CN115270999B

CN115270999B - Obstacle risk grade classification method and device, storage medium and vehicle

Info

Publication number: CN115270999B
Application number: CN202211171886.7A
Authority: CN
Inventors: 顾维灏; 艾锐; 刘方旭; 曹东璞; 王聪; 张凯; 刘任杰
Original assignee: Haomo Zhixing Technology Co Ltd
Current assignee: Haomo Zhixing Technology Co Ltd
Priority date: 2022-09-26
Filing date: 2022-09-26
Publication date: 2022-12-06
Anticipated expiration: 2042-09-26
Also published as: CN115270999A

Abstract

The application provides a method and a device for classifying risk levels of obstacles, a storage medium and a vehicle, and belongs to the technical field of automatic driving. According to the method and the device, only the barrier in the preset position range is subjected to risk assessment, so that the calculated amount caused by the barrier without prediction significance can be effectively reduced; determining the target risk level of the barrier to be evaluated of the pedestrian type and/or the bicycle type with high risk as the highest level; aiming at a second type of obstacle to be evaluated in a set position range, a target risk level corresponding to each obstacle to be evaluated is determined based on a position distribution relation between the obstacle to be evaluated and a vehicle, more accurate risk level evaluation of the obstacle to be evaluated is achieved, prediction of the obstacle with a low target risk level can be selectively avoided when the calculation power is insufficient or the calculation speed needs to be accelerated, the calculation time of a prediction algorithm can be effectively reduced, and the real-time performance of an automatic driving system is improved.

Description

Obstacle risk grade classification method and device, storage medium and vehicle

Technical Field

The application relates to the technical field of automatic driving, in particular to a method and a device for classifying risk levels of obstacles, a storage medium and a vehicle.

Background

In recent years, with the development of science and technology, especially the rapid development of intelligent computing, the research on the technology of the automatic driving vehicle becomes a focus of all industries. The automatic driving vehicle can judge the danger degree of other vehicles or pedestrians in the road to the self vehicle by predicting the collision probability of the self vehicle and other vehicles or pedestrians in the road, so that effective reference is provided for road dangerous collision early warning or obstacle avoidance route planning.

However, if the vehicle predicts the trajectories of all the perceived obstacles on the road, on one hand, the huge calculation amount will seriously increase the time consumption of the prediction algorithm; on the other hand, obstacles without prediction significance exist in a plurality of obstacles, and when prediction results of the obstacles are transmitted to a downstream regulation and control module to carry out decision planning of the motion of the vehicle, the time consumption of the regulation and control module is increased, so that the real-time performance of the whole automatic driving system is influenced.

Disclosure of Invention

The application provides a method and a device for classifying risk levels of obstacles, a storage medium and a vehicle, which are used for solving the problems that operation is long in time consumption and the real-time performance of an automatic driving system is poor due to the fact that the risk levels of the obstacles are not effectively classified in the prior art.

In order to solve the above problems, the present application adopts the following technical solutions:

in a first aspect, an embodiment of the present application provides a method for classifying risk levels of obstacles, where the method includes:

acquiring the positions of obstacles around the vehicle, and determining the obstacles in a preset position range as the obstacles to be evaluated;

acquiring the type of the obstacle to be evaluated, and determining the target risk level of the obstacle to be evaluated with the type as a first type as the highest level; the first type comprises a pedestrian type, and/or a bicycle type;

under the condition that the number of obstacles to be evaluated with the types of the second type is at least two, determining a target risk level corresponding to each obstacle to be evaluated of the second type based on the position distribution relation between the obstacles to be evaluated of the second type and the vehicle; the target risk level is less than or equal to the highest level.

In an embodiment of the present application, determining, based on a position distribution relationship between the second type of obstacle to be evaluated and the vehicle, a target risk level corresponding to each of the second type of obstacle to be evaluated includes:

determining any obstacle to be evaluated of the second type as a reference vehicle, and determining other obstacles to be evaluated of the second type except the reference vehicle as vehicles to be evaluated;

calculating the coverage rate corresponding to each vehicle to be evaluated, wherein the coverage rate represents the degree of the vehicle to be evaluated, which is shielded by the reference vehicle, by taking the vehicle as a viewpoint;

determining at least one initial risk level corresponding to each vehicle to be evaluated based on the coverage rate; the vehicles to be evaluated correspond to different coverage rates under the condition that the reference vehicles are different; different coverage rates correspond to different initial risk levels;

and determining a target risk level corresponding to each obstacle to be evaluated of the second type based on at least one initial risk level corresponding to each obstacle to be evaluated of the second type.

In an embodiment of the present application, calculating the coverage rate corresponding to each vehicle to be evaluated includes:

constructing a polar coordinate system taking the vehicle as a pole;

based on the polar coordinate system, four projection angles of four vertexes of a rectangular bounding box corresponding to each vehicle to be evaluated relative to the vehicle are calculated;

and determining the coverage rate of each vehicle to be evaluated based on the four projection angles corresponding to the reference vehicle and the vehicle to be evaluated respectively.

In an embodiment of the present application, determining, based on four projection angles corresponding to the reference vehicle and the vehicle to be evaluated, a coverage rate corresponding to each vehicle to be evaluated includes:

determining a maximum overlap angle based on a maximum value of the four projection angles of the reference vehicle and a maximum value of the four projection angles of the vehicle to be evaluated;

determining a minimum overlap angle based on the minimum value of the four projection angles of the reference vehicle and the minimum value of the four projection angles corresponding to the vehicle to be evaluated;

determining a target overlap angle based on the maximum overlap angle and the minimum overlap angle;

determining the coverage rate corresponding to each vehicle to be evaluated based on the target overlap angle and the frame projection angle of the vehicle to be evaluated; and the frame projection angle is the difference between the maximum value and the minimum value in the four projection angles of the vehicle to be evaluated.

In an embodiment of the present application, the initial risk levels include a low-to-high ignoring level, a normal level, an important level, and the highest level;

determining at least one initial risk level corresponding to each vehicle to be evaluated based on the coverage rate, wherein the determining comprises the following steps:

when the coverage rate is one hundred percent, determining the initial risk level of the vehicle to be evaluated as the neglected level; the ignore level indicates that trajectory prediction for an obstacle is not required;

when the coverage rate is greater than a coverage rate threshold value and less than one hundred percent, determining that the initial risk level of the vehicle to be evaluated is the common level; the ordinary level represents that simple trajectory prediction needs to be performed on the obstacle;

when the coverage rate is larger than zero and smaller than or equal to the coverage rate threshold value, determining the initial risk level of the vehicle to be evaluated as the importance level; the importance level represents that the common track prediction of the obstacle is required;

when the coverage rate is less than or equal to zero, determining the initial risk level of the vehicle to be evaluated as the highest level; the highest level indicates that accurate trajectory prediction of the obstacle is required.

In an embodiment of the present application, determining a target risk level corresponding to each obstacle to be evaluated of each second type based on at least one initial risk level corresponding to each obstacle to be evaluated of each second type includes:

and determining the highest level in at least one initial risk level corresponding to each obstacle to be evaluated of the second type as a target risk level corresponding to each obstacle to be evaluated of the second type.

