CN112747756A - Map construction method and device - Google Patents
Map construction method and device Download PDFInfo
- Publication number
- CN112747756A CN112747756A CN201911060372.2A CN201911060372A CN112747756A CN 112747756 A CN112747756 A CN 112747756A CN 201911060372 A CN201911060372 A CN 201911060372A CN 112747756 A CN112747756 A CN 112747756A
- Authority
- CN
- China
- Prior art keywords
- road
- unit
- map
- point cloud
- sub
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
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
- 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
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
Abstract
Disclosed are a map construction method, apparatus, computer-readable storage medium and electronic device, the method comprising: splitting a road into at least one road unit; acquiring sub-point cloud data corresponding to each road unit; aiming at the sub-point cloud data corresponding to each road unit, constructing a corresponding Gaussian mixture map; and obtaining a highly-mixed Gaussian map based on the mixed Gaussian map corresponding to each road unit. According to the method and the device, the actual scene of the road is fully considered in the process of constructing the map, so that the map can be loaded according to the actual scene of the road in the positioning process, the positioning accuracy is effectively improved, and the calculation cost in the positioning process is reduced.
Description
Technical Field
The present application relates to the field of electronic map technologies, and in particular, to a map construction method and apparatus.
Background
During the driving process of the vehicle, it is usually necessary to acquire road surface images and load a map in real time to realize the positioning and navigation of the vehicle. In order to locate the position of the vehicle well and to navigate the vehicle, it is necessary to perform map construction in advance.
At present, when a map is constructed, generally, a method is adopted to fit the height distribution of the whole point cloud within a certain horizontal range, so that the obtained map has no problem for roads positioned on the same plane, but for some three-dimensional scenes (such as overpasses, underlying garages, multi-floor garages and the like), because the same plane coordinate may correspond to a plurality of roads, the actual scene of each road cannot be considered when the map is constructed by adopting the existing method, so that the obtained map information cannot be suitable for positioning of each road, and the positioning effect is not ideal.
Disclosure of Invention
The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides a map construction method and device, a computer readable storage medium and electronic equipment, which can greatly reduce the data volume of a semantic Gaussian mixture map, thereby reducing the storage space occupied by the map, and can load the map according to the actual scene of a road in the positioning process because the actual scene of the road is fully considered in the map construction process, thereby effectively improving the positioning accuracy and reducing the calculation cost in the positioning process.
According to a first aspect of the present application, there is provided a map construction method including:
splitting a road into at least one road unit;
acquiring sub-point cloud data corresponding to each road unit;
aiming at the sub-point cloud data corresponding to each road unit, constructing a corresponding Gaussian mixture map;
and obtaining a highly-mixed Gaussian map based on the mixed Gaussian map corresponding to each road unit.
According to a second aspect of the present application, there is provided a map construction apparatus including:
the road unit acquisition module is used for splitting a road into at least one road unit;
the sub-point cloud acquisition module is used for acquiring sub-point cloud data corresponding to each road unit;
the first component module is used for constructing a corresponding Gaussian mixture map according to the sub-point cloud data corresponding to each road unit;
and the second component module is used for obtaining a highly-mixed Gaussian map based on the mixed Gaussian map corresponding to each road unit.
According to a third aspect of the present application, there is provided a computer-readable storage medium storing a computer program for executing the map construction method of the first aspect described above.
According to a fourth aspect of the present application, there is provided an electronic apparatus comprising:
a processor;
a memory for storing the processor-executable instructions;
the processor is configured to read the executable instructions from the memory and execute the instructions to implement the map building method according to the first aspect.
Compared with the prior art, the map construction method, the map construction device, the computer readable storage medium and the electronic equipment provided by the application have the following beneficial effects that:
on one hand, in the map construction process, the actual scene of the road is fully considered, the road is divided into the road units according to the actual scene, and the highly-mixed Gaussian map is obtained based on the mixed Gaussian maps respectively established for the road units, so that the map can be loaded according to the actual scene of the road in the positioning process, the loading of useless map parts is avoided, the positioning accuracy can be effectively improved, and the calculation cost in the positioning process can be reduced.
On the other hand, the height distribution can be stored in a mode of storing the parameters of the Gaussian function, so that the data volume of the mixed Gaussian map can be greatly reduced, the storage space occupied by the map is reduced, the reduction of the occupied storage space is favorable for the instant transmission of the point cloud map, the occupied bandwidth is reduced, the transmission delay is reduced, and meanwhile, the calculation amount during analysis is reduced.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.
FIG. 1 is a schematic view of a multi-level road scene to which an exemplary embodiment of the present application is applicable.
Fig. 2 is a first flowchart illustrating a map building method according to an exemplary embodiment of the present application.
Fig. 3 is a schematic flowchart illustrating a road being split into at least one road unit in a map building method according to an exemplary embodiment of the present application.
Fig. 4 is a schematic diagram of acquiring sub-point cloud data corresponding to each road unit in a map construction method according to an exemplary embodiment of the present application.
Fig. 5 is a first flowchart illustrating a gaussian mixture map constructed in a map construction method according to an exemplary embodiment of the present application.
Fig. 6 is a schematic flow chart illustrating a process of allocating data points to a unit space in a mapping method according to an exemplary embodiment of the present application.
