuous road line and then displaying them after visualization.
  • the roads at both ends of the intersection in the map are connected by an actual road segment (B->A, A->B, A->C in Figure 4), then Connection stores the starting segment of the connection relationship
  • Connection can also store the target road outgoingRoad; here outgoingRoad can be stored explicitly, or it can be determined by the road attribute of connectingRoad.
  • Connection will store incomingRoad and outgoingRoad, and the road level between incomingRoad and outgoingRoad Connect, and mark Connection as virtual at the same time.
  • the above method may further include: matching the connection relationship with the traffic light information of the intersection; and storing the traffic light and its matched connection relationship in the map in association.
  • connection relationship between the traffic lights at the intersection and the matching connection relationship can also be stored in the map when the map is generated.
  • an intersection may include one or more traffic lights, each traffic light may correspond to one or more connection relationships, and one connection relationship may correspond to one or more traffic lights, which is not limited here.
  • the map generating device may match the connection relationship with the traffic light information of the intersection according to the traffic rules or the traffic light indication rules, and then store the traffic lights and the matched connection relationship in the map in association.
  • the autonomous vehicle can determine the current traffic light state of the intersection through image recognition when driving to the intersection, and then perform corresponding driving control.
  • the traffic light includes at least one signal light.
  • the signal lamp in the map stores a signal lamp index correspondingly.
  • each Connection is used to describe a road-to-road one-way connection relationship. If there is a signal light that controls the driving of the vehicle on this connection relationship, a signal light index is extended and stored in the Connection for index control. The passage of the road turns.
  • the signal light index includes an index attribute, the content of which specifies the only traffic light stored above.
  • the signal light index includes a straightLightForWaitingZone attribute, which serves the special scenario of a left-turn waiting zone. If the Connection where the signal light is located indexes a left-turn waiting zone, and according to traffic rules, the left-turning vehicle needs to pay attention to the straight-going signal light and the left-turn signal light at the same time, then the Connection needs to store the signal light index that controls the left turn of the Connection, and the straightLightForWaitingZone value is false; At the same time, it is necessary to store the index of the signal light that controls straight travel, and the value of straightLightForWaitingZone is true.
  • FIG. 5 is a schematic flowchart of a positioning method provided by an embodiment of the present invention.
  • the positioning method is based on the map generated according to the map generation method of FIG. 1 for positioning. As shown in Figure 5, the method includes the following steps.
  • S501 Obtain from a map the coordinates of a location point corresponding to at least one road element within a first preset range around the vehicle, where each road element corresponds to a point set, and the point set includes a plurality of points that characterize the location of the road element.
  • the location point of the location, and the coordinates of the location point included in the point set corresponding to the road element are stored in the map.
  • the vehicle may be an automatic driving vehicle or a vehicle driven by a user, which is not limited here.
  • the map is a high-precision map used for positioning and navigation of autonomous vehicles or other vehicles.
  • the map provided by the embodiment of the present invention stores road elements in the form of point sets, each road element corresponds to a point set, and the map stores the coordinates of the location points included in the point set corresponding to the road element.
  • the coordinates of each location point are the three-dimensional coordinates of the location point in the global coordinates.
  • the at least one road element may include but is not limited to at least one of the following: left/right boundary line of the road, drivable boundary line, dashed lane line, start line, stop line, crosswalk, parking space, obstacle , Intersection boundaries, traffic lights, street lights and traffic signs.
  • the location point contained in the point set corresponding to the road element may be the location point of the road element itself, or the location point in the area where the road element is located, which is not limited here.
  • the left/right boundary line and the drivable boundary line each correspond to a point set composed of multiple location points.
  • the dashed lane line, the start line, and the stop line all correspond to a point set composed of two position points.
  • the crosswalk corresponds to a point set composed of multiple vertices of the polygon area to which the crosswalk belongs.
  • the parking space corresponds to a point set consisting of four vertices of the rectangular area to which the parking space belongs.
  • the obstacle corresponds to a point set composed of multiple vertices of the polyhedron to which the obstacle belongs.
  • the two end points of the line segments can be used as two position points to form a point set.
  • each vertex of the circumscribed polygonal area where the crosswalk is located can be used as each position point to form a point set.
  • the circumscribed polygon includes the pedestrian crossing within its range.
  • the polygon can be a rectangle, a hexagon, etc., which is not limited here.
  • each vertex of the circumscribed polyhedron where the obstacle is located can be used as each position point to form a point set.
  • the circumscribed polyhedron contains obstacles within its range.
  • it may be the smallest circumscribed cuboid, octahedron, etc. in the three-dimensional space where the obstacle is located, which is not limited here. It is easy to imagine that for a road element, other position points that can characterize the position, size, shape, etc. of the road element can also be selected to form a point set, which is not limited here.
  • the first preset range can be set according to actual needs, which is not limited here. For example, a circular area with a radius of 20 meters around the vehicle can be selected as the first preset range.
  • the point set corresponding to all or part of the road elements in the first preset range around the vehicle can be obtained from the data stored in the map, that is, at least one road element in the first preset range refers to all or part of the road elements in the first preset range. Part of the road element.
  • S501 may include: obtaining the original position of the vehicle using a first positioning method, where the positioning accuracy of the first positioning method is lower than a preset threshold; according to the original position, from the map Find the coordinates of the location point corresponding to at least one road element in the first preset range around the vehicle.
  • a low-precision positioning method may be used to locate the approximate range where the vehicle is located, and then the positioning method provided in this embodiment may be used to further locate the precise position of the vehicle.
  • the first positioning method is a low-precision positioning method.
  • the first positioning method may be a positioning method based on the Global Positioning System (Global Positioning System, GPS), communication base station, etc., which is not limited herein.
  • the preset threshold may be set according to actual requirements, and is used to characterize that the positioning accuracy of the first positioning method is lower than the positioning accuracy of positioning according to the map and the image in this embodiment.
  • the original location of the vehicle is located through the first positioning method, and then the location point coordinates corresponding to all or part of the road elements within the first preset range around the original location of the vehicle are searched from the map according to the original location.
  • S502 Acquire an image collected by the vehicle.
  • the image around the vehicle can be acquired through the image sensor.
  • the image sensor installed on the vehicle can be used to collect images around the vehicle.
  • S503 Determine the position of the vehicle according to a matching result between at least one road element within the first preset range and the image.
  • the above road elements can be projected onto the image respectively, and the second road contained in the image can be detected.
  • the first road element projected on the image is matched with the second road element on the detected image, and the location of the vehicle when the image is collected can be determined according to the matching result as the location of the vehicle.
  • a map can be used to store each road element in the form of a point set, where each road element corresponds to a point set, and the point set contains multiple location points that represent the location of the road element.
  • the map stores The coordinates of the location point contained in the point set corresponding to the road element.
  • the above method may further include: displaying the location of the vehicle.
  • the location of the vehicle can be displayed on the on-board terminal of the vehicle for the user to view, or the location of the vehicle can be sent to the user terminal such as the user’s mobile phone so that the user can view the vehicle’s location on the user terminal. position.
  • the located vehicle position it is also possible to use the located vehicle position to perform navigation and route planning for the vehicle, and to control the vehicle to travel along a preset driving trajectory, etc., which is not limited here.
  • FIG. 6 is a schematic flowchart of a positioning method provided by another embodiment of the present invention. This embodiment describes in detail the specific implementation process of matching at least one road element within the first preset range with the image. As shown in Figure 6, the method includes:
  • S601 is similar to S501 in the embodiment of FIG. 5, and will not be repeated here.
  • S602 is similar to S502 in the embodiment of FIG. 5, and will not be repeated here.
  • At least one road element within the first preset range is projected onto the image. It is possible that some road elements are located on the image after being projected, and other road elements are located outside the image after being projected, and will be in the first place here.
  • the road element within the preset range and located on the image after projection is called the first road element. There may be one or more first road elements, which are not limited here. Subsequently, the first road element is further matched with the image.
  • the coordinates of the location point corresponding to the at least one road element are coordinates in a global coordinate system.
  • S603 may include: according to a first coordinate system conversion matrix, transforming the coordinates of a position point corresponding to at least one road element within the first preset range from the global coordinate system to the pixel coordinate system of the image to obtain The at least one first road element is projected at a first position on the image, wherein the first coordinate system conversion matrix is a conversion matrix between the global coordinate system and the pixel coordinate system of the image.
  • the coordinates of the location point corresponding to each road element stored in the map are the coordinates in the global coordinate system
  • the coordinates of the pixels on the image are the coordinates in the image coordinate system.
  • the coordinates of the location point corresponding to each road element in the first preset range can be converted from the global coordinate system to the pixel coordinate system of the image through the first coordinate system conversion matrix, and the position of the road element on the image can be obtained. This is called the first position.
  • the internal parameters and external parameters of the camera on the vehicle can be calibrated in advance to obtain the first coordinate system conversion matrix.
  • the position point coordinates in the global coordinate system of the map can be sequentially converted to the vehicle positioning inertial navigation coordinate system, camera coordinate system, and image coordinate system, and finally each road element is projected to the image collected by the camera on.
  • the image can be detected through an image detection model or a traditional image processing method, and the position of each road element contained in the image can be detected.
  • each road element contained in the image is called a second road element
  • the position of the second road element on the image is called a second position. For example, if it is detected that an image contains two lane boundary lines and a traffic light, and there are three road elements in total, then these three road elements are all second road elements, and each second road element corresponds to one on the image Location, that is, the second location.
  • the image can be input to a detection and segmentation model based on deep learning, and the second road element in the image can be detected through the model, and the position of the second road element in the image can be located.
  • S605. Determine the position of the vehicle according to the matching result between the first position of the at least one first road element projected on the image and the second position of the at least one second road element on the image .
  • the first road element with the same type and the closest position is selected Match with the second road element, optimize the position of the vehicle according to the matching result, and finally determine the position of the vehicle.
  • the at least one road element includes multiple types.
  • S605 may include: taking a first road element and a second road element of the same type and closest to each other as a matching pair, possibly forming one or more matching pairs; for each matching pair, calculating the first in the matching pair The distance between the first position of the road element projected on the image and the second position of the second road element on the image is taken as the distance of the matching pair; the distance of each matching pair is taken as the projection error, and the minimum is calculated The position of the vehicle under the condition of transformed projection error is taken as the position of the vehicle.
  • a first road element and a second road element of the same type and closest to each other may be used as a matching pair, and one or more matching pairs may be formed.
  • For each matching pair calculate the Euclidean distance between the first position of the first road element projected on the image and the second position of the second road element on the image in the matching pair, and the distance is taken as The distance of the matched pair. Taking the distance of the matching pair as the projection error, and minimizing the projection error as the optimization goal, the position of the vehicle when the projection error is minimized is calculated as the location of the vehicle.
  • the calculated vehicle position is combined with the information of the first road element in the map and the information of the second road element in the image, and the positioning accuracy is high.
  • the first position obtained by the projection is matched with the second position of the detected second road element on the image, which can quickly and accurately Determine the location of the vehicle and improve the speed and accuracy of vehicle positioning.
  • FIG. 7 is a schematic flowchart of a vehicle navigation in a positioning method provided by another embodiment of the present invention.
  • at least one road element includes a road line.
  • the next driving point of the vehicle can be predicted based on the point set corresponding to the road line.
  • the method may further include:
  • the autonomous driving vehicle travels along a pre-designated or automatically planned travel path during the driving process.
  • the vehicle reaches a certain position, the coordinates of the next driving point of the vehicle are obtained.
  • the distance between the next driving point and the current position of the vehicle can be set according to actual needs, and is not limited here. For example, the next location point on the travel path that is 5 meters away from the current location may be used as the next travel point.
  • the position points corresponding to these road lines can be used as vertices to form a polygonal area, and the next driving point is located in the polygonal area.
  • the drivable area is an area where vehicles can pass; the non-driving area is an area where vehicles cannot pass. If a vehicle drives in this area, it may be dangerous or violate traffic rules. It can be determined whether the polygonal area is a drivable area or a non-drivable area according to the information of each area stored in the map or the related information of the road line. If the polygonal area is a drivable area, the vehicle is controlled to travel to the next driving point. If the polygonal area is a non-driving area, the next driving point of the vehicle is re-planned so that the next driving point of the vehicle is outside the polygonal area.
  • the coordinates of the location points corresponding to the road line in the area where the next driving point is located are obtained from the map, and these location points are formed into a polygonal area, and the next driving point is re-planned for the vehicle in time when the polygonal area is a non-driving area . Since the polygonal area formed by the location points can quickly determine whether the next driving point is in the non-driving area, sufficient time is reserved for replanning the next driving point, which can prevent the vehicle from entering the non-driving area and improve vehicle safety.
  • the map also stores the coordinates of the location points corresponding to at least one driving path, where each driving path corresponds to a point set, and the point set corresponding to each driving path is determined by multiple locations on the driving path.
  • Point composition
  • the travel path may be an empirical path obtained based on historical travel data of the vehicle or other vehicles, or may be a path planned based on the current traffic state, which is not limited here.
  • One or more driving routes can be saved in the map.
  • Each travel path can correspond to a point set composed of multiple location points on the travel path. For example, you can take a point every 5 meters, 10 meters, etc. on the travel path as a location point to form a point set.
  • the above method may further include: acquiring the coordinates of the location point corresponding to the first travel path from the map; and controlling the vehicle along the first travel path according to the coordinates of the location point corresponding to the first travel path. Drive on the driving path.
  • the first travel path is the path that the vehicle currently travels in the travel path stored in the map.
  • the coordinates of each location point corresponding to the first travel path can be obtained from the map, and then the vehicle is controlled to travel according to each location point corresponding to the first travel path.
  • the vehicle navigation can directly obtain the stored point set, and directly control the vehicle according to the stored point set, which can quickly realize the navigation of the vehicle.
  • the at least one road element includes at least one traffic light, and road turn information corresponding to the traffic light is also stored in the map.
  • traffic lights are used to indicate the turning of the vehicle at the intersection.
  • the road turn information corresponding to each traffic light is stored in the map.
  • the traffic light includes at least one signal light
  • each traffic light in the map corresponds to a point set consisting of four vertices of the rectangular area to which the traffic light belongs, and a set of road turning information, wherein each traffic light
  • the road turning information corresponding to the light includes at least one of the shape, color, and position of each signal light in the traffic light, and the indicated road turning information.
  • a traffic light may include three signal lights, namely a red light, a yellow light, and a green light.
  • the road turning information may include information such as going straight, turning left, and stopping, which is not limited here.
  • FIG. 8 is a schematic diagram of a flow of traffic light recognition in a positioning method provided by still another embodiment of the present invention.
  • the point set of the traffic light and the road turning information stored in the map can be used to control the vehicle.
  • the method includes:
  • S801 When it is detected that the image contains the first traffic light, obtain the current signal of the first traffic light and the third position of the first traffic light on the image through image recognition.
  • the traffic light included in the detected image is referred to as the first traffic light, and its position on the image is referred to as the third position.
  • Each first traffic light corresponds to a variety of signals, respectively indicating a variety of different road turns.
  • At least one second traffic light can be projected onto the image according to the coordinates of the location point corresponding to the at least one second traffic light within the second preset range, and then the projection result and the detected image The third position of the first traffic light is matched to obtain the target second traffic light that matches the first traffic light in the image.
  • S803 may include: projecting at least one second traffic light onto the image according to the second coordinate system conversion matrix and the coordinates of the location point corresponding to the at least one second traffic light to obtain at least one second traffic light.
  • the fourth position of the lamp projected on the image according to the third position of the first traffic light on the image and the fourth position of the at least one second traffic light projected on the image, the fourth position projected on the image is compared with The second traffic light with the closest position of the first traffic light is determined to be the target second traffic light matching the first traffic light.
  • the coordinates of the location point corresponding to the at least one second traffic light stored in the map are coordinates in the global coordinate system
  • the coordinates of the pixels on the image are coordinates in the image coordinate system.
  • the coordinates of the position points corresponding to the second traffic lights in the second preset range can be converted from the global coordinate system to the pixel coordinate system of the image through the second coordinate system conversion matrix, so as to realize that at least A second traffic light is projected onto the image.
  • At least one second traffic light within the second preset range is projected onto the image. Some of the second traffic lights may be located on the image after being projected, and other second traffic lights are located outside the image after being projected. For the second traffic light within the preset range and located on the image after projection, the position where it is projected onto the image is called the fourth position.
  • the fourth position projected on the image is compared with the first traffic light’s
  • the second traffic light closest to the three positions is determined to be the second traffic light matching the first traffic light, that is, the target second traffic light is determined.
