CN114913492A

CN114913492A - Intelligent vehicle road system scene generation method and system, storage medium and computing equipment

Info

Publication number: CN114913492A
Application number: CN202210409074.5A
Authority: CN
Inventors: 王祥鹤; 胡坚明; 张毅; 彭黎辉
Original assignee: Tsinghua University
Current assignee: Tsinghua University
Priority date: 2022-04-19
Filing date: 2022-04-19
Publication date: 2022-08-16

Abstract

The invention relates to a scene generating method, a scene generating system, a storage medium and computing equipment of an intelligent vehicle road system, wherein the scene generating method comprises the following steps: performing feature extraction on input traffic scene data to obtain scene elements; constructing a traffic scene unit cell type model according to the scene elements; forming an IVIS scene sample set by the traffic scene cellular model, and vectorizing and representing the scene elements in the IVIS scene sample set; clustering the vectorized scene elements to obtain clustered traffic scene data; and obtaining the IVIS typical and limit scenes according to the clustered traffic scene data. The invention can solve the problems that the real traffic scene is inexhaustible and difficult to reproduce trustinely, and realize high fidelity and flexible reconstruction of the IVIS scene. The invention can be widely applied to the field of intelligent vehicle-road system traffic.

Description

Intelligent vehicle road system scene generation method and system, storage medium and computing equipment

Technical Field

The invention belongs to the field of intelligent vehicle-road system traffic, and particularly relates to an IVIS scene generation method, system, storage medium and computing device based on actual traffic vehicle-road data.

Background

The introduction of the concept of Intelligent Vehicle-Infrastructure System (IVIS) represents a recent technological front problem in the traffic field. The intelligent vehicle-road system concept is different from the traditional method that the vehicle is used as a unique intelligent body and is also used for testing an intelligent vehicle, and the intelligent vehicle-road system concept represents that the existing test takes an intelligent driving vehicle as one element, and simultaneously takes the surrounding environment including the current road and the road testing equipment as another element, and the two elements are regarded as the same status, so that stronger linkage can be obtained during testing. However, because of the convenience of the preorder matching link in the traditional test, the IVIS product is difficult to realize large-scale popularization or application under the current situation; therefore, complete evaluation equipment for IVIS product admission authentication and a matched professional test site play a vital role in influencing the whole body. Accordingly, how to use lower time and money cost and utilize limited field resources to realize and optimize the test evaluation function coverage of the distributed scenes to the maximum degree also becomes one of the important bases for the actual development of the IVIS test.

The virtual test method based on the traffic scene data has great technical advantages in the aspects of test efficiency, test cost and the like. Because the real traffic scene has the characteristic of difficult exhaustion, the simulation test scene test is an important means adopted in the test and verification of the future automatic driving automobile and becomes a current research hotspot.

However, as far as the present, the industry has not yet provided a uniform data standard for defining the automatic driving test scenario, but has different definition modes; while the automatic driving test scene is constructed, the content of the automatic driving test scene also needs to be further clarified. In the invention, the element types of the automatic driving test scene are summarized into various minimum element units including people, vehicles, roads, environments, drive test equipment and the like by a scene element decomposition method; on the basis of the scene primitive, a subsequent test platform is constructed for the data processing method of the automatic driving test scene, so that the virtual test of the automatic driving vehicle test scene is further carried out.

Disclosure of Invention

In view of the above problems, an object of the present invention is to provide an IVIS scene generation method, system, storage medium and computing device based on actual traffic lane data, which can achieve high fidelity and flexible reconstruction of IVIS scenes for the problems that real traffic scenes are inexhaustible and difficult to reproduce trustiness.

In order to achieve the above object, on one hand, the technical scheme provided by the invention is as follows: an intelligent vehicle road system scene generation method comprises the following steps: performing feature extraction on input traffic scene data to obtain scene elements; constructing a traffic scene unit cell type model according to the scene elements; forming an IVIS scene sample set by the traffic scene cellular model, and vectorizing and representing the scene elements in the IVIS scene sample set; clustering the vectorized scene elements to obtain clustered traffic scene data; and obtaining the typical and extreme scenes of the IVIS according to the clustered traffic scene data.