Fig. 7 is a schematic flowchart illustrating a second process of constructing a gaussian mixture map in a map construction method according to an exemplary embodiment of the present application.
Fig. 8 is a schematic diagram of a mixture gaussian distribution in a mapping method according to an exemplary embodiment of the present application.
Fig. 9 is a flowchart illustrating a map building method according to an exemplary embodiment of the present application.
Fig. 10 is a first flowchart illustrating a road unit height map corresponding to each road unit in a map building method according to an exemplary embodiment of the present application.
Fig. 11 is a schematic flowchart illustrating a second process of constructing a road unit height map corresponding to each road unit in a map construction method according to an exemplary embodiment of the present application.
Fig. 12 is a first schematic diagram of a map building apparatus according to an exemplary embodiment of the present application.
Fig. 13 is a schematic diagram of a road unit obtaining module in a map building apparatus according to an exemplary embodiment of the present application.
Fig. 14 is a schematic diagram of a first building module in a map building apparatus according to an exemplary embodiment of the present application.
Fig. 15 is a schematic diagram of a map building apparatus according to an exemplary embodiment of the present application.
Fig. 16 is a block diagram of an electronic device provided in an exemplary embodiment of the present application.
Detailed Description
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.
It should be noted that: the relative arrangement of the components and steps, the numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.
It will be understood by those of skill in the art that the terms "first," "second," and the like in the embodiments of the present disclosure are used merely to distinguish one element from another, and are not intended to imply any particular technical meaning, nor is the necessary logical order between them.
It is also understood that in embodiments of the present disclosure, "a plurality" may refer to two or more and "at least one" may refer to one, two or more.
It is also to be understood that any reference to any component, data, or structure in the embodiments of the disclosure, may be generally understood as one or more, unless explicitly defined otherwise or stated otherwise.
In addition, the term "and/or" in the present disclosure is only one kind of association relationship describing an associated object, and means that three kinds of relationships may exist, for example, a and/or B may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" in the present disclosure generally indicates that the former and latter associated objects are in an "or" relationship.
It should also be understood that the description of the various embodiments of the present disclosure emphasizes the differences between the various embodiments, and the same or similar parts may be referred to each other, so that the descriptions thereof are omitted for brevity.
Meanwhile, it should be understood that the sizes of the respective portions shown in the drawings are not drawn in an actual proportional relationship for the convenience of description.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses.
Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
The disclosed embodiments may be applied to electronic devices such as terminal devices, computer systems, servers, etc., which are operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well known terminal devices, computing systems, environments, and/or configurations that may be suitable for use with electronic devices, such as terminal devices, computer systems, servers, and the like, include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, hand-held or laptop devices, microprocessor-based systems, set top boxes, programmable consumer electronics, network pcs, minicomputer systems, mainframe computer systems, distributed cloud computing environments that include any of the above systems, and the like.
Electronic devices such as terminal devices, computer systems, servers, etc. may be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, data structures, etc. that perform particular tasks or implement particular abstract data types. The computer system/server may be practiced in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computer system storage media including memory storage devices.
Summary of the application
In general, a map is required to be opened for navigation in a vehicle running process, and map construction is required in advance for good positioning of a vehicle position and navigation of the vehicle. At present, when a map is constructed, a method of fitting the height distribution of the whole point cloud within a certain horizontal range is generally adopted. However, the map obtained in this way is only suitable for roads located on the same plane, but is not suitable for some three-dimensional scenes because for some three-dimensional scenes, the same plane coordinate may correspond to multiple roads in the height direction, and the actual scene of each road cannot be considered when the map is constructed by the existing method, so that the obtained map information cannot be suitable for the positioning of each road.
As shown in fig. 1, when a map is constructed by using the conventional method, two roads a and B with different heights and perpendicular to each other are fitted according to the height distribution of the entire point cloud within a preset horizontal range, and the same map information is used for the road a and the road B during positioning. However, when the vehicle travels on the road B, the lidar on the vehicle does not scan an object below the road B, and therefore only the map information related to the road B needs to be loaded, while the map information of the road a and the map information of the road B are loaded on the existing map at the same time, and the actual scenes of the two roads cannot be considered, so that not only is the risk of positioning error increased, but also the calculation cost in the positioning process is increased.
The embodiment provides a completely different map construction method, the actual scene of the road is fully considered when the map is constructed, the road is split into a plurality of road units according to the actual scene, and the highly-mixed gaussian map is obtained based on the mixed gaussian maps respectively established for the plurality of road units, so that the map can be loaded according to the actual scene of the road in the positioning process, the positioning accuracy is effectively improved, and the calculation cost in the positioning process is reduced.
Exemplary method
Fig. 2 is a flowchart illustrating a mapping method according to an exemplary embodiment of the present application. The embodiment can be applied to electronic equipment, and particularly can be applied to a server or a general computer.
As shown in fig. 2, a map construction method provided in an exemplary embodiment of the present application includes the following steps:
step 10: the road is split into at least one road unit.
The road can be regarded as one or more road units connected in a certain sequence according to the actual scene. For example, the road includes at least two branches, and the branches are also provided with branches, at this time, the road can be divided into a plurality of road units according to the branch situation of the road, and the road can be regarded as being formed by splicing the plurality of road units.