  • S805 Determine the road turn indicated by the current signal of the first traffic light according to the current signal of the first traffic light and the road turning information corresponding to the target second traffic light.
  • the road turning information of the target second traffic light in the map is used as the road turning information of the first traffic light, combined with the current traffic light of the first traffic light. Signal, the direction of the road currently indicated by the first traffic light can be determined.
  • Figure 9 is a schematic diagram of a first traffic light provided by an embodiment of the present invention.
  • the road steering information corresponding to the first traffic light is as follows: when the red light is on, it indicates that the vehicle stops; when the green light is on, it indicates that the vehicle is going straight; when the left turn indicator is on, it indicates that the vehicle is turning left. If the current signal of the first traffic light is green, it is determined that the road turn indicated by the current signal of the first traffic light is straight.
  • the point set of the traffic light and the road turning information are stored in the map, and the current signal of the traffic light is detected in combination with the image, which can accurately identify the road turning indicated by the traffic light, and then accurately control the vehicle to perform corresponding turning.
  • high-precision maps use the OpenDrive format.
  • the OpenDrive format maps are too complex, and they focus on an overly ideal driving simulation system, and cannot flexibly and freely describe complex and changeable real road scenarios.
  • the OpenDrive format map is a vectorized map, which uses complex curve equations to fit road elements such as lane lines. The road elements in the map are stored in the form of curve equations. The storage structure is complicated, the analysis is difficult, and the use cost is high.
  • the OpenDrive format map only stores the location information of the traffic lights, and does not link the traffic light rules with the high-precision map, and cannot solve the decision-making and planning problems related to traffic lights in autonomous driving.
  • the embodiment of the present invention provides a lightweight high-precision map format serving the automatic driving system, and provides a positioning method based on the high-precision map format, which can support automatic driving positioning, decision-making planning control, traffic light understanding and Restore the static scene.
  • the high-precision map format stores the road elements of the road in the form of a point set and generates a map, avoiding storing road elements in the form of complex curve equations in the map, thereby reducing the amount of data for map analysis , Improve the resolution speed.
  • This high-precision map format improves the complex design of general vectorized maps, and uses a point set storage method that adapts to curvature, which simplifies map use while reducing map storage, transmission and analysis time.
  • the high-precision map format expands the storage method of traffic lights, stores the point set corresponding to the traffic lights and road turning information, supports the query and positioning of the traffic lights, and assists the automatic driving system to understand the traffic rules related to the traffic lights.
  • FIG. 10 is a schematic structural diagram of a map generating apparatus provided by an embodiment of the present invention.
  • the map generating device 100 includes: an acquisition module 1001, a generation module 1002, and a storage module 1003.
  • the acquiring module 1001 is configured to acquire the position information of at least one road element on the road collected by the sensor, and each road element corresponds to an identifier.
  • the generating module 1002 is configured to generate, for each road element in the at least one road element, a point set corresponding to the road element according to the position information of the road element, wherein the point set includes a plurality of points that characterize the location of the road element The location of the location point.
  • the storage module 1003 is used to store the identifiers of the road elements on the road and the corresponding point sets in association with the identifiers of the roads to generate a map.
  • the road element includes at least one of the following: a left/right boundary line of the road, a drivable boundary line, a dashed lane line, a start line, a stop line, a crosswalk, a parking space, an obstacle, an intersection boundary, Traffic lights, street lights and traffic signs.
  • the point set corresponding to the left/right boundary line is composed of multiple position points on the left/right boundary line.
  • the point set corresponding to the drivable boundary line is composed of multiple position points on the drivable boundary line.
  • the point set corresponding to the dashed lane line is composed of two end points on the dashed lane line.
  • the point set corresponding to the starting line is composed of two end points on the starting line.
  • the point set corresponding to the stop line is composed of two end points on the stop line.
  • the point set corresponding to the crosswalk is composed of multiple vertices of the polygon area to which the crosswalk belongs.
  • the point set corresponding to the parking space is composed of four vertices of the rectangular area to which the parking space belongs.
  • the point set corresponding to the obstacle is composed of multiple vertices of the polyhedron to which the obstacle belongs.
  • the density of the point concentration location corresponding to the road element of the curve type is determined according to its curvature.
  • the road element of the curve type may include the left/right boundary line or the drivable boundary line.
  • the density of the point concentration location corresponding to the left/right boundary line of the road is determined according to the curvature of the left/right boundary line of the road.
  • the road element includes a left/right boundary line
  • the storage module 1003 is further configured to: determine the position information of the reference line of the road according to the position information of the left/right boundary line of the road, and generate The identification of the reference line, wherein the reference line is located at the center of the road; a plurality of position points are selected from the reference line to form a point set corresponding to the reference line; the identification of the reference line and its The corresponding point set is stored in the map in association with the road identifier.
  • the road includes at least one lane, each lane corresponds to an identifier, and the road element includes the lane line of each lane; the storage module 1003 is also used to: for each lane, according to the lane The position information of the lane line determines the position information of the center line of the lane, where the center line is located at the center of the lane; for each lane, multiple position points are selected from the center line of the lane to form the center line of the lane Corresponding point set; the point set corresponding to the mark of each lane on the road and its centerline is stored in the map in association with the mark of the road.
  • the storage module 1003 is further configured to: obtain the speed limit value corresponding to each lane; and store the speed limit value corresponding to each lane and the identification of each lane in the map in association with each other.
  • the map further includes at least one intersection, and each intersection corresponds to an identifier;
  • the storage module 1003 is further configured to: for each intersection, obtain information about multiple roads associated with the intersection, and according to The information of the multiple roads determines every two roads in the multiple roads that have a connection relationship, wherein the connection relationship includes the starting road and the target road of the connection relationship; and the connection relationship and the identification of the intersection The association is stored in the map.
  • the storage module 1003 is further configured to: match the connection relationship with the traffic light information of the intersection; and store the traffic light and its matched connection relationship in the map in association with each other.
  • the map generating device provided in the embodiment of the present invention can be used to execute the above method embodiment, and its implementation principles and technical effects are similar, and details are not described in this embodiment here.
  • FIG. 11 is a schematic structural diagram of a positioning device provided by an embodiment of the present invention. As shown in FIG. 11, the positioning device 110 includes: an obtaining module 1101 and a positioning module 1102.
  • the obtaining module 1101 is configured to obtain the coordinates of a location point corresponding to at least one road element within a first preset range around the vehicle from the map, wherein each road element corresponds to a point set, and the point set includes multiple representations The location point where the road element is located, and the map stores the coordinates of the location point included in the point set corresponding to the road element.
  • the acquisition module 1101 is also used to acquire images collected by the vehicle.
  • the positioning module 1102 is configured to determine the position of the vehicle according to a matching result between at least one road element within the first preset range and the image.
  • the map stores each road element in the form of a point set.
  • Each road element corresponds to a point set.
  • the point set contains multiple location points that represent the location of the road element.
  • the map stores the coordinates of the location points included in the point set corresponding to the road element.
  • the acquiring module acquires the coordinates of the location point corresponding to at least one road element within the first preset range around the vehicle from the map through the acquiring module, and acquires the image collected by the vehicle, and then the positioning module corresponds to at least one road element within the first preset range Matching at least one road element within the first preset range with the image, and determining the position of the vehicle according to the matching result.
  • the point set of at least one road element can be directly extracted from the map and combined with the image for positioning, Avoid the analysis process of the curve equation of the road elements, thereby reducing the amount of data required for calculation and improving the positioning speed.
  • FIG. 12 is a schematic structural diagram of a positioning device provided by another embodiment of the present invention. As shown in FIG. 12, on the basis of the positioning device 110 provided in the embodiment shown in FIG. 11, the positioning device 120 provided in this embodiment may further include: a navigation module 1203, an identification module 1204 and a display module 1205.
  • the positioning module 1202 is specifically configured to: according to the coordinates of the location point corresponding to the at least one road element in the first preset range, direct at least one road element in the first preset range to Projecting on the image to obtain a first position where at least one first road element is projected on the image, wherein the first road element is within the first preset range and is located on the image after projection Image detection on the image to obtain the second position of at least one second road element on the image; according to the first position of the at least one first road element projected on the image and the The matching result between the second position of at least one second road element on the image determines the position of the vehicle.
  • the coordinates of the location point corresponding to the at least one road element are coordinates in a global coordinate system; the positioning module 1202 is specifically configured to: convert the coordinates within the first preset range according to a first coordinate system conversion matrix The coordinates of the location point corresponding to the at least one road element are converted from the global coordinate system to the pixel coordinate system of the image to obtain the first position of the at least one first road element projected on the image, wherein,
  • the first coordinate system conversion matrix is a conversion matrix between the global coordinate system and the pixel coordinate system of the image.
  • the at least one road element includes multiple types; the positioning module 1202 is specifically configured to: use a first road element and a second road element that are of the same type and closest to each other as a matching pair to obtain The at least one first road element and the at least one second road element determine at least one matching pair; for each matching pair, calculate the first position of the first road element projected on the image in the matching pair and the matching pair The distance between the second position of the second road element in the image in the image is used as the distance of the matching pair; the distance of each matching pair is used as the projection error, and the position of the vehicle under the condition of minimizing the projection error is calculated as the vehicle s position.
  • the acquisition module 1201 is specifically configured to: use a first positioning method to locate the original position of the vehicle, wherein the positioning accuracy of the first positioning method is lower than a preset threshold; according to the original position, Find the coordinates of the location point corresponding to at least one road element within the first preset range around the vehicle from the map.
  • the at least one road element includes a road line
  • the navigation module 1203 is configured to: when the vehicle is driving, obtain the coordinates of the next driving point of the vehicle, where the next driving point is The next location point of the current location on the travel path of the vehicle; obtaining the coordinates of the location point corresponding to the road line in the area where the next travel point is located from the map according to the coordinates of the next travel point; The position point corresponding to the road line in the area where the next travel point is located is used as the vertex to determine the polygon area corresponding to the next travel point; when the polygon area is a drivable area, the vehicle is controlled to move toward the Driving at the next driving point; when the polygonal area is a non-driving area, re-planning the next driving point of the vehicle so that the next driving point of the vehicle is outside the polygonal area.
  • the map also stores the coordinates of the location points corresponding to at least one driving path, where each driving path corresponds to a point set, and the point set corresponding to each driving path is determined by multiple locations on the driving path.
  • Point composition
  • the navigation module 1203 is further configured to: obtain the coordinates of the location point corresponding to the first travel path from the map; and control the vehicle to travel along the location point according to the coordinates of the location point corresponding to the first travel path.
  • the first driving path is used for driving.
  • the at least one road element includes at least one traffic light, and road turn information corresponding to the traffic light is also stored in the map.
  • the traffic light includes at least one signal light.
  • Each traffic light in the map corresponds to a point set composed of four vertices of the rectangular area to which the traffic light belongs, and a set of road turning information.
  • the road turning information corresponding to each traffic light includes at least one of the shape, color, and position of each signal light in the traffic light, and the indicated road turning information.
  • the recognition module 1204 is configured to: when detecting that the image contains the first traffic light, obtain the current signal of the first traffic light and the position of the first traffic light on the image through image recognition. A third location; obtain from the map the coordinates of the location point corresponding to at least one second traffic light within the second preset range around the vehicle; according to the coordinates of the location point corresponding to the at least one second traffic light and The third position of the first traffic light on the image is determined, and the target second traffic light that matches the first traffic light among the at least one second traffic light is determined; and the second traffic light is obtained from the map.
  • Road turn information corresponding to the target second traffic light determine the road turn indicated by the current signal of the first traffic light according to the current signal of the first traffic light and the road turn information corresponding to the target second traffic light ; According to the road turning indicated by the current signal of the first traffic light, the vehicle is controlled so that the vehicle drives in accordance with the road turning indicated by the current signal of the first traffic light.
  • the identification module 1204 is specifically configured to: direct the at least one second traffic light to the image according to the second coordinate system conversion matrix and the coordinates of the location point corresponding to the at least one second traffic light. Up projection to obtain the fourth position of at least one second traffic light projected on the image; according to the third position of the first traffic light on the image and the projection of the at least one second traffic light on the image For the fourth position on the image, the second traffic light whose fourth position projected on the image is closest to the position of the first traffic light is determined as the target second traffic light that matches the first traffic light.
  • the at least one road element includes at least one of the following: a left/right boundary line of the road, a drivable boundary line, a dashed lane line, a start line, a stop line, a crosswalk, a parking space, an obstacle, and an intersection Borders, traffic lights, street lights and traffic signs.
  • the left/right boundary line and the drivable boundary line all correspond to a point set consisting of multiple position points;
  • the dashed lane line, the start line, and the stop line all correspond to a set of points.
  • the crosswalk corresponds to a point set composed of multiple vertices of the polygon area to which the crosswalk belongs;
  • the parking space corresponds to four vertices in a rectangular area to which the parking space belongs A point set composed of;
  • the obstacle corresponds to a point set composed of multiple vertices of the polyhedron to which the obstacle belongs.
  • the display module 1205 is used to display the location of the vehicle after the location of the vehicle is determined.
  • the positioning device provided in the embodiment of the present invention may be used to execute the above method embodiment, and its implementation principles and technical effects are similar, and details are not described in this embodiment here.
  • FIG. 13 is a schematic diagram of the hardware structure of a map generating device provided by an embodiment of the present invention.
  • the map generating device 130 provided in this embodiment includes: at least one processor 1301 and a memory 1302.
  • the map generating device 130 also includes a communication component 1303. Among them, the processor 1301, the memory 1302, and the communication component 1303 are connected by a bus 1304.
  • At least one processor 1301 executes the computer executable instructions stored in the memory 1302, so that at least one processor 1301 executes the above map generation method.
  • the processor may be a central processing unit (Central Processing Unit, CPU), or other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), etc.
  • the general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in combination with the invention can be directly embodied as executed and completed by a hardware processor, or executed and completed by a combination of hardware and software modules in the processor.
  • the memory may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk memory.
  • NVM non-volatile storage
  • the bus can be an Industry Standard Architecture (ISA) bus, Peripheral Component Interconnect (PCI) bus, or Extended Industry Standard Architecture (EISA) bus, etc.
  • ISA Industry Standard Architecture
  • PCI Peripheral Component Interconnect
  • EISA Extended Industry Standard Architecture
  • the bus can be divided into address bus, data bus, control bus and so on.
  • the buses in the drawings of this application are not limited to only one bus or one type of bus.
  • FIG. 14 is a schematic diagram of the hardware structure of a positioning device provided by an embodiment of the present invention.
  • the positioning device 140 provided in this embodiment includes: at least one processor 1401 and a memory 1402.
  • the positioning device 140 further includes a communication component 1403. Among them, the processor 1401, the memory 1402, and the communication component 1403 are connected through a bus 1404.
  • At least one processor 1401 executes the computer executable instructions stored in the memory 1402, so that at least one processor 1401 executes the above positioning method.
  • the processor may be a central processing unit (English: Central Processing Unit, abbreviated as: CPU), or other general-purpose processors, digital signal processors (English: Digital Signal Processor, referred to as DSP), application specific integrated circuit (English: Application Specific Integrated Circuit, referred to as ASIC), etc.
  • the general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in combination with the invention can be directly embodied as executed and completed by a hardware processor, or executed and completed by a combination of hardware and software modules in the processor.
  • the memory may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk memory.
  • NVM non-volatile storage
  • the bus can be an Industry Standard Architecture (ISA) bus, Peripheral Component Interconnect (PCI) bus, or Extended Industry Standard Architecture (EISA) bus, etc.
  • ISA Industry Standard Architecture
  • PCI Peripheral Component Interconnect
  • EISA Extended Industry Standard Architecture
  • the bus can be divided into address bus, data bus, control bus and so on.
  • the buses in the drawings of this application are not limited to only one bus or one type of bus.
  • the present application also provides a computer-readable storage medium in which computer-executable instructions are stored.
  • the processor executes the computer-executable instructions, the above-mentioned map generation method or positioning method is realized.
  • the aforementioned computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM). ), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic Disk or Optical Disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EPROM Erasable Programmable Read-Only Memory
  • PROM Programmable Read-Only Memory
  • ROM Read-Only Memory
  • Magnetic Memory Flash Memory
  • Magnetic Disk Magnetic Disk or Optical Disk.
  • a readable storage medium may be any available medium that can be accessed by a general purpose or special purpose computer.
  • An exemplary readable storage medium is coupled to the processor, so that the processor can read information from the readable storage medium and can write information to the readable storage medium.
  • the readable storage medium may also be an integral part of the processor.
  • the processor and the readable storage medium may be located in the ASIC.
  • the processor and the readable storage medium may also exist in the device as discrete components.