In an embodiment of the application, the preset position range includes a front and rear rectangular area centered on the vehicle and a front fan-shaped area starting from the vehicle;

the method comprises the following steps of obtaining the positions of obstacles around a vehicle, determining the obstacles in a preset position range as the obstacles to be evaluated, and comprising the following steps:

judging whether the obstacles are in the front and rear rectangular areas and/or the front fan-shaped area or not based on the positions of the obstacles around the vehicle;

and determining the obstacle as the obstacle to be evaluated under the condition that the obstacle is in the front and back rectangular area and/or the front sector area.

In an embodiment of the application, the front and rear rectangular regions include a front rectangular region located in front of the vehicle and a rear rectangular region located in rear of the vehicle, widths of the front rectangular region and the rear rectangular region are both preset widths, a length of the rear rectangular region is a first preset length, and a length of the front rectangular region is a second preset length; wherein the second preset length increases with an increase in a current vehicle speed of the vehicle within a preset length upper limit;

the fan-shaped region in the place ahead is located the place ahead of vehicle, and fan-shaped region in the place ahead uses the vehicle is the centre of a circle, and the third is predetermine length and is the radius, predetermines the angle and is the central angle.

In a second aspect, based on the same inventive concept, an embodiment of the present application provides an obstacle risk level classification device, including:

the device comprises a to-be-evaluated obstacle determining module, a to-be-evaluated obstacle determining module and a judging module, wherein the to-be-evaluated obstacle determining module is used for acquiring the positions of obstacles around a vehicle and determining the obstacles in a preset position range as the to-be-evaluated obstacles;

the first risk level determination module is used for acquiring the type of the obstacle to be evaluated and determining the target risk level of the obstacle to be evaluated, which is of the first type, as the highest level; the first type comprises a pedestrian type, and/or a bicycle type;

the second risk level determination module is used for determining a target risk level corresponding to each obstacle to be evaluated of the second type based on the position distribution relation between the obstacle to be evaluated of the second type and the vehicle under the condition that the number of the obstacles to be evaluated of the second type is at least two; the target risk level is less than or equal to the highest level.

In an embodiment of the application, the second risk level determining module includes:

the vehicle category division submodule is used for determining any obstacle to be evaluated of the second type as a reference vehicle and determining other obstacles to be evaluated of the second type except the reference vehicle as vehicles to be evaluated;

the coverage rate calculation submodule is used for calculating the coverage rate corresponding to each vehicle to be evaluated, and the coverage rate represents the degree of the vehicle to be evaluated, which is shielded by the reference vehicle, by taking the vehicle as a viewpoint;

the initial risk level determining submodule is used for determining at least one initial risk level corresponding to each vehicle to be evaluated based on the coverage rate; the vehicles to be evaluated correspond to different coverage rates under the condition that the reference vehicles are different; different coverage rates correspond to different initial risk levels;

and the target risk level determining submodule is used for determining a target risk level corresponding to each barrier to be evaluated of the second type based on at least one initial risk level corresponding to each barrier to be evaluated of the second type.

In an embodiment of the present application, the coverage calculation sub-module includes:

the polar coordinate system constructing unit is used for constructing a polar coordinate system taking the vehicle as a pole;

the projection angle calculation unit is used for calculating four projection angles of four vertexes of a rectangular boundary box corresponding to each vehicle to be evaluated relative to the vehicle based on the polar coordinate system;

and the coverage rate determining unit is used for determining the coverage rate corresponding to each vehicle to be evaluated based on the four projection angles corresponding to the reference vehicle and the vehicle to be evaluated respectively.

In an embodiment of the present application, the coverage determining unit includes:

a maximum overlap angle determination subunit configured to determine a maximum overlap angle based on a maximum value of the four projection angles of the reference vehicle and a maximum value of the four projection angles of the vehicle to be evaluated;

the minimum overlap angle determining subunit is used for determining a minimum overlap angle based on the minimum value of the four projection angles of the reference vehicle and the minimum value of the four projection angles corresponding to the vehicle to be evaluated;

a target overlap angle determination subunit operable to determine a target overlap angle based on the maximum overlap angle and the minimum overlap angle;

the coverage rate determining subunit is used for determining the coverage rate corresponding to each vehicle to be evaluated based on the target overlap angle and the frame projection angle of the vehicle to be evaluated; and the frame projection angle is the difference between the maximum value and the minimum value in the four projection angles of the vehicle to be evaluated.

In an embodiment of the present application, the initial risk levels include a low-to-high ignoring level, a normal level, an important level, and the highest level; the initial risk level determination submodule includes:

an override level determination unit, configured to determine, when the coverage is one hundred percent, that an initial risk level of the vehicle to be evaluated is the override level; the ignore level indicates that trajectory prediction for an obstacle is not required;

a common level determination unit, configured to determine that the initial risk level of the vehicle to be evaluated is the common level when the coverage rate is greater than a coverage rate threshold and less than one hundred percent; the ordinary level represents that simple trajectory prediction needs to be performed on the obstacle;

the importance level determining unit is used for determining the initial risk level of the vehicle to be evaluated as the importance level when the coverage rate is greater than zero and less than or equal to the coverage rate threshold; the importance level represents that the common track prediction of the obstacle is required;

the highest level determining unit is used for determining that the initial risk level of the vehicle to be evaluated is the highest level when the coverage rate is less than or equal to zero; the highest level indicates that accurate trajectory prediction of the obstacle is required.

In an embodiment of the present application, the target risk level determining sub-module includes:

and the target risk level determining unit is used for determining the highest level in at least one initial risk level corresponding to each barrier to be evaluated of the second type as the target risk level corresponding to each barrier to be evaluated of the second type.

In an embodiment of the present application, the preset position range includes a front and rear rectangular region centered on the vehicle and a front fan-shaped region starting from the vehicle; the obstacle determination module to be evaluated includes:

the judging submodule is used for judging whether the obstacles are in the front and rear rectangular areas and/or the front fan-shaped area or not based on the positions of the obstacles around the vehicle;

and the obstacle to be evaluated determining submodule is used for determining the obstacle as the obstacle to be evaluated under the condition that the obstacle is positioned in the front and back rectangular area and/or the front fan-shaped area.

In an embodiment of the present application, the front and rear rectangular regions include a front rectangular region located in front of the vehicle and a rear rectangular region located behind the vehicle, widths of the front rectangular region and the rear rectangular region are both preset widths, a length of the rear rectangular region is a first preset length, and a length of the front rectangular region is a second preset length; wherein the second preset length is increased with the increase of the current speed of the vehicle within a preset length upper limit;

the fan-shaped area in the front is located in the front of the vehicle, the fan-shaped area takes the vehicle as a circle center, the third preset length is a radius, and the preset angle is a central angle.

In a third aspect, based on the same inventive concept, embodiments of the present application provide a storage medium, where machine-executable instructions are stored in the storage medium, and when the machine-executable instructions are executed by a processor, the method for classifying risk levels of obstacles according to the first aspect of the present application is implemented.