Step 20: and acquiring sub-point cloud data corresponding to each road unit.
The manner of acquiring the point cloud data may be selected as needed, and for example, the point cloud data may be acquired by a laser radar or a camera mounted on a vehicle. After each road unit is obtained, sub-point cloud data corresponding to each road unit can be obtained from the point cloud data, and therefore classification of the sub-point cloud data is achieved. It is understood that the sub-point cloud data herein includes road surface point cloud data corresponding to road units, and also includes other point cloud data. In this embodiment, the data point coordinates in each sub-point cloud data may be denoted as (X, Y, Z), where X and Y are plane coordinates and Z is a height coordinate.
Step 30: and constructing a corresponding Gaussian mixture map according to the sub-point cloud data corresponding to each road unit.
A Gaussian Mixture Model (abbreviated GMM) is a method of fitting a set of data using one or more Gaussian functions. In this embodiment, for the sub-point cloud data corresponding to each road unit, at least one gaussian function may be adopted to fit the heights of the data points in the sub-point cloud data, so that the height distribution of the data points may be represented in the form of a gaussian function, and the data storage amount is greatly reduced.
Step 40: and obtaining a highly-mixed Gaussian map based on the mixed Gaussian map corresponding to each road unit.
After the gaussian mixture maps corresponding to the sub-point cloud data are obtained, the gaussian mixture maps of the sub-point cloud data need to be merged to form a highly mixed gaussian map. When the height mixed Gaussian map is stored, the height distribution of each data point is not separately stored in a coordinate mode, but the mutual relation is established through a Gaussian function, and the height distribution is stored in a mode of storing parameters of the Gaussian function, so that the whole data storage amount can be reduced.
The map construction method provided by the embodiment has the beneficial effects that:
on one hand, in the embodiment, the actual scene of the road is fully considered when the map is constructed, the road is split into the road units according to the actual scene, and the highly-mixed gaussian map is obtained based on the mixed gaussian maps respectively established for the road units, so that the map can be loaded according to the actual scene of the road in the positioning process, the loading of useless map parts is avoided, the positioning accuracy can be effectively improved, and the calculation cost in the positioning process can be reduced.
On the other hand, the actual scene of the road is fully considered in the process of constructing the map, and the height distribution can be stored in a mode of storing parameters of the Gaussian function, so that the data volume of the semantic Gaussian mixture map can be greatly reduced, the storage space occupied by the map is reduced, the occupied storage space is reduced, the instant transmission of the point cloud map is facilitated, the bandwidth occupation is reduced, the transmission delay is reduced, and meanwhile, the calculation amount in the analysis process is also facilitated to be reduced.
Fig. 3 shows a schematic flowchart of the step of splitting the road into at least one road unit in the embodiment shown in fig. 2.
As shown in fig. 3, on the basis of the embodiment shown in fig. 2, in an exemplary embodiment of the present application, the step 10 of splitting the road may specifically include:
step 101: and acquiring a road in a track formed by the point cloud data and a running track of a carrier on the road.
In the present embodiment, the vehicle may be any movable object, for example, a vehicle traveling on a road. During the driving process of the vehicle, a plurality of images can be continuously obtained through a camera or a laser radar arranged on the vehicle, so that corresponding point cloud data can be obtained, the point cloud data comprises point clouds corresponding to roads, and the road point clouds in the continuous images can form the roads in the running process of the vehicle and the running tracks of the vehicle in the roads by processing the continuous point cloud data.
Step 102: according to the running track of the carrier, the road is split into at least one road unit, and topological information among the road units is obtained.
The pose of the vehicle in the running process can be acquired in real time, and the pose of the vehicle can be changed along with the difference of the running state of the vehicle. For example, the pose of the vehicle is adjusted along with the difference of the road, so that the pose of the vehicle is correspondingly acquired in the process of acquiring the running track of the vehicle, and the pose of the vehicle is associated with the road, so that the road can be split according to whether the pose of the vehicle changes or not.
Referring to fig. 4, for example, in the obtained vehicle running track diagram, the pose of the vehicle can be represented by arrows, and the pose of the vehicle is different if the direction of the arrows is different (as shown in fig. 4. a). Therefore, the road can be split into 6 road units according to the vehicle pose, and each road unit is numbered, each road unit corresponds to one rectangular frame, and the vehicle pose in each rectangular frame is the same (as shown in fig. 4. b). After the splitting is completed, the point cloud data corresponding to the road unit may be split to obtain sub-point cloud data corresponding to each road unit (as shown in fig. 4. c). Of course, in the process of splitting the road units, the topological information between the road units also needs to be acquired, so that the road units can be spliced according to the topological relation between the road units when the highly-mixed gaussian map is acquired in the following process. For example, the road units 1, 3, 4, 5 and 6 are all connected only to the road unit 2, but not to each other (as shown in fig. 4. d).
According to the embodiment, the road and the running track of the vehicle on the road in the point cloud data are obtained, and the road is split through the association between the pose of the vehicle and the road, so that the road can be effectively split, the topological relation among road units can be obtained, and the road units can be spliced when a highly-mixed Gaussian map is obtained subsequently.