Further, the performing feature extraction on the input traffic scene data to obtain scene elements includes:

and continuously dividing the traffic scene into four parts, reserving the minimum units which are divided into the minimum units and cannot be subdivided, and continuously repeating the current operation on the rest parts until the currently input existing scene picture completely becomes the minimum background composition unit to obtain the scene elements.

Further, the constructing a traffic scene unit cell model according to the scene element includes: the method takes a running main vehicle as a core, the decision space around the main vehicle is simplified into a unit cell pattern surrounded by hexagons, and the vehicle space right in front of, the vehicle space right behind, the vehicle space left in front of, the vehicle space right in front of, the vehicle space left behind and the vehicle space right behind of the core main vehicle are reserved.

Further, the traffic scene data vectorization is characterized by:

[ host vehicle element ], [ traffic participant element ], [ road element ], [ traffic signal element ], [ environmental element ], [ IVIS logical relationship/interaction pattern ], [ other parameters ] ].

Further, the [ host vehicle element ] includes a host vehicle speed, and there are no restrictions on a running state and an acceleration;

the [ traffic participant element ] includes a slave vehicle attribute and a non-motor vehicle and pedestrian attribute;

the [ road elements ] include road linearity, motorway attributes, road segment attributes, and road width;

the [ environmental element ] comprises a weather attribute, a scene closeness and a traffic flow control attribute;

the [ other parameters ] include a distance measure component.

Further, the clustering of the vectorized scene elements is realized by adopting an OIR-DBSCAN method; and the contour coefficient is used as the measurement standard of the accuracy of the scene clustering result.

Further, obtaining the IVIS typical and extreme scenes according to the clustered traffic scene data, including:

analyzing the parameter characteristics of the clustered traffic scene data, acquiring the data characteristics for generating scenes, and generating an IVIS typical field;

analyzing the parameter characteristics of the clustered traffic scene data, acquiring the data characteristics for generating the scene, judging whether the data characteristics meet preset conditions, and if so, generating a limit scene.

In another aspect, an intelligent vehicle road system scene generating system includes: the system comprises an acquisition module, a model construction module, a vector representation module, a clustering module and a scene generation module;

the characteristic extraction module is used for extracting the characteristics of the input traffic scene data to obtain scene elements;

the construction model module is used for constructing a traffic scene unit cell model according to the scene elements;

the vector characterization module is used for forming an IVIS scene sample set by the traffic scene cellular model and vectorizing and characterizing the scene elements in the IVIS scene sample set;

the clustering module is used for clustering the vectorized scene elements to obtain clustered traffic scene data;

and the scene generation module is used for obtaining the IVIS typical scene and the extreme scene according to the clustered traffic scene data.

In another aspect, a computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the above-described scene generation methods.

In another aspect, a computing device, comprising: one or more processors, memory, and one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the above-described scene generation methods.

Due to the adoption of the technical scheme, the invention has the following advantages:

1. the method provided by the invention refers to the idea of natural scene feature extraction and applies the primitive thought to solve the problems that the real scene is inexhaustible and difficult to reproduce credibly, realizes high-fidelity and flexible reconstruction of the IVIS scene, constructs a test platform, and performs virtual test on the test scene of the automatic driving vehicle.

2. The invention adopts a clustering algorithm, so that the IVIS scene library covers driving tests under more scenes with lower cost, simultaneously removes data redundancy, reduces the test cost and improves the test efficiency, thereby improving the universality and timeliness of the IVIS test.

3. The method solves the problem that the extreme test is difficult to develop in a real scene, and constructs a corresponding test scene through analysis and refinement of traffic data, so that the IVIS scene library meets the requirement of the extreme test, and simultaneously ensures the migration of the test result from the simulation scene side to the real road side, thereby improving the completeness and robustness of the IVIS test.