Abstract

一种地图生成方法、定位方法、装置、设备、存储介质及计算机程序，地图生成方法包括：获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识（S101）；针对至少一个道路元素中的每个道路元素，根据道路元素的位置信息，生成道路元素对应的点集，其中，点集包含多个表征道路元素所在位置的位置点（S102）；将道路上各道路元素的标识及对应的点集，与道路的标识进行关联存储，生成地图（S103）。

Description

地图生成方法、定位方法、装置、设备、存储介质及计算机程序
交叉引用声明
本申请要求于2019年11月29日提交中国专利局的申请号为201911207451.1的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明实施例涉及一种地图生成方法、定位方法、装置、设备、存储介质及计算机程序。
背景技术
地图，尤其是高精度地图(High Definition Map，HD Map)，是自动驾驶系统的核心技术。相比面向用户的静态电子地图而言，高精度地图不仅拥有更高的位置精度，且包含更丰富的道路和交通元素，甚至包括实时动态的交通和路况信息等。例如，高精度地图是自动驾驶车辆能够进行自动定位和行驶的基础之一。
发明内容
本发明实施例提供一种地图生成方法、定位方法、装置、设备、存储介质及计算机程序。
第一方面，本发明实施例提供一种地图生成方法，包括：获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识；针对所述至少一个道路元素中的每个道路元素，根据该道路元素的位置信息，生成该道路元素对应的点集，其中，所述点集包含多个表征该道路元素所在位置的位置点；将所述道路上各道路元素的标识及对应的点集，与所述道路的标识进行关联存储，生成地图。
第二方面，本发明实施例提供一种定位方法，包括：从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中，每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标；获取所述车辆采集的图像；根据所述第一预设范围内的所述至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置。
第三方面，本发明实施例提供一种地图生成装置，包括：获取模块，用于获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识；生成模块，用于针对所述至少一个道路元素中的每个道路元素，根据该道路元素的位置信息，生成该道路元素对应的点集，其中，所述点集包含多个表征该道路元素所在位置的位置点；存储模块，用于将所述道路上各道路元素的标识及对应的点集，与所述道路的标识进行关联存储，生成地图。
第四方面，本发明实施例提供一种定位装置，包括：获取模块，用于从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中，每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标；所述获取模块，还用于获取所述车辆采集的图像；定位模块，用于根据所述第一预设范围内的所述至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置。
第五方面，本发明实施例提供一种定位设备，包括：至少一个处理器和存储器；所述存储器存储计算机执行指令；所述至少一个处理器执行所述存储器存储的计算机执行指令，使得所述至少一个处理器执行如上第一方面以及第一方面各种可能的实施方式所述的地图生成方法。
第六方面，本发明实施例提供一种定位设备，包括：至少一个处理器和存储器；所述存储器存储计算机执行指令；所述至少一个处理器执行所述存储器存储的计算机执行指令，使得所述至少一个处理器执行如上第二方面以及第二方面各种可能的实施方式所述的定位方法。
第七方面，本发明实施例提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机执行指令，当处理器执行所述计算机执行指令时，实现如上第一方面以及第一方面各种可能的实施方式所述的地图生成方法，或者实现如上第二方面以及第二方面各种可能的实施方式所述的定位方法。
第八方面，本发明实施例提供一种计算机程序，包括计算机可读代码，当所述计算机可读代码在设备上运行时，使得所述设备中的处理器执行如上第一方面以及第一方面各种可能的实施方式所述的地图生成方法，或者执行如上第二方面以及第二方面各种可能的实施方式所述的定位方法。
本实施例提供的地图生成方法、定位方法、装置、设备、存储介质及计算机程序，获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识；针对至少一个 道路元素中的每个道路元素，根据该道路元素的位置信息，生成该道路元素对应的点集，其中，点集包含多个表征该道路元素所在位置的位置点；将道路上各道路元素的标识及对应的点集，与道路的标识进行关联存储，生成地图，从而能够根据各道路元素对应的点集，以点集的形式存储道路元素并生成地图，避免在地图中以复杂的曲线方程的形式存储道路元素，从而降低地图解析的数据量，提高解析速度。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明一实施例提供的地图生成方法的流程示意图；
图2为本发明实施例提供的地图中道路的参考线的示意图；
图3为本发明实施例提供的地图中路口的示意图；
图4为本发明实施例提供的道路连接关系的示意图；
图5为本发明一实施例提供的定位方法的流程示意图；
图6为本发明又一实施例提供的定位方法的流程示意图；
图7为本发明另一实施例提供的定位方法中车辆导航的流程示意图；
图8为本发明再一实施例提供的定位方法中交通灯识别的流程示意图；
图9为本发明实施例提供的交通灯的示意图；
图10为本发明一实施例提供的地图生成装置的结构示意图；
图11为本发明一实施例提供的定位装置的结构示意图；
图12为本发明又一实施例提供的定位装置的结构示意图；
图13为本发明一实施例提供的地图生成设备的硬件结构示意图；
图14为本发明一实施例提供的定位设备的硬件结构示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
高精度地图通常采用OpenDrive格式，OpenDrive格式是一套描述道路网络的开源的格式规范。OpenDrive格式的高精度地图多为矢量化地图，其中通过曲线方程拟合并存储车道线等道路元素。自动驾驶车辆在进行定位时从高精度地图中提取曲线方程进行解析，通过解析出的道路数据与采集的周围环境的信息对自身所在位置进行定位。
由于矢量化地图使用了复杂的曲线方程存储车道线等道路元素，而对复杂的曲线方程进行解析的数据计算量大，解析所花费的时间长，导致自动驾驶车辆定位速度慢，即定位存在延迟，因此难以满足自动驾驶车辆实时定位的需求。
本发明实施例中根据各道路元素对应的点集，以点集的形式存储道路元素并生成地图，避免在地图中以复杂的曲线方程的形式存储道路元素，从而降低地图解析的数据量，提高解析速度。进一步的，自动驾驶车辆进行定位或导航时直接从地图中提取道路元素的点集结合图像进行定位，可有效减少甚至避免对道路元素的曲线方程的解析过程，从而减少所需计算的数据量，提高定位速度。
图1为本发明一实施例提供的地图生成方法的流程示意图。如图1所示，该方法包括以下步骤。
S101、获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识。
本实施例中，传感器可以为图像传感器、激光传感器等，在此不作限定。道路元素包括以下中的至少一项：道路的左/右边界线(left/right boundary line)、可行驶边界线、虚线车道线、起始线(start line)、停止线(stop line)、人行横道(crosswalk)、停车位(park)、障碍物(obstacle)、路口边界、交通灯(traffic light)、路灯和交通牌。道路元素的位置信息可以是道路元素对应的点云数据的位置坐标。
例如，可以通过数据采集车辆在指定的道路上行驶，数据采集车辆上安装有图像传感器和 激光传感器。通过图像传感器采集包含道路上道路元素的图像，通过激光雷达等激光传感器扫描道路上道路元素的三维点云数据。进一步的，地图生成设备获取数据采集车辆采集的数据，可以通过对图像进行目标识别，识别出道路上的道路元素，给每个道路元素生成相应的标识。然后，地图生成设备结合数据采集车辆行驶过程中的定位信息、以及图像传感器、激光传感器在数据采集车辆上的安装参数等进行坐标系转换，确定道路上的道路元素的位置信息。
S102、根据各道路元素的位置信息，生成各道路元素对应的点集(PointSet)，其中，一个道路元素对应的点集包含多个表征该道路元素所在位置的位置点。
本实施例中，地图生成设备可以根据每个道路元素的位置信息生成其对应的点集。道路元素与点集一一对应。例如，某道路元素的位置信息包括该道路元素对应的三维点云数据，地图生成设备可以从该道路元素的三维点云数据中选取或计算出多个表征该道路元素所在位置的位置点，进而生成该道路元素对应的点集。
本实施例中，表征一个道路元素所在位置的位置点可以是该道路元素边缘轮廓的位置点，也可以是该道路元素的最小外接多边形或多面体上的点，在此不作限定。
例如，所述左/右边界线对应的点集由所述左/右边界线上多个位置点组成。
例如，所述可行驶边界线对应的点集由所述可行驶边界线上多个位置点组成。
例如，所述虚线车道线对应的点集由所述虚线车道线上两个端点组成。
例如，所述起始线对应的点集由所述起始线上两个端点组成。
例如，所述停止线对应的点集由所述停止线上两个端点组成。
例如，所述人行横道对应的点集由所述人行横道所属多边形区域的多个顶点组成。
例如，所述停车位对应的点集由所述停车位所属矩形区域的四个顶点组成。
例如，所述障碍物对应的点集由所述障碍物所属多面体的多个顶点组成。
本实施例中，对于线型的道路元素可以选取道路元素上的两个及两个以上的点构成点集，对于平面或立体空间的道路元素可以选取道路元素所在平面区域或空间区域的点构成点集，除上述生成点集的方式外，还可以有其他的点集生成方式，在此不作限定。
本实施例中，对于直线型的道路元素，如虚线车道线、起始线和停止线等，可以直接选取两个端点添加到点集中。
本实施例中，对于曲线类型的道路元素，如左/右边界线、可行驶边界线等，其对应点集中包含的位置点的密度由其曲率确定。例如，所述左/右边界线对应的点集中位置点的密度根据所述左/右边界线的曲率确定。其中，点集中位置点的密度与曲率呈正相关关系。以左/右边界线为例，对于曲率较小的左/右边界线，以较大的密度选取该边界线上的位置点添加到点集中；对于曲率较大的左/右边界线，以较小的密度选取该边界线上的位置点添加到点集中。这样根据曲线类型的道路元素的曲率确定其点集中位置点的选取密度和数量，能够使用尽量少的位置点来对不同曲率的道路元素都进行准确表征，避免由于位置点过少无法准确表征道路元素位置的问题。
S103、将道路上各道路元素的标识及对应的点集，与所述道路的标识进行关联存储，生成地图。
例如，道路的标识为A，该道路上包含三个道路元素，标识分别为A1，A2，A3，则地图生成设备在生成地图过程中，将A分别与A1，A2，A3进行关联存储，并将A1，A2，A3与各自对应的点集进行关联存储，以便自动驾驶车辆通过地图进行定位时可根据道路元素的标识快速从地图中查询到相应的点集。
本发明实施例能够根据各道路元素对应的点集，以点集的形式存储道路元素并生成地图，避免在地图中以复杂的曲线方程的形式存储道路元素，从而降低地图解析的数据量，提高解析速度。
可选地，本实施例生成的地图使用XML(eXtensible Markup Language，可扩展标记语言)格式加以组织。地图主要使用点集来进行各种信息的描述，点集的表达形式可为相邻点之间的连线，这样就避免了大量的矢量计算。
可选地，本实施例生成的地图的构成部分可以包括但不限于：头部部分Header、道路部分Road(s)、路口部分Junction(s)。
其中，Header用于描述地图信息。可选地，Header可以包括但不限于以下信息中的至少一种：版本信息、地图名称，地图版本号、生成日期、坐标系取值范围、地图厂商信息以及坐标系转化工具。另外，Header中还可以存储一个原点坐标(例如地图中心)，用于对地图中的数据进行坐标系转换。
Road(s)用于描述道路信息。其中，主要包括各种车道(lane)及可行驶轨迹的描述。例如，通过可行驶轨迹的描述，将现实意义中的路口进行简化，形成多条道路的集合。Road(s)还用于描述依托于道路信息存在的道路元素，如停止线、人行横道、交通灯等。
可选地，地图中每个道路部分Road具有一个属性信息，属性信息可以包括但不限于以下信息中的至少一个：名称(name_)，长度(len_)，标识符(id_)，隶属于哪个路口(junction_)。其中id_是道路的唯一标识符，name_及len_是可选的。可选地，每个道路部分Road的属性信息可以包括指示该道路是单行道还是双行道的信息。
可选地，地图中每个道路部分Road对应于一个链接信息，一个道路部分Road的链接信息记录了该道路所能到达的其他道路，以及能够进入该道路的其他道路。
Junction(s)用于描述多条道路如何进行连接，指引车辆轨迹，同时描述了整个路口的边界情况。
可选地，一条道路可以划分为多个道路段(section)。一条道路并不是一成不变的，车道数的增减、车道线属性的变化都表明了车道的复杂性。故本实施例的地图中将道路分段，使其可以进行复杂的道路信息描述。同时分段也将可行驶区域与道路边界进行区分，这样可以更好地描述一些特殊情况，如车行驶区域与道路边界中间存在特殊区域或通道等。可选地，道路段的划分依据可以包括以下中的至少一项：车道数量变化、车道属性变化、停止线、人行横道和交通灯。