Fig. 5 is a schematic flow chart illustrating a step of constructing a corresponding gaussian mixture map for the sub-point cloud data corresponding to each road unit in the embodiment shown in fig. 2.
As shown in fig. 5, on the basis of the embodiment shown in fig. 2, in an exemplary embodiment of the present application, the step of constructing the gaussian mixture map shown in step 30 may specifically include:
step 301: and aiming at the sub-point cloud data of each road unit, distributing the data points of the sub-point cloud data to at least one unit space according to a preset mode.
In this embodiment, the size of the unit space may be set as required, for example, a plane space (a plane perpendicular to the height of the data point) where the sub-point cloud data in one road unit is located may be regarded as one unit space, and all the data points are located in the unit space. For another example, the plane space in which the sub-point cloud data in a road unit is located may be divided into a plurality of unit spaces (each may be g)a、gb、gc、gdEtc.), and then load the data points so that the data points fall into each unit space, enabling assignment of the data points.
Step 302: and performing Gaussian fitting on the height of the data points in each unit space to obtain a mixed Gaussian distribution of the data points in the unit space.
In performing gaussian fitting, the heights of all data points therein are fitted per unit space to be expressed by a unified gaussian function expression. Referring to fig. 7, for example, in one embodiment, the number of gaussian functions is one for a unit space, and the height distribution of all data points in the unit space can be represented by only one gaussian distribution. As another example, in one embodiment, the number of gaussian functions is two for a unit space, and a mixed gaussian GMM can be formed by weighted superposition of two gaussian functions, each of which is a bell-shaped curve.
Each gaussian distribution contains three important parameters, namely, a mean (mean), a variance (variance) and a weight (weight), wherein the mean and the variance are basic parameters of the gaussian distribution, and the weight is the weight of the gaussian distribution, and when the number of gaussian functions is 1, the corresponding weight of the gaussian distribution is 1; when the number of Gaussian functions is two, the weight of each Gaussian distribution ranges from 0 to 1, and the sum of the weights of the two Gaussian distributions is 1 (for example, one is 0.3, and the other is 0.7). Of course, the number of the gaussian functions may be two or more, and is determined according to actual fitting.
Step 303: and obtaining a Gaussian mixture map corresponding to the sub-point cloud data of each road unit based on the Gaussian mixture distribution of each unit space.
After step 302, the height distribution of the data points in each unit space can be represented by the gaussian mixture distribution, and the gaussian mixture map corresponding to each sub-point cloud data can be obtained by merging the height distributions of all the data points in each sub-point cloud data. Referring to FIG. 8, for example, the plane space is divided into 4 unit spaces (denoted as g respectively)a、gb、gcAnd gd) In the case of (2), each unit space corresponds to a mixture gaussian distribution (for example: gaThe coordinate of the unit space is (1,1) and corresponds to GMM1 with Gaussian mixture distribution; gbThe coordinate of unit space is (1,2), corresponding to Gaussian mixture distributionGMM2；gcThe coordinate of the unit space is (2,1), and the unit space corresponds to GMM3 with Gaussian mixture distribution; gdThe coordinates of the unit space are (2,2) and correspond to the mixed Gaussian distribution GMM4, and then the Gaussian distributions are combined to form the mixed Gaussian map corresponding to the sub-point cloud data.
In the mixed Gaussian map obtained by the embodiment, the height data of each data point does not need to be stored, and the height distribution of all the data points in a unit space can be represented only by one mixed Gaussian distribution, so that the data quantity stored in the map is greatly reduced, and the storage space occupied by the map is reduced.
FIG. 6 shows a flow chart of the step of assigning data points to unit spaces in the embodiment shown in FIG. 5.
As shown in fig. 6, on the basis of the embodiment shown in fig. 5, in an exemplary embodiment of the present application, step 301 may specifically include:
step 3011: and dividing the map space into at least one unit space according to a preset unit size in a plane perpendicular to the height of the data point.
In this embodiment, the coordinates of each data point are (X, Y, Z), where the Z direction is the height direction of the data point, and the XY plane is a plane perpendicular to the height direction of the data point. And dividing the plane space according to a preset unit size along the XY direction. The number and size of the unit spaces after the plane space division are different according to the unit size. For example, if the unit size in both the X direction and the Y direction is set to 20 centimeters (cm), the size of the divided unit space is 20cm by 20 cm. Of course, in other embodiments, the unit space may have other values, and is not limited to the above. The unit sizes divided in the X direction and the Y direction may be the same or different, for example, the unit size in the X direction may be set to 20cm, the unit size in the Y direction may be set to 10cm, and the size of the resulting unit space is 20cm × 10 cm. The unit space obtained by the above division is rectangular, and in other embodiments, the unit space may be divided in other manners, and the unit space may also have other shapes, for example, a triangle, a hexagon, and the like, which is not limited herein.
After dividing the unit spaces, the position coordinates of each unit space may be marked. For example, the coordinates of a certain point in the unit space may be selected as the position coordinates of the unit space, and the point may be the center point of the unit space, a certain point on one side, or a certain vertex. In the present embodiment, it is preferable that the coordinates (X, Y) of the center point of the unit space be the position coordinates of the unit space.
Step 3012: and loading corresponding sub-point cloud data in a plurality of unit spaces. One data point or a plurality of data points can be distributed in each unit space according to the actual situation.