In a fourth aspect, based on the same inventive concept, embodiments of the present application provide a vehicle, including a processor and a memory, where the memory stores machine executable instructions capable of being executed by the processor, and the processor is configured to execute the machine executable instructions to implement the method for classifying risk levels of obstacles presented in the first aspect of the present application.

Compared with the prior art, the method has the following advantages:

according to the obstacle risk grade classification method provided by the embodiment of the application, the positions of obstacles around a vehicle are obtained, and only the obstacles in the preset position range are subjected to risk assessment, so that the calculated amount caused by the obstacles without prediction significance can be effectively reduced; meanwhile, based on the type of the obstacle to be evaluated, determining the target risk level of the obstacle to be evaluated, which is of a pedestrian type and/or a bicycle type with high risk, as the highest level; aiming at the second type of obstacle to be evaluated in the set position range, the target risk level corresponding to each second type of obstacle to be evaluated is determined based on the position distribution relation between the second type of obstacle to be evaluated and the vehicle, so that more accurate risk level evaluation of the second type of obstacle to be evaluated is realized, prediction of the obstacle with lower target risk level can be selectively avoided when the calculation power is insufficient or the calculation speed needs to be increased, the calculation time consumption of a prediction algorithm can be effectively reduced on the premise of ensuring the driving safety, and the real-time performance of an automatic driving system is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic road condition diagram of a vehicle and other vehicles according to an embodiment of the present application.

Fig. 2 is a flowchart illustrating steps of a method for classifying risk levels of an obstacle according to an embodiment of the present application.

FIG. 3 is a diagram illustrating a predetermined range of positions according to an embodiment of the present application.

FIG. 4 is a schematic diagram illustrating a situation where a vehicle to be evaluated is completely occluded in an embodiment of the present application

Fig. 5 is a schematic diagram of a rectangular coordinate system in an embodiment of the present application.

FIG. 6 is a schematic diagram of a polar coordinate system in an embodiment of the present application.

FIG. 7 is a schematic diagram of a situation in which a vehicle to be evaluated is partially occluded in an embodiment of the present application.

FIG. 8 is a schematic diagram of a situation in which a vehicle to be evaluated is not occluded in an embodiment of the present application.

Fig. 9 is a functional block diagram of an obstacle risk classification device according to an embodiment of the present application.

Reference numerals are as follows: 900-obstacle risk classification means; 901-obstacle determination module to be evaluated; 902-a first risk level determination module; 903 — a second risk level determination module.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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 invention.

It should be noted that the automatic driving system may be specifically divided into a sensing module, a prediction module, a regulation module and other modules, and in the actual situation of automatic driving, the automatic driving vehicle will simulate a decision process similar to that of a real driver: the sensing module senses barrier data, the prediction module predicts the self running track of the vehicle or predicts the track of the behaviors of other vehicles, pedestrians and other barriers on the basis of a prediction algorithm, and the prediction result is sent to the downstream regulation and control module to carry out decision planning on the motion of the vehicle. Therefore, the obstacle track can be predicted timely and accurately, the automatic driving vehicle can react to potential collision risks in advance, and driving safety is improved.

If the self vehicle predicts the track of all the vehicles which can be sensed on the road, huge calculation amount can seriously increase the time consumption of a prediction algorithm; meanwhile, when the number of obstacles to be predicted is large and obstacles without prediction significance often exist, the prediction results of the obstacles are transmitted to a downstream regulation and control module to carry out decision planning on the motion of the vehicle, so that the time consumption of the regulation and control module is increased, and the real-time performance of the whole automatic driving system is influenced.

Referring to fig. 1, which shows a schematic road condition diagram of a host vehicle and other vehicles, the inventor of the present application finds that when there are many obstacle vehicles on a road, the influence of the host vehicle on an autonomous driving vehicle at the current moment is low, and even if the host vehicle makes a lane change behavior or an acceleration/deceleration behavior at this moment, the obstacle vehicles marked in the diagram (obstacle vehicle No. 4, obstacle vehicle No. 5, and obstacle vehicle No. 8) all have low influence on the host vehicle. Therefore, prediction of future trajectories of these less-affected obstacle vehicles can be avoided to reduce the amount of computation generated in the prediction process. For example, the No. 4 obstacle vehicle and the No. 5 obstacle vehicle are on the rear side of the own vehicle, but three other obstacle vehicles (the No. 1 obstacle vehicle, the No. 2 obstacle vehicle and the No. 3 obstacle vehicle) are arranged between the No. 4 obstacle vehicle and the own vehicle to block, so that the influence of the No. 4 obstacle vehicle and the No. 5 obstacle vehicle on the own vehicle is transmitted to the own vehicle through the traveling state of the middle obstacle vehicle influenced by the obstacle vehicles, the risk level of the two obstacle vehicles is low, the prediction algorithm can ignore the obstacle according to the situation, and the No. 8 obstacle vehicle in front of the own vehicle can also be ignored similarly.

In the related art, there is a method depending on lane classification influence levels, which strongly depends on the relative position relationship between a lane line and an obstacle to perform the level classification of the obstacle, and which strongly depends on the relative position relationship between the lane line and the obstacle to perform the level classification of the obstacle, and which cannot be applied in an intersection scene where the lane line is unclear or the lane line is missing; in addition, there is a data-driven method, which outputs the importance level of the obstacle through a neural network, and can perform importance classification on the obstacle by using an attention mechanism, but this method is lack of interpretability and is greatly influenced by the quality of training data, and poor training data may result in unexpected output, and thus the effect is not stable.

Aiming at the problems in the prior art, the application aims to provide a method for classifying the risk levels of the obstacles to be evaluated, which does not depend on unstable information such as lane lines and a deep learning model with poor interpretability, is used for classifying different obstacles to be evaluated into different target risk levels based on the types of the obstacles to be evaluated in a preset position range and the position distribution relation between the obstacles and a vehicle, can selectively avoid the prediction of the obstacles with lower target risk levels when the calculation power is insufficient or the calculation speed needs to be accelerated, can effectively reduce the calculation time consumption of a prediction algorithm on the premise of ensuring the driving safety, and improves the real-time performance of an automatic driving system.

Referring to fig. 2, a flowchart illustrating steps of an obstacle risk level classification method according to the present application may include the steps of:

s101: the method comprises the steps of obtaining the positions of obstacles around a vehicle, and determining the obstacles in a preset position range as the obstacles to be evaluated.

In order to distinguish a vehicle from other vehicles around the vehicle, the following description will be made with a vehicle instead of the vehicle.

In the embodiment, the position of the obstacle around the vehicle can be acquired through a sensing module in the automatic driving system, in a specific implementation, the position information of the obstacle around the vehicle can be acquired in real time through a camera, a millimeter wave radar, a laser radar and other sensors, and the position information of the vehicle itself can be acquired in combination with a positioning module, so that the position distribution relationship between the vehicle and the obstacle around the vehicle can be obtained.

In the present embodiment, in order to reduce the amount of calculation caused by an obstacle with a long distance and without predictive significance, a preset position range is established around the own vehicle, and the preset position range and the own vehicle are in a relatively static relationship during the driving of the own vehicle, so that the preset position range can be referred to as a static area. Specifically, the obstacle outside the static area is far away from the vehicle, and the obstacle can be determined as an obstacle which does not need prediction; and the obstacles in the static area are determined as the obstacles to be evaluated, which need to be subjected to risk level evaluation, because the obstacles are closer to the own vehicle.