在一种可能的实施方式中，所述道路元素包括左/右边界线；上述方法还可以包括：根据所述道路的该左/右边界线的位置信息，确定所述道路的参考线(reference line)的位置信息，并生成所述参考线的标识，其中，所述参考线位于所述道路的中心；从所述参考线上选取多个位置点组成所述参考线对应的点集；将所述参考线的标识及其对应的点集，与所述道路的标识关联存储到所述地图中。
本实施例中，为便于自动驾驶车辆从地图中准确定位出自身在道路中的位置，在生成地图时额外为道路设置一个虚拟的参考线。该参考线在真实的道路场景中并不存在，是地图生成设备在生成地图时创建的，每条道路对应于一条参考线，参考线位于道路的中心，每条参考线对应于一个点集。
地图生成设备可以根据道路的左/右边界线的位置信息，确定出道路的中心位置作为参考线的位置信息，并生成参考线的标识，然后从参考线上选取多个位置点组成参考线对应的点集，将参考线的标识及其对应的点集，与道路的标识关联存储到地图中。
图2为本发明实施例提供的参考线的示意图。参照图2，图中的点划线为该道路的参考线，可以指定参考线的前进方向，如图中的S方向，沿S方向该参考线的左侧的车道可以依次编号为1,2,3，该参考线的右侧的车道可以依次编号为-1,-2,-3，将车道的编号存储到地图中。
在一种可能的实施方式中，所述道路上包括至少一个车道，每个车道对应于一个标识，所述道路元素包括各车道的车道线；上述方法还可以包括：针对每个车道，根据该车道的车道线的位置信息，确定该车道的中心线(center line)的位置信息，其中，一个车道的中心线位于该车道的中心；针对每个车道，从该车道的中心线上选取多个位置点组成该车道的中心线对应的点集；将道路上各车道的标识及其中心线对应的点集，与所述道路的标识关联存储到所述地图中。
本实施例中，道路可以包括一个或多个车道。一个车道包括两条车道线。两条车道线作为该车道的边界，两条车道线中间的区域为车辆在该车道行驶的区域。本实施例中为便于自动驾驶车辆能够在车道中央进行行驶，在生成地图时，为每个车道设置一个虚拟的中心线。该中心线在真实的车道中并不存在，是地图生成设备在生成地图时创建的，每个车道对应于一条中心线，中心线位于车道的中心，每条中心线对应于一个点集。
地图生成设备可以根据某车道的车道线的位置信息，确定出该车道的中心位置作为该车道的中心线的位置信息，然后从中心线上选取多个位置点组成其对应的点集，将车道的标识及其中心线对应的点集，与道路的标识关联存储到地图中。这样自动驾驶车辆在行驶过程中，可以从地图中获取当前所在车道的中心线对应的点集，然后沿该中心线在车道内行驶，可以保持车辆行驶在车道中心，避免事故发生。
可选地，地图中还可以为每个车道存储转弯信息(turn type)，转弯信息可以用于指示当前车道是否能够左转、右转、掉头等信息。
在一种可能的实施方式中，所述方法还包括：获取各车道对应的限速值；将各车道对应的 限速值与各车道的标识关联存储到所述地图中。
本实施例中，通常交通规则规定了一些车道的限速值，在地图生成时，可以将每个车道的限速值与该车道的标识关联存储到地图中。通过在地图中存储车道对应的限速值，能够使自动驾驶车辆在行驶过程中，可以从地图中获取当前行驶车道的限速值或者指定车道的限速值，进而根据限速值进行车速调整。
在一种可能的实施方式中，所述地图还包括至少一个路口，每个路口对应于一个标识，上述方法还可以包括：针对每个路口，获取与该路口相关联的多条道路的信息，并根据多条道路的信息确定所述多条道路中存在连接关系的每两条道路，其中，所述连接关系包括该连接关系的起始道路和目标道路；将所述连接关系与所述路口的标识关联存储到所述地图中。
本实施例中，与某路口相关联的多条道路均通过该路口与其他道路连接。道路的信息可以是道路的标识、位置等。地图生成设备在生成地图时，可以根据多条道路的信息确定出存在连接关系的每两条道路。其中，两条道路存在连接关系是指车辆可以在该路口从这两条道路中的一条道路行驶到另一条道路。地图生成设备可以将该路口对应的连接关系以及路口的标识关联存储到所述地图中。
以路口S为例，路口S连接四条道路，分别为A，B，C，D。其中，车辆沿道路A行驶到路口S时，右转则行驶到道路B，直行则行驶到道路C，左转则行驶到道路D。则路口S的连接关系可以包括但不限于A->B，A->C，A->D，具体的数据存储形式在此不作限定。
可以将路口S的连接关系与路口S关联存储到地图中。通过将连接关系与路口的标识关联存储到地图中，使自动驾驶车辆在行驶过程中可以从地图获取路口对应的连接关系，然后按照行驶路线和相应的连接关系进行车辆的行驶控制。
可选地，地图中路口部分Junction可以包括但不限于以下信息中的至少一个：名称(name_)，标识符(id_)。其中id_是Junction的唯一标识符，name_是可选的。Junction在一条道路可以连接到多条道路时使用，典型的使用场景是描述路口中各条道路之间的连接信息(Connection)，以及路口的形状(Boundary、Hole)。如图3所示为本实施例提供的地图中路口的示意图。其中，Junction Boundary表示该路口的边界，Junction Hole表示该路口中如花坛等障碍物。可选地，每一条道路间的连接关系由一个Connection表示。图4所示为本发明实施例提供的道路连接关系的示意图。图3和图4示出了部分道路元素的可视化效果，如将道路元素的点和相邻点连接形成连续的道路线等可视化处理后进行呈现。可选地，地图中路口两头的道路由一条实际存在的路段进行连接(如图4中的B->A，A->B，A->C)，则Connection存储该连接关系的起始段道路incomingRoad和所连接道路connectingRoad，以及从incomingRoad到connectingRoad的道路级别的连接。进一步的，Connection还可以存储目标段道路outgoingRoad；这里outgoingRoad可以显式存储，也可以由connectingRoad的道路属性来确定。如果地图中两个道路之间并不存在实际的道路连接(如图4中的B-C，C->A，C->B)，则Connection将存储incomingRoad和outgoingRoad，以及incomingRoad和outgoingRoad间道路级别的连接，同时将Connection标记为virtual。
在一种可能的实施方式中，上述方法还可以包括：匹配所述连接关系和所述路口的交通灯信息；将所述交通灯与其匹配的连接关系关联存储到所述地图中。
本实施例中，对于设置交通灯的路口，还可以在生成地图时将路口的交通灯与其匹配的连接关系存储到地图中。其中，一个路口可以包括一个或多个交通灯，每个交通灯可以对应一种或多种连接关系，一种连接关系可以对应一个或多个交通灯，在此不作限定。地图生成设备可以根据交通规则或者交通灯指示规则，匹配连接关系和路口的交通灯信息，然后将交通灯与其匹配的连接关系关联存储到地图中。
通过将交通灯与其匹配的连接关系关联存储到地图中，使得自动驾驶车辆在行驶到路口，可以通过图像识别确定路口当前的交通灯状态，然后进行相应的行驶控制。
可选地，交通灯包括至少一个信号灯。地图中信号灯对应存储一个信号灯索引。本实施例中，每个Connection用于描述一种道路到道路的单向连接关系，如果这个连接关系上存在一个控制车辆行驶的信号灯，则在Connection中扩展存储一个信号灯索引，用于索引控制该道路转向的通行。可选地，信号灯索引包括index属性，其内容指定了上文中存储的唯一的交通灯。
可选地，信号灯索引包括straightLightForWaitingZone属性，该属性服务于左转待转区这一特殊情景。如果信号灯所在Connection索引了一条左转待转区，并且根据交通规则，左转车辆需要同时关注直行信号灯和左转信号灯，则该Connection需要存储控制本Connection左转的信 号灯索引，straightLightForWaitingZone值为false；同时需要存储控制直行的信号灯索引，straightLightForWaitingZone值为true。
图5为本发明一实施例提供的定位方法的流程示意图。该定位方法基于根据图1的地图生成方法所生成的地图进行定位。如图5所示，该方法包括以下步骤。
S501、从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中，每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标。
在本实施例中，车辆可以为自动驾驶车辆，也可以是由用户驾驶的车辆，在此不作限定。地图为用于自动驾驶车辆或其他车辆进行定位和导航的高精度地图。本发明实施例提供的地图中以点集的形式存储道路元素，每个道路元素对应于一个点集，地图中存储有道路元素对应的点集所包含的位置点的坐标。其中，每个位置点的坐标为该位置点在全局坐标下的三维坐标。
可选地，至少一个道路元素可以包括但不限于以下中的至少一项：道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌。
其中，对于一个道路元素，该道路元素对应的点集所包含的位置点可以是该道路元素本身的位置点，也可以是该道路元素所在区域内的位置点，在此不作限定。
例如，所述左/右边界线、所述可行驶边界线均对应于一个由多个位置点组成的点集。
例如，所述虚线车道线、所述起始线、所述停止线均对应于一个由两个位置点组成的点集。
例如，所述人行横道对应于一个由所述人行横道所属多边形区域的多个顶点组成的点集。
例如，所述停车位对应于一个由所述停车位所属矩形区域的四个顶点组成的点集。
例如，所述障碍物对应于一个由所述障碍物所属多面体的多个顶点组成的点集。
在本实施例中，对于虚线车道线、起始线、停止线等由线段构成的道路元素，可以将线段的两个端点作为两个位置点，构成点集。对于人行横道，可以将人行横道所在的外接多边形区域的各个顶点作为各个位置点，构成点集。外接多边形将人行横道包含在其范围内，例如，多边形可以为矩形、六边形等，在此不作限定。对于花池、电线杆、护栏等障碍物，可以将障碍物所在的外接多面体的各个顶点作为各个位置点，构成点集。外接多面体将障碍物包含在其范围之内。例如，可以为障碍物所在立体空间中的最小外接长方体、八面体等，在此不作限定。容易想到的，对于一个道路元素，还可以选用其他能表征该道路元素的位置、大小、形状等的位置点构成点集，在此不作限定。
第一预设范围内可以根据实际需求进行设定，在此不作限定。例如，可以选取车辆周围半径为20米的圆形区域作为第一预设范围。可以从地图存储的数据中获取车辆周围第一预设范围内所有或部分的道路元素所对应的点集，即第一预设范围内的至少一个道路元素指第一预设范围内的所有或部分的道路元素。
可选地，S501可以包括：采用第一定位方式定位得到所述车辆的原始位置，其中，所述第一定位方式的定位精度低于预设阈值；根据所述原始位置，从所述地图中查找所述车辆周围所述第一预设范围内的至少一个道路元素对应的位置点的坐标。
在本实施例中，可以首先通过低精度的定位方式定位出车辆所在的大致范围，然后再采用本实施例提供的定位方法进一步定位出车辆的精准位置。其中，第一定位方式为低精度的定位方式，例如第一定位方式可以为基于全球定位系统(Global Positioning System，GPS)、通信基站等的定位方式，在此不作限定。预设阈值可以根据实际需求设定，用于表征第一定位方式的定位精度低于本实施例中按照地图和图像进行定位的定位精度。通过第一定位方式定位出车辆的原始位置，然后依据原始位置从地图查找车辆的原始位置周围第一预设范围内的所有或部分道路元素对应的位置点坐标。
S502、获取所述车辆采集的图像。
在本实施例中，可以通过图像传感器获取车辆周围的图像。具体而言，在车辆静止或行驶过程中，可以通过车辆上安装的图像传感器采集车辆周围的图像。
S503、根据所述第一预设范围内的至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置。
在本实施例中，可以根据第一预设范围内的至少一个道路元素对应的位置点的坐标以及坐标系转换矩阵，将上述道路元素分别向图像上投影，并且检测图像中包含的第二道路元素，将投影到图像上的第一道路元素与检测到的图像上的第二道路元素进行匹配，根据匹配结果可以 确定出车辆采集图像时所在的位置作为车辆的位置。
本发明实施例中，能够利用地图以点集形式存储每个道路元素，其中，每个道路元素对应于一个点集，点集包含多个表征该道路元素所在位置的位置点，地图中存储有该道路元素对应的点集所包含的位置点的坐标。通过从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，并获取车辆采集的图像，然后根据第一预设范围内的至少一个道路元素对应的位置点的坐标，将第一预设范围内的至少一个道路元素与图像进行匹配，并根据匹配结果确定车辆的位置，能够直接从地图中提取至少一个道路元素的点集结合图像进行定位，避免对道路元素的曲线方程的解析过程，从而减少所需计算的数据量，提高定位速度。
可选地，S503之后，上述方法还可以包括：显示所述车辆的位置。
在本实施例中，确定出车辆的位置之后，可以在车辆的车载终端上显示车辆的位置，以便用户查看，或者向用户的手机等用户终端发送车辆的位置，以便用户在用户终端查看车辆的位置。另外，还可以利用定位出的车辆位置对车辆进行导航和路径规划，控制车辆沿预设的行驶轨迹行驶等，在此不作限定。
图6为本发明又一实施例提供的定位方法的流程示意图。本实施例对第一预设范围内的至少一个道路元素与图像进行匹配的具体实现过程进行了详细说明。如图6所示，该方法包括：
S601、从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中，每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标。
