CN111026825B - Method and device for determining roundabout set - Google Patents
Method and device for determining roundabout set Download PDFInfo
- Publication number
- CN111026825B CN111026825B CN201911211602.0A CN201911211602A CN111026825B CN 111026825 B CN111026825 B CN 111026825B CN 201911211602 A CN201911211602 A CN 201911211602A CN 111026825 B CN111026825 B CN 111026825B
- Authority
- CN
- China
- Prior art keywords
- roundabout
- candidate
- road
- target
- determining
- 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.)
- Active
Links
Images
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Remote Sensing (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The application provides a method and a device for determining a rotary island set, and relates to the technical field of navigation. Firstly, a road network topological structure is obtained, wherein the road network topological structure comprises a plurality of road nodes, then road network data are constructed according to the road nodes, then a first candidate roundabout set is determined according to the road network data, the first candidate roundabout set is screened according to a preset rule to determine a second candidate roundabout set, then the second candidate roundabout set is screened by a machine learning mode to determine a third candidate roundabout set, the third candidate roundabout set is screened by a deep learning mode to determine a fourth candidate roundabout set, and finally the fourth candidate roundabout set is verified to determine a target roundabout set. The method and the device for determining the rotary island set have the advantages of being high in speed and high in precision.
Description
Technical Field
The application relates to the technical field of navigation, in particular to a method and a device for determining a rotary island set.
Background
At present, with the development of the automobile industry, the navigation technology is gradually improved.
The roundabout is also called ring traffic, is a special form of traffic node, and belongs to plane road intersection. The ring-shaped crossed section is also commonly called a rotary island, a rotary disc and the like. In order to accurately navigate for a user, the identification of the roundabout is crucial.
At present, a deep learning method commonly adopted for identifying the roundabout can achieve overall accuracy in identification capacity, but cannot achieve accurate identification of local details, and is low in identification speed.
Disclosure of Invention
The application aims to provide a method and a device for determining a roundabout set, so as to solve the problems that in the prior art, the roundabout identification cannot be accurately identified in local details and the identification speed is low.
In order to achieve the above object, the embodiments of the present application adopt the following technical solutions:
in one aspect, an embodiment of the present application provides a method for determining a roundabout set, where the method includes:
acquiring a road network topological structure, wherein the road network topological structure comprises a plurality of road nodes;
constructing road network data according to the road nodes;
determining a first candidate roundabout set according to the road network data;
screening the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set;
screening the second candidate roundabout set by using a machine learning mode to determine a third candidate roundabout set;
screening the third candidate roundabout set by utilizing a deep learning mode to determine a fourth candidate roundabout set;
and checking the fourth candidate roundabout set to determine a target roundabout set.
In another aspect, the present application further provides a roundabout set determining apparatus, including:
the system comprises a data acquisition unit, a data processing unit and a data processing unit, wherein the data acquisition unit is used for acquiring a road network topological structure, and the road network topological structure comprises a plurality of road nodes;
the data processing unit is used for constructing road network data according to the road nodes;
the data processing unit is further used for determining a first candidate roundabout set according to the road network data;
the data processing unit is further configured to screen the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set;
the data processing unit is further used for screening the second candidate roundabout set by utilizing a machine learning mode to determine a third candidate roundabout set;
the data processing unit is further configured to screen the third candidate roundabout set in a deep learning manner to determine a fourth candidate roundabout set;
the data processing unit is further configured to check the fourth candidate roundabout set to determine a target roundabout set.
Compared with the prior art, the method has the following beneficial effects:
the embodiment of the application provides a method and a device for determining a roundabout set, wherein a road network topological structure is firstly obtained, the road network topological structure comprises a plurality of road nodes, road network data is then constructed according to the road nodes, a first candidate roundabout set is determined according to the road network data, the first candidate roundabout set is then screened according to a preset rule to determine a second candidate roundabout set, the second candidate roundabout set is then screened in a machine learning mode to determine a third candidate roundabout set, the third candidate roundabout set is then screened in a deep learning mode to determine a fourth candidate roundabout set, and finally the fourth candidate roundabout set is verified to determine a target roundabout set. On the one hand, because this application adopts multiple screening mechanism, can accomplish the accurate discernment of local detail, the rotary island that consequently acquires is more accurate. On the other hand, because the deep learning mode screening is carried out after a plurality of times of screening, the data screened by the deep learning mode is greatly reduced, and the determining time of the rotary island can be effectively reduced.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and it will be apparent to those skilled in the art that other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a schematic block diagram of a server provided in an embodiment of the present application.