S102: the method comprises the steps of obtaining the type of an obstacle to be evaluated, and determining the target risk level of the obstacle to be evaluated with the type as the first type as the highest level.

It should be noted that the type of the obstacle to be evaluated may be classified into a first type and a second type. Wherein the first type represents obstacles that normally travel on non-motorized lanes, such as pedestrian types and bicycle types; the second type represents obstacles that are typically traveling in a motor vehicle lane, such as obstacle vehicle types including, but not limited to, motorcycles, cars, vans, trucks, fire trucks, and the like.

In the present embodiment, the obstacle to be evaluated of the first type is generally high in risk, and if the obstacle collides with the own vehicle, serious injury is caused to pedestrians. Therefore, when the type of the obstacle to be evaluated is identified as the first type, the target risk level of the obstacle to be evaluated of the first type is directly determined as the highest level. It should be noted that the highest level indicates that accurate trajectory prediction needs to be performed on the obstacle, so that it is ensured to the greatest extent that the obstacle does not collide with the first type of obstacle to be evaluated, and the driving safety of the first type of obstacle to be evaluated and the vehicle is ensured.

S103: under the condition that the number of the obstacles to be evaluated with the types of the second type is at least two, determining a target risk level corresponding to each obstacle to be evaluated of the second type based on the position distribution relation between the obstacles to be evaluated of the second type and the vehicle; the target risk level is less than or equal to the highest level.

In this embodiment, with reference to fig. 1, assuming that the vehicle with obstacle 1 to the vehicle with obstacle 8 in the figure are all the obstacles to be evaluated, that is, all the obstacles are within the preset position range, based on the position distribution relationship between the obstacle to be evaluated of the second type and the vehicle, the situation that each obstacle to be evaluated is blocked by other obstacles to be evaluated can be known, and based on the blocked situation of each obstacle to be evaluated, the target risk level corresponding to each obstacle to be evaluated of the second type can be determined.

For example, with reference to fig. 1, if the vehicle is taken as a self vehicle, and the obstacle vehicle No. 4 is completely covered by the obstacle vehicle No. 1, it is described that the influence of the obstacle vehicle No. 4 on the self vehicle is conducted to the self vehicle by the way that the influence affects the traveling state of the intermediate vehicle including the obstacle vehicle No. 1, so that the risk level of the obstacle vehicle No. 4 is low, the target risk level of the obstacle vehicle No. 4 can be determined as an ignoring level, and the prediction algorithm can ignore such an obstacle to be evaluated according to the situation.

For example, with reference to fig. 1, if the vehicle is taken as a viewpoint, the vehicle with the obstacle 6 is not covered by other obstacle vehicles at all, and the vehicle with the obstacle 6 can directly affect the driving state of the vehicle, so the risk level of the vehicle with the obstacle 6 is higher, and the target risk level of the vehicle with the obstacle 6 can be determined as the highest level.

In this embodiment, for an obstacle to be evaluated, which is only partially occluded, different target risk levels between the ignore level and the highest level may be assigned to the obstacle to be evaluated based on the degree of occlusion of the obstacle to be evaluated by other obstacles to be evaluated, for example, the risk levels may include an ignore level, a normal level, an important level, and a highest level from low to high. Wherein the ignore level indicates that trajectory prediction for the obstacle is not required; the common level represents that simple trajectory prediction needs to be performed on the obstacle; the importance level represents that the common track prediction of the obstacle is needed; the highest level indicates that accurate trajectory prediction of the obstacle is required.

In the embodiment, when the calculation power is insufficient or the calculation speed needs to be accelerated, the prediction of the obstacle with a low target risk level can be selectively avoided, for example, when the calculation power is insufficient or the vehicle travels to an area with a large traffic flow, the prediction of the obstacle with an overlooked level and an ordinary level can be avoided, so that the calculation time consumption of a prediction algorithm is effectively reduced, and the real-time performance of an automatic driving system is improved.

In one possible embodiment, the preset position range may include a front and rear rectangular region centered on the vehicle and a front fan-shaped region starting from the vehicle, and S101 may specifically include the following sub-steps:

s101-1: whether the obstacle is located in the front and rear rectangular regions and/or the front sectorial region is determined based on the position of the obstacle around the vehicle.

Referring to fig. 3, a schematic diagram of a preset position range is shown. The front and rear rectangular regions comprise a front rectangular region positioned in front of the vehicle and a rear rectangular region positioned behind the vehicle, and the widths of the front rectangular region and the rear rectangular region are preset widths d _width The length of the rear rectangular area is a first preset length d _backward The length of the front rectangular area is a second preset length d _forward . Wherein the second preset length d _forward The length of the vehicle body is increased along with the increase of the current vehicle speed of the vehicle body within the preset length upper limit.

Specifically, the second preset length d _forward The setting can be made according to the following formula:

d _forward =max（S，v*t）（1）；

wherein d is _forward Represents a second preset length; s represents a preset length upper limit; v represents the current vehicle speed; t represents a preset time.

Illustratively, the preset width d _width Can be provided withIs set to be 24 meters and has a first preset length d _backward Can be set to 50 meters; the upper limit of the preset length S may be set to 100 m, and the preset time t may be set to 5 seconds, that is, within 100 m, the second preset length d _forward The current speed of the vehicle increases when the second preset length d is larger than the first preset length d _forward Increasing to over 100 m, a second preset length d _forward The value of (c) is taken to be 100 meters.

In this embodiment, the front sector area is located in front of the vehicle, and the front sector area takes the vehicle as a center of the circle, the third preset length dr as a radius, and the preset angle θ as a central angle, where the third preset length dr may be set to 50 meters. It should be noted that the preset angle θ may be set according to an actual predicted task, and if an obstacle behind the host vehicle needs to be considered, θ may be set to be greater than 180 °.

The front and rear rectangular areas are mainly arranged for the second type of obstacles located on the motor vehicle lane, and can screen front and rear obstacle vehicles; the front sector area is mainly set for the first type of obstacles located on the non-motor vehicle lane, and obstacles such as pedestrians and/or bicycles can be screened.

S101-2: and in the case that the obstacle is in the front-back rectangular area and/or the front fan-shaped area, determining the obstacle as the obstacle to be evaluated.

In the present embodiment, by dividing the front and rear rectangular regions and the front fan-shaped region with the own vehicle as the center, it is possible to detect each type of obstacle and to eliminate the obstacles outside the front and rear rectangular regions and the front fan-shaped region, thereby reducing the amount of calculation due to an obstacle having no predictive significance.

In one possible embodiment, S103 may specifically include the following sub-steps:

s103-1: and determining any obstacle to be evaluated of the second type as a reference vehicle, and determining other obstacles to be evaluated of the second type except the reference vehicle as vehicles to be evaluated.

