Contour extraction method, intelligent device and storage medium for drivable area

By emitting rays from the center of the intelligent device to obtain the pillars and using a connectivity tree diagram to cluster the boundary points, the problem of low efficiency and low accuracy in obtaining the contour of the drivable area in the existing technology is solved, and efficient and accurate contour extraction is achieved.

CN118674744BActive Publication Date: 2024-11-22安徽蔚来智驾科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411139393.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-20
Publication Date
2024-11-22
Estimated Expiration
2044-08-20

AI Technical Summary

Technical Problem

Existing technologies are inefficient and inaccurate in acquiring the contours of drivable areas in autonomous driving, especially grid-based methods which suffer from quantization and boundary position errors.

Method used

By emitting multiple rays from the center of the smart device, the intersection points of the rays and the drivable area are obtained to form pillars. The boundary points of the pillars are then traversed sequentially, and the pillars are clustered using a connectivity tree diagram to obtain the outline of the drivable area.

Benefits of technology

It improves the efficiency and accuracy of acquiring the contours of drivable areas, reduces the amount of data transmission, and enhances the performance of autonomous driving systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118674744B_ABST
    Figure CN118674744B_ABST
Patent Text Reader

Abstract

The present application relates to the field of autonomous driving technology, and specifically provides a method for extracting the contour of a drivable area, an intelligent device, and a storage medium, aiming to solve the problem of efficiently and accurately obtaining the contour of a drivable area. To this end, the method provided by the present application includes obtaining the drivable area in the environment where the intelligent device is located, emitting multiple rays from the center within a range of a preset angle centered on the intelligent device and around the center; obtaining multiple columns according to the intersection of the multiple rays and the drivable area, the columns are sub-areas of the drivable area located in the fan-shaped area, and the fan-shaped area is the area between two adjacent rays; traversing the boundary points of each column in turn and sorting the boundary points according to the traversal order of the boundary points to obtain a set of boundary points; obtaining the contour of the drivable area according to the set of boundary points; the boundary points of the columns include the points at which the columns are farthest and closest to the center. Through the above method, the efficiency and accuracy of contour acquisition can be greatly improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a method for extracting the contour of a drivable area, an intelligent device, and a storage medium. Background Art

[0002] When the vehicle is controlled automatically, the sensors (such as radar) installed on the vehicle can be used to collect data frames of the vehicle's environment, and the perception model installed on the vehicle can be used to perceive the environment of the above data frames to obtain information such as the drivable area and obstacles in the vehicle's environment, and then the vehicle can be regulated and controlled based on this information (such as planning the vehicle's driving path and determining the vehicle's control decision). At present, the expression of the drivable area obtained by the perception model includes polygonal drivable areas and drivable areas in the form of grid occupation. Since the drivable area is obtained by perceiving a single data frame, the perception result is not very stable, and there are false detections or missed detections. Therefore, before the vehicle is regulated and controlled, it is necessary to perform time series fusion on the drivable area obtained by using a single data frame, and then regulate and control the vehicle based on the drivable area after time series fusion. For example, the contour of the drivable area after time series fusion is extracted, and the vehicle's driving path is planned according to the extracted contour.

[0003] At present, the conventional temporal fusion method and contour extraction method of the drivable area are mainly implemented based on the drivable area in the form of grid occupation. This method has the disadvantages of low efficiency and low precision. Specifically, the drivable area in the form of grid occupation obtained by the perception model is mainly to divide the rectangular ground area within a certain range centered on the vehicle into multiple grids, and identify the state of each grid; if the grid category is ground or obstacle, the grid state is identified as occupied; otherwise, the grid state is identified as unoccupied. Whether it is temporal fusion of the drivable area or contour extraction of the drivable area, all grids need to be traversed, resulting in low execution efficiency. Although quad tree and other methods can be used for optimization, the execution efficiency cannot be effectively improved. In addition, there is a quantization error in the drivable area based on the grid occupation form. For example, even if the drivable area only occupies a part of a grid, it is still considered that the drivable area occupies the entire grid. Based on this, after completing the temporal fusion or contour extraction, there will be a large error in the boundary position of the drivable area. The above quantization error can reach the grid resolution at most.

[0004] Accordingly, the art needs a new technical solution to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects, the present application is proposed to solve or at least partially solve the technical problem of how to efficiently and accurately obtain the contour of the drivable area.

[0006] In a first aspect, a method for extracting a contour of a drivable area is provided, the method comprising:

[0007] Acquire a drivable area in the environment where the smart device is located, where the drivable area is a polygonal drivable area obtained by sensing the environmental data through a perception model on the smart device;

[0008] Within a range of a preset angle around the smart device and centered at the center, emitting a plurality of rays from the center;

[0009] Obtain intersections of the plurality of rays and the drivable area, and obtain a plurality of pillars according to the intersections, wherein the pillars are sub-areas of the drivable area within a fan-shaped area, and the fan-shaped area is an area formed between two adjacent rays;

[0010] Traversing the boundary points of each of the plurality of columns in sequence, and sorting the boundary points of each column according to the traversal order of the boundary points of each column, so as to obtain a boundary point set;

[0011] Acquire the outline of the drivable area according to the set of boundary points;

[0012] in,

[0013] The boundary points of the column include an upper boundary point and a lower boundary point, and the upper boundary point and the lower boundary point are respectively the points of the column farthest and closest to the center.

[0014] In a technical solution of the above-mentioned method for extracting the contour of the drivable area, the step of obtaining a plurality of pillars according to the intersection points includes:

[0015] For each ray, if a line segment formed between two adjacent intersection points on the ray is located within the drivable area, the line segment is used as the column.

[0016] In a technical solution of the above-mentioned method for extracting the outline of the drivable area, the method includes determining whether the line segment is located in the drivable area by the following method:

[0017] The intersection points of the two adjacent intersection points that are farthest and closest to the center are respectively taken as the first point and the second point, and the target contour of the drivable area located between the first point and the second point is obtained along a first preset direction, where the first preset direction is a clockwise direction or a counterclockwise direction;

[0018] Acquire a third point closest to the first point on the target contour, and acquire a first vector with the first point and the third point as the starting point and the end point respectively;

[0019] Acquire a fourth point on the target contour that is closest to the second point, and acquire a second vector with the fourth point and the second point as the starting point and the end point respectively;

[0020] Obtaining the midpoint of the line segment formed by the two adjacent intersection points;

[0021] If the midpoint is located on the target side of the first vector and the second vector respectively, the line segment is located within the drivable area;

[0022] in,

[0023] The target sides of the first vector and the second vector are respectively: the sides to which the vector directions of the first vector and the second vector first point when they rotate along the first preset direction.

[0024] In a technical solution of the above-mentioned method for extracting the contour of a drivable area, there are multiple drivable areas, and the method comprises: clustering pillars belonging to the same drivable area into a pillar set;

[0025] The step of traversing the boundary points of each column in the plurality of columns in turn and sorting the boundary points of each column according to the traversal order of the boundary points of each column to obtain a boundary point set includes: for each column set, traversing the boundary points of each column in the column set in turn and sorting the boundary points of each column according to the traversal order of the boundary points of each column to obtain a boundary point set corresponding to the column set.

[0026] In a technical solution of the above-mentioned method for extracting the contour of the drivable area, clustering the pillars belonging to the same drivable area into a pillar set includes:

[0027] Establishing a connectivity dendrogram of each of the plurality of columns respectively;

[0028] According to the connectivity dendrogram of each column, searching for columns with connectivity and clustering the columns with connectivity into the same column set;

[0029] in,

[0030] The connectivity tree diagram of the first column includes: a root node, a right subtree, and a left subtree, wherein the right subtree includes a right node connected to the root node, and the left subtree includes a left node connected to the root node, and the first column is any one of the plurality of columns;

[0031] The root node represents the first column;

[0032] The right node represents a second column on the right side of the first column and connected with the first column, the left node represents a second column on the left side of the first column and connected with the first column, and the second column is the remaining column in the plurality of columns;

[0033] The right side and the left side of the first column are respectively: the side that the column direction of the first column points to first when rotating in the clockwise direction and the counterclockwise direction, and the column direction is the direction from the lower boundary point of the first column to the upper boundary point.

[0034] In a technical solution of the above-mentioned method for extracting the outline of the drivable area, the method includes: if the first pillar and the second pillar are adjacent to each other and overlap, then the first pillar and the second pillar are connected.

[0035] In a technical solution of the above-mentioned method for extracting the contour of the drivable area, searching for connected pillars and clustering the connected pillars into the same pillar set includes:

[0036] Step S1: Initialize all columns as unvisited;

[0037] Step S2: selecting an unvisited column as the third column, setting a column set corresponding to the third column, adding the third column to the column set, and modifying the third column to be visited;

[0038] Step S3: according to the connectivity tree diagram of the third column, obtain a fourth column, add the fourth column to the column set, and modify the fourth column to be visited, wherein the fourth column is the column represented by the right node and / or the left node in the connectivity tree diagram;

[0039] Step S4: acquiring a fifth column according to the connectivity tree diagram of the fourth column; if the fifth column is unvisited, adding the fifth column to the column set, and modifying the fifth column to be visited, the fifth column being the column represented by the right node and / or the left node in the connectivity tree diagram of the fourth column;

[0040] Step S5: taking the fifth column as a new fourth column, and repeating step S4 until no unvisited column is obtained, so as to complete clustering of the column set;

[0041] Step S6: Repeat steps S2 to S5 for the remaining unvisited columns until all columns are modified to have been visited.