Fig. 2 is a schematic flow chart of a method for determining a roundabout set according to an embodiment of the present disclosure.
Fig. 3 is a schematic flowchart of the sub-step of S104 in fig. 2 provided in an embodiment of the present application.
Fig. 4 is a schematic flowchart of the sub-step of S106 in fig. 2 provided in an embodiment of the present application.
Fig. 5 is a schematic flowchart of a sub-step of S108 in fig. 2 according to an embodiment of the present disclosure.
Fig. 6 is a schematic flowchart of another sub-step of S108 in fig. 2 according to an embodiment of the present disclosure.
Fig. 7 is a schematic flowchart of a sub-step of S110 in fig. 2 according to an embodiment of the present disclosure.
Fig. 8 is a schematic flowchart of a sub-step of S112 in fig. 2 according to an embodiment of the present disclosure.
Fig. 9 is a schematic block diagram of a roundabout set determination apparatus according to an embodiment of the present application.
In the figure: 100-a server; 101-a memory; 102-a processor; 103-a communication interface; 200-a rotary island set determination device; 210-a data acquisition unit; 220-data processing unit.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
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, it need not be further defined and explained in subsequent figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only for distinguishing the description, and are not to be construed as indicating or implying relative importance.
It should be noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
In the description of the present application, it should be noted that the terms "upper", "lower", "inner", "outer", and the like indicate orientations or positional relationships based on orientations or positional relationships shown in the drawings or orientations or positional relationships conventionally found in use of products of the application, and are used only for convenience in describing the present application and for simplification of description, but do not indicate or imply that the referred devices or elements must have a specific orientation, be constructed in a specific orientation, and be operated, and thus should not be construed as limiting the present application.
In the description of the present application, it is also to be noted that, unless otherwise explicitly specified or limited, the terms "disposed" and "connected" are to be interpreted broadly, e.g., as being either fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meaning of the above terms in the present application can be understood in a specific case by those of ordinary skill in the art.
Some embodiments of the present application will be described in detail below with reference to the accompanying drawings. The embodiments described below and the features of the embodiments can be combined with each other without conflict.
First embodiment
As described in the background art, the deep learning method commonly used for identifying the roundabout at present can achieve global accuracy in identification capability, but cannot achieve accurate identification of local details, and is slow in identification speed.
In view of this, the present application provides a method for determining a roundabout set, so as to achieve the effects of improving the screening efficiency and the recognition accuracy by a multiple screening method.
The water filling method determination method provided by the present application is exemplarily described below with a server as an execution subject.
Referring to fig. 1, the server includes a memory 101, a processor 102 and a communication interface 103, wherein the memory 101, the processor 102 and the communication interface 103 are electrically connected to each other directly or indirectly to implement data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The memory 101 may be used to store software programs and modules, such as program instructions/modules corresponding to the water filling manner determining apparatus 300 provided in the embodiments of the present application, and the processor 102 executes the software programs and modules stored in the memory 101, so as to execute various functional applications and data processing. The communication interface 103 may be used for communicating signaling or data with other node devices.
The Memory 101 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.
The processor 102 may be an integrated circuit chip having signal processing capabilities. The Processor 102 may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
It will be appreciated that the configuration shown in fig. 1 is merely illustrative and that the server may include more or fewer components than shown in fig. 1 or may have a different configuration than shown in fig. 1. The components shown in fig. 1 may be implemented in hardware, software, or a combination thereof.
Referring to fig. 2, the method for determining a roundabout set provided by the present application includes:
s102, a road network topological structure is obtained, wherein the road network topological structure comprises a plurality of road nodes.
And S104, constructing road network data according to the road nodes.
And S106, determining a first candidate roundabout set according to the road network data.
And S108, screening the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set.