在本实施例中，S601与图5实施例中的S501类似，此处不再赘述。
S602、获取所述车辆采集的图像。
在本实施例中，S602与图5实施例中的S502类似，此处不再赘述。
S603、根据所述第一预设范围内的至少一个道路元素对应的位置点的坐标，将所述第一预设范围内的至少一个道路元素分别向所述图像上投影，得到至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一道路元素为处于所述第一预设范围内且投影后位于所述图像上的道路元素。
在本实施例中，将第一预设范围内的至少一个道路元素向图像上进行投影，可能部分道路元素投影后位于图像上，其他道路元素投影后位于图像之外，在此将处于第一预设范围内的并且投影后位于图像上的道路元素称为第一道路元素。第一道路元素可能有一个或多个，在此不作限定。后续将第一道路元素与图像进一步进行匹配。
可选地，所述至少一个道路元素对应的位置点的坐标为全局坐标系下的坐标。S603可以包括：根据第一坐标系转换矩阵，将所述第一预设范围内的至少一个道路元素对应的位置点的坐标由所述全局坐标系转换到所述图像的像素坐标系下，得到所述至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一坐标系转换矩阵为所述全局坐标系与所述图像的像素坐标系之间的转换矩阵。
在本实施例中，地图中存储的每个道路元素对应的位置点的坐标为全局坐标系下的坐标，图像上像素的坐标为图像坐标系下的坐标。可以通过第一坐标系转换矩阵将第一预设范围内的每个道路元素对应的位置点的坐标由全局坐标系转换到图像的像素坐标系下，得到该道路元素在图像上的位置，在此称为第一位置。
例如，可以预先标定车辆上摄像头的内部参数和外部参数，得到第一坐标系转换矩阵。根据第一坐标系转换矩阵可以将地图的全局坐标系下的位置点坐标依次转换到车辆定位惯导坐标系、摄像机坐标系、图像坐标系上，最终将每个道路元素投影到摄像机采集的图像上。
S604、对所述图像进行图像检测，得到所述图像上至少一个第二道路元素的第二位置。
在本实施例中，可以通过图像检测模型或传统的图像处理方式对图像进行检测，检测出图像上包含的各道路元素的位置。在此将图像上所包含的各道路元素称为第二道路元素，第二道路元素在图像上的位置称为第二位置。例如，若检测出一张图像包含两条车道边界线和一个交通灯，共三个道路元素，则这三个道路元素均为第二道路元素，每个第二道路元素对应于图像上的一个位置，即第二位置。
可选地，可以将图像输入基于深度学习的检测与分割模型，通过模型可以检测出图像中的第二道路元素，并定位第二道路元素在图像中的位置。
S605、根据所述至少一个第一道路元素投影在所述图像上的第一位置与所述至少一个 第二道路元素在所述图像上的第二位置之间的匹配结果确定所述车辆的位置。
在本实施例中，根据至少一个第一道路元素投影在图像上的第一位置，以及检测出至少一个第二道路元素在图像上的第二位置，将类型相同且位置最近的第一道路元素和第二道路元素进行匹配，根据匹配结果对车辆的位置进行优化计算，可以最终确定出车辆的位置。
可选地，至少一个道路元素包括多种类型。S605可以包括：将类型相同且彼此位置最近的一个第一道路元素和一个第二道路元素作为一个匹配对，可能形成一个或多个匹配对；针对每个匹配对，计算该匹配对中第一道路元素投影在所述图像上的第一位置与第二道路元素在所述图像上的第二位置之间的距离，作为该匹配对的距离；将各匹配对的距离作为投影误差，计算最小化投影误差条件下车辆的位置作为所述车辆的位置。
在本实施例中，可以将类型相同且彼此位置最近的一个第一道路元素和一个第二道路元素作为一个匹配对，可能形成一个或多个匹配对。对于每个匹配对，分别计算该匹配对中第一道路元素投影在所述图像上的第一位置和第二道路元素在所述图像上的第二位置之间的欧式距离，将该距离作为该匹配对的距离。以匹配对的距离作为投影误差，以最小化投影误差为优化目标，计算最小化投影误差时车辆的位置即为定位出的车辆位置。由此计算出的车辆位置结合了地图中第一道路元素的信息以及图像中第二道路元素的信息，定位精度高。
本实施例通过将第一预设范围内的至少一个道路元素向图像上投影，将投影得到的第一位置与检测到的第二道路元素在图像上的第二位置进行匹配，能够快速准确的确定出车辆的位置，提高车辆定位的速度和精度。
图7为本发明另一实施例提供的定位方法中车辆导航的流程示意图。本实施例中至少一个道路元素包括道路线。可以根据道路线对应的点集对车辆的下一行驶点进行预判。如图7所示，该方法还可以包括：
S701、在所述车辆行驶时，获取所述车辆的下一行驶点的坐标，其中，所述下一行驶点为所述车辆的行驶路径上当前位置的下一个位置点。
在本实施例中，自动驾驶车辆在行驶过程中，沿着预先指定或自动规划的行驶路径进行行驶。当车辆行驶到某位置时，获取车辆的下一行驶点的坐标。下一行驶点与车辆当前位置的距离可以根据实际需求设定，在此不作限定。例如，可以将行驶路径上距离当前位置5米的下一个位置点作为下一行驶点。
S702、根据所述下一行驶点的坐标，从所述地图中获取所述下一行驶点所在区域内的道路线对应的位置点的坐标。
S703、以所述下一行驶点所在区域内的道路线对应的位置点为顶点，确定所述下一行驶点对应的多边形区域。
在本实施例中，下一行驶点所在区域内存在道路线，从地图中可以获取下一行驶点所在区域内的道路线对应的位置点坐标。其中，道路线可以是一条或多条。可以将这些道路线对应的位置点作为顶点，形成一个多边形区域，下一行驶点位于该多边形区域内。
S704、在所述多边形区域为可行驶区域时，控制所述车辆向所述下一行驶点行驶。
S705、在所述多边形区域为不可行驶区域时，重新规划所述车辆的下一行驶点，以使所述车辆的下一行驶点位于所述多边形区域之外。
在本实施例中，可行驶区域为车辆可以通行的区域；不可行驶区域为车辆不能通行的区域，如果车辆在该区域行驶可能会发生危险或违反交通规则。可以根据地图中存储的各个区域的信息或者道路线的相关信息确定多边形区域是可行驶区域，还是不可行驶区域。若该多边形区域为可行驶区域，则控制车辆向下一行驶点行驶。若该多边形区域为不可行驶区域，则重新规划车辆的下一行驶点，使车辆的下一行驶点位于多边形区域之外。
本实施例通过从地图中获取下一行驶点所在区域内的道路线对应的位置点的坐标，将这些位置点形成多边形区域，在多边形区域为不可行驶区域时及时为车辆重新规划下一行驶点。由于通过位置点构成的多边形区域能够很快确定出下一行驶点是否位于不可行驶区域，为重新规划下一行驶点留出充足的时间，能够防止车辆进入不可行驶区域，提高车辆安全性。
可选地，所述地图中还存储有至少一条行驶路径对应的位置点的坐标，其中，每条行驶路径对应一个点集，每条行驶路径对应的点集由该行驶路径上的多个位置点组成。
在本实施例中，行驶路径可以是根据该车辆或其他车辆的历史行驶数据得到的经验路径，也可以是根据当前交通状态规划出的路径，在此不作限定。地图中可以保存一条或多条行驶路径。每条行驶路径可以对应于一个由该行驶路径上多个位置点组成的点集。例如，可以将 行驶路径上每隔5米、10米等取一个点作为位置点，构成点集。
可选地，上述方法还可以包括：从所述地图中获取第一行驶路径对应的位置点的坐标；根据所述第一行驶路径对应的位置点的坐标，控制所述车辆沿所述第一行驶路径进行行驶。
在本实施例中，第一行驶路径为地图所存储的行驶路径中车辆当前要行驶的路径。可以从地图中获取该第一行驶路径对应的各个位置点的坐标，然后控制车辆按照该第一行驶路径对应的各个位置点进行行驶。
通过在地图中以点集的形式存储行驶路径，车辆导航可以直接获取存储的点集，直接按照存储的点集对车辆进行控制，能够快速实现对车辆的导航。
可选地，所述至少一个道路元素包括至少一个交通灯，所述地图中还存储有所述交通灯对应的道路转向信息。
在本实施例中，交通灯用于指示车辆在路口的转向。地图中存储有每个交通灯对应的道路转向信息。
其中，所述交通灯包括至少一个信号灯，所述地图中每个交通灯对应于一个由该交通灯所属矩形区域的四个顶点组成的点集、以及一组道路转向信息，其中，每个交通灯对应的道路转向信息包括该交通灯中每个信号灯的形状、颜色、位置、以及所指示的道路转向信息中的至少一项。
例如，一个交通灯可以包括三个信号灯，分别为红灯、黄灯和绿灯。道路转向信息可以包括直行、左转、停止等信息，在此不作限定。
图8为本发明再一实施例提供的定位方法中交通灯识别的流程示意图。本实施例中能够利用地图存储的交通灯的点集以及道路转向信息对车辆进行控制。如图8所示，该方法包括：
S801、在检测到所述图像中包含第一交通灯时，通过图像识别得到第一交通灯的当前信号和所述第一交通灯在所述图像上的第三位置。
在本实施例中，为便于描述，将检测出的图像中所包含的交通灯称为第一交通灯，其在图像上的位置称为第三位置。每个第一交通灯对应于多种信号，分别指示多种不同的道路转向。
S802、从所述地图中获取所述车辆周围第二预设范围内的至少一个第二交通灯对应的位置点的坐标。
S803、根据至少一个第二交通灯对应的位置点的坐标和第一交通灯在图像上的第三位置，确定至少一个第二交通灯中与所述第一交通灯相匹配的目标第二交通灯。
在本实施例中，可以根据第二预设范围内的至少一个第二交通灯对应的位置点的坐标将至少一个第二交通灯向图像上投影，然后将投影结果和检测到的图像中的第一交通灯的第三位置进行匹配，得到与图像中的第一交通灯相匹配的目标第二交通灯。
可选地，S803可以包括：根据第二坐标系转换矩阵和至少一个第二交通灯对应的位置点的坐标，将至少一个第二交通灯分别向所述图像上投影，得到至少一个第二交通灯投影在所述图像上的第四位置；根据第一交通灯在图像上的第三位置以及至少一个第二交通灯投影在图像上的第四位置，将投影在图像上的第四位置与该第一交通灯位置最近的第二交通灯确定为与第一交通灯相匹配的目标第二交通灯。
在本实施例中，地图中存储的至少一个第二交通灯对应的位置点的坐标为全局坐标系下的坐标，图像上像素的坐标为图像坐标系下的坐标。可以通过第二坐标系转换矩阵将第二预设范围内的第二交通灯对应的位置点的坐标由全局坐标系转换到图像的像素坐标系下，从而实现将第二预设范围内的至少一个第二交通灯向图像上投影。
将第二预设范围内的至少一个第二交通灯向图像上进行投影，可能部分第二交通灯投影后位于图像上，其他第二交通灯投影后位于图像之外，在此针对处于第二预设范围内的并且投影后位于图像上的第二交通灯，将其投影到图像上的位置称为第四位置。
通过计算检测出的第一交通灯在图像上的第三位置与第二交通灯投影在图像上的第四位置之间的距离，将投影在图像上的第四位置与第一交通灯的第三位置最接近的第二交通灯确定为与第一交通灯匹配的第二交通灯，即确定出目标第二交通灯。
S804、从所述地图中获取所述目标第二交通灯对应的道路转向信息。
S805、根据所述第一交通灯的当前信号和所述目标第二交通灯对应的道路转向信息，确定所述第一交通灯的当前信号所指示的道路转向。
在本实施例中，由于目标第二交通灯与第一交通灯相匹配，因此将地图中目标第二交 通灯的道路转向信息作为第一交通灯的道路转向信息，结合第一交通灯的当前信号，就能确定出第一交通灯当前指示的道路转向。
如图9所示为本发明实施例提供的第一交通灯的示意图，图中，从上到下依次为红灯、绿灯和左转指示灯。该第一交通灯对应的道路转向信息如下：红灯亮时，指示车辆停止；绿灯亮时，指示车辆直行；左转指示灯亮时，指示车辆左转。若第一交通灯的当前信号为绿灯亮，则确定第一交通灯的当前信号所指示的道路转向为直行。
S806、根据所述第一交通灯的当前信号所指示的道路转向，对所述车辆进行控制，以使所述车辆按照所述第一交通灯的当前信号所指示的道路转向进行行驶。
本发明实施例通过地图中存储交通灯的点集以及道路转向信息，结合图像检测交通灯的当前信号，能够准确识别交通灯指示的道路转向，进而准确控制车辆进行相应的转向。
一般地，高精度地图采用OpenDrive格式，OpenDrive格式的地图过于复杂，且其专注于过于理想的驾驶仿真系统，无法灵活自由地描述复杂多变的真实的道路情景。OpenDrive格式的地图为矢量化地图，使用了复杂的曲线方程拟合车道线等道路元素，地图中的道路元素以曲线方程的形式存储，存储结构复杂，解析困难，使用成本较高。并且，OpenDrive格式的地图只存储了交通灯的位置信息，没有将交通灯规则与高精度地图联系起来，无法解决自动驾驶中交通灯相关的决策规划问题。
本发明实施例提供了一种服务于自动驾驶系统的轻量级高精度地图格式，并且提供了基于该高精度地图格式的定位方法，能够支持自动驾驶的定位、决策规划控制、交通灯理解和静态场景还原。该高精度地图格式根据各道路元素对应的点集，以点集的形式存储道路的道路元素并生成地图，避免在地图中以复杂的曲线方程的形式存储道路元素，从而降低地图解析的数据量，提高解析速度。该高精度地图格式改进了一般矢量化地图复杂的设计，利用随曲率自适应的点集存储方式，在简化地图使用的同时降低了地图存储、传输和解析时间。并且该高精度地图格式扩展了交通灯存储方式，存储有交通灯对应的点集以及道路转向信息，支持交通灯的查询和定位、辅助自动驾驶系统理解交通灯相关交通规则。
图10为本发明一实施例提供的地图生成装置的结构示意图。如图10所示，该地图生成装置100包括：获取模块1001、生成模块1002及存储模块1003。
获取模块1001，用于获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识。
生成模块1002，用于针对所述至少一个道路元素中的每个道路元素，根据该道路元素的位置信息，生成该道路元素对应的点集，其中，该点集包含多个表征该道路元素所在位置的位置点。
存储模块1003，用于将该道路上各道路元素的标识及对应的点集，与所述道路的标识进行关联存储，生成地图。
可选地，所述道路元素包括以下中的至少一项：道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌。