In conclusion, the intelligent vehicle-road system traffic control method can be widely applied to the field of intelligent vehicle-road system traffic.

Drawings

Fig. 1 is a schematic flow chart of a scene generation method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a cellular traffic model according to an embodiment of the present invention;

FIG. 3 is a schematic view of a signal-less intersection scene in an embodiment of the invention;

FIG. 4 is a flowchart of the OIR-DBSCAN algorithm in an embodiment of the present invention;

FIG. 5 is a schematic diagram of a scene generation system according to an embodiment of the invention;

fig. 6 is a schematic structural diagram of a computing device according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the drawings of the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the described embodiments of the invention, are within the scope of the invention.

It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.

The invention provides a scene generation method, a scene generation system, a storage medium and computing equipment of an intelligent vehicle road system, which apply a primitive thought to realize high-fidelity and flexible reconstruction of an IVIS scene; by adopting a clustering algorithm, the IVIS scene library covers driving tests under more scenes with lower cost, meanwhile, data redundancy is removed, the test cost is reduced, and the test efficiency is improved, so that the universality and the timeliness of the IVIS test are improved; the problem that limit tests are difficult to develop in a real scene is solved, and corresponding test scenes are constructed through analysis and refinement of traffic data, so that an IVIS scene library meets the requirement of the limit tests, and meanwhile, the migration of test results from a simulation scene side to a real road side is guaranteed, and the completeness and robustness of the IVIS tests are improved; and providing an optimized scene library for the test of the automatic driving vehicle.

In an embodiment of the present invention, as shown in fig. 1, an intelligent vehicle road system scene generation method is provided to provide an optimized scene library for an automatic driving vehicle test. The embodiment is illustrated by applying the method to a terminal, and it can be understood that the method can also be applied to a server, and can also be applied to a system comprising the terminal and the server, and is implemented by interaction between the terminal and the server. The scene generation method provided by the embodiment can be used for generating the traffic scene of the intelligent vehicle and road, and can also be applied to other fields for generating other scenes. In this embodiment, the method includes the steps of:

step 1, performing feature extraction on input traffic scene data to obtain scene elements;

the method specifically comprises the following steps: for each input traffic scene data, the problem is first split into the generation of scene parameters, i.e. the following parameters are generated:

θ _env ＝(θ _static ,θ _dynamic )

wherein, theta _env I.e. the overall traffic scene environment parameter, theta _static And theta _dynamic The static part parameter and the dynamic part parameter are respectively. The static parameters include a start phaseIn time, the coordinates, speed, acceleration, lane conditions, environmental conditions, driving road conditions, weather conditions and other factors of each component in the scene; the dynamic parameters are not variable parameters such as speed, but represent intelligent parameters that can be dynamically changed according to observed traffic environment conditions, such as controllers of the autonomous vehicles, neural network parameters, and the like.

The method for extracting the features of the input traffic scene data to obtain the scene elements comprises the following steps:

and continuously dividing the existing scene into four parts, reserving the minimum unit which is divided into the minimum units and can not be subdivided, and continuously repeating the current operation on the rest parts until the currently input existing scene picture completely becomes the minimum background composition unit to obtain the scene element. At this time, these scene primitives formed by continuous division can be used as the core of the next step, and the operations of scene generation and expression are performed reversely.

Step 2, constructing a traffic scene unit cell type model according to scene elements;

the method comprises the following specific steps: as shown in fig. 2, for the current traffic scene, the running host vehicle is taken as a core, the surrounding decision space is simplified into a unit cell pattern surrounded by hexagons, and the vehicle space right in front of, the vehicle space right behind, the vehicle space left in front of, the vehicle space right in front of, the vehicle space left behind and the vehicle space right behind of the core host vehicle are reserved, so that not only is the interactive information between the current detected vehicle and the running traffic environment space not lost, but also most redundant data is removed.