And S110, screening the second candidate roundabout set by using a machine learning mode to determine a third candidate roundabout set.
And S112, screening the third candidate roundabout set by using a deep learning mode to determine a fourth candidate roundabout set.
And S114, checking the fourth candidate roundabout set to determine a target roundabout set.
It should be noted that the roundabout is also called ring traffic, which is a special form of traffic node and belongs to plane road intersection. The ring-shaped crossed section is also commonly called a rotary island, a rotary disc and the like. For example, if the method for determining a roundabout set is used for determining a roundabout of a whole world, the roundabout set includes the roundabout of the whole world, and then the roundabout can be accurately determined when navigating for a user.
When the roundabout set needs to be determined, firstly, a road network topological structure is obtained, and then road network data are constructed. And determining a first candidate roundabout set according to road network data, then screening the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set, screening the second candidate roundabout set by using a machine learning mode to determine a third candidate roundabout set, screening the third candidate roundabout set by using a deep learning mode to determine a fourth candidate roundabout set, and finally checking the fourth candidate roundabout set to determine a target roundabout set.
It can be understood that in the present application, multiple screening mechanisms need to be performed when determining the roundabout sets, so that the obtained roundabout sets are more accurate. Meanwhile, the deep learning mode screening is performed after a plurality of times of screening, so that the data screened by the deep learning mode is greatly reduced, and the determining time of the rotary island is effectively shortened. For example, when it is initially determined that the number of the roundabout in the roundabout set is ten million, and after the screening and the machine learning mode screening are performed according to the preset rule, only one million roundabout remains in the roundabout set, the number of the roundabout needing to be screened in the deep learning mode is greatly reduced, the time for screening the roundabout in the deep learning mode is effectively reduced, the screening precision is improved, and the screening time is effectively reduced.
As a possible implementation manner of the present application, the step S102 of obtaining the road network topology is to obtain road network data by reading OpenStreetMap (OSM for short, where chinese is a public map), which is a map service that is free and editable and is created by the public network. OpenStreetMap, which is a map-related geographic data improvement that exploits the power and gratuitous contributions of public communities. OSM is non-profit, feeding data back to the community for reuse with other products and services.
In other words, the OpenStreetMap can acquire a global road network structure, however, since the OpenStreetMap itself is not accurate enough for identifying the roundabout, the application utilizes the data of the OpenStreetMap to identify the roundabout.
It should be noted that the road network topology includes a plurality of road nodes, wherein each road node corresponds to its coordinate and ID of the street. For example, for a street with an ID of 001, the road nodes on the street include A, B, C, D, and for a street with an ID of 002, the road nodes on the street include A, E, F, G, i.e., each road node may correspond to an ID of one or more streets.
In addition, the road nodes provided by the application refer to longitude and latitude coordinates of some points on the OpenStreetMap, and the road nodes can be randomly determined or fixed as road nodes, and the application is not limited at all. In addition, the distance between every two road nodes is not limited, and the distance between every two road nodes can be determined according to actual conditions.
For example, because the roundabout is generally elliptical or circular in shape, when a certain street is a straight line, it may be determined that the roundabout is unlikely to occur, and the selection of the road nodes may be more sparse, for example, one road node is taken every 10 meters at a point on the straight line street. For some curved roads, the roundabout may appear, and therefore the road nodes are selected more densely, for example, one road node is taken every 1 meter on the curved street. Or, the condition of selecting a road node is determined according to the grade of the road, for example, when the grade of a certain street is 1 grade, it indicates that the certain street is relatively important, and it is necessary to take a road node every 1 meter; when the level of a street is level 4, it means that the street is relatively unimportant, and it is necessary to take a road node every 10 meters.
As a possible implementation manner of the present application, please refer to fig. 3, a manner of constructing the road network data in S104 is as follows:
and S1041, connecting the road nodes in the same road ID to construct road network data.
The server 100 reconstructs road nodes, reconstructs the road nodes according to the road IDs, and reconstructs the road network data that the server 100 can filter.