In this embodiment, the plane space is divided, so that at least one unit space can be obtained, and a data point is correspondingly loaded in each unit space, which is beneficial to performing subsequent gaussian fitting on the height of the data point in each unit space.
Further, in this embodiment, after the gaussian mixture map corresponding to each sub-point cloud data is obtained, the gaussian mixture map needs to be stored. According to different storage modes, the occupied space size of the mixed Gaussian map is different when the mixed Gaussian map is stored. The following description will be given taking an example in which 30 data points are associated with each unit space.
Currently, in point cloud data storage, each data point has X, Y, Z three coordinate parameters, each parameter occupies 4 bytes, and therefore, the space occupied by storing 30 data points is 4 × 3 × 30 — 360 bytes.
In one embodiment, when storing the gaussian mixture map, the height of a data point in a unit space is represented by two gaussian functions, each gaussian function has three parameters (mean, variance and weight), each parameter occupies 4 bytes, and therefore the space occupied by the gaussian function is: 4 × 3 × 2 ═ 24 bytes. The data point in the unit space is represented by a position coordinate (e.g., X, Y coordinate of the center point) of the unit space, each coordinate parameter occupies 4 bytes, and thus the coordinate occupation space of the data point is: 4 × 2 ═ 8 bytes. Therefore, 30 data points occupy 24+ 8-32 bytes when stored. When the mixed Gaussian map is stored in the mode, the height data of each data point is not required to be stored, the height distribution of all the data points in a unit space can be represented only by one mixed Gaussian distribution, and the data points in each unit space can be represented by only one position coordinate, so that the data quantity stored in the map is greatly reduced, and the storage space occupied by the map is reduced.
In this embodiment, when storing the gaussian mixture map, it is preferable to store the position coordinates of the unit space and the gaussian distribution information of the data points in the unit space. It should be understood that semantic categories of data points are also stored in the gaussian mixture map. Therefore, for the convenience of description, the gaussian mixture map includes a plurality of elements, each element including the following information: x, Y coordinates of the center point of the unit space, gaussian distribution information of the unit space (including mean, variance, and weight of each gaussian function), semantic category.
Further, after the sub-point cloud data corresponding to each road unit is obtained, on one hand, a corresponding gaussian mixture map can be constructed for the sub-point cloud data corresponding to each road unit, so as to construct a highly gaussian mixture map; on the other hand, a road surface height map can be constructed, and the map is beneficial to positioning by adopting the map subsequently.
Fig. 9 shows a schematic flow chart of the step of constructing the road height map after step 20, which may specifically include:
step 50: and constructing a road unit height map corresponding to each road unit based on the sub-point cloud data of each road unit.
After obtaining the sub-point cloud data, semantic segmentation may be performed on each data point in the sub-point cloud data to obtain semantic point cloud data. For example, an image obtained by the laser radar includes a vehicle, a pedestrian, a road, a zebra crossing and the like, at this time, the semantic types at least include the vehicle, the pedestrian, the road and the zebra crossing, each data point has a corresponding semantic type by labeling the semantic type of each data point in the point cloud data, and at this time, the data points having the semantic types constitute the semantic point cloud data. After the semantic point cloud data is obtained, the point cloud data with the semantic type as the road surface needs to be screened out from the semantic point cloud data so as to obtain the road surface point cloud data in the semantic point cloud data.
For a road map in a multi-layer road scene, different road units may be distributed at different heights, so that a height map of the road units can be correspondingly constructed according to the height distribution. In this embodiment, after obtaining the road surface point cloud data in the sub-point cloud data of each road unit, the height value of each data point may be obtained, so that a corresponding road unit height map may be constructed according to the height values of the data points.
Step 60: and merging the road unit height maps corresponding to each road unit to obtain a road surface height map.
After obtaining each road unit height map, merging the road unit height maps corresponding to each sub-point cloud data to form a road surface height map. During subsequent positioning, the road information with the corresponding height can be obtained by positioning in the road surface height map according to the height value of the acquired point cloud data, other irrelevant road information is not required to be obtained, the positioning accuracy is higher, and the data transmission amount and the calculation amount in the positioning process are smaller.
Fig. 10 is a flow chart illustrating a step of constructing a road unit height map corresponding to each road unit in the embodiment shown in fig. 9.
As shown in fig. 10 and fig. 11, based on the embodiment shown in fig. 9, in an exemplary embodiment of the present application, step 50 may specifically include:
step 501: and aiming at the road surface point cloud data in the sub-point cloud data of each road unit, distributing the data points of the road surface point cloud data to a plurality of unit spaces according to a preset mode.
In this embodiment, the size of the unit space may be set as required, for example, a plane space (a plane perpendicular to the height of the data point) where the road surface point cloud data of one road unit is located may be regarded as one unit space, and all the data points are located in the unit space. For another example, the plane space where the point cloud data of the road surface of one road unit is located may be divided into a plurality of unit spaces, and then the data points are loaded so that the data points fall into each unit space, thereby realizing the distribution of the data points.