Step 3, forming an IVIS scene sample set by the traffic scene crystal cell type model, and vectorizing and representing scene elements in the IVIS scene sample set;

specifically, on the basis of a traffic scene unit cell model, description and formation of an IVIS scene are performed to form a sample set, and a decomposed primitive is used for representing in a vector form; taking an IVIS scene as an example, traffic scene data vectorization is characterized as follows:

[ [ host vehicle element ], [ traffic participant element ], [ road element ], [ traffic signal element ], [ environmental element ], [ IVIS logical relationship/interaction pattern ], [ other parameters ] ].

Wherein, because each component has more than one element, each component is set as a vector, specifically:

[ vehicle element]The speed of the main vehicle is respectively the speed of the main vehicle, and the running state and the acceleration are not limited; however, considering that the influence of the host vehicle speed on the vector is expected to be as small as possible, and the influence on the scene vector becomes small after the vehicle speed reaches a large value, a larger measurement unit is used, and the actual vehicle speed v is measured _Real On the basis of (2), setting the speed value of the main vehicle as

For better unification, it is desirable to characterize the behavior of backing, steering, etc. and the driving state in the same mode during the scene vectorization, so the driving state is represented by the longitudinal acceleration value a by introducing the acceleration vector group _Vertical With a value of lateral acceleration a _Horizontal Is subjected to vector characterization [ a ] _Vertical ,a _Horizontal ]Where both values can be defined as the actual acceleration value a _Real Decomposing to obtain the product.

If the acceleration is not limited, the value of the component that the acceleration has no limitation in the element of the vehicle is [0,0]]If there is a limit, the component is the actual lower limit a of acceleration _Lower With the upper limit a _Upper Combinations of values [ a ] _Lower ,a _Upper ]；

The [ traffic participant element ] includes two components, respectively, a slave vehicle attribute, a non-motor vehicle and a pedestrian attribute. Consider the classification from vehicle attributes as an example:

for collision avoidance, the rear vehicle should be considered as a slave vehicle; when the vehicle is going straight or backing, one vehicle at the front side and the left and right sides is considered, when the vehicle at the left side or the right side is considered, a transverse relative position threshold value is set, the vehicle reaching the threshold value is preferentially considered, and when the threshold value is not reached, the vehicle closest to the relative position at the front side relative to the driving direction of the vehicle is preferentially considered; when changing track to the left or the left, only two vehicles at most on the front side and the left side need to be considered, and the three vehicles are taken as the slave vehicles, and the track is changed to the right or the vehicle turns around simultaneously.

Thus, in [ traffic participant element ]: the slave vehicles are arranged according to the influence on the driving condition of the master vehicle, and the slave vehicles in the same lane are preferably considered when specific reference is made, and then from left to right, wherein the attributes of the slave vehicles are [ the speed of the slave vehicle, the acceleration of the slave vehicle, the position of the slave vehicle, and the preferential traffic condition of the slave vehicle ].

Similarly to the master car: the value of the speed of the vehicle is

The acceleration value is the actual value, and the corresponding relation of the slave vehicle positions is [ transverse relative position, longitudinal relative position ]]The horizontal relative positions are { left-1, right-1, same lane-0 }, the vertical relative positions are { front-1, rear-1, side-by-side-0 }, and if [0,0 ═ 0}]It means that there is no corresponding slave vehicle; if the slave vehicle has the priority right of passage, the value is inf, otherwise, the value is 0.