[0042] In a technical solution of the above-mentioned method for extracting the contour of the drivable area, the step of sequentially traversing the boundary points of each pillar in the pillar set includes:

[0043] According to the connectivity tree diagram of each column in the column set and the preset traversal rule, traverse the boundary points of each column in sequence along the second preset direction;

[0044] in,

[0045] The preset traversal rules include a first traversal rule and a second traversal rule;

[0046] The first traversal rule includes:

[0047] When the second preset direction is clockwise: if the currently traversed boundary point is the upper boundary point of the first target column, then the next traversal is the second target column represented by the right node in the target tree diagram, and the target tree diagram is the connectivity tree diagram of the first target column; if the currently traversed boundary point is the lower boundary point of the first target column, then the next traversal is the third target column represented by the left node in the target tree diagram;

[0048] When the second preset direction is counterclockwise: if the currently traversed boundary point is the upper boundary point of the first target column, the next traversal is the third target column; if the currently traversed boundary point is the lower boundary point of the first target column, the next traversal is the second target column;

[0049] The second traversal rule includes: if the currently traversed boundary point is the upper boundary point of the first target column, then the next traversal is the upper boundary point; if the currently traversed boundary point is the lower boundary point of the first target column, then the next traversal is the lower boundary point.

[0050] In a technical solution of the above-mentioned method for extracting the contour of the drivable area, the preset traversal rule includes a third traversal rule, and the third traversal rule includes:

[0051] When the second preset direction is clockwise: if the currently traversed boundary point is the upper boundary point of the first target column and the target tree diagram has no right node, the next traversal is the lower boundary point of the first target column; if the currently traversed boundary point is the lower boundary point of the first target column and the target tree diagram has no left node, the next traversal is the upper boundary point of the first target column;

[0052] When the second preset direction is counterclockwise: if the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has no left node, the next traversal is the lower boundary point of the first target column; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has no right node, the next traversal is the upper boundary point of the first target column.

[0053] In a technical solution of the above-mentioned method for extracting the outline of the drivable area, in the connectivity tree diagram of the first pillar, the right nodes are arranged sequentially from left to right in the right subtree, and after the sequential arrangement, the distance between the pillar represented by the right node and the center is sequentially reduced; the left nodes are arranged sequentially from left to right in the left subtree, and after the sequential arrangement, the distance between the pillar represented by the left node and the center is sequentially reduced;

[0054] The preset traversal rule includes a fourth traversal rule, and the fourth traversal rule includes:

[0055] When the second preset direction is clockwise: if the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has multiple right nodes, then the upper boundary point of the fourth target column represented by the first right node in the target tree diagram is traversed next, and the first right node is the right node arranged at the leftmost side of the right subtree; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has multiple left nodes, then the lower boundary point of the fifth target column represented by the last left node in the target tree diagram is traversed next, and the last left node is the left node arranged at the rightmost side of the left subtree;

[0056] When the second preset direction is counterclockwise: if the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has multiple left nodes, the next traversal is the upper boundary point of the fifth target column; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has multiple right nodes, the next traversal is the lower boundary point of the fourth target column.

[0057] In a technical solution of the above-mentioned method for extracting the outline of the drivable area, in the connectivity tree diagram of the first pillar, the right nodes are arranged sequentially from left to right in the right subtree, and after the sequential arrangement, the distance between the pillar represented by the right node and the center is sequentially reduced; the left nodes are arranged sequentially from left to right in the left subtree, and after the sequential arrangement, the distance between the pillar represented by the left node and the center is sequentially reduced;

[0058] The preset traversal rule includes a fifth traversal rule, and the fifth traversal rule includes:

[0059] When the second preset direction is clockwise:

[0060] If the currently traversed boundary point is the upper boundary point of the first target column and the next traversed sixth target column is a one-to-many column, then in the connectivity tree diagram of the sixth target column, it is detected whether there is a second node on the left side of the first node representing the first target column; if there is a second node, the lower boundary point of the seventh target column represented by the second node is traversed; if there is no second node, the boundary point of the sixth target column is traversed; the one-to-many column is a column whose connectivity tree diagram has multiple right nodes and / or multiple left nodes;

[0061] If the currently traversed boundary point is the lower boundary point of the first target column and the next traversed sixth target column is the one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether there is a third node on the right side of the first node; if there is a third node, traverse the upper boundary point of the eighth target column represented by the third node; if there is no third node, traverse the boundary points of the sixth target column;

[0062] When the second preset direction is counterclockwise:

[0063] If the currently traversed boundary point is the upper boundary point of the first target column and the next traversed sixth target column is the one-to-many column, then in the connectivity tree diagram of the sixth target column, it is detected whether there is the third node on the right side of the first node; if there is the third node, then the lower boundary point of the eighth target column represented by the third node is traversed; if there is no third node, then the boundary points of the sixth target column are traversed;

[0064] If the currently traversed boundary point is the lower boundary point of the first target column and the next traversed sixth target column is the one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether the second node exists on the left side of the first node; if there is the second node, traverse the upper boundary point of the seventh target column represented by the second node; if there is no second node, traverse the boundary points of the sixth target column.

[0065] In a second aspect, an intelligent device is provided, comprising at least one processor; and a memory communicatively connected to the at least one processor; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method described in any one of the technical solutions of the above-mentioned method for extracting the contour of the drivable area is implemented.

[0066] In a third aspect, a computer-readable storage medium is provided, wherein a plurality of program codes are stored therein, wherein the program codes are suitable for being loaded and run by a processor to execute the method described in any one of the technical solutions of the above-mentioned method for extracting the contour of the drivable area.

[0067] The above one or more technical solutions of the present application have at least one or more of the following beneficial effects:

[0068] In a technical solution for implementing the contour extraction method of the drivable area provided in the present application, the drivable area in the environment where the smart device is located can be obtained, and the drivable area is a polygonal drivable area obtained by the perception model on the smart device perceiving the environmental data; within the range of a preset angle centered on the smart device and around the center, multiple rays are emitted from the center, and the angles between all two adjacent rays are the same; the intersection points of the multiple rays and the drivable area are obtained, and multiple columns are obtained according to the intersection points, and the columns are sub-areas of the drivable area located in the fan-shaped area, and the fan-shaped area is the area formed between two adjacent rays; the boundary points of each column in the multiple columns are traversed in turn, and the boundary points of each column are sorted according to the traversal order of the boundary points of each column to obtain a boundary point set; according to the boundary point set, the contour of the drivable area is obtained; wherein the boundary points of the column include an upper boundary point and a lower boundary point, and the upper boundary point and the lower boundary point are the points at which the column is farthest and closest to the center, respectively.

[0069] The above implementation scheme uses rays to perform spatial division of the polygonal drivable area, while the prior art uses grids to perform spatial division of the rectangular ground area when extracting contours of the drivable area based on grid occupation. The number of rays obtained after spatial division in the above implementation scheme is much smaller than the number of grids obtained after spatial division in the above prior art. For example, in some scenes, the number of rays is 1,200 and the number of grids is 250,000. Since the number of rays is much smaller than the number of grids, the efficiency of traversing rays when extracting contours is much higher than the efficiency of traversing grids, so that the efficiency of extracting contours in the above implementation scheme is much higher than the efficiency of extracting contours in the above prior art.

[0070] In addition, the above implementation scheme traverses the upper and lower boundary points of the column when extracting the contour, rather than traversing all the boundary points of the column. Based on this, the efficiency of contour extraction can be further improved.

[0071] In addition, the accuracy of the contour can be flexibly adjusted by adjusting the density of the rays. If higher contour accuracy is required, the density of the rays can be increased. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] The disclosure of the present application will become easier to understand with reference to the accompanying drawings. It is easy for those skilled in the art to understand that these drawings are only for illustrative purposes and are not intended to limit the scope of protection of the present application. Among them:

[0073] Figure 1 It is a schematic flow chart of the main steps of a method for extracting the contour of a drivable area according to an embodiment of the present application;

[0074] Figure 2 is a schematic diagram of a drivable area, rays and pillars according to an embodiment of the present application;

[0075] Figure 3 is a schematic diagram of a drivable area, rays and intersection points according to an embodiment of the present application;

[0076] Figure 4 is a flowchart of main steps for determining whether a line segment formed between two adjacent intersection points on a ray is within a drivable area according to an embodiment of the present application;

[0077] Figure 5 According to one embodiment of the present application, Figure 3 Indication of whether the line segment between intersection point 1 and intersection point 2 is within the drivable area Figure 1 ;

[0078] Figure 6 According to an embodiment of the present application Figure 3 Indication of whether the line segment between intersection point 1 and intersection point 2 is within the drivable area Figure 2 ;

[0079] Figure 7 is a schematic diagram of determining whether a line segment between two adjacent intersection points is located in a drivable area according to an embodiment of the present application Figure 1 ;

[0080] Figure 8 is a schematic diagram of determining whether a line segment between two adjacent intersection points is located in a drivable area according to an embodiment of the present application Figure 2 ;

[0081] Fig. 9 It is a flowchart of the main steps of clustering pillars belonging to the same drivable area into a pillar set according to an embodiment of the present application;

[0082] Fig.10 is a schematic diagram of a connected column according to an embodiment of the present application;

[0083] Fig.11 is a schematic diagram of a connectivity tree diagram of a column according to an embodiment of the present application;

[0084] Fig.12 This is a schematic flow chart of the main steps of obtaining a set of pillars according to a drivable area according to an embodiment of the present application;

[0085] Fig.13 This is a schematic diagram of a column assembly according to an embodiment of the present application. Figure 1 ;

[0086] Fig.14 This is a schematic diagram of a column assembly according to an embodiment of the present application. Figure 2 ;

[0087] Fig.15 This is a schematic diagram of a column assembly according to an embodiment of the present application. Figure 3 ;

[0088] Fig.16 is a schematic diagram of a polygonal drivable area according to an embodiment of the present application;

[0089] Fig.17 yes Fig.16 A schematic diagram of the intersection of the drivable area with the vehicle as the center and the rays emitted from the center within a range of 360° around the center;

[0090] Fig.18 is based on Fig.17 Schematic diagram of the outline of the drivable area obtained by the rays in;

[0091] Fig.19 It is a schematic diagram of the main structure of a smart device according to an embodiment of the present application.

[0092] Reference numerals:

[0093] 11: memory; 12: processor. DETAILED DESCRIPTION

[0094] Some embodiments of the present application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application and are not intended to limit the protection scope of the present application.

[0095] In the description of this application, "processor" may include hardware, software or a combination of the two. The processor may be a central processing unit, a microprocessor, an image processor, a digital signal processor or any other suitable processor. The processor has data and / or signal processing functions. The processor may be implemented in software, hardware or a combination of the two. Computer-readable storage media include any suitable medium that can store program code, such as a magnetic disk, a hard disk, an optical disk, a flash memory, a read-only memory, a random access memory, etc. The term "A and / or B" means all possible combinations of A and B, such as just A, just B or A and B.

[0096] The relevant user personal information that may be involved in the various embodiments of this application is strictly in accordance with the requirements of laws and regulations, following the principles of legality, legitimacy and necessity, based on the reasonable purposes of business scenarios, to process the personal information that users actively provide during the use of products / services or generated due to the use of products / services, as well as the personal information obtained with the user's authorization.

[0097] The user personal information processed by this application will vary depending on the specific product / service scenario, and shall be based on the specific scenario in which the user uses the product / service, and may involve the user's account information, device information, driving information, vehicle information or other related information. This application will treat the user's personal information and its processing with a high degree of diligence.

[0098] This application attaches great importance to the security of user personal information and has taken reasonable and feasible security protection measures that meet industry standards to protect user information and prevent personal information from being accessed, disclosed, used, modified, damaged or lost without authorization.

[0099] Some of the terms involved in this application are explained below.

[0100] The smart device described in this application may include driving devices, smart cars, robots and other devices. In some embodiments of the present application, the smart device may also include at least one sensor, which is used to sense information. The sensor is communicatively connected to any type of processor mentioned in this application. Optionally, the smart device may also include an automatic driving system, which is used to guide the smart device to drive itself or assist in driving. The sensor may include a camera, a radar (such as a lidar), etc. The sensor can be used to collect environmental data of the environment in which the smart device is located. For example, an image of the environment in which the smart device is located can be collected by a camera, and a point cloud of the environment can be collected by a radar.

[0101] A perception model may be provided on the smart device, and the perception model may be used to perceive the environmental data collected by the sensor to obtain a polygonal drivable area in the environment where the smart device is located. For example, the point cloud frame collected by the laser radar is perceived by the perception model to obtain a polygonal drivable area, and the polygonal drivable area is composed of point clouds in the point cloud frame; the image frame collected by the camera is perceived by the perception model to obtain a polygonal drivable area, and the polygonal drivable area is composed of pixels in the image frame. In this embodiment, a conventional perception model may be used, and the specific structure and working principle of the perception model will not be described in detail.

[0102] The following is a description of an embodiment of the method for extracting the contour of a drivable area provided in the present application.

[0103] See attached Figure 1 , Figure 1 FIG. 1 is a flow chart showing the main steps of a method for extracting the contour of a drivable area according to an embodiment of the present application. Figure 1 As shown, the contour extraction method of the drivable area in the embodiment of the present application mainly includes the following steps S101 to S105.

[0104] Step S101: obtaining a drivable area in the environment where the smart device is located, where the drivable area is a polygonal drivable area obtained by sensing environmental data through a perception model on the smart device.

[0105] Step S102: within a range of a preset angle around the smart device as the center, multiple rays are emitted from the center, and the angles between all two adjacent rays are the same, that is, the range is equally divided by the multiple rays.

[0106] When setting the value of the preset angle, those skilled in the art may set the value according to the range of the drivable area that they want to focus on. For example, in some embodiments, the preset angle may be 360°, that is, multiple rays are emitted from the center within a circle around the center.

[0107] The smaller the angle between two adjacent rays, the denser the rays are, and the more accurate the contour position of the drivable area obtained by using the rays is. Therefore, when determining the value of the angle between two adjacent rays, those skilled in the art can set the value according to the desired contour accuracy of the drivable area. For example, in some embodiments, the angle between two adjacent rays can be 0.3°; if the above preset angle is 360°, then 1200 rays can be emitted from the center within the range of the preset angle around the center.

[0108] Step S103: Obtain intersections of a plurality of rays and the drivable area, and obtain a plurality of pillars according to the intersections. The pillars are sub-areas of the drivable area within the fan-shaped area. The fan-shaped area is an area formed between two adjacent rays.

[0109] like Figure 2 As shown, the smart device is a vehicle, and the environment where the vehicle is located has drivable area 1 and drivable area 2. Multiple rays are emitted from the center within a range of 360° with the vehicle as the center. All rays will pass through drivable area 2, and some rays will pass through drivable area 1. When passing through the drivable area, the rays will intersect or be tangent to the drivable area. The intersection and tangency points can be understood as the intersection points of the rays and the drivable area. Two adjacent rays will form a fan-shaped area, and the sub-area of ​​the drivable area within this fan-shaped area is the column. Figure 2 Taking ray 1 and ray 2 in the figure as an example, the drivable area has three sub-areas in the fan-shaped area formed by rays 1 and 2. Figure 2 One of the three sub-regions is labeled, namely Figure 2 The pillar in the.

[0110] Step S104: traverse the boundary points of each column in the plurality of columns in turn, and sort the boundary points of each column according to the traversal order of the boundary points of each column to obtain a boundary point set. The boundary points of the column include an upper boundary point and a lower boundary point, and the upper boundary point and the lower boundary point are the points of the column farthest and closest to the center, respectively.

[0111] Specifically, a boundary point (upper boundary point or lower boundary point) can be selected as the starting point, and starting from this starting point, the remaining boundary points are traversed in a fixed traversal direction (such as clockwise or counterclockwise). The traversal order of each boundary point (upper boundary point or lower boundary point) can be determined by traversal, and each boundary point (upper boundary point or lower boundary point) is sorted according to the traversal order, and a boundary point set is obtained according to the sorted boundary points. In other words, the boundary point set includes the boundary points of each column, and these boundary points are arranged in sequence according to the traversal order.

[0112] Step S105: Obtain the outline of the drivable area according to the boundary point set.

[0113] Specifically, the boundary points of the boundary point set may be used as contour points.

[0114] Taking a vehicle as an example, the present application is compared with the method of obtaining contours using a grid in the prior art. The method of obtaining contours using a grid is a method of extracting contours of a drivable area occupied by a grid.

[0115] When the existing technology uses grids to obtain contours, for a rectangular ground area centered on the vehicle and 50 meters in front, back, left, and right, 250,000 grids can be obtained by dividing the grid according to a grid resolution of 0.2m. When obtaining the grid corresponding to the drivable area (grids with the category of ground and grid status of occupied state), 250,000 grids need to be traversed, which seriously reduces the efficiency of contour acquisition; in addition, when the drivable area only occupies a part of a grid, the entire grid will be used to extract the contour, resulting in a large error between the extracted contour and the actual contour.

[0116] When the method provided by the present application is used to obtain the contour, 1200 rays are emitted within a range of 360° around the center with the vehicle as the center, and the angle between two adjacent rays is 0.3. When obtaining the contour, only 1200 rays need to be traversed, which can greatly improve the efficiency of contour acquisition compared to traversing 250,000 grids. In addition, when the distance from the vehicle is 50 meters, the lateral resolution of the two rays is 0.26m, which is basically the same as the above-mentioned grid resolution of 0.2m; and the longitudinal resolution of the rays is much smaller than the above-mentioned grid resolution of 0.2m, which can accurately indicate the position of the drivable area. In addition, when the prior art uses the grid to obtain the contour, if the contour data (such as the occupancy status of the grid) is to be transmitted, 250,000 bytes of data need to be transmitted, while when the method provided by the present application is used, if the contour data of the drivable area is to be transmitted, about 5000 bytes of data are required, and the bandwidth pressure of the present application for transmitting data is only 2% of the above-mentioned prior art. Among them, the lateral resolution can be understood as the distance between two adjacent rays. Since the distance between two adjacent rays is fan-shaped, the lateral resolution is not fixed. The farther away from the vehicle, the larger the lateral resolution. However, since the rays are dense, the lateral resolution is much smaller than the above-mentioned 0.2m grid resolution even when the distance from the vehicle is 50 meters. The ray is composed of a series of points, and the longitudinal resolution can be understood as the distance between two adjacent points on the ray.

[0117] It can be seen that the method based on the above steps S101 to S105 can greatly improve the efficiency and accuracy of obtaining the contour of the drivable area compared to the prior art.

[0118] The above-mentioned step S103 and step S104 are described below.

[0119] 1. Explain step S103.

[0120] In some implementations of the above step S103, when the rays are dense, the pillars between two adjacent rays can be approximated as a section or line segment on the rays. Based on this, in this implementation, for each ray, the pillars can be obtained by the following method:

[0121] Determine whether the line segment formed between two adjacent intersection points on the ray is within the drivable area; if so, use this line segment as a pillar; otherwise, do not use this line segment as a pillar. Figure 3 As shown, the ray intersects with the drivable area to generate multiple intersections. Intersection 1 and intersection 2 are two adjacent intersections. The line segment formed by the two is located in the drivable area. Therefore, this line segment can be used as a column. In this embodiment, the column can be represented by [ray_index, rad_min, rad_max], where ray_index represents the index number of the ray to which the column belongs, rad_min represents the position of the lower boundary point of the column, and rad_max represents the position of the upper boundary point of the column.

[0122] Through this implementation, the pillars can be acquired quickly and accurately.

[0123] In some implementations of the above step S103, Figure 4 The following steps S201 to S205 are shown to determine whether a line segment formed between two adjacent intersection points on the ray is located within the drivable area.

[0124] Step S201: The intersection point farthest and closest to the center among two adjacent intersection points are respectively taken as the first point and the second point, and the target contour of the drivable area between the first point and the second point is obtained along a first preset direction.

[0125] The first preset direction is clockwise or counterclockwise. Figure 3 As shown, intersection 1 is the first point and intersection 2 is the second point. Taking intersection 1 as the starting point and moving from this starting point to intersection 2 in a clockwise direction, the target contour can be obtained; taking intersection 2 as the starting point and moving from this starting point to intersection 1 in a counterclockwise direction, the target contour can also be obtained.

[0126] Since the target contour is the contour between two intersection points, it is much shorter than the contour of the entire drivable area, so obtaining the target contour does not waste a lot of time, thereby affecting efficiency. In addition, in some embodiments, it is not necessary to obtain each contour point of the target contour, and only the two contour points closest to the first point and the second point on the target contour can be obtained. Therefore, the above process can be understood as first determining the area where the target contour is located, and then obtaining two contour points on the target contour respectively, which will not waste time and will not affect efficiency.

[0127] Step S202: Acquire the third point closest to the first point on the target contour, and acquire a first vector with the first point and the third point as the starting point and the end point respectively.

[0128] Step S203: Acquire a fourth point closest to the second point on the target contour, and acquire a second vector with the fourth point and the second point as the starting point and the end point respectively.

[0129] Specifically, a conventional vector acquisition method may be used to acquire a first vector according to the positions of the first point and the third point, and to acquire a second vector according to the positions of the fourth point and the second point.

[0130] Step S204: Obtain the midpoint of the line segment formed by two adjacent intersection points.

[0131] Step S205: If the midpoints are located on the target sides of the first vector and the second vector respectively, the line segment is located in the drivable area; otherwise, the line segment is not located in the drivable area.

[0132] The target side of the first vector is the side that the vector direction of the first vector first points to when rotating along the first preset direction. The target side of the second vector is the side that the vector direction of the second vector first points to when rotating along the first preset direction. The first preset direction is the first preset direction in step S201.

[0133] In this embodiment, the side that the vector directions of the first and second vectors point to first when rotating in the clockwise direction is described as the right side, and the side that the vector directions of the first and second vectors point to first when rotating in the counterclockwise direction is described as the left side. Therefore, when the first preset direction is the clockwise direction, if the midpoint of the line segment is located on the right side of both the first and second vectors, the line segment is located in the drivable area; when the first preset direction is the counterclockwise direction, if the midpoint of the line segment is located on the left side of both the first and second vectors, the line segment is located in the drivable area.

[0134] The following is combined with Figure 5 To Attachment Figure 8 , the above-mentioned implementation mode is explained.

[0135] 1. A scenario in which the first preset direction is clockwise and the line segment is within the drivable area is described.

[0136] First, refer to the attached Figure 5 , Figure 5 To determine Figure 3 Indication of whether the line segment between intersection point 1 and intersection point 2 is within the drivable area Figure 1 The first and second points represent Figure 3 The first preset direction is clockwise. Based on the target contour obtained by the first preset direction, the third point closest to the first point and the fourth point closest to the second point on the target contour can be obtained. The first vector can be obtained according to the first and third points, and the second vector can be obtained according to the fourth and second points. The vector directions of the first and second vectors are as follows: Figure 5As shown by the arrow in . Since the first preset direction is clockwise, the target side of the first and second vectors is the right side. Figure 5 As shown, the midpoint of the line segment formed between the first point and the second point is located on the right side of both the first and second vectors. Therefore, this line segment is within the drivable area.

[0137] See attached Figure 6 , Figure 6 To determine Figure 3 Indication of whether the line segment between intersection point 1 and intersection point 2 is within the drivable area Figure 2 .like Figure 5 As shown, after obtaining the fourth point, the fourth point can be moved along the vector direction of the second vector to move the fourth point in front of the second point to obtain Figure 6 The fifth point shown in the figure is then used as the starting point and the end point to obtain the third vector, and the third vector is used to replace the second vector. Then, it is determined that the midpoint of the line segment formed between the first point and the second point is located on the right side of both the first and third vectors, as shown in the figure. Figure 6 As shown, the midpoint is on the right side of the first and third vectors at the same time, so this line segment is within the drivable area.

[0138] 2. Explain the scenarios where the first preset direction is counterclockwise and the line segment is located within and not located within the drivable area.

[0139] (1) Describe the scenario where the line segment is within the drivable area.

[0140] See attached Figure 7 , Figure 7 Indication for determining whether the line segment between two adjacent intersection points is within the drivable area Figure 1 .like Figure 7 As shown, since the first preset direction is counterclockwise, the target side of the first and second vectors is the left side. The midpoint of the line segment formed between the first point and the second point is located on the left side of both the first and second vectors, so this line segment is located in the drivable area.

[0141] (2) Explain the scenario where the line segment is not within the drivable area.

[0142] See attached Figure 8 , Figure 8 Indication for determining whether the line segment between two adjacent intersection points is within the drivable area Figure 2 .like Figure 8 As shown, since the first preset direction is counterclockwise, the target side of the first and second vectors is the left side. The midpoint of the line segment formed between the first point and the second point is located on the right side of both the first and second vectors, not on the left side. Therefore, this line segment is not located in the drivable area.

[0143] 2. Explain step S104.

[0144] In some implementations of the above step S104, when there are multiple drivable areas, in order to improve the efficiency of obtaining the contour of the drivable area, the pillars belonging to the same drivable area can be clustered into a pillar set, and a pillar set represents a drivable area. On this basis, when traversing the boundary points, each pillar set can be traversed separately, so that the contour of each drivable area can be obtained in parallel, thereby improving the efficiency of contour acquisition. Specifically, when traversing the boundary points of the pillars, for each pillar set, the boundary points of each pillar in the pillar set are traversed in turn, and the boundary points of each pillar are sorted according to the traversal order of the boundary points of each pillar, so as to obtain the boundary point set corresponding to the pillar set. Further, when executing step S105, the contours of the drivable areas corresponding to each pillar set can be obtained respectively according to the boundary point set corresponding to each pillar set.

[0145] In some implementations of the above step S104, Fig. 9 The following steps S301 to S302 are shown to cluster the pillars belonging to the same drivable area into a pillar set.

[0146] Step S301: Establish a connectivity tree diagram for each of the plurality of columns.

[0147] In this embodiment, any one of the plurality of columns is described as a first column, and the remaining columns except the first column are described as second columns. The connectivity tree diagram is described below in combination with the first and second columns.

[0148] The connectivity tree diagram of the first column includes: a root node, a right subtree, and a left subtree, wherein the right subtree includes a right node connected to the root node, and the left subtree includes a left node connected to the root node. The right subtree is located on the right side of the root node, and the left subtree is located on the left side of the root node.

[0149] The root node represents the first column.

[0150] The right node represents the second column that is on the right side of the first column and is connected to the first column. The right side of the first column is the side that the column direction of the first column points to first when rotating in the clockwise direction. The column direction is the direction from the lower boundary point of the first column to the upper boundary point. Fig.10 As shown, column 2 is on the right side of column 1.

[0151] The left node represents a second column that is on the left side of the first column and is connected to the first column. The left side of the first column is the side that the column direction of the first column points to first when rotating in the counterclockwise direction. Fig.10As shown, column 1 is on the left side of column 2.

[0152] The first and second columns are connected, which means that the first and second columns are connectable. In some embodiments, if it is determined that the first and second columns are adjacent and overlap, it can be determined that the first and second columns are connected. Fig.10 As shown, column 1 and column 2 are adjacent and overlapped, so column 1 and column 2 are connected. The above adjacent means that the fan-shaped areas where the columns are located are adjacent. For different columns in the same fan-shaped area, it is not necessary to determine whether they are connected.

[0153] See attached Fig.11 , Fig.11 The connectivity tree diagram of a column is shown as an example. Fig.11 As shown, sectorpillar is the root node, left and right represent the left subtree and right subtree of the connectivity tree diagram respectively. The left subtree includes three left nodes Lpillar0, Lpillar1, and Lpillar2, which are connected to the root node sector pillar. The right subtree includes a right node Rpillar0, which is also connected to the root node sector pillar.

[0154] In this embodiment, in order to distinguish each column, the columns can be coded to form column numbers, and each column number is different. The coding method of the columns is described below, where the columns are line segments on the ray obtained by the method described in the above step S103.

[0155] For each ray emitted from the center, an index number ray_index is set. There may be one or more pillars on a ray, and an index number pillar_index is set for each pillar on the ray. The pillar number pillar_encoder can include two parts: the ray index number ray_index and the pillar index number pillar_index. The index number ray_index is the index number of the ray to which the pillar belongs, and the index number pillar_index is the index number of the pillar on the ray to which it belongs.

[0156] For example, pillar_encoder=ray_index+pillar_index. In some implementations, to ensure the uniqueness of pillar_encoder, a weight w with a larger value may be set for ray_index, pillar_encoder=ray_index×w+pillar_index. The value of the weight w is greater than the maximum value of the pillars on a ray, for example, w=10000.

[0157] In some implementations, for different pillars on the same ray, the farther the distance between the pillar and the center, the smaller the pillar index number pillar_index; the closer the distance between the pillar and the center, the larger the pillar index number pillar_index.

[0158] The following is an explanation of the arrangement of the left / right nodes in the left / right subtree in the connectivity tree diagram. Specifically, in the connectivity tree diagram, the right nodes are arranged from left to right in the right subtree, and after the arrangement, the distance between the column represented by the right node and the center decreases; the left nodes are arranged from left to right in the left subtree, and after the arrangement, the distance between the column represented by the left node and the center decreases. Fig.11 ,In the left subtree, the left nodes Lpillar0, Lpillar1, and Lpillar2 are arranged from left to right, and the distances between the pillars represented by the left nodes Lpillar0, Lpillar1, and Lpillar2 and the center decrease successively.

[0159] Step S302: searching for connected columns according to the connectivity dendrogram of each column and clustering the connected columns into the same column set.

[0160] For the first column, the left and right nodes in its connectivity tree diagram can be searched, and the second column represented by the left and right nodes is the column that is connected to the first column. Therefore, the connected columns can be obtained by searching the nodes of the connectivity tree diagram.

[0161] For example, the connectivity tree diagram of column A includes right node 1, which represents column B. Columns A and B are connected, so columns A and B can be clustered into the same column set m. Furthermore, the connectivity tree diagram of column B includes right nodes 1 and 2, which represent columns C and D respectively. Columns B, C, and D are connected, so columns B, C, and D can be clustered into the same column set, and column B is already in column set m, so columns C and D are also put into column set m. Furthermore, the connectivity tree diagrams of columns C and D each have only one left node 1, and their left nodes 1 both represent column B. There are no new columns, so clustering of column set m can be stopped, and column set m ultimately includes columns A, B, C, and D.

[0162] Based on the method described in steps S301 to S302 above, the connectivity and tree diagram between the pillars can be used to quickly and accurately cluster the pillars in different drivable areas.

[0163] The above step S302 is described below.

[0164] In some implementations of the above step S302, the following steps S3021 to S3026 can be used to search for connected columns and cluster the connected columns into the same column set according to the connectivity dendrogram of each column.

[0165] Step S3021: Initialize all columns as unvisited.

[0166] Step S3022: select an unvisited column as the third column, set a column set m corresponding to the third column, add the third column to the column set, and modify the third column to visited.

[0167] Step S3023: According to the connectivity tree diagram of the third column, obtain the fourth column, add the fourth column to the above column set m, and modify the fourth column to visited, where the fourth column is the column represented by the right node and / or left node in the connectivity tree diagram.

[0168] Specifically, the left subtree may be traversed first and then the right subtree, or the right subtree may be traversed first and then the left subtree, and the columns represented by the traversed left / right nodes may be added to the above-mentioned column set m.

[0169] Step S3024: Obtain the fifth column according to the connectivity tree diagram of the fourth column; if the fifth column is unvisited, add the fifth column to the above column set m, and modify the fifth column to visited. The fifth column is the column represented by the right node and / or left node in the connectivity tree diagram of the fourth column.

[0170] Similar to the fourth column, you can also traverse the left subtree first and then the right subtree, or traverse the right subtree first and then the left subtree, and add the columns represented by the traversed left / right nodes to the above column set m.

[0171] Step S3025: taking the fifth column as the new fourth column, and repeating step S4 until no unvisited columns are obtained, thereby completing the clustering of the column set m.

[0172] For example, a total of 26 columns are obtained through the method described in step S103, namely, columns A to Z. First, columns A to Z are all set to unvisited, and then column A is used as the third column and is placed in column set m1.

[0173] According to the connectivity tree diagram of column A, column B is obtained as the fourth column, and column B is placed in the column set m1.

[0174] According to the connectivity tree diagram of column B, column C is obtained as the fifth column, and column C is placed in column set m1.

[0175] Furthermore, column C is used as the new fourth column, column D is obtained as the fifth column according to the connectivity tree diagram of column C, and column D is placed in the column set m1.

[0176] Column D is used as the new fourth column, column E is obtained as the fifth column according to the connectivity tree diagram of column D, and column E is placed in the column set m1.

[0177] Column E is used as the new fourth column. According to the connectivity tree diagram of column E, no unvisited columns are obtained, and the clustering of column set m1 is completed. Column set m1 finally includes columns A, B, C, D, and E.

[0178] Step S3026: Repeat steps S3022 to S3025 for the remaining unvisited columns until all columns are modified to visited.

[0179] Referring to the example of step S3025, for pillars F to Z, steps S3022 to S3025 are repeatedly executed until pillars A to Z are all modified to visited.

[0180] Based on the method described in steps S3021 to S3025 above, by setting the visit status of the column (unvisited, visited), the connectivity tree diagram of all columns can be quickly and accurately traversed, and the connected columns can be clustered into the same column set. The above search process is similar to Depth First Search.

[0181] The following is combined with Fig.12 , the process of obtaining the drivable area to obtaining the column set in this application is described. Among them, the smart device is a vehicle. Fig.12 As shown, the column set can be obtained through the following steps S401 to S407.

[0182] Step S401: input any number and shape of drivable areas. The drivable areas are obtained by the perception model on the smart device sensing the environmental data.

[0183] Step S402: emitting multiple rays from the center within a range of 360° around the center and centered on the vehicle.

[0184] Step S403: obtaining a plurality of pillars according to the intersection points of the plurality of rays and the drivable area.

[0185] Step S404: Establish a connectivity tree diagram for each column.

[0186] Step S405: According to the connectivity tree diagram of each column, search for connected columns and assign the connected columns to the same column set.

[0187] Step S406: Output the column set.

[0188] The methods of the above steps may adopt the relevant methods in the above method embodiments.

[0189] In some implementations of the above step S104, when traversing the boundary points of each column in the column set in sequence, the connectivity tree diagram obtained when the columns are clustered can be reused to traverse the boundary points. Specifically, in this implementation, according to the connectivity tree diagram of each column in the column set and the preset traversal rule, the boundary points of each column can be traversed in sequence along the second preset direction, and the second preset direction is a clockwise direction or a counterclockwise direction.

[0190] The preset traversal rules include a first traversal rule and a second traversal rule. The first and second traversal rules are described below.

[0191] 1. Explain the first traversal rule.

[0192] (1) The first traversal rule in which the second preset direction is the clockwise direction is described.

[0193] If the boundary point currently traversed is the upper boundary point of the first target column, the next traversal target tree diagram represents the second target column represented by the right node, and the target tree diagram is the connectivity tree diagram of the first target column; if the boundary point currently traversed is the lower boundary point of the first target column, the next traversal target tree diagram represents the third target column represented by the left node.

[0194] See attached Fig.13 , Fig.13 Indication of column assembly Figure 1 .like Fig.13 As shown, if the currently traversed boundary point is the upper boundary point of column A (i.e., the first target column), the connectivity tree diagram of column A (i.e., the target tree diagram) has a right node, and the right node represents column B (i.e., the second target column). Therefore, the next boundary point traversed is column B.

[0195] If the currently traversed boundary point is the lower boundary point of column A (i.e., the first target column), the connectivity tree diagram of column A (i.e., the target tree diagram) has a left node, and the left node represents column C (i.e., the third target column). Therefore, the next boundary point traversed is column C.

[0196] (2) The first traversal rule in which the second preset direction is counterclockwise is described.

[0197] If the currently traversed boundary point is the upper boundary point of the first target column, the next traversal is the third target column; if the currently traversed boundary point is the lower boundary point of the first target column, the next traversal is the second target column.

[0198] The first traversal rule in the counterclockwise direction is similar to the first traversal rule in the clockwise direction. The main difference is that if the next node to be traversed in the clockwise direction is the right node, then the next node to be traversed in the counterclockwise direction is changed from the right node to the left node, that is, the left and right are reversed; if the next node to be traversed in the clockwise direction is the left node, then the next node to be traversed in the counterclockwise direction is changed from the left node to the right node, that is, the left and right are reversed.

[0199] See again the attached Fig.13 , if the current traversed boundary point is the upper boundary point of column A (i.e., the first target column), according to the above example, we know that column C is the third target column, so the next traversed boundary point is column C. If the current traversed boundary point is the lower boundary point of column A (i.e., the first target column), according to the above example, we know that column B is the second target column, so the next traversed boundary point is column B.

[0200] 2. Explain the second traversal rule.

[0201] If the currently traversed boundary point is the upper boundary point of the first target column, the next traversal is the upper boundary point; if the currently traversed boundary point is the lower boundary point of the first target column, the next traversal is the lower boundary point.

[0202] See again the attached Fig.13 , if the current traversed boundary point is the upper boundary point of column A (i.e., the first target column), then the next traversal is the upper boundary point. At the same time, combined with the first traversal rule, if the second preset direction is clockwise, then the next traversal is the upper boundary point of column B, i.e. Fig.13 Path 1 in holds true; if the second preset direction is counterclockwise, the next traversal is the upper boundary point of column C.

[0203] If the current traversed boundary point is the lower boundary point of column A (i.e., the first target column), then the next traversal is the lower boundary point. At the same time, combined with the first traversal rule, if the second preset direction is clockwise, then the next traversal is the lower boundary point of column B; if the second preset direction is counterclockwise, then the next traversal is the lower boundary point of column C, that is, Fig.13 Path 5 in holds.

[0204] In some implementations of the above step S104, the preset traversal rule includes a third traversal rule, and the third traversal rule is described below.

[0205] 1. The third traversal rule in which the second preset direction is clockwise is described.

[0206] If the currently traversed boundary point is the upper boundary point of the first target column and the target tree diagram has no right node, the next traversal is the lower boundary point of the first target column; if the currently traversed boundary point is the lower boundary point of the first target column and the target tree diagram has no left node, the next traversal is the upper boundary point of the first target column.

[0207] See again the attached Fig.13 , if the currently traversed boundary point is the upper boundary point of column B (i.e., the first target column), according to the first and second traversal rules, the next traversal should be the upper boundary point of the right node in the connectivity tree diagram of column B (i.e., the target tree diagram), but the connectivity tree diagram of column B does not have a right node, i.e. Fig.13 Path 2 in does not hold, so the next traversal is the lower boundary point of column B, that is, Fig.13 Path 3 in is established. After traversing to the lower boundary point of column B, according to the first and second traversal rules, the next traversal is the lower boundary point of column A, that is, Fig.13 Path 4 in holds.

[0208] If the currently traversed boundary point is the lower boundary point of column C (i.e., the first target column), according to the first and second traversal rules, the next point to be traversed is the lower boundary point of the left node in the connectivity tree diagram of column C (i.e., the target tree diagram), but the connectivity tree diagram of column C has no left node, i.e. Fig.13 Path 6 in does not hold, so the next traversal is the upper boundary point of column C, that is, Fig.13 Path 7 in holds.

[0209] 2. Explain the third traversal rule where the second preset direction is counterclockwise.

[0210] If the currently traversed boundary point is the upper boundary point of the first target column and the target tree diagram has no left node, the next traversal is the lower boundary point of the first target column; if the currently traversed boundary point is the lower boundary point of the first target column and the target tree diagram has no right node, the next traversal is the upper boundary point of the first target column.

[0211] The third traversal rule in the counterclockwise direction is similar to the third traversal rule in the clockwise direction. The main difference is: if the next node to be traversed in the clockwise direction is the right node, then the next node to be traversed in the counterclockwise direction is changed from the right node to the left node, that is, the left and right are reversed; if the next node to be traversed in the clockwise direction is the left node, then the next node to be traversed in the counterclockwise direction is changed from the left node to the right node, that is, the left and right are reversed.

[0212] In some implementations of the above step S104, the preset traversal rule includes a fourth traversal rule, and the fourth traversal rule is described below.

[0213] 1. The fourth traversal rule in which the second preset direction is clockwise is described.

[0214] (1) Explain sub-rule 1.

[0215] If the currently traversed boundary point is the upper boundary point of the first target column and the target tree diagram has multiple right nodes, the next traversal target tree diagram is the upper boundary point of the fourth target column represented by the first right node, and the first right node is the leftmost right node arranged in the right subtree.

[0216] According to the implementation of the aforementioned step S301, in the connectivity tree diagram, the right nodes are arranged from left to right in the right subtree, and after the arrangement, the distance between the column represented by the right node and the center decreases successively. Therefore, the distance between the column represented by the leftmost right node and the center is the farthest. Fig.11 As shown, the leftmost right node is Rpillar0.

[0217] See attached Fig.14 , Fig.14 Indication of column assembly Figure 2 If the current traversed boundary point is the upper boundary point of column B (i.e., the first target column), since the connectivity tree diagram of column B (i.e., the target tree diagram) has two right nodes, these two right nodes represent columns C and D respectively, among which the right node representing column C is on the leftmost side of the right subtree. Therefore, the next traversed boundary point is the upper boundary point of column C, such as Fig.14 Indicated by the arrow between B and C.

[0218] (2) Explain sub-rule 2.

[0219] If the currently traversed boundary point is the lower boundary point of the first target column and the target tree diagram has multiple left nodes, then the next traversal target tree diagram will represent the lower boundary point of the fifth target column represented by the last left node, and the last left node is the rightmost left node arranged in the left subtree. According to the implementation of the aforementioned step S301, in the connectivity tree diagram, the left nodes are arranged from left to right in the left subtree, and after the arrangement, the distance between the column represented by the left node and the center decreases. Therefore, the distance between the column represented by the rightmost left node and the center is the shortest. Fig.11 As shown, the rightmost left node is Lpillar2.

[0220] See again the attached Fig.14 , if the current traversed boundary point is the lower boundary point of column L (i.e., the first target column), since the connectivity tree diagram of column L (i.e., the target tree diagram) has two left nodes, these two left nodes represent column J and column K respectively, among which, the left node representing column K is on the far right of the left subtree, therefore, the next traversed lower boundary point of column K is, as Fig.14 As shown by the arrow between L and K in FIG.

[0221] 2. The fourth traversal rule in which the second preset direction is counterclockwise is described.

[0222] If the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has multiple left nodes, the next traversal is the upper boundary point of the fifth target column; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has multiple right nodes, the next traversal is the lower boundary point of the fourth target column.

[0223] The fourth traversal rule in the counterclockwise direction is similar to the fourth traversal rule in the clockwise direction. The main difference is: if the next node to be traversed in the clockwise direction is the right node, then the next node to be traversed in the counterclockwise direction is changed from the right node to the left node, that is, the left and right are reversed; if the next node to be traversed in the clockwise direction is the left node, then the next node to be traversed in the counterclockwise direction is changed from the left node to the right node, that is, the left and right are reversed.

[0224] In some implementations of the above step S104, the preset traversal rule includes a fifth traversal rule, and the fifth traversal rule is described below.

[0225] 1. The fifth traversal rule in which the second preset direction is clockwise is described.

[0226] (1) Explain sub-rule 1.

[0227] If the currently traversed boundary point is the upper boundary point of the first target column and the next traversed sixth target column is a one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether there is a second node on the left side of the first node representing the first target column; if there is a second node, traverse the lower boundary point of the seventh target column represented by the second node; if there is no second node, traverse the boundary point of the sixth target column; the one-to-many column is a column whose connectivity tree diagram has multiple right nodes and / or multiple left nodes.

[0228] See attached Fig.15 , Fig.15 Indication of column assembly Figure 3 . If the currently traversed boundary point is the upper boundary point of column E (i.e., the first target column), according to the first and second traversal rules, the next traversal is the upper boundary point of column I (i.e., the sixth target column), but column I is a one-to-many column, and its connectivity tree diagram has two left nodes, one left node (i.e., the first node) represents column E, and the other left node represents column F. The node representing column F is on the right side of the first node, that is, there is no second node on the left side of the first node. Therefore, the next traversal is still the upper boundary point of column I (i.e., the sixth target column).

[0229] (2) Explain sub-rule 2.

[0230] If the currently traversed boundary point is the lower boundary point of the first target column and the next traversed sixth target column is a one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether there is a third node on the right side of the first node; if there is a third node, traverse the upper boundary point of the eighth target column represented by the third node; if there is no third node, traverse the boundary points of the sixth target column.

[0231] See again the attached Fig.15 If the currently traversed boundary point is the lower boundary point of column G (i.e., the first target column), according to the first and second traversal rules, the next traversal point is the lower boundary point of column D (i.e., the sixth target column). However, column D is a one-to-many column, and its connectivity tree diagram has two right nodes, one right node (i.e., the first node) represents column G, and the other right node represents column H. The node of column H is on the right side of the first node, that is, there is a third node (i.e., the node of column H) on the right side of the first node. Then the next traversal point is the upper boundary point of column H.

[0232] 2. Explain the fifth traversal rule in which the second preset direction is counterclockwise.

[0233] (1) Explain sub-rule 1.

[0234] If the currently traversed boundary point is the upper boundary point of the first target column and the next traversed sixth target column is a one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether there is the third node on the right side of the first node; if there is a third node, traverse the lower boundary point of the eighth target column represented by the third node; if there is no third node, traverse the boundary points of the sixth target column.

[0235] (2) Explain sub-rule 2.

[0236] If the currently traversed boundary point is the lower boundary point of the first target column and the next traversed sixth target column is a one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether there is the second node on the left side of the first node; if there is a second node, traverse the upper boundary point of the seventh target column represented by the second node; if there is no second node, traverse the boundary points of the sixth target column.

[0237] The fifth traversal rule in the counterclockwise direction is similar to the fifth traversal rule in the clockwise direction. The main difference is that if the next node to be traversed in the clockwise direction is the right node, then the next node to be traversed in the counterclockwise direction is changed from the right node to the left node, that is, the left and right are reversed; if the next node to be traversed in the clockwise direction is the left node, then the next node to be traversed in the counterclockwise direction is changed from the left node to the right node, that is, the left and right are reversed.

[0238] The following is combined with Fig.15 , taking the starting point of the traversal as the lower boundary point of column H as an example, Fig.15 The traversal process of the column set shown is illustrated, and the second preset direction is clockwise.

[0239] (1) The boundary point currently traversed is the lower boundary point of column H (i.e., the first target column): According to the first and second traversal rules, the next point to traverse is the lower boundary point of column D (i.e., the sixth target column). However, column D (i.e., the sixth target column) is a one-to-many column. According to the fifth traversal rule, since there is no other node (i.e., the third node) to the right of the right node (i.e., the first node) representing column H in the connectivity tree diagram of column D, the next point to traverse is still the lower boundary point of column D.

[0240] (2) The current traversed boundary point is the lower boundary point of column D (i.e., the first target column): According to the first and second traversal rules, the next point to be traversed is the lower boundary point of column B.

[0241] (3) The current traversed boundary point is the lower boundary point of column B (i.e., the first target column): According to the first and second traversal rules, the next point to be traversed is the lower boundary point of column A.

[0242] (4) The boundary point currently traversed is the lower boundary point of column A (i.e., the first target column): According to the first traversal rule, the next point to traverse is the left node in the connectivity tree diagram of column A. However, there is no left node in the connectivity tree diagram of column A. According to the third traversal rule, the next point to traverse is the upper boundary point of column A.

[0243] (5) The current traversed boundary point is the upper boundary point of column A (i.e., the first target column): According to the first and second traversal rules, the next boundary point to be traversed is the upper boundary point of column B.

[0244] (6) The current traversed boundary point is the upper boundary point of column B (i.e., the first target column): According to the first and second traversal rules, the next boundary point to be traversed is the upper boundary point of column C.

[0245] (7) The boundary point currently traversed is the upper boundary point of column C (i.e., the first target column): According to the first traversal rule, the next node to be traversed in the connectivity tree diagram of column C is the right node, but the connectivity tree diagram of column C has two right nodes. According to the fourth traversal rule, the column represented by the first right node is obtained, i.e., column E. The next node to be traversed is the upper boundary point of column E.

[0246] (8) The boundary point currently traversed is the upper boundary point of column E (i.e., the first target column): According to the first and second traversal rules, the next one needs to traverse the upper boundary point of column I. However, column I (i.e., the sixth target column) is a one-to-many column. According to the fifth traversal rule, since there is no other node (i.e., the second node) to the left of the left node (i.e., the first node) representing column E in the connectivity tree diagram of column I, the next one still traverses the upper boundary point of column I.

[0247] (9) The boundary point currently traversed is the upper boundary point of column I (i.e., the first target column): According to the first traversal rule, the next point to traverse is the right node in the connectivity tree diagram of column I. However, there is no right node in the connectivity tree diagram of column I. According to the third traversal rule, the next point to traverse is the lower boundary point of column I.

[0248] (10) The boundary point currently traversed is the lower boundary point of column I (i.e., the first target column): According to the first traversal rule, the next left node in the connectivity tree diagram of column I needs to be traversed, but the connectivity tree diagram of column I has two left nodes. According to the fourth traversal rule, the column represented by the last left node is obtained, i.e., column F. The next lower boundary point of column F is traversed.

[0249] (11) The boundary point currently traversed is the lower boundary point of column F (i.e., the first target column): According to the first and second traversal rules, the next one to traverse is the lower boundary point of column C (i.e., the sixth target column). However, column C (i.e., the sixth target column) is a one-to-many column. According to the fifth traversal rule, since there is no other node (i.e., the third node) to the right of the right node (i.e., the first node) representing column F in the connectivity tree diagram of column C, the next one to traverse is still the lower boundary point of column C.

[0250] (12) The boundary point currently traversed is the lower boundary point of column C (i.e., the first target column): According to the first and second traversal rules, the next point to be traversed is the lower boundary point of column B (i.e., the sixth target column). However, column B (i.e., the sixth target column) is a one-to-many column. According to the fifth traversal rule, since there are other nodes (i.e., the third node) to the right of the right node (i.e., the first node) representing column C in the connectivity tree diagram of column B, and this third node represents column D, the next point to be traversed is the upper boundary point of column D.

[0251] (13) The boundary point currently traversed is the upper boundary point of column D (i.e., the first target column): According to the first and second traversal rules, the next sequence traverses the upper boundary point of column G.

[0252] (14) The boundary point currently traversed is the upper boundary point of column G (i.e., the first target column): According to the first traversal rule, the next point to traverse is the right node in the connectivity tree diagram of column G. However, there is no right node in the connectivity tree diagram of column G. According to the third traversal rule, the next point to traverse is the lower boundary point of column G.

[0253] (15) The boundary point currently traversed is the lower boundary point of column G (i.e., the first target column): According to the first and second traversal rules, the next point to be traversed is the lower boundary point of column D (i.e., the sixth target column). However, column D (i.e., the sixth target column) is a one-to-many column. According to the fifth traversal rule, since there are other nodes (i.e., the third node) to the right of the right node (i.e., the first node) representing column G in the connectivity tree diagram of column D, and this third node represents column H, the next point to be traversed is the upper boundary point of column H.

[0254] (16) The boundary point currently traversed is the upper boundary point of column H (i.e., the first target column): According to the first traversal rule, the next point to traverse is the right node in the connectivity tree diagram of column H, but there is no right node in the connectivity tree diagram of column H. According to the third traversal rule, the next point to traverse is the lower boundary point of column H, but the lower boundary point of H has been traversed. Therefore, the traversal of the entire column set is stopped.

[0255] According to the traversal process of (1)-(16) above, it can be determined that the traversal order of the boundary points is: the lower boundary point of column H, the lower boundary point of column D, the lower boundary point of column B, the lower boundary point of column A, the upper boundary point of column A, the upper boundary point of column B, the upper boundary point of column C, the upper boundary point of column E, the upper boundary point of column I, the lower boundary point of column I, the lower boundary point of column F, the lower boundary point of column C, the upper boundary point of column D, the upper boundary point of column G, the lower boundary point of column G, and the upper boundary point of column H.

[0256] The following is combined with Fig.16 To Attachment Fig.18 , the effect of the method for obtaining the contour of the polygonal drivable area provided in this application is explained.

[0257] Fig.16 An example of a polygonal drivable area is shown. Fig.17 The example shows the situation where the rays emitted from the center within a range of 360° around the center and the drivable area intersect with the vehicle as the center. Fig.18 The outline of the drivable area obtained by using these rays is shown as an example. Fig.16 and Fig.18 , it can be determined that the contour obtained by using the present application has a higher accuracy.

[0258] It should be pointed out that although the various steps in the above embodiments are described in a specific order, those skilled in the art can understand that in order to achieve the effect of the present application, different steps do not have to be executed in such an order, and they can be executed simultaneously (in parallel) or in other orders. These adjusted schemes are equivalent to the technical schemes described in this application, and therefore will also fall within the scope of protection of this application.

[0259] It is understood by those skilled in the art that all or part of the processes in the method for implementing the above-mentioned embodiment of the present application can also be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium may include: any entity or device, medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium, etc. that can carry the computer program code.

[0260] Another aspect of the present application also provides a computer-readable storage medium.

[0261] In an embodiment of a computer-readable storage medium according to the present application, the computer-readable storage medium may be configured to store a program for executing the contour extraction method of the drivable area of ​​the above-mentioned method embodiment, and the program may be loaded and run by a processor to implement the contour extraction method of the drivable area. For ease of explanation, only the parts related to the embodiment of the present application are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present application. The computer-readable storage medium may be a storage device formed by various electronic devices. Optionally, the computer-readable storage medium in the embodiment of the present application is a non-temporary computer-readable storage medium.

[0262] Another aspect of the present application also provides a smart device.

[0263] In an embodiment of an intelligent device according to the present application, the intelligent device may include at least one processor; and a memory connected to the at least one processor in communication; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method described in any of the above embodiments is implemented. Fig.19 , Fig.19 FIG. 4 exemplarily shows that the memory 11 and the processor 12 are communicatively connected via a bus.

[0264] So far, the technical solution of the present application has been described in conjunction with an embodiment shown in the accompanying drawings, but it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments. Without departing from the principles of the present application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present application.

Claims

1. A method for extracting the contour of a drivable area, characterized in that: The method comprises: Acquire a drivable area in the environment where the smart device is located, where the drivable area is a polygonal drivable area obtained by sensing the environmental data through a perception model on the smart device; Within a range of a preset angle around the smart device and centered at the center, emitting a plurality of rays from the center; Obtain intersections of the plurality of rays and the drivable area, and obtain a plurality of pillars according to the intersections, wherein the pillars are sub-areas of the drivable area within a fan-shaped area, and the fan-shaped area is an area formed between two adjacent rays; Traversing the boundary points of each of the plurality of columns in sequence, and sorting the boundary points of each column according to the traversal order of the boundary points of each column, so as to obtain a boundary point set; Acquire the outline of the drivable area according to the set of boundary points; in, The boundary points of the column include an upper boundary point and a lower boundary point, and the upper boundary point and the lower boundary point are respectively the points of the column farthest and closest to the center.

2. The method according to claim 1, characterized in that The step of acquiring a plurality of columns according to the intersection point comprises: For each ray, if a line segment formed between two adjacent intersection points on the ray is located within the drivable area, the line segment is used as the column.

3. The method according to claim 2, characterized in that The method comprises determining whether the line segment is within the drivable area by: The intersection points of the two adjacent intersection points that are farthest and closest to the center are respectively taken as the first point and the second point, and the target contour of the drivable area located between the first point and the second point is obtained along a first preset direction with the first point as the starting point, where the first preset direction is a clockwise direction or a counterclockwise direction; Acquire a third point closest to the first point on the target contour, and acquire a first vector with the first point and the third point as the starting point and the end point respectively; Acquire a fourth point on the target contour that is closest to the second point, and acquire a second vector with the fourth point and the second point as the starting point and the end point respectively; Obtaining the midpoint of the line segment formed by the two adjacent intersection points; If the midpoint is located on the target side of the first vector and the second vector respectively, the line segment is located within the drivable area; in, The target sides of the first vector and the second vector are respectively: the sides to which the vector directions of the first vector and the second vector first point when they rotate along the first preset direction.

4. The method according to claim 1, characterized in that: There are multiple drivable areas, and the method includes: clustering pillars belonging to the same drivable area into a pillar set; The step of traversing the boundary points of each column in the plurality of columns in turn and sorting the boundary points of each column according to the traversal order of the boundary points of each column to obtain a boundary point set includes: for each column set, traversing the boundary points of each column in the column set in turn and sorting the boundary points of each column according to the traversal order of the boundary points of each column to obtain a boundary point set corresponding to the column set.

5. The method according to claim 4, characterized in that The clustering of pillars belonging to the same drivable area into a pillar set includes: Establishing a connectivity dendrogram of each of the plurality of columns respectively; According to the connectivity dendrogram of each column, searching for columns with connectivity and clustering the columns with connectivity into the same column set; in, The connectivity tree diagram of the first column includes: a root node, a right subtree, and a left subtree, wherein the right subtree includes a right node connected to the root node, and the left subtree includes a left node connected to the root node, and the first column is any one of the plurality of columns; The root node represents the first column; The right node represents a second column on the right side of the first column and connected with the first column, the left node represents a second column on the left side of the first column and connected with the first column, and the second column is the remaining column among the plurality of columns except the first column; The right side and the left side of the first column are respectively: the side that the column direction of the first column points to first when rotating in the clockwise direction and the counterclockwise direction, and the column direction is the direction from the lower boundary point of the first column to the upper boundary point.

6. The method according to claim 5, characterized in that The method comprises: If the first column and the second column are adjacent to each other and overlap, then the first column and the second column are connected.

7. The method according to claim 5, characterized in that The step of searching for connected pillars and clustering the connected pillars into the same pillar set includes: Step S1: Initialize all columns as unvisited; Step S2: selecting an unvisited column as the third column, setting a column set corresponding to the third column, adding the third column to the column set, and modifying the third column to be visited; Step S3: according to the connectivity tree diagram of the third column, obtain a fourth column, add the fourth column to the column set, and modify the fourth column to be visited, wherein the fourth column is the column represented by the right node and / or the left node in the connectivity tree diagram; Step S4: acquiring a fifth column according to the connectivity tree diagram of the fourth column; if the fifth column is unvisited, adding the fifth column to the column set, and modifying the fifth column to be visited, the fifth column being the column represented by the right node and / or the left node in the connectivity tree diagram of the fourth column; Step S5: taking the fifth column as a new fourth column, and repeating step S4 until no unvisited column is obtained, so as to complete clustering of the column set; Step S6: Repeat steps S2 to S5 for the remaining unvisited columns until all columns are modified to have been visited.

8. The method according to claim 5, characterized in that The step of sequentially traversing the boundary points of each column in the column set includes: According to the connectivity tree diagram of each column in the column set and the preset traversal rule, traverse the boundary points of each column in sequence along the second preset direction; in, The preset traversal rules include a first traversal rule and a second traversal rule; The first traversal rule includes: When the second preset direction is clockwise: if the currently traversed boundary point is the upper boundary point of the first target column, then the next traversal is the second target column represented by the right node in the target tree diagram, and the target tree diagram is the connectivity tree diagram of the first target column; if the currently traversed boundary point is the lower boundary point of the first target column, then the next traversal is the third target column represented by the left node in the target tree diagram; When the second preset direction is counterclockwise: if the currently traversed boundary point is the upper boundary point of the first target column, the next traversal is the third target column; if the currently traversed boundary point is the lower boundary point of the first target column, the next traversal is the second target column; The second traversal rule includes: if the currently traversed boundary point is the upper boundary point of the first target column, then the next traversal is the upper boundary point; if the currently traversed boundary point is the lower boundary point of the first target column, then the next traversal is the lower boundary point.

9. The method according to claim 8, characterized in that The preset traversal rule includes a third traversal rule, and the third traversal rule includes: When the second preset direction is clockwise: if the currently traversed boundary point is the upper boundary point of the first target column and the target tree diagram has no right node, the next traversal is the lower boundary point of the first target column; if the currently traversed boundary point is the lower boundary point of the first target column and the target tree diagram has no left node, the next traversal is the upper boundary point of the first target column; When the second preset direction is counterclockwise: if the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has no left node, the next traversal is the lower boundary point of the first target column; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has no right node, the next traversal is the upper boundary point of the first target column.

10. The method according to claim 8, characterized in that In the connectivity tree diagram of the first column, the right nodes are arranged sequentially from left to right in the right subtree, and after the sequential arrangement, the distance between the column represented by the right node and the center is sequentially reduced; the left nodes are arranged sequentially from left to right in the left subtree, and after the sequential arrangement, the distance between the column represented by the left node and the center is sequentially reduced; The preset traversal rule includes a fourth traversal rule, and the fourth traversal rule includes: When the second preset direction is clockwise: if the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has multiple right nodes, then the upper boundary point of the fourth target column represented by the first right node in the target tree diagram is traversed next, and the first right node is the right node arranged at the leftmost side of the right subtree; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has multiple left nodes, then the lower boundary point of the fifth target column represented by the last left node in the target tree diagram is traversed next, and the last left node is the left node arranged at the rightmost side of the left subtree; When the second preset direction is counterclockwise: if the boundary point currently traversed is the upper boundary point of the first target column and the target tree diagram has multiple left nodes, the next traversal is the upper boundary point of the fifth target column; if the boundary point currently traversed is the lower boundary point of the first target column and the target tree diagram has multiple right nodes, the next traversal is the lower boundary point of the fourth target column.

11. The method according to claim 8, characterized in that In the connectivity tree diagram of the first column, the right nodes are arranged sequentially from left to right in the right subtree, and after the sequential arrangement, the distance between the column represented by the right node and the center is sequentially reduced; the left nodes are arranged sequentially from left to right in the left subtree, and after the sequential arrangement, the distance between the column represented by the left node and the center is sequentially reduced; The preset traversal rule includes a fifth traversal rule, and the fifth traversal rule includes: When the second preset direction is clockwise: If the currently traversed boundary point is the upper boundary point of the first target column and the next traversed sixth target column is a one-to-many column, then in the connectivity tree diagram of the sixth target column, it is detected whether there is a second node on the left side of the first node representing the first target column; if there is a second node, the lower boundary point of the seventh target column represented by the second node is traversed; if there is no second node, the boundary point of the sixth target column is traversed; the one-to-many column is a column whose connectivity tree diagram has multiple right nodes and / or multiple left nodes; If the currently traversed boundary point is the lower boundary point of the first target column and the next traversed sixth target column is the one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether there is a third node on the right side of the first node; if there is a third node, traverse the upper boundary point of the eighth target column represented by the third node; if there is no third node, traverse the boundary points of the sixth target column; When the second preset direction is counterclockwise: If the currently traversed boundary point is the upper boundary point of the first target column and the next traversed sixth target column is the one-to-many column, then in the connectivity tree diagram of the sixth target column, it is detected whether there is the third node on the right side of the first node; if there is the third node, then the lower boundary point of the eighth target column represented by the third node is traversed; if there is no third node, then the boundary points of the sixth target column are traversed; If the currently traversed boundary point is the lower boundary point of the first target column and the next traversed sixth target column is the one-to-many column, then in the connectivity tree diagram of the sixth target column, check whether the second node exists on the left side of the first node; if there is the second node, traverse the upper boundary point of the seventh target column represented by the second node; if there is no second node, traverse the boundary points of the sixth target column.

12. A smart device, characterized in that: include: at least one processor; and, a memory communicatively coupled to the at least one processor; The memory stores a computer program, and when the computer program is executed by the at least one processor, the method for extracting the contour of the drivable area according to any one of claims 1 to 11 is implemented.

13. A computer-readable storage medium storing a plurality of program codes, characterized in that: The program code is suitable for being loaded and run by a processor to execute the contour extraction method of the drivable area according to any one of claims 1 to 11.