After the unit spaces are set, the position coordinates of each unit space may be marked. For example, the coordinates of a certain point in the unit space may be selected as the position coordinates of the unit space, and the point may be the center point of the unit space, a certain point on one side, or a certain vertex. In the present embodiment, it is preferable that the coordinates (X, Y) of the center point of the unit space be the position coordinates of the unit space.
Step 502: and acquiring the height average value of all data points in each unit space to obtain the height value of the unit space.
When only one data point exists in the unit space, the height value of the data point is the height value of the unit space; when there are more than two data points in the unit space, the height value of the unit space is the average value of the height values of the data points. Of course, in other embodiments, the height average of the data points may be obtained in other ways, and is not limited to the above.
Step 503: and combining the height values of each unit space to obtain a road unit height map corresponding to each road unit. Here, the height value of each unit space corresponds to its position coordinates (X, Y).
In this embodiment, after the road unit height map corresponding to the road point cloud data in each piece of sub-point cloud data is obtained, the road unit height map needs to be stored. According to different storage modes, the occupied space size of the mixed Gaussian map is different when the mixed Gaussian map is stored. The following description will be given taking an example in which 30 data points are associated with each unit space.
Currently, in point cloud data storage, each data point has X, Y, Z three coordinate parameters, each parameter occupies 4 bytes, and therefore, the space occupied by storing 30 data points is 4 × 3 × 30 — 360 bytes.
In one embodiment, when the road unit height map is stored, the height of the data point in the unit space is represented by one coordinate parameter (i.e. height average value), and each coordinate parameter occupies 4 bytes, so the space occupied by the coordinate parameter in the height direction is: 4 × 1 ═ 4 bytes. The data point in the unit space is represented by a position coordinate (e.g., X, Y coordinate of the center point) of the unit space, each coordinate parameter occupies 4 bytes, and thus the coordinate occupation space of the data point is: 4 × 2 ═ 8 bytes. Therefore, 30 data points occupy 4+ 8-12 bytes of space when stored. When the road unit height map is stored in the mode, the data amount of map storage is greatly reduced, and therefore the storage space occupied by the map is reduced.
After the highly-mixed gaussian map is constructed, the highly-mixed gaussian map can be used for positioning. For example, in the driving process of an automobile, three-dimensional point cloud data needs to be acquired in real time, and positioning is carried out according to the acquired point cloud data and a constructed semantic Gaussian mixture map. Therefore, after the step 40, the method may further include:
step 70: and determining the matching degree of the real-time point cloud data and the highly-mixed Gaussian map.
The real-time point cloud data may be obtained in various ways, for example, by a laser radar or a camera mounted on a vehicle, where the vehicle may be an automobile or other equipment, and is not limited herein. In the moving process of the carrier, real-time point cloud data can be obtained through a laser radar according to a preset frequency (for example, 10Hz), the real-time point cloud data comprises a plurality of data points (of course, according to actual conditions, only 1 data point can be included), the data points are matched with the highly-mixed Gaussian map, and the current position of the carrier is determined according to the matching degree, so that the carrier is positioned.
The highly-mixed Gaussian map constructed by the embodiment can be stored in the cloud, and the highly-mixed Gaussian map can be acquired in real time when the carrier is positioned, so that positioning is realized. It should be understood that, in the positioning process, the height position of the vehicle can be determined according to the road surface height map, so as to determine the height mixed gaussian map at the height position to be loaded; meanwhile, when the highly-mixed Gaussian map is loaded, other road units connected with the current road unit can be loaded according to the topological relation among the road units, and the road units disconnected with the current road unit are not loaded, so that the introduction of upper-layer or lower-layer roads which are completely irrelevant to the current driving route is avoided, and the stability and the positioning precision of real-time positioning are improved.
Exemplary devices
Fig. 12 is a schematic diagram of a map building apparatus provided in an exemplary embodiment of the present application, and includes a road unit obtaining module 81, a sub-point cloud obtaining module 82, a first component module 83, and a second component module 84. The road unit obtaining module 81 is configured to split a road into at least one road unit; the sub-point cloud obtaining module 82 is configured to obtain sub-point cloud data corresponding to each road unit; the first component module 83 is configured to construct a corresponding gaussian mixture map for the sub-point cloud data corresponding to each road unit; the second component module 84 is configured to obtain a highly-mixed gaussian map based on the mixed gaussian map corresponding to each road unit.
Further, referring to fig. 13, the road unit obtaining module 81 includes a track obtaining unit 811 and a splitting unit 812. The track acquiring unit 811 is configured to acquire a road in a track formed by the point cloud data and a running track of a vehicle on the road; the splitting unit 812 is configured to split the road into at least one road unit according to the running track of the vehicle, and acquire topology information between the road units.
Further, referring to fig. 14, the first component module 83 includes a data distribution unit 831, a first obtaining unit 832 and a second obtaining unit 833. The data distribution unit 831 is configured to, for the sub-point cloud data of each road unit, distribute data points of the sub-point cloud data to at least one unit space in a preset manner; the first obtaining unit 832 is configured to perform gaussian fitting on the height of each data point in the unit space to obtain a gaussian mixture distribution of the data points in the unit space; the second obtaining unit 833 is configured to obtain a gaussian mixture map corresponding to the sub-point cloud data of each road unit based on the gaussian mixture distribution of each unit space.