The attributes of the non-motor vehicles and the pedestrians are { the positions of the non-motor vehicles and the pedestrians, and the relative speed directions of the non-motor vehicles and the pedestrians }, wherein:

the position corresponding relationship is the same as that of the slave vehicle, namely [ the transverse relative position and the longitudinal relative position ], wherein the transverse relative position is { left side is-1, right side is 1, the same lane is 0}, the longitudinal relative position is { front is 1, rear is-1, and side by side is 0}, and if [0,0] is the vehicle, the corresponding non-motor vehicle and the pedestrian do not exist;

the relative speed direction is [ lateral relative speed direction, longitudinal relative speed direction ], where the lateral relative speed direction is { relative left-1, relative right-1, relative lane-0 }, and the longitudinal relative speed direction is { relative forward-1, relative backward-1, relative side-by-side-0 }, where [0,0] indicates a relative standstill or absence, and is a case where it is not considered by a vehicle or a pedestrian.

The [ road elements ] include four components, respectively road linearity, motorway attributes, link attributes and road width, wherein:

the road linear property is represented by radian value of road curvature;

the attribute of the motor lane is [ single/double direction, lane number ], wherein, the single direction is 1, the double direction is 2, and the lane number is the real number;

the road section attribute is { common road section is equal to 1, the road section entrance and exit is equal to 5, and the four branch fork is equal to 10 };

the road width value is an actual road width value;

the value is 0 if no traffic signal light exists, and is 10 if the traffic signal light exists.

The environment elements include three components, namely, a weather attribute, a scene closure property, and a traffic flow control attribute, where the weather attributes are enumerated and assigned in consideration of the influence of weather on the sight line and the road condition, and then the weather attributes are { clear day is 0, strong wind day is 0.5, cloudy day is 1, fog day is 1.5, sand and dust day is 3, rain day is 4, snow day is 5, hail day is 9 };

the time interval attribute is [ day 1/night 0, with illumination 1/without illumination 0, with tunnel-1/without tunnel 0], then in the actual data processing, the term may be considered to be summed, that is, in the same actual case, with illumination at night approximately equal to day;

[ other parameters]Comprising a component characterizing a measure of the agent

D is the distance between the device under test and the agent, v _max Namely the highest speed limit of the current road section.

As shown in fig. 3, in the crossroad section without traffic lights, the main vehicle is HV, the rest are the auxiliary vehicles RV, and the drive test equipment is RSU; for the scene, the scene vectorization rule is adopted to perform data expression on the scene, and the specific data expression is as follows:

the speed of the main vehicle is 20km/h, the main vehicle moves straight, and the acceleration limit is-4 to 2m/s ² ；

The number of the slave vehicles is 1, and the acceleration is limited to 2m/s ² The right back of the main vehicle has no priority right of passage;

no influence of other non-motor vehicles and pedestrians exists;

the road width of the parallel two-way single-lane crossroads is 3.5m, and no roadblock or traffic light exists;

the traveling in snowy days is possible, the illumination is good, and the distance between roadside equipment is more than 100 m.

Step 4, clustering the scene elements after the opposite quantization to obtain clustered traffic scene data;

in this embodiment, the clustering method is implemented by using an OIR-DBSCAN method. Reverse k neighbor R noting a certain point x _k (x) R, satisfying the reverse k neighbor

Where X is the set of population points, and for the reverse nearest neighbor y of X,

∈R:x∈N _k (y) wherein N _k Y is the nearest neighbor of y, usually k is 1, i.e. y is the reverse nearest neighbor of x

Is the nearest neighbor of y; given the global parameter epsilon, the ratio of the kernel distance epsilon' of a certain point to epsilon is called as the kernel distance ratio, and the density of the data point distribution can be reflected.