In the present embodiment, referring to fig. 4, a schematic diagram of a situation in which a vehicle to be evaluated is completely occluded is shown. The vehicle with the obstacle number 11 is a reference vehicle, and the vehicle with the obstacle number 12 is a vehicle to be evaluated. It should be noted that, in the case of a fixed reference vehicle, the vehicle to be evaluated may be one or more. If the number of the second type of obstacles to be evaluated is only two, the two obstacles to be evaluated can be mutually used as a reference vehicle and a vehicle to be evaluated to judge the risk level; if the number of the second type of obstacles to be evaluated exceeds two, after one of the obstacles to be evaluated is taken as a reference vehicle, the remaining two obstacles to be evaluated are both the vehicles to be evaluated.

In one example, the obstacles to be evaluated include obstacle vehicle number 1, obstacle vehicle number 2, and obstacle vehicle number 3. When the No. 1 obstacle vehicle is used as a reference vehicle, the No. 2 obstacle vehicle and the No. 3 obstacle vehicle are both vehicles to be evaluated; when the No. 2 obstacle vehicle is used as a reference vehicle, the No. 1 obstacle vehicle and the No. 3 obstacle vehicle are both vehicles to be evaluated; when the obstacle vehicle number 3 is used as a reference vehicle, the obstacle vehicle number 1 and the obstacle vehicle number 2 are both vehicles to be evaluated.

In this embodiment, each obstacle to be evaluated of the second type is used as a primary reference vehicle to determine a situation that another vehicle to be evaluated is blocked by the reference vehicle, so that the risk level of the vehicle to be evaluated can be evaluated according to the blocked situation of the vehicle to be evaluated obtained each time.

S103-2: and calculating the coverage rate corresponding to each vehicle to be evaluated.

In this embodiment, it should be noted that the coverage characterization takes the vehicle as a viewpoint, and the degree of shielding of the vehicle to be evaluated by the reference vehicle is represented. By calculating the coverage rate, the degree of the sheltering of the vehicle to be evaluated by the reference vehicle can be effectively quantified, and the accurate risk grade division can be realized.

S103-3: and determining at least one initial risk level corresponding to each vehicle to be evaluated respectively based on the coverage rate.

It should be noted that, in the present embodiment, each obstacle to be evaluated of the second type is used as a primary reference vehicle to determine a situation that another vehicle to be evaluated is shielded by the reference vehicle, so that for the same vehicle to be evaluated, the coverage rate corresponding to the vehicle to be evaluated will be different when the selected reference vehicle is different, and different coverage rates correspond to different initial risk levels.

For example, if M obstacle vehicles in a preset position range exist around the vehicle, the M obstacle vehicles jointly form a vehicle set, one obstacle vehicle is taken from the vehicle set every time, and the coverage rate of other M-1 obstacle vehicles relative to a reference vehicle is calculated by taking the obstacle vehicle as the reference vehicle; and traversing the M obstacle vehicles, namely each obstacle vehicle is taken as a primary reference vehicle, performing M (M-1) times of calculation in total, obtaining M-1 different coverage rate results by each obstacle vehicle, and correspondingly obtaining at least one initial risk grade based on the M-1 different coverage rate results.

Specifically, the coverage may be divided into different intervals, each interval corresponding to one initial risk level, and therefore, when several coverage rates are in the same interval, the several coverage rates correspond to the same initial risk level.

S103-4: and determining a target risk level corresponding to each obstacle to be evaluated of each second type based on at least one initial risk level corresponding to each obstacle to be evaluated of each second type.

In this embodiment, the higher the initial risk level of the obstacle to be evaluated, the greater the risk of the obstacle to be evaluated. Therefore, the highest level of the at least one initial risk level corresponding to each obstacle to be evaluated of the second type can be directly determined as the target risk level corresponding to each obstacle to be evaluated of the second type.

In one possible embodiment, S103-2 may include the following sub-steps:

s103-2-1: and constructing a polar coordinate system taking the vehicle as a pole.

In the present embodiment, first, a rectangular coordinate system is established for the own vehicle, as shown in fig. 5. With the origin of the center of the vehicle (0,0), the direction of travel of the traffic flow is the x-axis forward direction, and the direction perpendicular to the x-axis is the y-axis forward direction, and then the rectangular coordinates of the four vertices of the rectangular bounding box of the vehicle with the obstacle number 7 can be obtained. Then, the rectangular coordinates of the four vertexes are transferred to a polar coordinate system as shown in fig. 6, the center of the vehicle is taken as a pole, the positive direction of the x axis is taken as a polar axis (i.e. 0 degree of the polar coordinate), the counterclockwise direction is taken as the positive direction of the angle, and the angle range of the polar coordinate is (-pi, pi).

In the present embodiment, for the No. 7 obstacle vehicle shown in fig. 6, the polar coordinates of the vertices of its rectangular bounding box are expressed as follows:

（ρ，θ）（2）；

where ρ represents the diameter of the pole from the pole to the vertex, and θ is the angle between the diameter and the axis of the pole.

S103-2-2: and calculating four projection angles of four vertexes of the rectangular bounding box corresponding to each vehicle to be evaluated relative to the vehicle based on the polar coordinate system.

In this embodiment, four projection angles can be obtained in the polar coordinate system according to four vertices of the rectangular bounding box of each obstacle to be evaluated, where the four projection angles are:

（θ _{vertex 1} ，θ _{Vertex 2} ，θ _{Vertex 3} ，θ _{Vertex 4} ）（3）；

Defining the frame projection angle of each barrier to be evaluated under the polar coordinate as a theta frame projection angle, then:

θ _H =max（θ _{vertex 1} ，θ _{Vertex 2} ，θ _{Vertex 3} ，θ _{Vertex 4} ）（4）；

θ _L =mix（θ _{Vertex 1} ，θ _{Vertex 2} ，θ _{Vertex 3} ，θ _{Vertex 4} ）（5）；

θ frame projection angle = θ _H -θ _L （6）；

It should be noted that the frame projection angle θ is an included angle formed by projecting the boundary of the obstacle to be evaluated to two boundary lines formed by the vehicle.

For example, continuing to refer to fig. 4, an included angle formed by the solid lines on both sides of the No. 11 obstacle vehicle is a frame projection angle corresponding to the No. 11 obstacle vehicle, and an included angle formed by the dashed lines on both sides of the No. 12 obstacle vehicle is a frame projection angle corresponding to the No. 12 obstacle vehicle.

S103-2-3: and determining the coverage rate of each vehicle to be evaluated based on the four projection angles corresponding to the reference vehicle and the vehicle to be evaluated respectively.

In the embodiment, the obstacle to be evaluated is divided into a front obstacle and a rear obstacle to be calculated respectively based on the position of the obstacle to be evaluated according to the second type in the rectangular coordinate system. The front obstacle refers to an obstacle to be evaluated, the x coordinate of which in the rectangular coordinate system is larger than 0, and the rear obstacle refers to an obstacle to be evaluated, the x coordinate of which in the rectangular coordinate system is smaller than 0.

In the present embodiment, a front obstacle set is generated for obstacle vehicles whose current position is in front of the own vehicle, one obstacle vehicle is taken as a reference vehicle at a time, a frame projection angle of the reference vehicle is generated, and a relationship between the frame projection angle and frame projection angles of other vehicles to be evaluated is determined, and there are three cases: full coverage, partial coverage, no coverage.