For example, the road ID of the road node a is 001, the road IDs of the road nodes B are 001 and 002, the road ID of the road node C is 001, the road ID of the road node D is 001, the road ID of the road node E is 002, the road ID of the road node F is 002, and the road ID of the road node G is 002. When the road network data is constructed, A, B, C, D and A, E, F, G are connected to construct two pieces of road network data.
That is, when the processed data is city X data, reconstruction is required via a road node of the city X. It is understood that after data reconstruction, the server 100 is actually capable of obtaining a plan view of the city X, and there are a plurality of connected road nodes on the plan view.
Referring to fig. 4, S106 includes:
and S1061, determining intersection road nodes in the road network data.
And S1062, scanning points of each intersection road node according to a preset radius to determine whether a loop exists in a preset range, and taking all loops as a first candidate roundabout set.
As a possible implementation manner, the road node includes an intersection road node, for example, when the server 100 acquires a road node from the OpenStreetMap, it marks whether the road node is an intersection node. As another implementation manner of the present application, when the server 100 acquires a road node from the openstreet map, it is not marked whether the road node is an intersection node. Then, however, each road node corresponds to a road ID, and when the number of road IDs that certain road nodes share is more than one, the server 100 may also determine that the road node is an intersection road node.
After determining the intersection road nodes, the server 100 performs point sweeping on each intersection road node according to a preset radius, and further determines whether a loop exists in a preset range.
For example, if the preset radius is set to 100m, then for each intersection road node, it will sweep the point by the radius of 100 m. In the application, the sweeping points are connected outwards in sequence according to the longitude and latitude of the road nodes by taking the road nodes of the intersection as the original points, and then the connection lines between the points are obtained. It can be understood that, since the road network data is already constructed in S1041, in S1061, it is only necessary to actually divide an area from the determined intersection road node, and it can be further determined whether a loop exists within the radius range of the intersection road node.
As a possible implementation manner, whether a loop exists in the radius range of the intersection road node is determined, and whether the loop starts from an initial road node and can pass through other road nodes is determined, and then the loop returns to the initial road node again.
For example, if a loop of a-B-C-D-a can be formed from road node a as the initial road node, it means that the loop may be a roundabout. Wherein whether a loop can be formed can be determined by the angle of two lines formed by the three points.
It can be understood that, since the roundabout necessarily includes an exit, an entrance, and the like, that is, the roundabout necessarily includes intersection road nodes, all loops that may be roundabout can be determined by determining whether the roundabout exists within a preset radius of all the intersection road nodes, and all the loops are taken as the first candidate roundabout set.
Since one roundabout includes a plurality of intersections, a large number of repeated intersections may occur in the first candidate roundabout set, for example, in A, B, C, D, four road nodes, where a and B are both intersection road nodes, a loop determined with a as a starting point is a-B-C-D-a, and a loop with B as a starting point is B-C-D-a-B, but the two loops are substantially equal to each other, and thus after the first candidate roundabout set is obtained, the server 100 may also reject the same loop.
After the loop is obtained, the loop is subjected to preliminary impurity removal, as a possible implementation manner, referring to fig. 5, S108 includes:
s1081, acquiring the number of intersections corresponding to each loop.
S1082, when the number of intersections corresponding to the loop is larger than a threshold value, placing the loop into a second candidate roundabout set.
In view of the above, in the present application, the server 100 can obtain the number of intersections corresponding to each loop, because the roundabout generally needs to include at least three roads so that the automobile can exit the roundabout from a suitable exit after entering the roundabout. Then, the threshold is set to 3, and the server 100 determines whether or not to place the loop in the second candidate roundabout set.
For example, if a roundabout includes A, B, C, D four road nodes, and A, B, C road nodes are intersection road nodes, it means that the road nodes include at least three roads, which can be placed in the second candidate roundabout set.
Referring to fig. 6 as another implementation manner of the present application, S108 of the present application includes:
s1083, calculating a midpoint of each loop.
S1084, determining the distance between each road node and the middle point in the loop.
S1085, when the distance of the road node farthest from the midpoint is smaller than a preset multiple of the distance of the road node closest to the midpoint, placing the loop into a second candidate roundabout set.
Since some loops actually have a rectangular shape or an irregular shape, it is necessary to reject this part of the data through the filtering rule. In the present application, since the server 100 obtains the longitude and latitude coordinates of each road node, after determining a loop, the server 100 can calculate the midpoint of each loop, that is, the longitude and latitude coordinates of each road node in the loop are summed to calculate the average value.
For example, the coordinates of the road node A are (X) 1 ，Y 1 ) B is (X) 2 ，Y 2 ) C is (X) 3 ，Y 3 ) D is (X) 4 ，Y 4 ) Then A, B, C, D the middle point of the loop formed by the four road nodes is ((X) 1 +X 2 +X 3 +X 4 ）/4，（Y 1 +Y 2 +Y 3 +Y 4 ）/4）。
After the midpoint coordinates are determined, the distance between each road node and the midpoint can be determined, and meanwhile, the distance between the road node farthest from the midpoint and the distance between the road node closest to the midpoint are determined. Therefore, in the present application, when the distance of the road node farthest from the midpoint is less than the preset multiple of the distance of the road node closest to the midpoint, the server 100 may determine that the loop is likely to be a roundabout, and then place the loop into the second candidate roundabout set.
For example, the preset multiple is set to be 4 times, and generally, when the distance of the road node farthest from the midpoint is greater than 4 times the distance of the road node closest to the midpoint, the shape of the loop is generally a rectangle or some irregular shape, which is directly eliminated.
As a possible implementation manner of the present application, please refer to fig. 7, wherein S110 includes:
s1101, obtaining a label of a target roundabout in the second candidate roundabout set, and taking the target roundabout as a training set, wherein the target roundabout is a roundabout in different areas.
And S1102, training the preset first model by using the training set and the received training rule to determine a target first model.
And S1103, screening other roundabouts except the target roundabouts in the second candidate roundabond set by using the target first model to determine a third candidate roundabond set.
The method comprises the steps of screening a second candidate roundabout set by adopting a machine learning mode, and further determining a third candidate roundabout set.
For example, when a global roundabout needs to be determined, if the determined first candidate roundabout set includes 1 million roundabout, after the screening in S108, the second candidate roundabout set includes only ten million roundabout.
On this basis, 10 ten thousand roundabouts in the second candidate roundabouts set can be determined as target roundabouts, and the target roundabouts can be used as a training set, and meanwhile, the label of each target roundabouts is obtained. For example, the tag may be a tag that is manually screened by a worker, for example, when the worker identifies that a certain target roundabout is indeed a roundabout, the tag is modified to 1; when the staff identifies that a certain target rotary island is a non-rotary island, the label of the target rotary island is modified to be 0, and then all target rotary islands in the training set are set with labels.
Moreover, it can be understood that, in order to make the data more accurate, when determining the target islands, different regions of the islands may be selected, for example, among the 10 ten thousand target islands, 1 ten thousand target islands are located in china, 1 ten thousand target islands are located in japan, and 1 ten thousand target islands are located in the united states ….
After the training set is determined, the preset first model can be trained by using the training set and the received training rules to determine the target first model. Since the method of training a model using a training set is the prior art, it is not described in detail herein.
When the third candidate roundabout set is determined, the training rule may be input by a worker, for example, when the worker performs labeling on the target roundabout, the worker may find a place where there is a large difference between the roundabout and the non-roundabout, for example, a distance between every two road nodes in the roundabout, an angle between every three road nodes, a perimeter of a roundabout loop, a roundabout center, and the like, and may use the place as the training rule to perform enhanced training on the first model, and then screen, by using the first model of the target obtained through training, the roundabout other than the target roundabout in the second candidate roundabout set to determine the third candidate roundabout set.
As a possible implementation manner of the present application, please refer to fig. 8, S112 includes:
and S1121, acquiring a label of a target roundabout in the second candidate roundabout set, and taking the target roundabout as a training set, wherein the target roundabout is a roundabout in different areas.
And S1122, training the preset second model by using the training set to determine a target second model.
And S1123, screening the third candidate roundabout set by using the target second model to determine a fourth candidate roundabout set.