其中，所述左/右边界线对应的点集由所述左/右边界线上多个位置点组成。
其中，所述可行驶边界线对应的点集由所述可行驶边界线上多个位置点组成。
其中，所述虚线车道线对应的点集由所述虚线车道线上两个端点组成。
其中，所述起始线对应的点集由所述起始线上两个端点组成。
其中，所述停止线对应的点集由所述停止线上两个端点组成。
其中，所述人行横道对应的点集由所述人行横道所属多边形区域的多个顶点组成。
其中，所述停车位对应的点集由所述停车位所属矩形区域的四个顶点组成。
其中，所述障碍物对应的点集由所述障碍物所属多面体的多个顶点组成。
可选地，曲线类型的道路元素对应的点集中位置点的密度根据其曲率确定。曲线类型的道路元素可包括所述左/右边界线或所述可行驶边界线。例如，所述道路的左/右边界线对应的点集中位置点的密度根据所述道路的左/右边界线的曲率确定。
可选地，所述道路元素包括左/右边界线；所述存储模块1003还用于：根据所述道路的该左/右边界线的位置信息，确定所述道路的参考线的位置信息，并生成所述参考线的标识，其中，所述参考线位于所述道路的中心；从所述参考线上选取多个位置点组成所述参考线对应的点集；将所述参考线的标识及其对应的点集，与所述道路的标识关联存储到所述地图中。
可选地，所述道路上包括至少一个车道，每个车道对应于一个标识，所述道路元素包括各车道的车道线；所述存储模块1003还用于：针对每个车道，根据该车道的车道线的位置信息，确定该车道的中心线的位置信息，其中，该中心线位于该车道的中心；针对每个车道，从该车道的中心线上选取多个位置点组成该车道的中心线对应的点集；将该道路上各车道的标识及其中心线对应的点集，与所述道路的标识关联存储到所述地图中。
可选地，所述存储模块1003还用于：获取各车道对应的限速值；将各车道对应的限速值与各车道的标识关联存储到所述地图中。
可选地，所述地图还包括至少一个路口，每个路口对应于一个标识；所述存储模块1003还用于：针对每个路口，获取与该路口相关联的多条道路的信息，并根据多条道路的信息确定所述多条道路中存在连接关系的每两条道路，其中，所述连接关系包括该连接关系的起始道路和目标道路；将所述连接关系以及所述路口的标识关联存储到所述地图中。
可选地，所述存储模块1003还用于：匹配所述连接关系和所述路口的交通灯信息；将所述交通灯与其匹配的连接关系关联存储到所述地图中。
本发明实施例提供的地图生成装置，可用于执行上述的方法实施例，其实现原理和技术效果类似，本实施例此处不再赘述。
图11为本发明一实施例提供的定位装置的结构示意图。如图11所示，该定位装置110包括：获取模块1101和定位模块1102。
获取模块1101，用于从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中，每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标。
所述获取模块1101还用于获取所述车辆采集的图像。
定位模块1102，用于根据所述第一预设范围内的至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置。
本发明实施例中，地图以点集的形式存储每个道路元素。每个道路元素对应于一个点集，点集包含多个表征该道路元素所在位置的位置点，地图中存储有该道路元素对应的点集所包含的位置点的坐标。通过获取模块从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，并获取车辆采集的图像，然后定位模块根据第一预设范围内的至少一个道路元素对应的位置点的坐标，将第一预设范围内的至少一个道路元素与图像进行匹配，并根据匹配结果确定车辆的位置，能够直接从地图中提取至少一个道路元素的点集结合图像进行定位，避免对道路元素的曲线方程的解析过程，从而减少所需计算的数据量，提高定位速度。
图12为本发明又一实施例提供的定位装置的结构示意图。如图12所示，本实施例提供的定位装置120在图11所示实施例提供的定位装置110的基础上，还可以包括：导航模块1203、识别模块1204和显示模块1205。
可选地，所述定位模块1202具体用于：根据所述第一预设范围内的至少一个道路元素对应的位置点的坐标，将所述第一预设范围内的至少一个道路元素分别向所述图像上投影，得到至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一道路元素为处于所述第一预设范围内且投影后位于所述图像上的道路元素；对所述图像进行图像检测，得到所述图像上至少一个第二道路元素的第二位置；根据所述至少一个第一道路元素投影在所述图像上的第一位置与所述至少一个第二道路元素在所述图像上的第二位置之间的匹配结果确定所述车辆的位置。
可选地，所述至少一个道路元素对应的位置点的坐标为全局坐标系下的坐标；所述定位模块1202具体用于：根据第一坐标系转换矩阵，将所述第一预设范围内的至少一个道路元素对应的位置点的坐标由所述全局坐标系转换到所述图像的像素坐标系下，得到所述至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一坐标系转换矩阵为所述全局坐标系与所述图像的像素坐标系之间的转换矩阵。
可选地，所述至少一个道路元素包括多种类型；所述定位模块1202具体用于：将类型相同且彼此位置最近的一个第一道路元素和一个第二道路元素作为一个匹配对，以从所述至少一个第一道路元素和所述至少一个第二道路元素确定至少一个匹配对；针对每个匹配对，计算该匹配对中第一道路元素投影在图像上的第一位置与该匹配对中第二道路元素在所述图像上的第二位置之间的距离，作为该匹配对的距离；将各匹配对的距离作为投影误差，计算最小化投影误差条件下车辆的位置作为所述车辆的位置。
可选地，所述获取模块1201具体用于：采用第一定位方式定位得到所述车辆的原始位置，其中，所述第一定位方式的定位精度低于预设阈值；根据所述原始位置，从所述地图中查找所述车辆周围所述第一预设范围内的至少一个道路元素对应的位置点的坐标。
可选地，所述至少一个道路元素包括道路线，所述导航模块1203用于：在所述车辆行驶时，获取所述车辆的下一行驶点的坐标，其中，所述下一行驶点为所述车辆的行驶路径上当前位置的下一个位置点；根据所述下一行驶点的坐标，从所述地图中获取所述下一行驶点所在区域内的道路线对应的位置点的坐标；以所述下一行驶点所在区域内的道路线对应的位置点为顶点，确定所述下一行驶点对应的多边形区域；在所述多边形区域为可行驶区域时，控制所述车辆向所述下一行驶点行驶；在所述多边形区域为不可行驶区域时，重新规划所述车辆的下一行驶点，以使所述车辆的下一行驶点位于所述多边形区域之外。
可选地，所述地图中还存储有至少一条行驶路径对应的位置点的坐标，其中，每条行驶路径对应一个点集，每条行驶路径对应的点集由该行驶路径上的多个位置点组成。
可选地，所述导航模块1203还用于：从所述地图中获取第一行驶路径对应的位置点的坐标；根据所述第一行驶路径对应的位置点的坐标，控制所述车辆沿所述第一行驶路径进行行驶。
可选地，所述至少一个道路元素包括至少一个交通灯，所述地图中还存储有所述交通灯对应的道路转向信息。
可选地，所述交通灯包括至少一个信号灯。所述地图中每个交通灯对应于一个由该交通灯所属矩形区域的四个顶点组成的点集、以及一组道路转向信息。其中，每个交通灯对应的道路转向信息包括该交通灯中每个信号灯的形状、颜色、位置、以及所指示的道路转向信息中的至少一项。
可选地，所述识别模块1204用于：在检测到所述图像中包含第一交通灯时，通过图像识别得到第一交通灯的当前信号和所述第一交通灯在所述图像上的第三位置；从所述地图中获取所述车辆周围第二预设范围内的至少一个第二交通灯对应的位置点的坐标；根据所述至少一个第二交通灯对应的位置点的坐标和所述第一交通灯在所述图像上的第三位置，确定所述至少一个第二交通灯中与所述第一交通灯相匹配的目标第二交通灯；从所述地图中获取所述目标第二交通灯对应的道路转向信息；根据所述第一交通灯的当前信号和所述目标第二交通灯对应的道路转向信息，确定所述第一交通灯的当前信号所指示的道路转向；根据所述第一交通灯的当前信号所指示的道路转向，对所述车辆进行控制，以使所述车辆按照所述第一交通灯的当前信号所指示的道路转向进行行驶。
可选地，所述识别模块1204具体用于：根据第二坐标系转换矩阵和所述至少一个第二交通灯对应的位置点的坐标，将所述至少一个第二交通灯分别向所述图像上投影，得到至少一个第二交通灯投影在所述图像上的第四位置；根据所述第一交通灯在所述图像上的第三位置以及所述至少一个第二交通灯投影在所述图像上的第四位置，将投影在所述图像上的第四位置与所述第一交通灯位置最近的第二交通灯确定为与所述第一交通灯相匹配的目标第二交通灯。
可选地，所述至少一个道路元素包括以下中的至少一项：道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌。
其中，所述左/右边界线、所述可行驶边界线均对应于一个由多个位置点组成的点集；所述虚线车道线、所述起始线、所述停止线均对应于一个由两个位置点组成的点集；所述人行横道对应于一个由所述人行横道所属多边形区域的多个顶点组成的点集；所述停车位对应于一个由所述停车位所属矩形区域的四个顶点组成的点集；所述障碍物对应于一个由所述障碍物所属多面体的多个顶点组成的点集。
可选地，所述显示模块1205用于：确定所述车辆的位置之后，显示所述车辆的位置。
本发明实施例提供的定位装置，可用于执行上述的方法实施例，其实现原理和技术效果类似，本实施例此处不再赘述。
图13为本发明一实施例提供的地图生成设备的硬件结构示意图。如图13所示，本实施例提供的地图生成设备130包括：至少一个处理器1301和存储器1302。该地图生成设备130还包括通信部件1303。其中，处理器1301、存储器1302以及通信部件1303通过总线1304连接。
在具体实现过程中，至少一个处理器1301执行所述存储器1302存储的计算机可执行 指令，使得至少一个处理器1301执行如上的地图生成方法。
处理器1301的具体实现过程可参见上述方法实施例，其实现原理和技术效果类似，本实施例此处不再赘述。
在上述的图13所示的实施例中，应理解，处理器可以是中央处理单元(Central Processing Unit，CPU)，还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)等。通用处理器可以是微处理器或者是任何常规的处理器等。结合发明所公开的方法的步骤可以直接体现为硬件处理器执行完成，或者用处理器中的硬件及软件模块组合执行完成。
存储器可能包含高速RAM存储器，也可能还包括非易失性存储NVM，例如至少一个磁盘存储器。
总线可以是工业标准体系结构(Industry Standard Architecture，ISA)总线、外部设备互连(Peripheral Component Interconnect，PCI)总线或扩展工业标准体系结构(Extended Industry Standard Architecture，EISA)总线等。总线可以分为地址总线、数据总线、控制总线等。为便于表示，本申请附图中的总线并不限定仅有一根总线或一种类型的总线。
图14为本发明一实施例提供的定位设备的硬件结构示意图。如图14所示，本实施例提供的定位设备140包括：至少一个处理器1401和存储器1402。该定位设备140还包括通信部件1403。其中，处理器1401、存储器1402以及通信部件1403通过总线1404连接。
在具体实现过程中，至少一个处理器1401执行所述存储器1402存储的计算机可执行指令，使得至少一个处理器1401执行如上的定位方法。
处理器1401的具体实现过程可参见上述方法实施例，其实现原理和技术效果类似，本实施例此处不再赘述。
在上述的图14所示的实施例中，应理解，处理器可以是中央处理单元(英文：Central Processing Unit，简称：CPU)，还可以是其他通用处理器、数字信号处理器(英文：Digital Signal Processor，简称：DSP)、专用集成电路(英文：Application Specific Integrated Circuit，简称：ASIC)等。通用处理器可以是微处理器或者是任何常规的处理器等。结合发明所公开的方法的步骤可以直接体现为硬件处理器执行完成，或者用处理器中的硬件及软件模块组合执行完成。
存储器可能包含高速RAM存储器，也可能还包括非易失性存储NVM，例如至少一个磁盘存储器。
总线可以是工业标准体系结构(Industry Standard Architecture，ISA)总线、外部设备互连(Peripheral Component Interconnect，PCI)总线或扩展工业标准体系结构(Extended Industry Standard Architecture，EISA)总线等。总线可以分为地址总线、数据总线、控制总线等。为便于表示，本申请附图中的总线并不限定仅有一根总线或一种类型的总线。
本申请还提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机可执行指令，当处理器执行所述计算机可执行指令时，实现如上的地图生成方法或定位方法。
上述的计算机可读存储介质可以是由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。可读存储介质可以是通用或专用计算机能够存取的任何可用介质。
一种示例性的可读存储介质耦合至处理器，从而使处理器能够从该可读存储介质读取信息，且可向该可读存储介质写入信息。当然，可读存储介质也可以是处理器的组成部分。处理器和可读存储介质可以位于ASIC中。当然，处理器和可读存储介质也可以作为分立组件存在于设备中。
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (50)