As shown in fig. 4, the OIR-DBSCAN method includes the following steps:

step 4.1, judging whether the kernel-to-kernel ratio epsilon'/epsilon of a certain point is smaller than a preset first threshold value alpha or not; if the distance is smaller than alpha, k +2 is carried out, and the core distance epsilon' of the point is obtained; if not, entering a step 4.4;

in the present embodiment, the core distance ∈' ═ KNNMatrix (init, k); an init table clustering initial point;

step 4.2, judging whether the kernel distance ratio of the point after k +2 is within a preset threshold range; if the cluster radius is within the threshold range, assigning the cluster radius r as the core distance epsilon' of the point, otherwise, entering the step 4.3;

in the present embodiment, the threshold range is α ≦ ε'/ε ≦ 1;

4.3, judging whether the nuclear distance ratio epsilon'/epsilon of the point is more than 1; if the r is larger than 1, assigning r to KNNMatrix (init, k-2); if not, returning to the step 4.2 for recalculation;

step 4.4, continuously judging whether the kernel distance ratio epsilon'/epsilon of the point is in the threshold range; if the cluster radius is within the threshold range, assigning the cluster radius r as a global parameter epsilon of the point, otherwise, entering the step 4.5;

in the present embodiment, the threshold range is α ≦ ε'/ε ≦ 1;

step 4.5, judging whether the number of nearest neighbors of each point in the k nearest neighbors of the point is more than or equal to k; if the value is larger than k, the clustering radius r is assigned as the core distance epsilon' of the point, otherwise, the point is a noise point.

In each of the above steps, for a specific point, m is recorded _init The sum of the distances from the k neighbor of the point to the k neighbor is

And taking the point with the minimum Value in the list as the initial point of the cluster, wherein:

and after the clustering from the first initial point is finished, deleting the samples with the labels in the clustering process from the list, selecting the next clustering initial point, and repeating the clustering and the iteration until the clustering is finished.

In the clustering process, cosine similarity is adopted to measure the distance between the vectorized traffic scenes, and the specific formula is as follows:

in the formula, theta represents an included angle between two traffic scene vectors;

and

individual watchTwo traffic scene vectors are shown;

for the obtained scene clustering result, using the contour coefficient s (i) as the measurement standard of the accuracy thereof, wherein the specific formula is as follows:

wherein b (i) represents that the average distance from a vector i to all points in a cluster not containing the vector i is taken as the minimum value, and the minimum value is used for measuring the average dissimilarity degree of the vector i to other clusters; a (i) means averaging the distance of a vector i to all other points in the cluster to which it belongs.

Step 5, analyzing the parameter characteristics of the traffic scene data obtained by clustering to obtain an IVIS typical scene and a limit scene;

the method specifically comprises the following steps: analyzing the parameter characteristics of the clustered traffic scene data, acquiring the data characteristics for generating scenes, and generating an IVIS typical scene;

analyzing the parameter characteristics of the clustered traffic scene data, acquiring the data characteristics for generating the scene, judging whether the data characteristics meet preset conditions, namely the scene has limit numerical values or the coupling between scene elements reaches the limit condition, and if so, generating the limit scene.

As shown in fig. 5, in one embodiment of the present invention, there is provided an IVIS scene generation system based on actual transportation vehicle path data, including: the system comprises a feature extraction module, a construction model module, a vector representation module, a clustering module and a scene generation module;

the construction model module is used for constructing a traffic scene unit cell model according to scene elements;

the vector characterization module is used for forming an IVIS scene sample set by the traffic scene cellular model and vectorizing and characterizing scene elements in the IVIS scene sample set;

the clustering module is used for clustering the scene elements subjected to the opposite quantization to obtain clustered traffic scene data;

The system provided in this embodiment is used for executing the above method embodiments, and for details of the process and the details, reference is made to the above embodiments, which are not described herein again.

As shown in fig. 6, which is a schematic structural diagram of a computing device provided in an embodiment of the present invention, the computing device may be a terminal, and may include: a processor (processor), a communication Interface (communication Interface), a memory (memory), a display screen and an input device. The processor, the communication interface and the memory are communicated with each other through a communication bus. The processor is used to provide computing and control capabilities. The memory includes a nonvolatile storage medium, an internal memory, the nonvolatile storage medium storing an operating system and a computer program that when executed by a processor implements a scene generation method; the internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The communication interface is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, a manager network, NFC (near field communication) or other technologies. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on a shell of the computing equipment, an external keyboard, a touch pad or a mouse and the like. The processor may call logic instructions in memory to perform the following method:

performing feature extraction on input traffic scene data to obtain scene elements; constructing a traffic scene unit cell type model according to scene elements; forming an IVIS scene sample set by the traffic scene crystal cell type model, and performing vectorization representation on scene elements in the IVIS scene sample set; clustering the scene elements after the opposite quantization to obtain traffic scene data after clustering; and obtaining the IVIS typical and limiting scenes according to the clustered traffic scene data.