Further, referring to fig. 15, the map building apparatus further includes a unit height map building module 85 and a road height map obtaining module 86. The unit height map building module 85 is configured to build a road unit height map corresponding to each road unit based on the sub-point cloud data of each road unit; the road surface height map obtaining module 86 is configured to combine the road unit height maps corresponding to each road unit to obtain a road surface height map.
Further, the map building device further comprises a matching module 87, and the matching module 87 is used for determining the matching degree of the real-time point cloud data and the height mixed Gaussian map.
Exemplary electronic device
Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 16. FIG. 16 illustrates a block diagram of an electronic device in accordance with an embodiment of the present application.
As shown in fig. 16, the electronic device 90 includes one or more processors 91 and a memory 92.
The processor 91 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 90 to perform desired functions.
Memory 92 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, Random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, Read Only Memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium and executed by the processor 91 to implement the mapping method of the various embodiments of the application described above and/or other desired functions.
In one example, the electronic device 90 may further include: an input device 93 and an output device 94, which are interconnected by a bus system and/or other form of connection mechanism (not shown).
The input device 93 may also include, for example, a keyboard, a mouse, and the like. The output device 94 can output various information to the outside. The output devices 94 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices.
Of course, for the sake of simplicity, only some of the components related to the present application in the electronic device 90 are shown in fig. 16, and components such as a bus, an input/output interface, and the like are omitted. In addition, the electronic device 90 may include any other suitable components, depending on the particular application.
Exemplary computer program product and computer-readable storage Medium
In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in a mapping method according to various embodiments of the present application described in the "exemplary methods" section of this specification, supra.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform steps in a map construction method according to various embodiments of the present application described in the "exemplary methods" section above of this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.
The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.
Claims (10)
1. A map construction method, comprising:
splitting a road into at least one road unit;
acquiring sub-point cloud data corresponding to each road unit;
aiming at the sub-point cloud data corresponding to each road unit, constructing a corresponding Gaussian mixture map;
and obtaining a highly-mixed Gaussian map based on the mixed Gaussian map corresponding to each road unit.
2. The method of claim 1, wherein the splitting the road into at least one road unit comprises:
acquiring a road in a track formed by point cloud data and a running track of a carrier on the road;
according to the running track of the carrier, the road is split into at least one road unit, and topological information among the road units is obtained.
3. The method of claim 1, wherein the constructing a corresponding Gaussian mixture map for the sub-point cloud data corresponding to each of the road units comprises:
for the sub-point cloud data of each road unit, distributing data points of the sub-point cloud data to at least one unit space according to a preset mode;
performing Gaussian fitting on the height of the data points in each unit space to obtain a Gaussian mixture distribution of the data points in the unit space;
and obtaining a Gaussian mixture map corresponding to the sub-point cloud data of each road unit based on the Gaussian mixture distribution of each unit space.
4. The method of claim 3, wherein the assigning data points of the sub-point cloud data to at least one unit space in a preset manner for the sub-point cloud data of each road comprises:
dividing the map space into at least one unit space according to a preset unit size in a plane vertical to the height of the data point;
and loading corresponding sub-point cloud data in a plurality of unit spaces.
5. The method of claim 4, wherein the Gaussian mixture map comprises coordinates corresponding to a central point of each unit space, Gaussian distribution information of data points in the unit space, and semantic types.
6. The method according to any one of claims 1-5, wherein after the step of obtaining the sub-point cloud data corresponding to each road unit, the method further comprises:
constructing a road unit height map corresponding to each road unit based on the sub-point cloud data of each road unit;
and merging the road unit height maps corresponding to each road unit to obtain a road surface height map.
7. The method of claim 6, wherein said constructing a road unit height map for each said road unit based on said sub-point cloud data for each said road unit comprises:
aiming at the sub-point cloud data of each road unit, distributing data points of the sub-point cloud data to a plurality of unit spaces according to a preset mode;
acquiring the height average value of all data points in each unit space to obtain the height value of the unit space;
and combining the height values of each unit space to obtain a road unit height map corresponding to the sub-point cloud data of each road unit.
8. A map building apparatus comprising:
the road unit acquisition module is used for splitting a road into at least one road unit;
the sub-point cloud acquisition module is used for acquiring sub-point cloud data corresponding to each road unit;
the first component module is used for constructing a corresponding Gaussian mixture map according to the sub-point cloud data corresponding to each road unit;
and the second component module is used for obtaining a highly-mixed Gaussian map based on the mixed Gaussian map corresponding to each road unit.
9. A computer-readable storage medium storing a computer program for executing the map construction method according to any one of claims 1 to 7.