  1. 一种地图生成方法，包括：
    获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识；
    针对所述至少一个道路元素中的每个道路元素，根据该道路元素的位置信息，生成该道路元素对应的点集，其中，所述点集包含多个表征该道路元素所在位置的位置点；
    将所述道路上各道路元素的标识及对应的点集，与所述道路的标识进行关联存储，生成地图。
  2. 根据权利要求1所述的方法，其特征在于，所述道路元素包括以下中的至少一项：所述道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌；
    所述左/右边界线对应的点集由所述左/右边界线上多个位置点组成；
    所述可行驶边界线对应的点集由所述可行驶边界线上多个位置点组成；
    所述虚线车道线对应的点集由所述虚线车道线上两个端点组成；
    所述起始线对应的点集由所述起始线上两个端点组成；
    所述停止线对应的点集由所述停止线上两个端点组成；
    所述人行横道对应的点集由所述人行横道所属多边形区域的多个顶点组成；
    所述停车位对应的点集由所述停车位所属矩形区域的四个顶点组成；
    所述障碍物对应的点集由所述障碍物所属多面体的多个顶点组成。
  3. 根据权利要求2所述的方法，其特征在于，
    曲线类型的道路元素对应的点集中位置点的密度根据其曲率确定；
    所述曲线类型的道路元素包括所述左/右边界线或所述可行驶边界线。
  4. 根据权利要求1-3任一项所述的方法，其特征在于，所述道路元素包括所述左/右边界线；所述方法还包括：
    根据所述道路的所述左/右边界线的位置信息，确定所述道路的参考线的位置信息，并生成所述参考线的标识，其中，所述参考线位于所述道路的中心；
    从所述参考线上选取多个位置点组成所述参考线对应的点集；
    将所述参考线的标识及其对应的点集，与所述道路的标识关联存储到所述地图中。
  5. 根据权利要求1-4任一项所述的方法，其特征在于，所述道路上包括至少一个车道，每个车道对应于一个标识，所述道路元素包括各车道的车道线；所述方法还包括：
    针对每个车道，
    根据该车道的车道线的位置信息，确定该车道的中心线的位置信息，其中，所述中心线位于该车道的中心；
    从该车道的中心线上选取多个位置点组成该车道的中心线对应的点集；
    将所述道路上各车道的标识及其中心线对应的点集，与所述道路的标识关联存储到所述地图中。
  6. 根据权利要求5所述的方法，其特征在于，所述方法还包括：
    获取各车道对应的限速值；
    将各车道对应的限速值与各车道的标识关联存储到所述地图中。
  7. 根据权利要求1-6任一项所述的方法，其特征在于，所述地图还包括至少一个路口，每个路口对应于一个标识，所述方法还包括：
    针对每个路口，
    获取与该路口相关联的多条道路的信息，
    根据所述多条道路的信息确定所述多条道路中存在连接关系的每两条道路，其中，所述连接关系包括该连接关系的起始道路和目标道路；
    将所述连接关系与所述路口的标识关联存储到所述地图中。
  8. 根据权利要求7所述的方法，其特征在于，所述方法还包括：
    匹配所述连接关系和所述路口的交通灯信息；
    将所述交通灯与其匹配的连接关系关联存储到所述地图中。
  9. 一种定位方法，包括：
    从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中， 每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标；
    获取所述车辆采集的图像；
    根据所述第一预设范围内的所述至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置。
  10. 根据权利要求9所述的方法，其特征在于，根据所述第一预设范围内的所述至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置，包括：
    根据所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标，将所述第一预设范围内的所述至少一个道路元素分别向所述图像上投影，得到至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一道路元素为处于所述第一预设范围内且投影后位于所述图像上的道路元素；
    对所述图像进行图像检测，得到所述图像上至少一个第二道路元素的第二位置；
    根据所述至少一个第一道路元素投影在所述图像上的第一位置与所述至少一个第二道路元素在所述图像上的第二位置之间的匹配结果确定所述车辆的位置。
  11. 根据权利要求10所述的方法，其特征在于，所述至少一个道路元素对应的位置点的坐标为全局坐标系下的坐标；根据所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标，将所述第一预设范围内的所述至少一个第一道路元素分别向所述图像上投影，得到所述至少一个第一道路元素投影在所述图像上的第一位置，包括：
    根据第一坐标系转换矩阵，将所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标由所述全局坐标系转换到所述图像的像素坐标系下，得到所述至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一坐标系转换矩阵为所述全局坐标系与所述图像的像素坐标系之间的转换矩阵。
  12. 根据权利要求10所述的方法，其特征在于，所述至少一个道路元素包括多种类型；根据所述至少一个第一道路元素投影在所述图像上的第一位置与所述至少一个第二道路元素在所述图像上的第二位置之间的匹配结果确定所述车辆的位置，包括：
    将类型相同且彼此位置最近的一个第一道路元素和一个第二道路元素作为一个匹配对，以从所述至少一个第一道路元素和所述至少一个第二道路元素确定至少一个匹配对；
    针对每个匹配对，计算该匹配对中所述第一道路元素投影在所述图像上的第一位置与该匹配对中所述第二道路元素在所述图像上的第二位置之间的距离，作为该匹配对的距离；
    将各匹配对的距离作为投影误差，计算最小化投影误差条件下车辆的位置作为所述车辆的位置。
  13. 根据权利要求9-12任一项所述的方法，其特征在于，从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，包括：
    采用第一定位方式定位得到所述车辆的原始位置，其中，所述第一定位方式的定位精度低于预设阈值；
    根据所述原始位置，从所述地图中查找所述车辆周围所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标。
  14. 根据权利要求9-13任一项所述的方法，其特征在于，所述至少一个道路元素包括道路线，所述方法还包括：
    在所述车辆行驶时，获取所述车辆的下一行驶点的坐标，其中，所述下一行驶点为所述车辆的行驶路径上当前位置的下一个位置点；
    根据所述下一行驶点的坐标，从所述地图中获取所述下一行驶点所在区域内的道路线对应的位置点的坐标；
    以所述下一行驶点所在区域内的道路线对应的位置点为顶点，确定所述下一行驶点对应的多边形区域；
    在所述多边形区域为可行驶区域时，控制所述车辆向所述下一行驶点行驶；
    在所述多边形区域为不可行驶区域时，重新规划所述车辆的下一行驶点，以使所述车辆的下一行驶点位于所述多边形区域之外。
  15. 根据权利要求9-14任一项所述的方法，其特征在于，所述地图中还存储有至少一条行驶路径对应的位置点的坐标，其中，每条行驶路径对应一个点集，每条行驶路径对应的点集由该行驶路径上的多个位置点组成。
  16. 根据权利要求15所述的方法，其特征在于，所述方法还包括：
    从所述地图中获取第一行驶路径对应的位置点的坐标；
    根据所述第一行驶路径对应的位置点的坐标，控制所述车辆沿所述第一行驶路径进行行驶。
  17. 根据权利要求9-16任一项所述的方法，其特征在于，所述至少一个道路元素包括至少一个交通灯，所述地图中还存储有所述交通灯对应的道路转向信息。
  18. 根据权利要求17所述的方法，其特征在于，所述交通灯包括至少一个信号灯，所述地图中每个交通灯对应于一个由该交通灯所属矩形区域的四个顶点组成的点集、以及一组道路转向信息，其中，每个交通灯对应的道路转向信息包括该交通灯中每个信号灯的形状、颜色、位置、以及所指示的道路转向信息中的至少一项。
  19. 根据权利要求17或18所述的方法，其特征在于，所述方法还包括：
    在检测到所述图像中包含第一交通灯时，通过图像识别得到所述第一交通灯的当前信号和所述第一交通灯在所述图像上的第三位置；
    从所述地图中获取所述车辆周围第二预设范围内的至少一个第二交通灯对应的位置点的坐标；
    根据所述至少一个第二交通灯对应的位置点的坐标和所述第一交通灯在所述图像上的第三位置，确定所述至少一个第二交通灯中与所述第一交通灯相匹配的目标第二交通灯；
    从所述地图中获取所述目标第二交通灯对应的道路转向信息；
    根据所述第一交通灯的当前信号和所述目标第二交通灯对应的道路转向信息，确定所述第一交通灯的当前信号所指示的道路转向；
    根据所述第一交通灯的当前信号所指示的道路转向，对所述车辆进行控制。
  20. 根据权利要求19所述的方法，其特征在于，根据所述至少一个第二交通灯对应的位置点的坐标和所述第一交通灯在所述图像上的第三位置，确定所述至少一个第二交通灯中与所述第一交通灯相匹配的目标第二交通灯，包括：
    根据第二坐标系转换矩阵和所述至少一个第二交通灯对应的位置点的坐标，将所述至少一个第二交通灯分别向所述图像上投影，得到至少一个第二交通灯投影在所述图像上的第四位置；
    根据所述第一交通灯在所述图像上的第三位置以及所述至少一个第二交通灯投影在所述图像上的第四位置，将投影在所述图像上的第四位置与所述第一交通灯位置最近的第二交通灯确定为与所述第一交通灯相匹配的目标第二交通灯。
  21. 根据权利要求9-20任一项所述的方法，其特征在于，所述至少一个道路元素包括以下中的至少一项：所述道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌。
  22. 根据权利要求21所述的方法，其特征在于，
    所述左/右边界线、所述可行驶边界线均对应于一个由多个位置点组成的点集；
    所述虚线车道线、所述起始线、所述停止线均对应于一个由两个位置点组成的点集；
    所述人行横道对应于一个由所述人行横道所属多边形区域的多个顶点组成的点集；
    所述停车位对应于一个由所述停车位所属矩形区域的四个顶点组成的点集；
    所述障碍物对应于一个由所述障碍物所属多面体的多个顶点组成的点集。
  23. 根据权利要求9-22任一项所述的方法，其特征在于，确定所述车辆的位置之后，所述方法还包括：
    显示所述车辆的位置。
  24. 一种地图生成装置，包括：
    获取模块，用于获取传感器采集的道路上至少一个道路元素的位置信息，每个道路元素对应一个标识；
    生成模块，用于针对所述至少一个道路元素中的每个道路元素，根据该道路元素的位置信息，生成该道路元素对应的点集，其中，所述点集包含多个表征该道路元素所在位置的位置点；
    存储模块，用于将所述道路上各道路元素的标识及对应的点集，与所述道路的标识进行关联存储，生成地图。
  25. 根据权利要求24所述的装置，其特征在于，所述道路元素包括以下中的至少一项：所述道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌；
    所述左/右边界线对应的点集由所述左/右边界线上多个位置点组成；
    所述可行驶边界线对应的点集由所述可行驶边界线上多个位置点组成；
    所述虚线车道线对应的点集由所述虚线车道线上两个端点组成；
    所述起始线对应的点集由所述起始线上两个端点组成；
    所述停止线对应的点集由所述停止线上两个端点组成；
    所述人行横道对应的点集由所述人行横道所属多边形区域的多个顶点组成；
    所述停车位对应的点集由所述停车位所属矩形区域的四个顶点组成；
    所述障碍物对应的点集由所述障碍物所属多面体的多个顶点组成。
  26. 根据权利要求25所述的装置，其特征在于，曲线类型的道路元素对应的点集中位置点的密度根据其曲率确定；所述曲线类型的道路元素包括所述左/右边界线或所述可行驶边界线。
  27. 根据权利要求24-26任一项所述的装置，其特征在于，所述道路元素包括所述左/右边界线；所述存储模块还用于：
    根据所述道路的所述左/右边界线的位置信息，确定所述道路的参考线的位置信息，并生成所述参考线的标识，其中，所述参考线位于所述道路的中心；
    从所述参考线上选取多个位置点组成所述参考线对应的点集；
    将所述参考线的标识及其对应的点集，与所述道路的标识关联存储到所述地图中。
  28. 根据权利要求24-27任一项所述的装置，其特征在于，所述道路上包括至少一个车道，每个车道对应于一个标识，所述道路元素包括各车道的车道线；所述存储模块还用于：
    针对每个车道，
    根据该车道的车道线的位置信息，确定该车道的中心线的位置信息，其中，所述中心线位于该车道的中心；
    从该车道的中心线上选取多个位置点组成该车道的中心线对应的点集；
    将所述道路上各车道的标识及其中心线对应的点集，与所述道路的标识关联存储到所述地图中。
  29. 根据权利要求28所述的装置，其特征在于，所述存储模块还用于：
    获取各车道对应的限速值；
    将各车道对应的限速值与各车道的标识关联存储到所述地图中。
  30. 根据权利要求24-29任一项所述的装置，其特征在于，所述地图还包括至少一个路口，每个路口对应于一个标识，所述存储模块还用于：
    针对每个路口，
    获取与该路口相关联的多条道路的信息，
    根据所述多条道路的信息确定所述多条道路中存在连接关系的每两条道路，其中，所述连接关系包括该连接关系的起始道路和目标道路；
    将所述连接关系以及所述路口的标识关联存储到所述地图中。
  31. 根据权利要求30所述的装置，其特征在于，所述存储模块还用于：
    匹配所述连接关系和所述路口的交通灯信息；
    将所述交通灯与其匹配的连接关系关联存储到所述地图中。
  32. 一种定位装置，包括：
    获取模块，用于从地图中获取车辆周围第一预设范围内的至少一个道路元素对应的位置点的坐标，其中，每个道路元素对应于一个点集，所述点集包含多个表征该道路元素所在位置的位置点，所述地图中存储有该道路元素对应的点集所包含的位置点的坐标；
    所述获取模块，还用于获取所述车辆采集的图像；
    定位模块，用于根据所述第一预设范围内的所述至少一个道路元素与所述图像之间的匹配结果确定所述车辆的位置。
  33. 根据权利要求32所述的装置，其特征在于，所述定位模块具体用于：
    根据所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标，将所述第一预设范围内的所述至少一个道路元素分别向所述图像上投影，得到至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一道路元素为处于所述第一预设范围内且投影后位于所述图像上的道路元素；
    对所述图像进行图像检测，得到所述图像上至少一个第二道路元素的第二位置；
    根据所述至少一个第一道路元素投影在所述图像上的第一位置与所述至少一个第二道路元素在所述图像上的第二位置之间的匹配结果确定所述车辆的位置。
  34. 根据权利要求33所述的装置，其特征在于，所述至少一个道路元素对应的位置点的坐标为全局坐标系下的坐标；所述定位模块具体用于：
    根据第一坐标系转换矩阵，将所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标由所述全局坐标系转换到所述图像的像素坐标系下，得到所述至少一个第一道路元素投影在所述图像上的第一位置，其中，所述第一坐标系转换矩阵为所述全局坐标系与所述图像的像素坐标系之间的转换矩阵。
  35. 根据权利要求33所述的装置，其特征在于，所述至少一个道路元素包括多种类型；所述定位模块具体用于：
    将类型相同且彼此位置最近的一个第一道路元素和一个第二道路元素作为一个匹配对，以从所述至少一个第一道路元素和所述至少一个第二道路元素确定至少一个匹配对；
    针对每个匹配对，计算该匹配对中所述第一道路元素投影在所述图像上的第一位置与该匹配对中所述第二道路元素在所述图像上的第二位置之间的距离，作为该匹配对的距离；
    将各匹配对的距离作为投影误差，计算最小化投影误差条件下车辆的位置作为所述车辆的位置。
  36. 根据权利要求32-35任一项所述的装置，其特征在于，所述获取模块具体用于：
    采用第一定位方式定位得到所述车辆的原始位置，其中，所述第一定位方式的定位精度低于预设阈值；
    根据所述原始位置，从所述地图中查找所述车辆周围所述第一预设范围内的所述至少一个道路元素对应的位置点的坐标。
  37. 根据权利要求32-36任一项所述的装置，其特征在于，所述至少一个道路元素包括道路线，所述装置还包括导航模块，所述导航模块用于：
    在所述车辆行驶时，获取所述车辆的下一行驶点的坐标，其中，所述下一行驶点为所述车辆的行驶路径上当前位置的下一个位置点；
    根据所述下一行驶点的坐标，从所述地图中获取所述下一行驶点所在区域内的道路线对应的位置点的坐标；
    以所述下一行驶点所在区域内的道路线对应的位置点为顶点，确定所述下一行驶点对应的多边形区域；
    在所述多边形区域为可行驶区域时，控制所述车辆向所述下一行驶点行驶；
    在所述多边形区域为不可行驶区域时，重新规划所述车辆的下一行驶点，以使所述车辆的下一行驶点位于所述多边形区域之外。
  38. 根据权利要求32-37任一项所述的装置，其特征在于，所述地图中还存储有至少一条行驶路径对应的位置点的坐标，其中，每条行驶路径对应一个点集，每条行驶路径对应的点集由该行驶路径上的多个位置点组成。
  39. 根据权利要求38所述的装置，其特征在于，所述导航模块还用于：
    从所述地图中获取第一行驶路径对应的位置点的坐标；
    根据所述第一行驶路径对应的位置点的坐标，控制所述车辆沿所述第一行驶路径进行行驶。
  40. 根据权利要求32-39任一项所述的装置，其特征在于，所述至少一个道路元素包括至少一个交通灯，所述地图中还存储有所述交通灯对应的道路转向信息。
  41. 根据权利要求40所述的装置，其特征在于，所述交通灯包括至少一个信号灯，所述地图中每个交通灯对应于一个由该交通灯所属矩形区域的四个顶点组成的点集、以及一组道路转向信息，其中，每个交通灯对应的道路转向信息包括该交通灯中每个信号灯的形状、颜色、位置、以及所指示的道路转向信息中的至少一项。
  42. 根据权利要求40或41所述的装置，其特征在于，所述装置还包括识别模块，所述识别模块用于：
    在检测到所述图像中包含第一交通灯时，通过图像识别得到所述第一交通灯的当前信号和所述第一交通灯在所述图像上的第三位置；
    从所述地图中获取所述车辆周围第二预设范围内的至少一个第二交通灯对应的位置点的坐标；
    根据所述至少一个第二交通灯对应的位置点的坐标和所述第一交通灯在所述图像上的第三位置，确定所述至少一个第二交通灯中与所述第一交通灯相匹配的目标第二交通灯；
    从所述地图中获取所述目标第二交通灯对应的道路转向信息；
    根据所述第一交通灯的当前信号和所述目标第二交通灯对应的道路转向信息，确定所述第一交通灯的当前信号所指示的道路转向；
    根据所述第一交通灯的当前信号所指示的道路转向，对所述车辆进行控制。
  43. 根据权利要求42所述的装置，其特征在于，所述识别模块具体用于：
    根据第二坐标系转换矩阵和所述至少一个第二交通灯对应的位置点的坐标，将所述至少一个第二交通灯分别向所述图像上投影，得到至少一个第二交通灯投影在所述图像上的第四位置；
    根据所述第一交通灯在所述图像上的第三位置以及所述至少一个第二交通灯投影在所述图像上的第四位置，将投影在所述图像上的第四位置与所述第一交通灯位置最近的第二交通灯确定为与所述第一交通灯相匹配的目标第二交通灯。
  44. 根据权利要求32-43任一项所述的装置，其特征在于，所述至少一个道路元素包括以下中的至少一项：所述道路的左/右边界线、可行驶边界线、虚线车道线、起始线、停止线、人行横道、停车位、障碍物、路口边界、交通灯、路灯和交通牌。
  45. 根据权利要求44所述的装置，其特征在于，
    所述左/右边界线、所述可行驶边界线均对应于一个由多个位置点组成的点集；
    所述虚线车道线、所述起始线、所述停止线均对应于一个由两个位置点组成的点集；
    所述人行横道对应于一个由所述人行横道所属多边形区域的多个顶点组成的点集；
    所述停车位对应于一个由所述停车位所属矩形区域的四个顶点组成的点集；
    所述障碍物对应于一个由所述障碍物所属多面体的多个顶点组成的点集。
  46. 根据权利要求32-45任一项所述的装置，其特征在于，所述装置还包括显示模块，所述显示模块用于：
    确定所述车辆的位置之后，显示所述车辆的位置。
  47. 一种地图生成设备，包括：至少一个处理器和存储器；
    所述存储器存储计算机可执行指令；
    所述至少一个处理器执行所述存储器存储的计算机可执行指令，使得所述至少一个处理器执行如权利要求1-8任一项所述的地图生成方法。
  48. 一种定位设备，包括：至少一个处理器和存储器；
    所述存储器存储计算机可执行指令；
    所述至少一个处理器执行所述存储器存储的计算机可执行指令，使得所述至少一个处理器执行如权利要求9-23任一项所述的定位方法。
  49. 一种计算机可读存储介质，存储有计算机可执行指令，当处理器执行所述计算机可执行指令时，实现如权利要求1-8任一项所述的地图生成方法，或者实现如权利要求9-23任一项所述的定位方法。
  50. 一种计算机程序，包括计算机可读代码，当所述计算机可读代码在设备上运行时，使得所述设备中的处理器执行权利要求1-8任一项所述的地图生成方法，或者执行权利要求9-23任一项所述的定位方法。
PCT/CN2020/130627 2019-11-29 2020-11-20 地图生成方法、定位方法、装置、设备、存储介质及计算机程序 WO2021104180A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2022506803A JP2022542712A (ja) 2019-11-29 2020-11-20 地図生成方法、測位方法、装置、機器、記憶媒体及びコンピュータプログラム
KR1020227000811A KR20220018594A (ko) 2019-11-29 2020-11-20 지도 생성 방법, 포지셔닝 방법, 장치, 디바이스, 저장 매체 및 컴퓨터 프로그램