When the data is not provided in S110, the training rule is not artificially limited when the second model is trained in S1122, and the autonomy of machine training is further fully invoked, and a larger part of data can be removed from the second candidate roundabout set through the training method provided in S110, the number of roundabout in the second candidate roundabout set is 1000 ten thousand, and after the machine learning method of S110 is trained, the number of roundabout in the third candidate roundabout set is 100 ten thousand, so that the data traffic is greatly reduced, the data size of data screening by using the deep learning method is smaller, and the data processing is faster.
Meanwhile, when it is necessary to explain, the step of S114 includes:
and S1141, checking the roundabout in the fourth candidate roundabout set according to a preset shape to determine a target roundabout set.
After the screening in S112, the verification may be performed manually, and when a non-ring island exists in the fourth candidate ring island set, a worker is required to edit a corresponding code to remove the relevant non-ring island data.
For example, generally, after the above processing, some part of the roundabout in the fourth candidate roundabout set may be concave and not be a standard roundabout, so that a worker may edit a corresponding program, so that the server 100 checks the roundabout in the fourth candidate roundabout set according to a preset shape to determine the target roundabout set.
Through the implementation mode, secondary inspection can be realized, and the screening precision is higher.
Second embodiment
Referring to fig. 9, a functional unit of the apparatus 200 for determining a set of islands shown in fig. 1 according to a preferred embodiment of the present invention is shown. It should be noted that the basic principle and the resulting technical effect of the apparatus 200 for determining a set of rotary islands provided in the present embodiment are the same as those of the above embodiments, and for the sake of brief description, reference may be made to the corresponding contents in the above embodiments for parts that are not mentioned in the embodiments of the present invention. The roundabout set determination apparatus 200 includes:
a data obtaining unit 210, configured to obtain a road network topology, where the road network topology includes a plurality of road nodes.
It is understood that S102 may be performed by the data acquisition unit 210.
The data processing unit 220 is configured to construct road network data according to a plurality of road nodes.
It is understood that S104 may be performed by the data processing unit 220.
The data processing unit 220 is specifically configured to connect road nodes in the same road ID to construct road network data.
The data processing unit 220 is further configured to determine a first candidate roundabout set according to the road network data.
It is understood that S106 may be performed by the data processing unit 220.
The data processing unit 220 is further configured to filter the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set.
It is understood that S108 may be performed by the data processing unit 220.
The data processing unit 220 is further configured to filter the second candidate roundabout set by using a machine learning manner to determine a third candidate roundabout set.
It is understood that S110 may be performed by the data processing unit 220.
The data processing unit 220 is further configured to filter the third candidate roundabout set by using a deep learning manner to determine a fourth candidate roundabout set.
It is understood that S112 may be performed by the data processing unit 220.
The data processing unit 220 is further configured to check the fourth candidate roundabout set to determine the target roundabout set.
It is understood that S114 may be performed by the data processing unit 220.
In summary, an embodiment of the present application provides a method and an apparatus for determining a roundabout set, where the method includes obtaining a road network topology, where the road network topology includes a plurality of road nodes, then constructing road network data according to the road nodes, then determining a first candidate roundabout set according to the road network data, then screening the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set, then screening the second candidate roundabout set by using a machine learning manner to determine a third candidate roundabout set, then screening the third candidate roundabout set by using a deep learning manner to determine a fourth candidate roundabout set, and finally verifying the fourth candidate roundabout set to determine a target roundabout set. On the one hand, because this application adopts multiple screening mechanism, can accomplish the accurate discernment of local detail, the rotary island that consequently acquires is more accurate. On the other hand, because the deep learning mode screening is carried out after a plurality of times of screening, the data screened by the deep learning mode is greatly reduced, and the determining time of the rotary island can be effectively reduced.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (6)
1. A method for determining a set of rotary islands, the method comprising:
acquiring a road network topological structure, wherein the road network topological structure comprises a plurality of road nodes;
constructing road network data according to the road nodes;
determining a first candidate roundabout set according to the road network data;
screening the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set;
screening the second candidate roundabout set by using a machine learning mode to determine a third candidate roundabout set;
screening the third candidate roundabout set by utilizing a deep learning mode to determine a fourth candidate roundabout set;
checking the fourth candidate roundabout set to determine a target roundabout set;