10. An electronic device, the electronic device comprising:
a processor;
a memory for storing the processor-executable instructions;
the processor is configured to read the executable instructions from the memory and execute the instructions to implement the mapping method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060372.2A CN112747756B (en) | 2019-10-30 | 2019-10-30 | Map construction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060372.2A CN112747756B (en) | 2019-10-30 | 2019-10-30 | Map construction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112747756A true CN112747756A (en) | 2021-05-04 |
| CN112747756B CN112747756B (en) | 2023-04-07 |
Family
ID=75644979
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911060372.2A Active CN112747756B (en) | 2019-10-30 | 2019-10-30 | Map construction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112747756B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101122465A (en) * | 2006-08-11 | 2008-02-13 | 株式会社电装 | Navigation apparatus |
| CN101685017A (en) * | 2008-09-27 | 2010-03-31 | 阿尔派株式会社 | Navigation apparatus and display method thereof |
| WO2011023247A1 (en) * | 2009-08-25 | 2011-03-03 | Tele Atlas B.V. | Generating raster image representing road existence probability based on probe measurements |
| CN102782452A (en) * | 2010-05-24 | 2012-11-14 | 三菱电机株式会社 | Navigation system |
| CN104916133A (en) * | 2015-06-09 | 2015-09-16 | 福建工程学院 | Road altitude information extraction method and system based on traffic track data |
| CN108765487A (en) * | 2018-06-04 | 2018-11-06 | 百度在线网络技术（北京）有限公司 | Rebuild method, apparatus, equipment and the computer readable storage medium of three-dimensional scenic |
| CN109186625A (en) * | 2018-10-24 | 2019-01-11 | 北京奥特贝睿科技有限公司 | Intelligent vehicle carries out pinpoint method and system using mixing sampling filter |
| CN110226186A (en) * | 2016-12-26 | 2019-09-10 | 宝马股份公司 | A kind of method and apparatus for indicating map element and vehicle/robot method and apparatus are positioned based on this |
| CN110221616A (en) * | 2019-06-25 | 2019-09-10 | 清华大学苏州汽车研究院（吴江） | A kind of method, apparatus, equipment and medium that map generates |
-
2019
- 2019-10-30 CN CN201911060372.2A patent/CN112747756B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101122465A (en) * | 2006-08-11 | 2008-02-13 | 株式会社电装 | Navigation apparatus |
| CN101685017A (en) * | 2008-09-27 | 2010-03-31 | 阿尔派株式会社 | Navigation apparatus and display method thereof |
| WO2011023247A1 (en) * | 2009-08-25 | 2011-03-03 | Tele Atlas B.V. | Generating raster image representing road existence probability based on probe measurements |
| CN102782452A (en) * | 2010-05-24 | 2012-11-14 | 三菱电机株式会社 | Navigation system |
| CN104916133A (en) * | 2015-06-09 | 2015-09-16 | 福建工程学院 | Road altitude information extraction method and system based on traffic track data |
| CN110226186A (en) * | 2016-12-26 | 2019-09-10 | 宝马股份公司 | A kind of method and apparatus for indicating map element and vehicle/robot method and apparatus are positioned based on this |
| CN108765487A (en) * | 2018-06-04 | 2018-11-06 | 百度在线网络技术（北京）有限公司 | Rebuild method, apparatus, equipment and the computer readable storage medium of three-dimensional scenic |
| CN109186625A (en) * | 2018-10-24 | 2019-01-11 | 北京奥特贝睿科技有限公司 | Intelligent vehicle carries out pinpoint method and system using mixing sampling filter |
| CN110221616A (en) * | 2019-06-25 | 2019-09-10 | 清华大学苏州汽车研究院（吴江） | A kind of method, apparatus, equipment and medium that map generates |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112747756B (en) | 2023-04-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111694903A (en) | Map construction method, map construction device, map construction equipment and readable storage medium | |
| CN104683405B (en) | The method and apparatus of cluster server distribution map matching task in car networking | |
| CN110378175B (en) | Method and device for recognizing road edge | |
| US20180293687A1 (en) | Ridesharing management for autonomous vehicles | |
| US20210381846A1 (en) | Methods and apparatuses for navigation guidance and establishing a three-dimensional real scene model, device and medium | |
| JP2023545423A (en) | Point cloud segmentation method, device, equipment and storage medium | |
| CN114764778A (en) | Target detection method, target detection model training method and related equipment | |
| KR102567632B1 (en) | Automatically pair GPS data to the planned driving route of a moving object | |
| CN114219276A (en) | Warehouse logistics simulation method, device, equipment and storage medium | |
| CN111383355A (en) | Three-dimensional point cloud completion method and device and computer readable storage medium | |
| CN110766793B (en) | Map construction method and device based on semantic point cloud | |
| CN112747756B (en) | Map construction method and device | |
| CN113111787A (en) | Target detection method, device, equipment and storage medium | |
| US11823039B2 (en) | Safe and fast exploration for reinforcement learning using constrained action manifolds | |
| CN113804196B (en) | Unmanned vehicle path planning method and related equipment | |
| CN115062240A (en) | Parking lot sorting method and device, electronic equipment and storage medium | |
| CN113269827B (en) | Method and computing device for realizing automatic calibration | |
| CN113516013A (en) | Target detection method and device, electronic equipment, road side equipment and cloud control platform | |
| US10743142B2 (en) | Constructing a map of a physical space | |
| CN113804208A (en) | Unmanned vehicle path optimization method and related equipment | |
| CN115294234B (en) | Image generation method and device, electronic equipment and storage medium | |
| US20220405999A1 (en) | Uv map using weight painting | |
| CN113479198B (en) | Unmanned vehicle control method and device | |
| CN113806862B (en) | Unmanned vehicle simulation method and device, storage medium and electronic equipment | |
| US20240119615A1 (en) | Tracking three-dimensional geometric shapes |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |