CN106228602A - Determination method, device and the mobile unit of three-dimensional vehicle alternative area - Google Patents

Abstract

The invention provides a method, device and vehicle-mounted equipment for determining a three-dimensional vehicle candidate area. The method includes: acquiring multiple point clouds in a road scene, using the multiple point clouds to determine a ground plane in the road scene; generating multiple three-dimensional areas on the ground plane according to a preset vehicle size; Divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometry, determine one or more of the second geometry contained in each three-dimensional area; The second geometric body is used to determine whether the three-dimensional area is a three-dimensional vehicle candidate area. The embodiment of the present invention utilizes the prior knowledge that the vehicle is located near the ground level and the template of the known vehicle size, which greatly reduces the number of three-dimensional areas to be searched, improves the search efficiency, and meets the real-time requirements in practical applications.

Description

Method, device, and vehicle-mounted equipment for determining a three-dimensional vehicle candidate area

technical field

The invention relates to the technical field of intelligent vehicle environment perception, in particular to a method, device and vehicle-mounted equipment for determining a three-dimensional vehicle candidate area.

Background technique

The point cloud is a collection of massive points on the surface characteristics of the target. The dense point cloud information of the road scene can be obtained through the vehicle-mounted lidar or binocular camera and other sensors. A point cloud can be viewed as a collection of 3D coordinates. The three-dimensional candidate area of the vehicle is generated through the point cloud data, which is convenient for the subsequent algorithm to quickly and accurately identify and locate the vehicle in the scene. This technology is one of the core modules for applications such as front collision warning and adaptive cruise control.

In related technologies, generating the 3D candidate areas of the vehicle through point cloud data requires traversing all possible vehicle candidate areas in the 3D point cloud data, and the search complexity is exponential, which cannot meet the real-time requirements of the algorithm in practical applications. , this technical problem needs to be solved urgently.

Contents of the invention

In view of the above problems, the present invention is proposed to provide a method and device for determining a three-dimensional vehicle candidate area and corresponding on-board equipment that overcome the above problems or at least partially solve the above problems.

According to an aspect of the present invention, a method for determining a three-dimensional vehicle candidate area is provided, including:

Obtaining multiple point clouds in the road scene, using the multiple point clouds to determine the ground plane in the road scene;

generating a plurality of three-dimensional areas on the ground plane according to the preset vehicle size;

Divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometry, and determine one or more of the second geometry contained in each three-dimensional area;

Whether the three-dimensional area is a three-dimensional vehicle candidate area is determined according to the one or more second geometric bodies included in each three-dimensional area.

Optionally, the acquiring multiple point clouds in the road scene includes:

Acquisition of multiple point clouds in a road scene based on laser measurement principles; and/or

According to the principle of photogrammetry, multiple point clouds in the road scene are obtained.

Optionally, using the multiple point clouds to determine the ground plane in the road scene includes:

A point cloud among the plurality of point clouds belonging to the ground plane in the road scene is determined through a random sampling consistent RANSAC algorithm.

Optionally, according to the preset vehicle size, multiple three-dimensional areas are generated on the ground plane, including:

According to the preset vehicle size, a plurality of three-dimensional areas are randomly generated on the ground plane, wherein the three-dimensional areas are determined by specified parameters.

Optionally, the specified parameters include at least one of the following:

The three-dimensional coordinates of the center of the vehicle, the size of the vehicle, and the orientation angle of the vehicle on the ground plane.

Optionally, dividing the first geometry containing the plurality of point clouds into a set of a specified number of second geometry includes:

The cuboid containing the plurality of point clouds is discretized according to a fixed interval, and divided into a set of a specified number of cubes.

Optionally, determining whether the three-dimensional area is a three-dimensional vehicle candidate area according to one or more second geometric bodies included in each three-dimensional area includes:

According to one or more of the second geometric bodies contained in each three-dimensional area, calculate the number of point clouds contained in each three-dimensional area and/or the average height of point cloud by using a three-dimensional integral map;

Determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud.

Optionally, according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud, determining whether the three-dimensional area is a three-dimensional vehicle candidate area includes:

According to the specified weight, the number of point clouds contained in each three-dimensional area and the average height of the point cloud are summed and weighted to obtain the comprehensive score value of each three-dimensional area;

According to the comprehensive score value of each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area.

Optionally, according to the comprehensive scoring value of each three-dimensional area, determining whether the three-dimensional area is a three-dimensional vehicle candidate area includes:

Sorting the comprehensive scoring values of the various three-dimensional regions from large to small;

Select the specified number of 3D areas ranked first as the 3D vehicle candidate areas.

Optionally, the method also includes:

In the process of driving, according to the determined three-dimensional vehicle candidate area, identify driving-related information;

The identified driving-related information is analyzed to generate corresponding driving navigation prompt information and provided to the user.

Optionally, the driving-related information includes at least one of the following:

The distance between the vehicle in front and/or the vehicle behind and the vehicle, road condition identification information, and road environment information.

Optionally, the method is applied in a vehicle-mounted intelligent control unit.

According to another aspect of the present invention, a device for determining a three-dimensional vehicle candidate area is also provided, including:

The point cloud acquisition module is suitable for acquiring multiple point clouds in road scenes;

a ground plane determination module adapted to determine the ground plane in the road scene by using the plurality of point clouds;

A three-dimensional area generation module, adapted to generate multiple three-dimensional areas on the ground plane according to the preset vehicle size;

A geometry determination module, adapted to divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometries, and determine one or more of the second geometries contained in each three-dimensional area;

The candidate area determination module is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to one or more of the second geometric bodies included in each three-dimensional area.

Optionally, the point cloud acquisition module is also suitable for:

Acquisition of multiple point clouds in a road scene based on laser measurement principles; and/or

According to the principle of photogrammetry, multiple point clouds in the road scene are obtained.

Optionally, the ground plane determination module is also adapted to:

A point cloud among the plurality of point clouds belonging to the ground plane in the road scene is determined through a random sampling consistent RANSAC algorithm.

Optionally, the three-dimensional area generation module is also suitable for:

According to the preset vehicle size, a plurality of three-dimensional areas are randomly generated on the ground plane, wherein the three-dimensional areas are determined by specified parameters.

Optionally, the specified parameters include at least one of the following:

The three-dimensional coordinates of the center of the vehicle, the size of the vehicle, and the orientation angle of the vehicle on the ground plane.

Optionally, the geometry determination module is also adapted to:

The cuboid containing the plurality of point clouds is discretized according to a fixed interval, and divided into a set of a specified number of cubes.

Optionally, the candidate area determination module includes:

The calculation unit is adapted to calculate the number of point clouds and/or the average height of point clouds contained in each three-dimensional area by using a three-dimensional integral map according to the one or more second geometric bodies contained in each three-dimensional area;

The determination unit is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud.

Optionally, the determining unit is further adapted to:

According to the specified weight, the number of point clouds contained in each three-dimensional area and the average height of the point cloud are summed and weighted to obtain the comprehensive score value of each three-dimensional area;

According to the comprehensive score value of each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area.

Optionally, the determining unit is further adapted to:

Sorting the comprehensive scoring values of the various three-dimensional regions from large to small;

Select the specified number of 3D areas ranked first as the 3D vehicle candidate areas.

Optionally, the device also includes:

The driving navigation prompt information generation module is adapted to identify driving-related information according to the determined three-dimensional vehicle candidate area during the driving process; analyze the identified driving-related information to generate corresponding driving navigation prompt information, and provided to users.

Optionally, the driving-related information includes at least one of the following:

The distance between the vehicle in front and/or the vehicle behind and the vehicle, road condition identification information, and road environment information.

Optionally, the device is applied in a vehicle-mounted intelligent control unit.

According to yet another aspect of the present invention, there is also provided a vehicle-mounted device, including any one of the devices for determining a three-dimensional vehicle candidate area described above.

In the technical solution provided by the embodiments of the present invention, firstly, multiple point clouds in the road scene are obtained, and the ground plane in the road scene is determined by using the multiple point clouds. Subsequently, according to the preset vehicle size, multiple three-dimensional regions are generated on the ground plane, and then the first geometry containing multiple point clouds is divided into a set of a specified number of second geometry, and one or more a second geometry. Afterwards, according to one or more second geometric bodies contained in each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area. It can be seen that the embodiment of the present invention utilizes the prior knowledge that the vehicle is located near the ground level, and at the same time utilizes the template of the known vehicle size, which greatly reduces the number of three-dimensional areas to be searched, improves the search efficiency, and satisfies the real-time requirements in practical applications. requirements. Moreover, since each second geometric body contains at least one point cloud in the road scene, the embodiment of the present invention determines whether the three-dimensional area is a three-dimensional vehicle candidate area based on one or more second geometric bodies contained in each three-dimensional area, ensuring three-dimensional Region recall.

The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the description, and in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable , the specific embodiments of the present invention are enumerated below.

Those skilled in the art will be more aware of the above and other objects, advantages and features of the present invention according to the following detailed description of specific embodiments of the present invention in conjunction with the accompanying drawings.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiment. The drawings are only for the purpose of illustrating a preferred embodiment and are not to be considered as limiting the invention. Also throughout the drawings, the same reference numerals are used to designate the same components. In the attached picture:

FIG. 1 shows a flowchart of a method for determining a three-dimensional vehicle candidate area according to an embodiment of the present invention;

FIG. 2 shows a flowchart of a method for determining a three-dimensional vehicle candidate area according to another embodiment of the present invention;

FIG. 3 shows a schematic structural diagram of a device for determining a three-dimensional vehicle candidate area according to an embodiment of the present invention; and

Fig. 4 shows a schematic structural diagram of an apparatus for determining a three-dimensional vehicle candidate area according to another embodiment of the present invention.

detailed description

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

In order to solve the above technical problem, an embodiment of the present invention provides a method for determining a three-dimensional vehicle candidate area. Fig. 1 shows a flowchart of a method for determining a three-dimensional vehicle candidate area according to an embodiment of the present invention. As shown in Fig. 1, the method may at least include the following steps S102 to S108.

Step S102, acquiring multiple point clouds in the road scene, and using the multiple point clouds to determine the ground plane in the road scene.

Step S104, generating multiple three-dimensional regions on the ground plane according to the preset vehicle size.

Step S106, dividing the first geometry including multiple point clouds into a set of a specified number of second geometry, and determining one or more second geometry contained in each three-dimensional area.

Step S108, according to one or more second geometric bodies contained in each three-dimensional area, determine whether the three-dimensional area is a three-dimensional vehicle candidate area.

The embodiment of the present invention implements a method for determining a three-dimensional candidate area of a vehicle based on a point cloud. First, multiple point clouds in a road scene are obtained, and the ground plane in the road scene is determined by using the multiple point clouds. Subsequently, according to the preset vehicle size, multiple three-dimensional regions are generated on the ground plane, and then the first geometry containing multiple point clouds is divided into a set of a specified number of second geometry, and one or more a second geometry. Afterwards, according to one or more second geometric bodies contained in each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area. It can be seen that the embodiment of the present invention utilizes the prior knowledge that the vehicle is located near the ground level, and at the same time utilizes the template of the known vehicle size, which greatly reduces the number of three-dimensional areas to be searched, improves the search efficiency, and satisfies the real-time requirements in practical applications. requirements. Moreover, since each second geometric body contains at least one point cloud in the road scene, the embodiment of the present invention determines whether the three-dimensional area is a three-dimensional vehicle candidate area based on one or more second geometric bodies contained in each three-dimensional area, ensuring three-dimensional Region recall.

The multiple point clouds in the road scene are acquired in step S102 above. The embodiment of the present invention provides a variety of optional solutions. For example, according to the principle of laser measurement, multiple point clouds in the road scene can be collected, and the Three-dimensional coordinates (XYZ) and laser reflection intensity (Intensity); or according to the principle of photogrammetry to obtain multiple point clouds in the road scene, the three-dimensional coordinates (XYZ) and color information (RGB) of the point cloud can be collected; or, you can also The point cloud is obtained by combining the principles of laser measurement and photogrammetry, and the three-dimensional coordinates (XYZ), laser reflection intensity (Intensity) and color information (RGB) of the point cloud can be obtained.

First, taking the principle of laser measurement as an example, 3D lidar can be used to collect point clouds. Three-dimensional lidar is one of the important sensors for intelligent vehicles to obtain external information. It has the advantages of high reliability, real-time performance and high accuracy, so it is widely used in the research of intelligent vehicle environment perception. The radar has multiple laser sensors, and the discrete data points measured by each sensor can be expressed as Pi(x, y, z, s), where x, y, and z represent three-dimensional physical distances in meters, and s represents the reflection intensity, which is infinite Outline value. The point cloud data is a collection of large-scale discrete measurement point data {P1, P2, P3,..., PN}, which provide sufficient information for restoring the basic shape characteristics and structural details of the measurement object.

Secondly, taking the principle of photogrammetry as an example, binocular cameras can be used to collect point clouds. Place two cameras on the same plane to shoot the target object, and get many sets of corresponding object image pairs. Since the camera and the object conform to the principle of triangulation, and use the parallax of the corresponding points to calculate the stereoscopic information of the object, so as to obtain the object 3D point cloud data.

In an optional embodiment of the present invention, in step S102 above, multiple point clouds are used to determine the ground plane in the road scene, and the RANSAC (RANdom Sample Consensus, random sampling consistency) algorithm can be used to determine which of the multiple point clouds belongs to A point cloud of the ground plane in a road scene.

RANSAC can iteratively estimate the parameters of a mathematical model from a set of observation data sets containing "outlier points". It is an indeterminate algorithm that has a certain probability of producing a reasonable result. The basic assumptions of RANSAC are:

(1) The data consists of "inside points", for example: the distribution of data can be explained by some model parameters;

(2) "Outlier points" are data that cannot adapt to the model;

(3) Data other than this is noise.

The reasons for outliers are: noise extremes, wrong measurement methods, and wrong assumptions about the data. RANSAC also makes the following assumption: given a set of (usually small) inlier points, there is a process that can estimate the parameters of a model that can explain or be applicable to the inlier points.

The input of the RANSAC algorithm is a set of observation data, a parametric model that can explain or adapt to the observation data, and some credible parameters. RANSAC achieves its goal by iteratively selecting a set of random subsets in the data. The selected subsets are assumed to be interior points and verified by the following method:

1) There is a model adapted to the hypothetical interior point, that is, all unknown parameters can be calculated from the hypothetical interior point;

2) Use the model obtained in 1) to test all other data, and if a certain point is suitable for the estimated model, consider it to be an internal point;

3) If enough points are classified as hypothetical inliers, then the estimated model is reasonable enough;

4) Then, re-estimate the model with all hypothetical inlier points, because it is only estimated by the initial hypothetical inlier points;

5) Finally, the model is evaluated by estimating the error rate of the inlier points and the model.

This process is repeated a fixed number of times, and each time the resulting model is either discarded due to too few inliers, or selected because it is better than the existing model.

In the embodiment of the present invention, a large number of point clouds in the input point cloud data belong to the ground plane, and the ground plane can be estimated relatively robustly through the RANSAC algorithm. The ground plane π=(a,b,c,d) ^T is expressed parametrically by a four-dimensional vector, and its reference coordinate system is the observation coordinate system centered on the sensor.

In step S104 above, multiple three-dimensional areas are generated on the ground plane according to the preset vehicle size. The embodiment of the present invention provides an optional solution. In this solution, according to the preset vehicle size, The plane randomly generates multiple 3D regions, where the 3D regions are determined by specifying parameters. The specified parameters here may be the three-dimensional coordinates of the center of the vehicle, the size of the vehicle, the orientation angle of the vehicle on the ground plane, etc., and the present invention is not limited thereto.

Specifically, in the natural lane scenario, assuming that the vehicles are all located near the road surface (that is, the ground level) (regardless of extreme cases such as suspended vehicles), the size template (length, width, height) of the vehicle is input according to the known vehicle type, Make a slight disturbance near the road surface and traverse the possible orientation angles of the vehicle template at the same time, that is, M possible three-dimensional areas can be obtained. These three-dimensional areas can be determined by seven parameters (x, y, z, l, w, h, θ) Determine, where (x, y, z) represents the three-dimensional coordinates of the vehicle center; (l, w, h) are the length, width, and height templates of the vehicle; θ represents the orientation angle of the vehicle on the horizontal plane.

In an optional embodiment of the present invention, in step S106 above, the first geometric body containing multiple point clouds is divided into a set of a specified number of second geometric bodies. The present invention provides an optional solution, that is, the A cuboid containing multiple point clouds is discretized according to a fixed interval, divided into a set of the specified number of cubes. Specifically, divide the cuboid containing all input 3D point clouds into a set of several small cubes according to the fixed interval discretization, then for the M 3D regions obtained above, each 3D region will correspond to several Collection of small cubes. Here, the designated number can be set according to actual needs, for example, if it is necessary to divide smaller cubes, a larger designated number is set.

After determining one or more second geometric bodies contained in each three-dimensional area in step S106, since each second geometric body contains at least one point cloud in the road scene, step S108 may be based on the one or more second geometric bodies contained in each three-dimensional area According to the characteristics of each three-dimensional area, it can be determined whether the three-dimensional area is a three-dimensional vehicle candidate area according to the characteristics of each three-dimensional area. The characteristics of the three-dimensional area here may be the number of point clouds contained in the three-dimensional area, the average height of the point cloud, the ratio of the number of point clouds, up, down, left, and right, etc., and the present invention is not limited thereto. In an optional embodiment of the present invention, the feature of each three-dimensional area can be calculated by using the three-dimensional integral map, and the calculation complexity is O(1), which saves the amount of calculation and can quickly calculate the features of a large number of three-dimensional areas.

Further, step S108 can determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to the number of point clouds contained in the three-dimensional area; or determine whether the three-dimensional area is a three-dimensional vehicle according to the average height of point clouds contained in the three-dimensional area Candidate area; or determine whether the three-dimensional area is a three-dimensional vehicle candidate area by combining the number of point clouds contained in the three-dimensional area and the average height of the point cloud. Here, the number of point clouds contained in each three-dimensional area and the average height of the point cloud can be summed and weighted according to the specified weight to obtain the comprehensive score value of each three-dimensional area, and then according to the comprehensive score value of each three-dimensional area, determine whether the three-dimensional area is 3D vehicle candidate regions.

In an optional embodiment of the present invention, the comprehensive score values of each three-dimensional area may be sorted from large to small, and a specified number of three-dimensional areas ranked higher are selected as three-dimensional vehicle candidate areas.

The implementation process of the method for determining a three-dimensional vehicle candidate area of the present invention will be described in detail below through a specific embodiment. In this embodiment, firstly, the dense point cloud information of the road scene can be obtained through sensors such as a vehicle-mounted lidar or a binocular camera. A point cloud can be viewed as a collection of 3D coordinates. As shown in FIG. 2, the method may at least include the following steps S202 to S208.

In step S202, the RANSAC algorithm searches for the ground plane.

In this step, a large number of points in the input point cloud data belong to the ground plane, and the ground plane can be estimated more robustly through the RANSAC algorithm. The ground plane π=(a,b,c,d) ^T is expressed parametrically by a four-dimensional vector, and its reference coordinate system is the observation coordinate system centered on the sensor.

Step S204, randomly generating a three-dimensional area near the ground plane according to the vehicle template.

In the natural lane scenario, assuming that the vehicles are all located near the road surface (that is, the ground level) (regardless of extreme cases such as suspended vehicles), the size template (length, width, height) of the vehicle is input according to the known vehicle type, and the vehicle near the road surface Doing a slight disturbance and traversing the possible orientation angles of the vehicle template at the same time can obtain M possible three-dimensional areas, which can be determined by seven parameters (x, y, z, l, w, h, θ), where (x, y, z) represent the three-dimensional coordinates of the vehicle center; (l, w, h) are the length, width, and height templates of the vehicle; θ represents the orientation angle of the vehicle on the horizontal plane.

In step S206, the three-dimensional space is discretized and the characteristics of each three-dimensional area are obtained.

In this step, the cuboid containing all input 3D point clouds is discretized according to a fixed interval and divided into a set of several small cubes, then for the M 3D regions obtained in step S204, each 3D region will correspond to the A collection of several small cubes. Using the three-dimensional integral map, the number of point clouds and the average height of point clouds contained in each three-dimensional area can be quickly obtained, and the number of point clouds and the average height of point cloud can be used as the characteristics of the three-dimensional area. In other embodiments of the present invention, other features of the three-dimensional area can also be calculated using the three-dimensional integral map, such as the ratio of the number of point clouds, up, down, left, and right, etc., and the present invention is not limited thereto.

Step S208, combining the features linearly to obtain the 3D area score, and sorting to obtain the 3D vehicle candidate area.

In this step, the number of point clouds and the average height of point clouds in each three-dimensional area are weighted and summed according to a certain weight, and the obtained result is used as the score of the three-dimensional area, and the M three-dimensional areas are sorted according to the scores from large to small, Take the first N three-dimensional vehicle candidate areas as the final algorithm output.

The embodiment of the present invention utilizes the prior knowledge that the vehicle is located near the ground level and the template of the known vehicle size, which greatly reduces the number of three-dimensional areas to be searched, improves the search efficiency, and meets the real-time requirements in practical applications. Moreover, since each second geometric body contains at least one point cloud in the road scene, the embodiment of the present invention determines whether the three-dimensional area is a three-dimensional vehicle candidate area based on one or more second geometric bodies contained in each three-dimensional area, ensuring three-dimensional Region recall. In addition, in the embodiment of the present invention, when calculating the features of each three-dimensional area, the three-dimensional integral graph is used, and the calculation complexity is O(1), which saves the amount of calculation and can quickly calculate the features of a large number of three-dimensional areas.

In an optional embodiment of the present invention, during the driving process, the driving-related information can be identified according to the determined three-dimensional vehicle candidate area, and then the identified driving-related information can be analyzed to generate corresponding driving navigation prompt information , and provide it to the user. Here, the driving-related information may include the distance between the preceding vehicle and/or the following vehicle and the own vehicle, driving road condition identification information, road environment information, etc., and the present invention is not limited thereto. Specifically, the road condition identification information, for example, speed limit identification information, height limit identification information, road guidance identification information, instruction identification information, road construction safety identification information, prohibition identification information, warning identification information, etc. of the driving road conditions. Road environment information, for example, building information around the road and construction information around the road.

In an optional embodiment of the present invention, if the recognized distance between the preceding vehicle and the own vehicle is less than a predetermined safe driving distance threshold, corresponding driving navigation prompt information, such as warning information, is generated to inform the user. If the identified distance between the vehicle behind and the own vehicle is less than the predetermined safe driving distance threshold, corresponding driving navigation prompt information, such as warning information, is generated to inform the user. In practical applications, the driving-related information and driving navigation prompt information can be highlighted in association on the display screen, and the driving navigation prompt information can be output through voice playback.

In addition, after identifying the distance between the vehicle in front and/or the vehicle behind and the own vehicle, in addition to generating driving navigation prompt information, the speed of the own vehicle can also be adjusted in real time. For example, if the recognized distance between the vehicle in front and the vehicle in front is less than the predetermined safe driving distance threshold, the vehicle speed will be reduced as soon as possible to avoid collision with the vehicle in front.

In an optional embodiment of the present invention, information such as the driving speed of the vehicle in front or behind can also be statistically analyzed according to the determined three-dimensional vehicle candidate area, so that the vehicle can perform adaptive cruise.

In an optional embodiment of the present invention, the method mentioned above can be applied in a vehicle-mounted intelligent control unit.

It should be noted that, in practical applications, all the above optional implementation manners may be combined arbitrarily in combination to form optional embodiments of the present invention, which will not be repeated here.

Based on the method for determining a three-dimensional vehicle candidate area provided in the above embodiments, and based on the same inventive concept, an embodiment of the present invention also provides a device for determining a three-dimensional vehicle candidate area.

Fig. 3 shows a schematic structural diagram of an apparatus for determining a three-dimensional vehicle candidate area according to an embodiment of the present invention. As shown in FIG. 3 , the device may at least include a point cloud acquisition module 310 , a ground plane determination module 320 , a three-dimensional area generation module 330 , a geometry determination module 340 and a candidate area determination module 350 .

The function of each component or device and the connection relationship between the various parts of the device for determining the three-dimensional vehicle candidate area in the embodiment of the present invention are now introduced:

The point cloud acquisition module 310 is adapted to acquire multiple point clouds in the road scene;

The ground plane determination module 320, coupled with the point cloud acquisition module 310, is adapted to use the multiple point clouds to determine the ground plane in the road scene;

The three-dimensional area generation module 330, coupled with the ground plane determination module 320, is adapted to generate multiple three-dimensional areas on the ground plane according to the preset vehicle size;

The geometry determination module 340, coupled with the three-dimensional region generation module 330, is adapted to divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometry, and determine one or more of the three-dimensional regions contained in each Describe the second geometry;

The candidate area determination module 350 is coupled with the geometry determination module 340 and is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to one or more second geometric bodies included in each three-dimensional area.

In an embodiment of the present invention, the point cloud acquisition module 310 acquires multiple point clouds in the road scene, and this embodiment of the present invention provides a variety of optional solutions, for example, collecting multiple points in the road scene according to the principle of laser measurement Cloud, which can collect the three-dimensional coordinates (XYZ) and laser reflection intensity (Intensity) of the point cloud; or obtain multiple point clouds in the road scene according to the principle of photogrammetry, and can collect the three-dimensional coordinates (XYZ) and color information of the point cloud (RGB); or, the point cloud can also be obtained by combining the principles of laser measurement and photogrammetry, and the three-dimensional coordinates (XYZ), laser reflection intensity (Intensity) and color information (RGB) of the point cloud can be obtained.

First, taking the principle of laser measurement as an example, the point cloud acquisition module 310 may use a three-dimensional laser radar to collect point clouds. Three-dimensional lidar is one of the important sensors for intelligent vehicles to obtain external information. It has the advantages of high reliability, real-time performance and high accuracy, so it is widely used in the research of intelligent vehicle environment perception. The radar has multiple laser sensors, and the discrete data points measured by each sensor can be expressed as Pi(x, y, z, s), where x, y, and z represent three-dimensional physical distances in meters, and s represents the reflection intensity, which is infinite Outline value. The point cloud data is a collection of large-scale discrete measurement point data {P1, P2, P3,..., PN}, which provide sufficient information for restoring the basic shape characteristics and structural details of the measurement object.

Secondly, taking the principle of photogrammetry as an example, the point cloud acquisition module 310 may use binocular cameras to collect point clouds. Place two cameras on the same plane to shoot the target object, and get many sets of corresponding object image pairs. Since the camera and the object conform to the principle of triangulation, and use the parallax of the corresponding points to calculate the stereoscopic information of the object, so as to obtain the object 3D point cloud data.

In an embodiment of the present invention, when the ground plane determining module 320 uses multiple point clouds to determine the ground plane in the road scene, it may use the RANSAC algorithm to determine the point cloud belonging to the ground plane in the road scene among the multiple point clouds. For a detailed introduction to the RANSAC algorithm, please refer to the previous article, and will not repeat it here.

In the embodiment of the present invention, a large number of point clouds in the input point cloud data belong to the ground plane, and the ground plane determining module 320 can estimate the ground plane relatively robustly through the RANSAC algorithm. The ground plane π=(a,b,c,d) ^T is expressed parametrically by a four-dimensional vector, and its reference coordinate system is the observation coordinate system centered on the sensor.

In an embodiment of the present invention, the three-dimensional area generating module 330 generates multiple three-dimensional areas on the ground plane according to the preset vehicle size. This embodiment of the present invention provides an optional solution, in which, it can be based on The preset vehicle size randomly generates multiple three-dimensional areas on the ground plane, where the three-dimensional areas are determined by specifying parameters. The specified parameters here may be the three-dimensional coordinates of the center of the vehicle, the size of the vehicle, the orientation angle of the vehicle on the ground plane, etc., and the present invention is not limited thereto.

Specifically, in the natural lane scenario, assuming that the vehicles are all located near the road surface (i.e., the ground level) (without considering extreme cases such as suspended vehicles), the 3D area generation module 330 inputs the size template of the vehicle according to the known vehicle type (length, Width, height), make a slight disturbance near the road surface, and traverse the possible orientation angles of the vehicle template at the same time, that is, M possible three-dimensional areas can be obtained. These three-dimensional areas can be passed through 7 parameters (x, y, z, l, w, h, θ), where (x, y, z) represents the three-dimensional coordinates of the vehicle center; (l, w, h) are the length, width, and height templates of the vehicle; θ represents the orientation angle of the vehicle on the horizontal plane.

In an embodiment of the present invention, the geometry determination module 340 divides the first geometry containing multiple point clouds into a set of a specified number of second geometry. The cuboid of the point cloud is discretized according to a fixed interval and divided into a set of specified number of cubes. Specifically, divide the cuboid containing all input 3D point clouds into a set of several small cubes according to the fixed interval discretization, then for the M 3D regions obtained above, each 3D region will correspond to several Collection of small cubes. Here, the designated number can be set according to actual needs, for example, if it is necessary to divide smaller cubes, a larger designated number is set.

In an embodiment of the present invention, the candidate area determination module 350 can determine the characteristics of each three-dimensional area according to one or more second geometric bodies contained in each three-dimensional area, and then can determine whether the three-dimensional area is based on the characteristics of each three-dimensional area Alternate regions for 3D vehicles. The characteristics of the three-dimensional area here may be the number of point clouds contained in the three-dimensional area, the average height of the point cloud, the ratio of the number of point clouds, up, down, left, and right, etc., and the present invention is not limited thereto. In an optional embodiment of the present invention, the feature of each three-dimensional area can be calculated by using the three-dimensional integral map, and the calculation complexity is O(1), which saves the amount of calculation and can quickly calculate the features of a large number of three-dimensional areas.

In an embodiment of the present invention, as shown in FIG. 4, the above-mentioned candidate area determination module 350 may include:

The calculation unit 410 is adapted to calculate the number of point clouds and/or the average height of point clouds contained in each three-dimensional area by using a three-dimensional integral graph according to the one or more second geometric bodies contained in each three-dimensional area;

The determination unit 420, coupled with the calculation unit 410, is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud.

In an embodiment of the present invention, the determining unit 420 is further adapted to:

According to the specified weight, the number of point clouds contained in each three-dimensional area and the average height of the point cloud are summed and weighted to obtain the comprehensive score value of each three-dimensional area;

According to the comprehensive score value of each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area.

In an embodiment of the present invention, the determining unit 420 is further adapted to:

Sorting the comprehensive scoring values of the various three-dimensional regions from large to small;

Select the specified number of 3D areas ranked first as the 3D vehicle candidate areas.

In an embodiment of the present invention, as shown in FIG. 4, the device shown in FIG. 3 above may further include:

The driving navigation prompt information generation module 360 is coupled with the candidate area determination module 350, and is suitable for identifying driving-related information according to the determined three-dimensional vehicle candidate areas during driving; for the identified driving-related information Perform analysis to generate corresponding driving navigation prompt information and provide it to users.

In an embodiment of the present invention, the driving-related information may include the distance between the preceding vehicle and/or the following vehicle and the own vehicle, driving road condition identification information, road environment information, etc., and the present invention is not limited thereto. Specifically, the road condition identification information, for example, speed limit identification information, height limit identification information, road guidance identification information, instruction identification information, road construction safety identification information, prohibition identification information, warning identification information, etc. of the driving road conditions. Road environment information, for example, building information around the road and construction information around the road.

In an embodiment of the present invention, if the identified distance between the vehicle in front and the vehicle is less than the predetermined safe driving distance threshold, the driving navigation prompt information generation module 360 generates corresponding driving navigation prompt information, such as warning information, to inform user. If the identified distance between the vehicle behind and the own vehicle is less than the predetermined safe driving distance threshold, corresponding driving navigation prompt information, such as warning information, is generated to inform the user. In practical applications, the driving-related information and driving navigation prompt information can be highlighted in association on the display screen, and the driving navigation prompt information can be output through voice playback.

In an embodiment of the present invention, after recognizing the distance between the preceding vehicle and/or the following vehicle and the own vehicle, in addition to generating driving navigation prompt information, the vehicle speed of the own vehicle can also be adjusted in real time. For example, if the recognized distance between the vehicle in front and the vehicle in front is less than the predetermined safe driving distance threshold, the vehicle speed will be reduced as soon as possible to avoid collision with the vehicle in front.

In an optional embodiment of the present invention, information such as the driving speed of the vehicle in front or behind can also be statistically analyzed according to the determined three-dimensional vehicle candidate area, so that the vehicle can perform adaptive cruise.

In an optional embodiment of the present invention, the above-mentioned device can be applied in a vehicle-mounted intelligent control unit.

Based on the same inventive concept, an embodiment of the present invention also provides a vehicle-mounted device, including any one of the devices for determining a three-dimensional vehicle candidate area described above.

According to any one of the above preferred embodiments or a combination of multiple preferred embodiments, the embodiments of the present invention can achieve the following beneficial effects:

In the technical solution provided by the embodiments of the present invention, firstly, multiple point clouds in the road scene are obtained, and the ground plane in the road scene is determined by using the multiple point clouds. Subsequently, according to the preset vehicle size, multiple three-dimensional regions are generated on the ground plane, and then the first geometry containing multiple point clouds is divided into a set of a specified number of second geometry, and one or more a second geometry. Afterwards, according to one or more second geometric bodies contained in each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area. It can be seen that the embodiment of the present invention utilizes the prior knowledge that the vehicle is located near the ground level, and at the same time utilizes the template of the known vehicle size, which greatly reduces the number of three-dimensional areas to be searched, improves the search efficiency, and satisfies the real-time requirements in practical applications. requirements. Moreover, since each second geometric body contains at least one point cloud in the road scene, the embodiment of the present invention determines whether the three-dimensional area is a three-dimensional vehicle candidate area based on one or more second geometric bodies contained in each three-dimensional area, ensuring three-dimensional Region recall.

In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure the understanding of this description.

Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, in order to streamline this disclosure and to facilitate an understanding of one or more of the various inventive aspects, various features of the invention are sometimes grouped together in a single embodiment, figure, or its description. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the Detailed Description are hereby expressly incorporated into this Detailed Description, with each claim standing on its own as a separate embodiment of this invention.

Those skilled in the art can understand that the modules in the device in the embodiment can be adaptively changed and arranged in one or more devices different from the embodiment. Modules or units or components in the embodiments may be combined into one module or unit or component, and furthermore may be divided into a plurality of sub-modules or sub-units or sub-assemblies. All features disclosed in this specification (including accompanying claims, abstract and drawings) and any method or method so disclosed may be used in any combination, except that at least some of such features and/or processes or units are mutually exclusive. All processes or units of equipment are combined. Each feature disclosed in this specification (including accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

Furthermore, those skilled in the art will understand that although some embodiments described herein include some features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention. and form different embodiments. For example, in the claims, any one of the claimed embodiments can be used in any combination.

The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to implement the device for determining a three-dimensional vehicle candidate area and some or all of the components in the vehicle-mounted equipment according to the embodiment of the present invention. Some or all functions. The present invention can also be implemented as an apparatus or an apparatus program (for example, a computer program and a computer program product) for performing a part or all of the methods described herein. Such a program for realizing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet site, or provided on a carrier signal, or provided in any other form.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means can be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. does not indicate any order. These words can be interpreted as names.

So far, those skilled in the art should appreciate that, although a number of exemplary embodiments of the present invention have been shown and described in detail herein, without departing from the spirit and scope of the present invention, the disclosed embodiments of the present invention can still be used. Many other variations or modifications consistent with the principles of the invention are directly identified or derived from the content. Accordingly, the scope of the present invention should be understood and deemed to cover all such other variations or modifications.

An aspect of the embodiments of the present invention provides A1, a method for determining a three-dimensional vehicle candidate area, including:

Obtaining multiple point clouds in the road scene, using the multiple point clouds to determine the ground plane in the road scene;

generating a plurality of three-dimensional areas on the ground plane according to the preset vehicle size;

Divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometry, and determine one or more of the second geometry contained in each three-dimensional area;

Whether the three-dimensional area is a three-dimensional vehicle candidate area is determined according to the one or more second geometric bodies included in each three-dimensional area.

A2. The method according to A1, wherein said obtaining multiple point clouds in the road scene includes:

Acquisition of multiple point clouds in a road scene based on laser measurement principles; and/or

According to the principle of photogrammetry, multiple point clouds in the road scene are obtained.

A3. The method according to A1 or A2, wherein, using the plurality of point clouds to determine the ground plane in the road scene includes:

A point cloud among the plurality of point clouds belonging to the ground plane in the road scene is determined through a random sampling consistent RANSAC algorithm.

A4. The method according to any one of A1-A3, wherein a plurality of three-dimensional areas are generated on the ground plane according to a preset vehicle size, including:

According to the preset vehicle size, a plurality of three-dimensional areas are randomly generated on the ground plane, wherein the three-dimensional areas are determined by specified parameters.

A5. The method according to A4, wherein the specified parameters include at least one of the following:

The three-dimensional coordinates of the center of the vehicle, the size of the vehicle, and the orientation angle of the vehicle on the ground plane.

A6. The method according to any one of A1-A5, wherein dividing the first geometry containing the plurality of point clouds into a set of a specified number of second geometry includes:

The cuboid containing the plurality of point clouds is discretized according to a fixed interval, and divided into a set of a specified number of cubes.

A7. The method according to A6, wherein, according to the one or more second geometric bodies included in each three-dimensional area, determining whether the three-dimensional area is a three-dimensional vehicle candidate area includes:

According to one or more of the second geometric bodies contained in each three-dimensional area, calculate the number of point clouds contained in each three-dimensional area and/or the average height of point cloud by using a three-dimensional integral map;

Determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud.

A8. The method according to A7, wherein, according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud, determining whether the three-dimensional area is a three-dimensional vehicle candidate area includes:

According to the specified weight, the number of point clouds contained in each three-dimensional area and the average height of the point cloud are summed and weighted to obtain the comprehensive score value of each three-dimensional area;

According to the comprehensive score value of each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area.

A9. The method according to A8, wherein, according to the comprehensive scoring value of each three-dimensional area, determining whether the three-dimensional area is a three-dimensional vehicle candidate area includes:

Sorting the comprehensive scoring values of the various three-dimensional regions from large to small;

Select the specified number of 3D areas ranked first as the 3D vehicle candidate areas.

A10. The method according to any one of A1-A9, further comprising:

In the process of driving, according to the determined three-dimensional vehicle candidate area, identify driving-related information;

The identified driving-related information is analyzed to generate corresponding driving navigation prompt information and provided to the user.

A11. The method according to A10, wherein the driving-related information includes at least one of the following:

The distance between the vehicle in front and/or the vehicle behind and the vehicle, road condition identification information, and road environment information.

A12. The method according to any one of A1-A11, wherein the method is applied in a vehicle-mounted intelligent control unit.

Another aspect of the embodiments of the present invention also provides B13, a device for determining a three-dimensional vehicle candidate area, including:

The point cloud acquisition module is suitable for acquiring multiple point clouds in road scenes;

a ground plane determination module adapted to determine the ground plane in the road scene by using the plurality of point clouds;

A three-dimensional area generation module, adapted to generate multiple three-dimensional areas on the ground plane according to the preset vehicle size;

A geometry determination module, adapted to divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometries, and determine one or more of the second geometries contained in each three-dimensional area;

The candidate area determination module is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to one or more of the second geometric bodies included in each three-dimensional area.

B14. The device according to B13, wherein the point cloud acquisition module is also suitable for:

Acquisition of multiple point clouds in a road scene based on laser measurement principles; and/or

According to the principle of photogrammetry, multiple point clouds in the road scene are obtained.

B15. The device according to B13 or B14, wherein the ground plane determination module is further adapted to:

A point cloud among the plurality of point clouds belonging to the ground plane in the road scene is determined through a random sampling consistent RANSAC algorithm.

B16. The device according to any one of B13-B15, wherein the three-dimensional area generating module is further adapted to:

According to the preset vehicle size, a plurality of three-dimensional areas are randomly generated on the ground plane, wherein the three-dimensional areas are determined by specified parameters.

B17. The device according to B16, wherein the specified parameters include at least one of the following:

The three-dimensional coordinates of the center of the vehicle, the size of the vehicle, and the orientation angle of the vehicle on the ground plane.

B18. The device according to any one of B13-B17, wherein the geometry determination module is further adapted to:

The cuboid containing the plurality of point clouds is discretized according to a fixed interval, and divided into a set of a specified number of cubes.

B19. The device according to B18, wherein the candidate area determination module includes:

The calculation unit is adapted to calculate the number of point clouds and/or the average height of point clouds contained in each three-dimensional area by using a three-dimensional integral map according to the one or more second geometric bodies contained in each three-dimensional area;

The determination unit is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud.

B20. The device according to B19, wherein the determining unit is further adapted to:

According to the specified weight, the number of point clouds contained in each three-dimensional area and the average height of the point cloud are summed and weighted to obtain the comprehensive score value of each three-dimensional area;

According to the comprehensive score value of each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area.

B21. The device according to B20, wherein the determining unit is further adapted to:

Sorting the comprehensive scoring values of the various three-dimensional regions from large to small;

Select the specified number of 3D areas ranked first as the 3D vehicle candidate areas.

B22. The device according to any one of B13-B21, further comprising:

The driving navigation prompt information generation module is adapted to identify driving-related information according to the determined three-dimensional vehicle candidate area during the driving process; analyze the identified driving-related information to generate corresponding driving navigation prompt information, and provided to users.

B23. The device according to B22, wherein the driving-related information includes at least one of the following:

The distance between the vehicle in front and/or the vehicle behind and the vehicle, road condition identification information, and road environment information.

B24. The device according to any one of B13-B23, wherein the device is applied in a vehicle-mounted intelligent control unit.

Still another aspect of the embodiments of the present invention also provides C25, a vehicle-mounted device, including the device for determining a three-dimensional vehicle candidate area described in any one of B13-B24.

Claims (10)

1. A method for determining a three-dimensional vehicle candidate area, comprising: Obtaining multiple point clouds in the road scene, using the multiple point clouds to determine the ground plane in the road scene; generating a plurality of three-dimensional areas on the ground plane according to the preset vehicle size; Divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometry, and determine one or more of the second geometry contained in each three-dimensional area; Whether the three-dimensional area is a three-dimensional vehicle candidate area is determined according to the one or more second geometric bodies included in each three-dimensional area. 2. The method according to claim 1, wherein said obtaining a plurality of point clouds in the road scene comprises: Acquisition of multiple point clouds in a road scene based on laser measurement principles; and/or According to the principle of photogrammetry, multiple point clouds in the road scene are obtained. 3. The method according to claim 1 or 2, wherein, using the plurality of point clouds to determine the ground plane in the road scene comprises: A point cloud among the plurality of point clouds belonging to the ground plane in the road scene is determined through a random sampling consensus RANSAC algorithm. 4. The method according to any one of claims 1-3, wherein, according to a preset vehicle size, generating a plurality of three-dimensional areas on the ground plane, comprising: According to the preset vehicle size, a plurality of three-dimensional areas are randomly generated on the ground plane, wherein the three-dimensional areas are determined by specified parameters. 5. The method of claim 4, wherein the specified parameters include at least one of the following: The three-dimensional coordinates of the center of the vehicle, the size of the vehicle, and the orientation angle of the vehicle on the ground plane. 6. The method according to any one of claims 1-5, wherein dividing the first geometry comprising the plurality of point clouds into a set of specified numbers of second geometry comprises: The cuboid containing the plurality of point clouds is discretized according to a fixed interval, and divided into a set of a specified number of cubes. 7. The method according to claim 6, wherein, according to one or more of the second geometric bodies included in each three-dimensional area, determining whether the three-dimensional area is a three-dimensional vehicle candidate area comprises: According to one or more of the second geometric bodies contained in each three-dimensional area, calculate the number of point clouds contained in each three-dimensional area and/or the average height of point cloud by using a three-dimensional integral map; According to the number of point clouds contained in each three-dimensional area and/or the average height of the point cloud, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area. 8. The method according to claim 7, wherein, according to the number of point clouds contained in each of the three-dimensional areas and/or the average height of the point clouds, determining whether the three-dimensional area is a three-dimensional vehicle candidate area comprises: According to the specified weight, the number of point clouds contained in each three-dimensional area and the average height of the point cloud are summed and weighted to obtain the comprehensive score value of each three-dimensional area; According to the comprehensive score value of each three-dimensional area, it is determined whether the three-dimensional area is a three-dimensional vehicle candidate area. 9. A device for determining a three-dimensional vehicle candidate area, comprising: The point cloud acquisition module is suitable for acquiring multiple point clouds in road scenes; a ground plane determination module adapted to determine the ground plane in the road scene by using the plurality of point clouds; A three-dimensional area generation module, adapted to generate multiple three-dimensional areas on the ground plane according to the preset vehicle size; A geometry determination module, adapted to divide the first geometry containing the plurality of point clouds into a set of a specified number of second geometries, and determine one or more of the second geometries contained in each three-dimensional area; The candidate area determining module is adapted to determine whether the three-dimensional area is a three-dimensional vehicle candidate area according to one or more of the second geometric bodies included in each three-dimensional area. 10. A vehicle-mounted device, comprising the device for determining a three-dimensional vehicle candidate area according to claim 9.