In addition, the logic instructions in the memory may be implemented in the form of software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

Those skilled in the art will appreciate that the architecture shown in fig. 6 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects may be applied, and that a particular computing device may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment of the invention, a computer program product is provided, the computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the method provided by the above-described method embodiments, for example comprising: performing feature extraction on input traffic scene data to obtain scene elements; constructing a traffic scene unit cell type model according to scene elements; forming an IVIS scene sample set by the traffic scene crystal cell type model, and performing vectorization representation on scene elements in the IVIS scene sample set; clustering the scene elements after the opposite quantization to obtain traffic scene data after clustering; and obtaining the IVIS typical and limiting scenes according to the clustered traffic scene data.

In one embodiment of the invention, a non-transitory computer-readable storage medium is provided, which stores server instructions that cause a computer to perform the methods provided by the above embodiments, for example, including: performing feature extraction on input traffic scene data to obtain scene elements; constructing a traffic scene unit cell type model according to scene elements; forming an IVIS scene sample set by the traffic scene crystal cell type model, and performing vectorization representation on scene elements in the IVIS scene sample set; clustering the scene elements after the opposite quantization to obtain traffic scene data after clustering; and obtaining the IVIS typical and limiting scenes according to the clustered traffic scene data.

The implementation principle and technical effect of the computer-readable storage medium provided by the above embodiments are similar to those of the above method embodiments, and are not described herein again.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. An intelligent vehicle road system scene generation method is characterized by comprising the following steps:

performing feature extraction on input traffic scene data to obtain scene elements;

constructing a traffic scene unit cell type model according to the scene elements;

forming an IVIS scene sample set by the traffic scene cellular model, and vectorizing and representing the scene elements in the IVIS scene sample set;

clustering the vectorized scene elements to obtain clustered traffic scene data;

and obtaining the IVIS typical and limit scenes according to the clustered traffic scene data.

2. The scene generation method of claim 1, wherein the performing feature extraction on the input traffic scene data to obtain scene primitives comprises:

3. The scene generation method of claim 1, wherein the constructing a traffic scene unit cell model from the scene primitives comprises: the method takes a running main vehicle as a core, the decision space around the main vehicle is simplified into a unit cell pattern surrounded by hexagons, and the vehicle space right in front of, the vehicle space right behind, the vehicle space left in front of, the vehicle space right in front of, the vehicle space left behind and the vehicle space right behind of the core main vehicle are reserved.

4. The scene generation method of claim 1, wherein the traffic scene data vectorization is characterized by:

5. The scene generating method according to claim 4, wherein the [ host-vehicle element ] includes a host-vehicle speed, and there are no restrictions on a running state and an acceleration;

the [ other parameters ] include a distance measure component.

6. The scene generation method according to claim 1, wherein said clustering of said vectorized scene primitives is implemented using an OIR-DBSCAN method; and the contour coefficient is used as the measurement standard of the accuracy of the scene clustering result.

7. The scene generation method of claim 1, wherein obtaining IVIS typical and extreme scenes from the clustered traffic scene data comprises:

8. The utility model provides an intelligence car road system scene generation system which characterized in that includes: the system comprises an acquisition module, a model construction module, a vector representation module, a clustering module and a scene generation module;

9. A computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the scene generation methods of claims 1-7.

10. A computing device, comprising: one or more processors, memory, and one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the scene generation methods of claims 1-7.