Specifically, the projection angle of the frame of the reference vehicle is set to θ _A The maximum value of the four projection angles corresponding to the reference vehicle is theta _A-H The maximum value of the four projection angles corresponding to the reference vehicle is theta _A-L (ii) a The projection angle of the frame of the vehicle to be evaluated is theta _B The maximum value of the four projection angles corresponding to the vehicle to be evaluated is theta _B-H The maximum value of the four projection angles corresponding to the vehicle to be evaluated is theta _B-L Then, S103-2-3 may specifically include the following sub-steps:

s103-2-3-1: based on the maximum value of the four projection angles of the reference vehicle and the maximum value of the four projection angles of the vehicle to be evaluated, the maximum overlap angle may be determined according to the following formula.

Specifically, the maximum overlap angle can be obtained according to the following formula:

φ _H =min（θ _A-H ，θ _B-H ）（7）；

wherein phi is _H Represents the maximum overlap angle; theta.theta. _A-H Representing the maximum value of the four projection angles corresponding to the reference vehicle; theta _B-H Representing the maximum of the four projection angles corresponding to the vehicle to be evaluated.

That is, the smaller value of the maximum value of the four projection angles of the reference vehicle and the maximum value of the four projection angles of the vehicle to be evaluated may be determined as the maximum overlap angle.

S103-2-3-2: and determining the minimum overlap angle based on the minimum value of the four projection angles of the reference vehicle and the minimum value of the four projection angles corresponding to the vehicle to be evaluated.

Specifically, the minimum overlap angle can be obtained according to the following formula:

φ _L =max（θ _A-L ，θ _B-L ）（8）；

wherein phi is _L Represents a minimum overlap angle; theta _A-L Represents the minimum value of the four projection angles corresponding to the reference vehicle; theta _B-L Represents the minimum of the four projection angles corresponding to the vehicle to be evaluated.

That is, the larger of the minimum value of the four projection angles of the reference vehicle and the minimum value of the four projection angles of the vehicle to be evaluated may be determined as the minimum overlap angle.

S103-2-3-3: based on the maximum and minimum overlap angles, a target overlap angle is determined.

Specifically, the target overlap angle may be obtained according to the following formula:

φ=φ _H -φ _L （9）；

where φ represents a target overlap angle; phi is a _H Represents the maximum overlap angle; phi is a _L The minimum overlap angle is indicated.

S103-2-3-4: and determining the coverage rate corresponding to each vehicle to be evaluated based on the target overlap angle and the frame projection angle of the vehicle to be evaluated.

Specifically, the coverage rate corresponding to each vehicle to be evaluated can be calculated according to the following formula:

P=φ/θ _B （10）；

wherein P represents the coverage rate of the vehicle to be evaluated relative to a reference vehicle; phi denotes a target overlap angle; theta _B Representing the projected angle of the frame of the vehicle to be evaluated.

It should be noted that the target overlap angle indicates a portion where the projection angles of the vehicle to be evaluated and the reference vehicle overlap. Therefore, based on the formula (10), the proportion of the overlapped part to the projection angle of the frame corresponding to the vehicle to be evaluated can be obtained, and the proportion is the coverage rate corresponding to the vehicle to be evaluated.

In a possible embodiment, in the case of including the initial risk level as a low-to-high ignore level, a normal level, an important level and a highest level, S103-3 may specifically include the following sub-steps:

s103-3-1: when the coverage rate is one hundred percent, the initial risk level of the vehicle to be evaluated is determined as the neglect level.

In the present embodiment, the ignore level indicates that the vehicle does not need to predict the trajectory of the obstacle.

For example, with reference to fig. 4, if the vehicle with the obstacle 11 is the reference vehicle and the vehicle with the obstacle 12 is the vehicle to be evaluated, the frame projection angle of the vehicle with the obstacle 12 is completely blocked by the vehicle with the obstacle 11 from the viewpoint of the vehicle, and at this time, the coverage rate P =100% corresponding to the vehicle with the obstacle 12, and the vehicle with the frame projection angle completely covered is classified into the risk class of the neglect class.

S103-3-2: and when the coverage rate is greater than the coverage rate threshold value and less than one hundred percent, determining that the initial risk level of the vehicle to be evaluated is a common level.

In the present embodiment, the normal level indicates that the vehicle needs to simply predict the trajectory of the obstacle.

For example, referring to fig. 7, a schematic diagram of a situation where a vehicle to be evaluated is partially occluded is shown. If the coverage rate threshold is set to be 50%, if the coverage rate P corresponding to the obstacle vehicle number 12 is calculated, the following relationship exists: 50% < P <100%, barrier vehicle No. 12 will be classified under the risk classification of normal class.

S103-3-3: and when the coverage rate is greater than zero and less than or equal to the coverage rate threshold value, determining the initial risk level of the vehicle to be evaluated as the importance level.

In the present embodiment, the importance level indicates that the vehicle needs to predict the general trajectory of the obstacle.

For example, with reference to fig. 7, if the coverage threshold is set to 50%, the following relationship exists if the coverage P corresponding to the obstacle vehicle No. 12 is calculated: 0<P ≦ 50%, barrier vehicle # 12 will be classified under the risk level of importance.

S103-3-4: and when the coverage rate is less than or equal to zero, determining the initial risk level of the vehicle to be evaluated as the highest level.

In the present embodiment, the highest level indicates that the vehicle needs to accurately predict the trajectory of the obstacle.

Illustratively, referring to fig. 8, a schematic diagram of a situation in which the vehicle to be evaluated is not occluded is shown. At this time, the coverage rate P corresponding to the obstacle vehicle number 12 will be a negative value, and the obstacle vehicle number 12 will be classified into the highest risk level.

In a specific implementation, if the target overlap angle Φ <0 calculated for a certain vehicle to be evaluated, the initial risk level of the vehicle to be evaluated can be directly set as the highest level without calculating the coverage rate P.

It should be noted that more coverage threshold values may be set, and the risk level may be further subdivided to improve the detection accuracy, and the number and the value of the coverage threshold values are not specifically limited in this embodiment, and may be set according to actual requirements.

It should be noted that, for an obstacle to be evaluated, which is located behind the host vehicle at the current position, since the projection angle of the corner point of the obstacle to be evaluated crosses pi, the calculation result of the projection angle of the frame is a reflex angle, that is, an angle larger than 180 °, which is inconsistent with the actual situation, therefore, when calculating the projection angle, the projection angle θ of four vertices of each rectangular bounding box is added with a 3/2 × pi, and the projection angle θ is limited to a positive number, thereby preventing misjudgment of the projection angle direction due to the crossing of the pi angle.

In the embodiment, the calculation of the coverage rate may be obtained by calculating the projection arc length of the target overlap angle at the vehicle to be evaluated, in addition to calculating the ratio of the target overlap angle to the frame projection angle corresponding to the vehicle to be evaluated. Specifically, when the projection arc length is zero, it indicates that the vehicle to be evaluated is completely blocked, and at this time, the initial risk level can be divided into an ignorance level; when the projection arc length is larger than the preset arc length threshold, it is indicated that the vehicle to be evaluated is not shielded, and at the moment, the initial risk level can be divided into the highest level; when the projection arc length is between zero and the preset arc length threshold value, the longer the projection arc length is, the less the vehicle to be evaluated is shielded, and the higher the corresponding initial risk level is.