wherein the road nodes comprise intersection road nodes; the step of determining a first candidate roundabout set according to the road network data comprises:
determining intersection road nodes in the road network data;
sweeping points of each intersection road node according to a preset radius to determine whether a loop exists in the preset range, and taking all the loops as a first candidate roundabout set;
the step of screening the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set comprises the following steps:
calculating the midpoint of each loop;
determining a distance of each road node in the loop from the midpoint;
when the distance of the road node farthest from the midpoint is smaller than a preset multiple of the distance of the road node closest to the midpoint, putting the loop into the second candidate roundabout set;
the step of screening the second candidate roundabout set by using a machine learning manner to determine a third candidate roundabout set comprises:
acquiring a label of a target roundabout in the second candidate roundabout set, and taking the target roundabout as a training set, wherein the target roundabout is a roundabout in different areas;
training a preset first model by using the training set and the received training rules to determine a target first model;
screening other roundabouts except the target roundabouts in the second candidate roundabouts set by using the target first model to determine a third candidate roundabond set;
the step of screening the third candidate roundabout set by using a deep learning manner to determine a fourth candidate roundabout set comprises:
acquiring a label of a target roundabout in the second candidate roundabout set, and taking the target roundabout as a training set, wherein the target roundabout is a roundabout in different areas;
training a preset second model by using the training set to determine a target second model;
and screening the third candidate roundabout set by using the target second model to determine the fourth candidate roundabout set.
2. The method for determining a roundabout set according to claim 1, wherein each of the road nodes corresponds to longitude and latitude information and a road ID, and the step of constructing the road network data according to the plurality of road nodes comprises:
and connecting the road nodes in the same road ID to construct the road network data.
3. The method for determining a set of rotary islands according to claim 1, wherein the step of screening the first set of candidate rotary islands according to a predetermined rule to determine a second set of candidate rotary islands comprises:
acquiring the number of intersections corresponding to each loop;
and when the number of intersections corresponding to the loop is greater than a threshold value, putting the loop into the second candidate roundabout set.
4. The method for determining a set of roundabouts according to claim 1, wherein the step of checking the fourth candidate set of roundabouts to determine a target set of roundabouts comprises:
and checking the roundabout in the fourth candidate roundabout set according to a preset shape to determine a target roundabout set.
5. An apparatus for determining a set of rotary islands, the apparatus comprising:
the system comprises a data acquisition unit, a data processing unit and a data processing unit, wherein the data acquisition unit is used for acquiring a road network topological structure, and the road network topological structure comprises a plurality of road nodes;
the data processing unit is used for constructing road network data according to the road nodes;
the data processing unit is further used for determining a first candidate roundabout set according to the road network data;
the data processing unit is further configured to screen the first candidate roundabout set according to a preset rule to determine a second candidate roundabout set;
the data processing unit is further configured to screen the second candidate roundabout set in a machine learning manner to determine a third candidate roundabout set;
the data processing unit is further configured to screen the third candidate roundabout set in a deep learning manner to determine a fourth candidate roundabout set;
the data processing unit is further configured to verify the fourth candidate roundabout set to determine a target roundabout set; wherein the content of the first and second substances,
wherein the data processing unit is specifically configured to:
determining intersection road nodes in the road network data;
sweeping points of each intersection road node according to a preset radius to determine whether a loop exists in the preset range, and taking all the loops as a first candidate roundabout set;
calculating the midpoint of each loop;
determining a distance of each road node in the loop from the midpoint;
when the distance of the road node farthest from the midpoint is smaller than a preset multiple of the distance of the road node closest to the midpoint, putting the loop into the second candidate roundabout set;
acquiring a label of a target roundabout in the second candidate roundabout set, and taking the target roundabout as a training set, wherein the target roundabout is a roundabout in different areas;
training a preset first model by using the training set and the received training rules to determine a target first model;
screening other roundabouts except the target roundabouts in the second candidate roundabout set by using the target first model to determine a third candidate roundabout set;
acquiring a label of a target roundabout in the second candidate roundabout set, and taking the target roundabout as a training set, wherein the target roundabout is a roundabout in different areas;
training a preset second model by using the training set to determine a target second model;
and screening the third candidate roundabout set by utilizing the target second model to determine the fourth candidate roundabout set.