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911207451.1 2019-11-29
CN201911207451.1A CN112880693A (zh) 2019-11-29 2019-11-29 地图生成方法、定位方法、装置、设备及存储介质

Publications (1)

Publication Number Publication Date
WO2021104180A1 true WO2021104180A1 (zh) 2021-06-03

Family

ID=76039190

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/130627 WO2021104180A1 (zh) 2019-11-29 2020-11-20 地图生成方法、定位方法、装置、设备、存储介质及计算机程序

Country Status (4)

Country Link
JP (1) JP2022542712A (zh)
KR (1) KR20220018594A (zh)
CN (1) CN112880693A (zh)
WO (1) WO2021104180A1 (zh)

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113551679A (zh) * 2021-07-23 2021-10-26 杭州海康威视数字技术股份有限公司 一种示教过程中的地图信息构建方法、构建装置
CN113674358A (zh) * 2021-08-09 2021-11-19 浙江大华技术股份有限公司 一种雷视设备的标定方法、装置、计算设备及存储介质
CN113778102A (zh) * 2021-09-22 2021-12-10 重庆长安汽车股份有限公司 Avp全局路径规划系统、方法、车辆及存储介质
CN113822932A (zh) * 2021-08-30 2021-12-21 湖北亿咖通科技有限公司 设备定位方法、装置、非易失性存储介质及处理器
CN113990064A (zh) * 2021-09-14 2022-01-28 北京无限感测科技有限公司 一种交通数据采集方法、系统和计算机可读存储介质
CN114485610A (zh) * 2021-12-27 2022-05-13 广州极飞科技股份有限公司 矢量地图构建方法及装置、路径规划方法及装置
CN114526722A (zh) * 2021-12-31 2022-05-24 易图通科技（北京）有限公司 地图对齐处理方法、装置及可读存储介质
CN114705180A (zh) * 2022-06-06 2022-07-05 中汽创智科技有限公司 高精地图的数据修正方法、装置、设备及存储介质
CN114777799A (zh) * 2022-06-15 2022-07-22 中汽创智科技有限公司 一种路口高精地图生成方法、装置、电子设备及存储介质
CN114924680A (zh) * 2022-05-27 2022-08-19 北京斯年智驾科技有限公司 基于web前端可视化界面的港口无人驾驶地图生成方法
CN115223358A (zh) * 2022-06-01 2022-10-21 北京四维图新科技股份有限公司 一种交通信息动态更新方法、装置及车路协同系统
CN115346012A (zh) * 2022-08-19 2022-11-15 腾讯科技（深圳）有限公司 路口面的生成方法、装置、设备、存储介质和程序产品
CN115602041A (zh) * 2021-07-09 2023-01-13 华为技术有限公司（Cn） 信息生成方法和装置、信息使用方法和装置
TWI792474B (zh) * 2021-08-05 2023-02-11 和碩聯合科技股份有限公司 地圖建立方法與地圖建立系統
CN116977950A (zh) * 2023-09-22 2023-10-31 腾讯科技（深圳）有限公司 路口面的确定方法、装置、设备和计算机可读存储介质
CN117456496B (zh) * 2023-12-25 2024-03-22 合众新能源汽车股份有限公司 地图车道边界线的处理方法和系统

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114034295A (zh) * 2021-11-05 2022-02-11 阿波罗智能技术(北京)有限公司 高精地图生成方法、装置、电子设备、介质及程序产品
WO2023107002A2 (en) * 2021-12-09 2023-06-15 Grabtaxi Holdings Pte. Ltd. System and method for adaptively predicting a road segment attribute based on a graph indicative of relationship between a road segment and a detection
CN114440861B (zh) * 2022-04-07 2022-06-24 智道网联科技(北京)有限公司 交通综合杆的生成方法、装置及设备
CN114705204B (zh) * 2022-05-09 2024-01-30 浙江大学 一种基于道路基础设计资料的高精度地图生成方法
CN114994673B (zh) * 2022-08-04 2022-10-21 南京隼眼电子科技有限公司 用于雷达的道路地图生成方法、装置及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130279760A1 (en) * 2012-04-23 2013-10-24 Electronics And Telecommunications Research Institute Location correction apparatus and method
CN107451526A (zh) * 2017-06-09 2017-12-08 蔚来汽车有限公司 地图的构建及其应用
CN109143259A (zh) * 2018-08-20 2019-01-04 北京主线科技有限公司 面向港口无人集卡的高精地图制作方法
CN109141444A (zh) * 2018-08-28 2019-01-04 北京三快在线科技有限公司 定位方法、装置、存储介质及移动设备
CN110108288A (zh) * 2019-05-27 2019-08-09 北京史河科技有限公司 一种场景地图构建方法及装置、场景地图导航方法及装置
CN110501018A (zh) * 2019-08-13 2019-11-26 广东星舆科技有限公司 一种服务于高精度地图生产的交通标志牌信息采集方法

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104281746B (zh) * 2014-09-28 2018-04-17 同济大学 一种基于点云的交通安全道路特征图成图方法
CN108267747B (zh) * 2017-01-03 2021-10-22 中交宇科（北京）空间信息技术有限公司 基于激光点云的道路特征提取方法和装置
CN108700428B (zh) * 2017-03-01 2021-06-01 华为技术有限公司 电子地图中道路的绘制方法及装置
CN107928566A (zh) * 2017-12-01 2018-04-20 深圳市沃特沃德股份有限公司 视觉扫地机器人及障碍物检测方法
US11131550B2 (en) * 2018-03-29 2021-09-28 WeRide Corp. Method for generating road map for vehicle navigation and navigation device
CN108955702B (zh) * 2018-05-07 2021-09-07 西安交通大学 基于三维激光和gps惯性导航系统的车道级地图创建系统
CN109978946A (zh) * 2019-03-20 2019-07-05 安徽宽广科技有限公司 一种基于互联网的车辆定位分析系统
CN110379007B (zh) * 2019-07-25 2022-01-18 厦门大学 基于车载移动激光扫描点云的三维公路曲线重建方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130279760A1 (en) * 2012-04-23 2013-10-24 Electronics And Telecommunications Research Institute Location correction apparatus and method
CN107451526A (zh) * 2017-06-09 2017-12-08 蔚来汽车有限公司 地图的构建及其应用
CN109143259A (zh) * 2018-08-20 2019-01-04 北京主线科技有限公司 面向港口无人集卡的高精地图制作方法
CN109141444A (zh) * 2018-08-28 2019-01-04 北京三快在线科技有限公司 定位方法、装置、存储介质及移动设备
CN110108288A (zh) * 2019-05-27 2019-08-09 北京史河科技有限公司 一种场景地图构建方法及装置、场景地图导航方法及装置
CN110501018A (zh) * 2019-08-13 2019-11-26 广东星舆科技有限公司 一种服务于高精度地图生产的交通标志牌信息采集方法

Cited By (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115602041A (zh) * 2021-07-09 2023-01-13 华为技术有限公司（Cn） 信息生成方法和装置、信息使用方法和装置
CN115602041B (zh) * 2021-07-09 2024-04-09 华为技术有限公司 信息生成方法和装置、信息使用方法和装置
CN113551679A (zh) * 2021-07-23 2021-10-26 杭州海康威视数字技术股份有限公司 一种示教过程中的地图信息构建方法、构建装置
TWI792474B (zh) * 2021-08-05 2023-02-11 和碩聯合科技股份有限公司 地圖建立方法與地圖建立系統
CN113674358A (zh) * 2021-08-09 2021-11-19 浙江大华技术股份有限公司 一种雷视设备的标定方法、装置、计算设备及存储介质
CN113822932A (zh) * 2021-08-30 2021-12-21 湖北亿咖通科技有限公司 设备定位方法、装置、非易失性存储介质及处理器
CN113822932B (zh) * 2021-08-30 2023-08-18 亿咖通(湖北)技术有限公司 设备定位方法、装置、非易失性存储介质及处理器
CN113990064A (zh) * 2021-09-14 2022-01-28 北京无限感测科技有限公司 一种交通数据采集方法、系统和计算机可读存储介质
CN113778102A (zh) * 2021-09-22 2021-12-10 重庆长安汽车股份有限公司 Avp全局路径规划系统、方法、车辆及存储介质
CN113778102B (zh) * 2021-09-22 2023-05-12 重庆长安汽车股份有限公司 Avp全局路径规划系统、方法、车辆及存储介质
CN114485610A (zh) * 2021-12-27 2022-05-13 广州极飞科技股份有限公司 矢量地图构建方法及装置、路径规划方法及装置
CN114485610B (zh) * 2021-12-27 2023-07-18 广州极飞科技股份有限公司 矢量地图构建方法及装置、路径规划方法及装置
CN114526722A (zh) * 2021-12-31 2022-05-24 易图通科技（北京）有限公司 地图对齐处理方法、装置及可读存储介质
CN114924680A (zh) * 2022-05-27 2022-08-19 北京斯年智驾科技有限公司 基于web前端可视化界面的港口无人驾驶地图生成方法
CN114924680B (zh) * 2022-05-27 2023-08-01 北京斯年智驾科技有限公司 基于web前端可视化界面的港口无人驾驶地图生成方法
CN115223358A (zh) * 2022-06-01 2022-10-21 北京四维图新科技股份有限公司 一种交通信息动态更新方法、装置及车路协同系统
CN115223358B (zh) * 2022-06-01 2024-04-02 北京四维图新科技股份有限公司 一种交通信息动态更新方法、装置及车路协同系统
CN114705180A (zh) * 2022-06-06 2022-07-05 中汽创智科技有限公司 高精地图的数据修正方法、装置、设备及存储介质
CN114777799B (zh) * 2022-06-15 2022-10-18 中汽创智科技有限公司 一种路口高精地图生成方法、装置、电子设备及存储介质
CN114777799A (zh) * 2022-06-15 2022-07-22 中汽创智科技有限公司 一种路口高精地图生成方法、装置、电子设备及存储介质
CN115346012A (zh) * 2022-08-19 2022-11-15 腾讯科技（深圳）有限公司 路口面的生成方法、装置、设备、存储介质和程序产品
CN115346012B (zh) * 2022-08-19 2023-10-17 腾讯科技（深圳）有限公司 路口面的生成方法、装置、设备、存储介质和程序产品
CN116977950A (zh) * 2023-09-22 2023-10-31 腾讯科技（深圳）有限公司 路口面的确定方法、装置、设备和计算机可读存储介质
CN116977950B (zh) * 2023-09-22 2024-02-23 腾讯科技（深圳）有限公司 路口面的确定方法、装置、设备和计算机可读存储介质
CN117456496B (zh) * 2023-12-25 2024-03-22 合众新能源汽车股份有限公司 地图车道边界线的处理方法和系统

Also Published As

Publication number Publication date
JP2022542712A (ja) 2022-10-06
CN112880693A (zh) 2021-06-01
KR20220018594A (ko) 2022-02-15

Similar Documents

Publication Publication Date Title
WO2021104180A1 (zh) 地图生成方法、定位方法、装置、设备、存储介质及计算机程序
US11933627B2 (en) High-definition map building method and apparatus
US10628671B2 (en) Road modeling from overhead imagery
KR20200101855A (ko) 차량의 주행 궤적 예측 방법 및 장치
US20170343374A1 (en) Vehicle navigation method and apparatus
US20120065878A1 (en) Lane-based road transport information generation
JP6870475B2 (ja) 車線情報出力方法および車線情報出力装置
CN112325896B (zh) 导航方法、装置、智能行驶设备及存储介质
CN111076731A (zh) 自动驾驶高精定位与路径规划方法
CN110837092A (zh) 车辆定位、车道级路径规划的方法及装置
CN109491378A (zh) 自动驾驶车辆的基于道路分段的路线引导系统
US20140249748A1 (en) Route guidance at intersections
US20230266141A1 (en) Method and Apparatus for Automatically Marking U-Turn Lane Line, Computer-Readable Storage Medium, and Map
JP2003123197A (ja) 道路標示等認識装置
WO2022227878A1 (zh) 一种车道线标注方法及装置
JP2008020604A (ja) 道路地図作成プログラム、地図作成装置及び位置表示システム
WO2023179028A1 (zh) 一种图像处理方法、装置、设备及存储介质
US20230018996A1 (en) Method, device, and computer program for providing driving guide by using vehicle position information and signal light information
CN113008260A (zh) 一种导航信息处理方法、装置、电子设备及存储介质
CN114509065B (zh) 地图构建方法、系统、车辆终端、服务器端及存储介质
JP2018173762A (ja) 車線情報生成装置及び車線情報生成方法並びに車線情報生成用プログラム
JP2018173512A (ja) 車線情報生成装置及び車線情報生成方法並びに車線情報生成用プログラム
JP5435294B2 (ja) 画像処理装置及び画像処理プログラム
CN111238500B (zh) 道路地图区域的道路线段的地图生成方法、装置及系统
CN116952278A (zh) 一种去除感知地图抖动的仿真测试方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20893570

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 20227000811

Country of ref document: KR

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 2022506803

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20893570

Country of ref document: EP

Kind code of ref document: A1