Overall, according to the obstacle risk level classification method provided by the embodiment of the application, by acquiring the positions of obstacles around a vehicle and determining the obstacles to be evaluated in a rectangular area and/or a fan-shaped area, a large number of obstacles without prediction significance can be filtered when the number of the obstacles is large, and thus the calculated amount brought by the obstacles is effectively reduced; meanwhile, based on the type of the obstacle to be evaluated, determining the target risk level of the first type (pedestrian type and/or bicycle type) of the obstacle to be evaluated with higher risk as the highest level, further avoiding collision with the first type of obstacle to be evaluated to the greatest extent, and ensuring the driving safety of the first type of obstacle to be evaluated and the self-vehicle; aiming at a second type of obstacle to be evaluated in a set position range, the coverage rate corresponding to each second type of obstacle to be evaluated is calculated based on the position distribution relation between the second type of obstacle to be evaluated and the vehicle, the target risk level corresponding to each second type of obstacle to be evaluated is determined based on the coverage rate, more precise and accurate risk evaluation of the second type of obstacle to be evaluated is achieved, prediction of the obstacle with lower target risk level can be selectively avoided when the vehicle is insufficient in computing power or the calculation speed needs to be accelerated, computing time consumption of a prediction algorithm can be effectively reduced on the premise of ensuring driving safety, and real-time performance of an automatic driving system is improved.

For example, with reference to fig. 1, assuming that, during the driving process of the host vehicle, the obstacle vehicles 1 to 7 are all within the preset position range, and the obstacle vehicle 8 is outside the preset position range, it is determined that the obstacle vehicles 1 to 7 are the obstacles to be evaluated, and the obstacle vehicle 8 is ignored, and the trajectory prediction is not performed; for the No. 1 to No. 7 obstacle vehicles, calculating the coverage rate corresponding to each obstacle vehicle according to the position distribution relationship between the No. 1 to No. 7 obstacle vehicles and the self vehicle, wherein when the No. 1 obstacle vehicle is used as a reference vehicle, the No. 1 obstacle vehicle, the No. 3 obstacle vehicle, the No. 6 obstacle vehicle and the No. 7 obstacle vehicle are not shielded by the No. 1 obstacle vehicle, the coverage rates are all smaller than 0 at the moment, the initial risk levels of the four obstacle vehicles can be determined to be the highest level, the action tracks of the four obstacle vehicles are predicted accurately, and alarm processing is carried out after collision risks exist; for another example, when the No. 1 obstacle vehicle is used as the reference vehicle, both the No. 4 obstacle vehicle and the No. 5 obstacle vehicle are shielded by the No. 1 obstacle vehicle, wherein the No. 4 obstacle vehicle is completely shielded, the coverage rate of the No. 4 obstacle vehicle is 100%, the initial risk level of the No. 4 obstacle vehicle can be determined as the neglect level, the trajectory prediction is not needed, the No. 5 obstacle vehicle is partially shielded, the coverage rate of the No. 5 obstacle vehicle is between 0 and 100%, in the case that the coverage rate threshold is 50%, if the coverage rate corresponding to the No. 5 obstacle vehicle is 20%, the initial risk level can be determined as the importance level, and if the coverage rate corresponding to the No. 5 obstacle vehicle is 80%, the initial risk level can be determined as the normal level. Furthermore, when the vehicle is insufficient in computing power or runs to an area with large traffic flow, the obstacles of the neglect level and the common level can be avoided, and only the obstacles of the importance level and the highest level are subjected to track prediction, so that the computing time consumption of a prediction algorithm is effectively reduced, and the real-time performance of an automatic driving system is improved.

In a second aspect, based on the same inventive concept, referring to fig. 9, an embodiment of the present application provides an obstacle risk level classification apparatus 900, where the obstacle risk level classification apparatus 900 includes:

the obstacle determination module 901 is configured to acquire positions of obstacles around the vehicle, and determine an obstacle in a preset position range as an obstacle to be evaluated;

a first risk level determining module 902, configured to obtain a type of an obstacle to be evaluated, and determine a target risk level of the obstacle to be evaluated, which is of the first type, as a highest level; the first type includes a pedestrian type, and/or a bicycle type;

a second risk level determining module 903, configured to determine, when the number of the to-be-evaluated obstacles of the second type is at least two, a target risk level corresponding to each of the to-be-evaluated obstacles of the second type based on a position distribution relationship between the to-be-evaluated obstacles of the second type and the vehicle; the target risk level is less than or equal to the highest level.

In an embodiment of the present application, the second risk level determining module 903 includes:

the vehicle type division submodule is used for determining any second type of obstacle to be evaluated as a reference vehicle and determining other second types of obstacles to be evaluated except the reference vehicle as vehicles to be evaluated;

the coverage rate calculation submodule is used for calculating the coverage rate corresponding to each vehicle to be evaluated, and the coverage rate represents the degree of the shielding of the vehicle to be evaluated by a reference vehicle by taking the vehicle as a viewpoint;

the initial risk level determining submodule is used for determining at least one initial risk level corresponding to each vehicle to be evaluated based on the coverage rate; the method comprises the following steps that a vehicle to be evaluated corresponds to different coverage rates under the condition that reference vehicles are different; different coverage rates correspond to different initial risk levels;

and the target risk level determining submodule is used for determining a target risk level corresponding to each barrier to be evaluated of each second type based on at least one initial risk level corresponding to each barrier to be evaluated of each second type.

the projection angle calculation unit is used for calculating four projection angles of four vertexes of the rectangular bounding box corresponding to each vehicle to be evaluated relative to the vehicle based on the polar coordinate system;

a maximum overlap angle determining subunit configured to determine a maximum overlap angle based on a maximum value of the four projection angles of the reference vehicle and a maximum value of the four projection angles of the vehicle to be evaluated;

the coverage rate determining subunit is used for determining the coverage rate corresponding to each vehicle to be evaluated based on the target overlap angle and the frame projection angle of the vehicle to be evaluated; and the frame projection angle is the difference between the maximum value and the minimum value of the four projection angles of the vehicle to be evaluated.

In an embodiment of the present application, the initial risk level includes a low-to-high ignoring level, a normal level, an important level, and a highest level; the initial risk level determination submodule includes:

the neglected level determining unit is used for determining the initial risk level of the vehicle to be evaluated as the neglected level when the coverage rate is one hundred percent; the ignore level indicates that trajectory prediction for the obstacle is not needed;

the system comprises a common level determining unit, a risk level judging unit and a risk level judging unit, wherein the common level determining unit is used for determining that the initial risk level of a vehicle to be evaluated is a common level when the coverage rate is greater than a coverage rate threshold value and less than one hundred percent; the common level represents that simple track prediction needs to be carried out on the obstacles;

the importance level determining unit is used for determining the initial risk level of the vehicle to be evaluated as an importance level when the coverage rate is greater than zero and less than or equal to a coverage rate threshold value; the importance level represents that the common track prediction of the obstacle is needed;

and the target risk level determining unit is used for determining the highest level in at least one initial risk level corresponding to each barrier to be evaluated of each second type as the target risk level corresponding to each barrier to be evaluated of each second type.