6. The rotary island set determination apparatus according to claim 5, wherein each of the road nodes corresponds to latitude and longitude information and a road ID;
the data processing unit is further configured to connect road nodes in the same road ID to construct the road network data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911211602.0A CN111026825B (en) | 2019-12-02 | 2019-12-02 | Method and device for determining roundabout set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911211602.0A CN111026825B (en) | 2019-12-02 | 2019-12-02 | Method and device for determining roundabout set |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111026825A CN111026825A (en) | 2020-04-17 |
| CN111026825B true CN111026825B (en) | 2023-04-18 |
Family
ID=70207645
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911211602.0A Active CN111026825B (en) | 2019-12-02 | 2019-12-02 | Method and device for determining roundabout set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111026825B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112923935B (en) * | 2021-01-28 | 2022-03-11 | 腾讯科技（深圳）有限公司 | Method, device and equipment for drawing roundabout route and storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104328723A (en) * | 2014-11-04 | 2015-02-04 | 吴小平 | Design scheme of clockwise rotary road plane intersection |
| CN105260699B (en) * | 2015-09-10 | 2018-06-26 | 百度在线网络技术（北京）有限公司 | A kind of processing method and processing device of lane line data |
| CN107729915A (en) * | 2017-09-08 | 2018-02-23 | 第四范式（北京）技术有限公司 | For the method and system for the key character for determining machine learning sample |
| US11144681B2 (en) * | 2017-11-10 | 2021-10-12 | Autodesk, Inc. | Generative design pipeline for urban and neighborhood planning |
| EP3514494A1 (en) * | 2018-01-19 | 2019-07-24 | Zenuity AB | Constructing and updating a behavioral layer of a multi layered road network high definition digital map |
-
2019
- 2019-12-02 CN CN201911211602.0A patent/CN111026825B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111026825A (en) | 2020-04-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021217859A1 (en) | Target anomaly identification method and apparatus, and electronic device and storage medium | |
| Longley et al. | The size, shape and dimension of urban settlements | |
| EP1614087A1 (en) | System, apparatus and method for mapping | |
| CN111026825B (en) | Method and device for determining roundabout set | |
| US10943474B2 (en) | Method, apparatus, and computer program product for detecting changes in road traffic condition | |
| CN113779430B (en) | Road network data generation method and device, computing equipment and storage medium | |
| CN105574313B (en) | A kind of determination method and apparatus of category of roads | |
| US11238291B2 (en) | Method, apparatus, and computer program product for determining if probe data points have been map-matched | |
| CN110809284B (en) | Positioning method, system, device and readable storage medium based on MR data | |
| CN116069889B (en) | Method and device for judging single-line topological state of linear element | |
| CN117611879A (en) | Defect detection method, device, equipment and computer readable medium | |
| CN112257970A (en) | Automatic city functional area dividing method based on interest point big data | |
| WO2020135141A1 (en) | Traffic road network description method and device | |
| Mandlburger et al. | Derivation of a countrywide river network based on Airborne Laser Scanning DEMs-results of a pilot study | |
| CN107729308A (en) | Large scale DLG quality examinations statistical method and device | |
| Ahmed et al. | How accurate are locations in Malaysian accident data? Development of a rectification procedure based on nested filtered search technique | |
| US10467467B2 (en) | Method, system and computer program for automatically detecting traffic circles on digital maps | |
| CN111352964A (en) | Method, device and equipment for acquiring interest point information and storage medium | |
| WO2021262082A1 (en) | Method and device for correcting errors in map data | |
| CN106780431B (en) | Pipe inner stuffing method for detecting position and device | |
| CN111858543B (en) | Quality assessment method and device for commercial map and computing equipment | |
| Kashian et al. | OSM POI Analyzer: A Platform for Assessing Position of POIs in OpenStreetMap | |
| CN113990064B (en) | Traffic data acquisition method, system and computer readable storage medium | |
| CN117289355B (en) | Underground pipeline detection data processing method | |
| CN111476033B (en) | Bus stop name generation method and device |
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 |