In one embodiment of the present application, the preset position range includes a front and rear rectangular region centered on the vehicle and a front fan-shaped region starting from the vehicle; the obstacle to be evaluated determination module 901 includes:

a judgment submodule for judging whether the obstacle is located in the front and rear rectangular regions and/or the front sectorial region based on the position of the obstacle around the vehicle;

In an embodiment of the application, the front and rear rectangular regions include a front rectangular region located in front of the vehicle and a rear rectangular region located in rear of the vehicle, the widths of the front rectangular region and the rear rectangular region are both preset widths, the length of the rear rectangular region is a first preset length, and the length of the front rectangular region is a second preset length; the second preset length is increased along with the increase of the current speed of the vehicle within the upper limit of the preset length;

the fan-shaped region in place ahead is located the place ahead of vehicle, and fan-shaped region in place ahead uses the vehicle as the centre of a circle, and the third is preset length and is the radius, and the angle of predetermineeing is the central angle.

It should be noted that, for the specific implementation of the obstacle risk level classification device 900 according to the embodiment of the present application, reference is made to the specific implementation of the obstacle risk level classification method provided in the first aspect of the embodiment of the present application, and details are not repeated here.

In a third aspect, based on the same inventive concept, an embodiment of the present application provides a storage medium, where the storage medium stores therein machine-executable instructions, and when the machine-executable instructions are executed by a processor, the method for classifying risk levels of obstacles according to the first aspect of the present application is implemented.

It should be noted that, for a specific implementation of the storage medium according to the embodiment of the present application, reference is made to the specific implementation of the obstacle risk level classification method provided in the first aspect of the embodiment of the present application, and details are not repeated here.

It should be noted that, for a specific implementation of the vehicle according to the embodiment of the present application, reference is made to the specific implementation of the obstacle risk level classification method provided in the first aspect of the embodiment of the present application, and details are not repeated herein.

As will be appreciated by one of skill in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. 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 terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, 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 terminal 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 terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present invention have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.

Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be 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 terminal 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 terminal. 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 terminal apparatus that comprises the element.

The method, the device, the storage medium and the vehicle for classifying the risk level of the obstacle provided by the invention are described in detail, a specific example is applied in the text to explain the principle and the implementation mode of the invention, and the description of the embodiment is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed, and in summary, the content of the present specification should not be construed as a limitation to the present invention.

Claims

1. An obstacle risk level classification method, characterized by comprising:

acquiring the type of the obstacle to be evaluated, and determining the target risk level of the obstacle to be evaluated with the type as the first type as the highest level; the first type comprises a pedestrian type, and/or a bicycle type;

under the condition that the number of obstacles to be evaluated with the types of the second type is at least two, determining a target risk level corresponding to each obstacle to be evaluated of the second type based on the position distribution relation between the obstacles to be evaluated of the second type and the vehicle; the target risk level is less than or equal to the highest level;

determining a target risk level corresponding to each obstacle to be evaluated of the second type based on the position distribution relation between the obstacle to be evaluated of the second type and the vehicle, including:

determining at least one initial risk level corresponding to each vehicle to be evaluated based on the coverage rate; the vehicles to be evaluated correspond to different coverage rates under the condition that the reference vehicles are different; different coverage rates correspond to different initial risk levels; and the higher the coverage rate is, the lower the initial risk level corresponding to the coverage rate is;

2. The method for classifying the risk level of an obstacle according to claim 1, wherein calculating the coverage rate corresponding to each vehicle to be evaluated comprises:

constructing a polar coordinate system taking the vehicle as a pole;

based on the polar coordinate system, four projection angles of four vertexes of the rectangular bounding box corresponding to each vehicle to be evaluated relative to the vehicle are calculated;

3. The obstacle risk level classification method according to claim 2, wherein determining the coverage rate of each vehicle to be evaluated based on the four projection angles corresponding to the reference vehicle and the vehicle to be evaluated respectively comprises:

4. The obstacle risk level classification method according to claim 1, wherein the initial risk level includes a low-to-high ignore level, a normal level, an importance level, and the highest level;

when the coverage rate is one hundred percent, determining the initial risk level of the vehicle to be evaluated as the neglected level;

when the coverage rate is greater than a coverage rate threshold value and less than one hundred percent, determining that the initial risk level of the vehicle to be evaluated is the common level;

when the coverage rate is larger than zero and smaller than or equal to the coverage rate threshold value, determining the initial risk level of the vehicle to be evaluated as the importance level;

and when the coverage rate is less than or equal to zero, determining the initial risk level of the vehicle to be evaluated as the highest level.

5. The method for classifying the risk level of an obstacle according to claim 1, wherein determining the target risk level corresponding to each obstacle to be evaluated of the second type based on at least one initial risk level corresponding to each obstacle to be evaluated of the second type comprises:

6. The obstacle risk level classification method according to claim 1, wherein the preset position range includes a front and rear rectangular area centered on the vehicle and a front fan-shaped area starting from the vehicle;

judging whether the obstacle is in the front and rear rectangular areas and/or the front fan-shaped area or not based on the position of the obstacle around the vehicle;

and determining the obstacle as the obstacle to be evaluated under the condition that the obstacle is in the front and back rectangular area and/or the front fan-shaped area.

7. The obstacle risk level classification method according to claim 6,

the front rectangular area and the rear rectangular area are both preset widths, the length of the rear rectangular area is a first preset length, and the length of the front rectangular area is a second preset length; wherein the second preset length is increased with the increase of the current speed of the vehicle within a preset length upper limit;

the fan-shaped area in the place ahead of the vehicle, just fan-shaped area in the place ahead uses the vehicle is the centre of a circle, and the third is predetermine length and is the radius, predetermines the angle and is the central angle.

8. An obstacle risk level classification device, characterized in that the obstacle risk level classification device comprises:

the second risk level determination module is used for determining a target risk level corresponding to each obstacle to be evaluated of the second type based on the position distribution relation between the obstacle to be evaluated of the second type and the vehicle under the condition that the number of the obstacles to be evaluated of the second type is at least two; the target risk level is less than or equal to the highest level;

the second risk level determination module comprises:

the initial risk level determining submodule is used for determining at least one initial risk level corresponding to each vehicle to be evaluated based on the coverage rate; the vehicles to be evaluated correspond to different coverage rates under the condition that the reference vehicles are different; different coverage rates correspond to different initial risk levels; and the higher the coverage rate is, the lower the initial risk level corresponding to the coverage rate is;

9. A storage medium having stored therein machine executable instructions which, when executed by a processor, implement the obstacle risk level classification method of any one of claims 1-7.

10. A vehicle comprising a processor and a memory, the memory storing machine executable instructions executable by the processor for executing the machine executable instructions to implement the obstacle risk level classification method of any one of claims 1-7.