Automatic driving boundary refining method based on SL coordinate system

Abstract

The invention relates to an automatic driving boundary refining method based on an SL coordinate system, and the method comprises the steps: S1, obtaining driving boundaries which allow a vehicle to drive when there is no barrier in a lane, and the driving boundaries comprise a left side boundary, a right side boundary and a center line; s2, an SL coordinate system is established based on the center line of the lane, S represents the longitudinal position in the direction of the center line, L represents the transverse position perpendicular to the direction of the center line, the position of the center line in the longitudinal direction is 0, the direction deviating from the left side of the center line is the positive direction, and the direction deviating from the right side of the center line is the negative direction; s3, a data structure expression mode of the boundary is confirmed and is (S, L, R), S represents a longitudinal position, L represents a transverse position which can be driven at the corresponding S position in the forward direction to the maximum extent, and R represents a transverse position which can be driven at the corresponding S position in the negative direction to the maximum extent; s4, an obstacle detection boundary is set, and obstacles in the detection boundary are traversed, classified and recognized; and S5, carrying out iterative traversal on the obstacle boundary according to a certain resolution to obtain an SL coordinate boundary of the obstacle boundary, and synthesizing the driving boundary under the SL coordinate system and the obstacle boundary to obtain a final maximum boundary. Through the steps, the drivable boundary of the automatic driving vehicle is optimized, and the passing efficiency of the automatic driving vehicle is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle autonomous driving planning, and in particular to a boundary refinement method for autonomous driving based on an SL coordinate system. Background Art

[0002] Autonomous driving technology has achieved rapid development in the past decade, and planning is a crucial part of it. At present, the time-space decoupled planning scheme commonly used in the industry decomposes planning into path planning and speed planning, where path planning is responsible for outputting a smooth and collision-free path. In urban structured roads, optimization-based path planning relies on lane information provided by high-precision maps, and constructs the initial path boundary in the SL coordinate system to ensure that the autonomous driving vehicle travels within the lane. In addition, path planning also requires the perception and prediction modules to provide obstacle information to construct collision constraints to ensure that there will be no collision with other traffic participants; however, this method will cause a large amount of empty space to be occupied when the obstacle is not parallel to the road, thereby reducing the passing performance of the autonomous driving vehicle.

[0003] Therefore, there is an urgent need to provide an autonomous driving boundary refinement method based on the SL coordinate system to solve the above problems. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings and defects of the prior art and provide a boundary refinement method for autonomous driving based on the SL coordinate system, which optimizes the drivable boundaries of autonomous driving vehicles and improves the traffic efficiency of autonomous driving vehicles.

[0005] The objective of the present invention is achieved through the following technical solutions:

[0006] A boundary refinement method for autonomous driving based on an SL coordinate system, characterized by comprising the following steps:

[0007] S1: Obtaining the driving boundary within which the vehicle is allowed to drive when there is no obstacle in the lane, wherein the driving boundary includes a left boundary, a right boundary and a center line;

[0008] S2: The SL coordinate system is established based on the centerline of the lane. S represents the longitudinal position along the centerline, and L represents the lateral position perpendicular to the centerline. The position of the centerline in the longitudinal direction is 0, the direction deviating from the left side of the centerline is the positive direction, and the direction deviating from the right side of the centerline is the negative direction.

[0009] S3: The data structure representation of the confirmed boundary is (S, L, R), where S represents the longitudinal position, L represents the maximum traversable lateral position in the positive direction at the corresponding S position, and R represents the maximum traversable lateral position in the negative direction at the corresponding S position;

[0010] S4: Set the obstacle detection boundary, and traverse and classify obstacles within the detection boundary;

[0011] S5: Iterate the obstacle boundary at a certain resolution to obtain the SL coordinate boundary of the obstacle boundary, and combine the driving boundary in the SL coordinate system with the obstacle boundary to obtain the final maximum boundary.

[0012] Optionally, it also includes:

[0013] S6: Determine whether the obstacle is tilted;

[0014] S61: If the angle between the boundary of the obstacle and the road is 0 or tends to 0, the obstacle is deemed not to be inclined, and the boundary of the obstacle is the maximum driving boundary of the vehicle;

[0015] S62: If the angle between the boundary of the obstacle and the road is greater than a certain threshold, the obstacle is set to be inclined, and the longitudinal starting point S is obtained according to the four vertices of the obstacle. S , longitudinal end point E S , horizontal starting point S L , lateral end point E L , according to the SL coordinates of the four vertices, the corresponding left and right boundaries at the boundary S are obtained.

[0016] Optionally, it also includes:

[0017] S63: If (S L +E L ) / 2>0, the center point of the obstacle is located on the left side of the lane centerline, and the left boundary corresponding to the boundary S includes the first left boundary and the second left boundary;

[0018] S64: The first left boundary is determined by the left vertex and the lower vertex of the obstacle; the second left boundary is determined by the lower vertex and the right vertex of the obstacle.

[0019] Optionally, it also includes:

[0020] S65: If (S L +E L ) / 2<0, the center point of the obstacle is located on the right side of the lane centerline, and the right boundary corresponding to the boundary S includes the first right boundary and the second right boundary;

[0021] S66: The first right boundary is determined by the left vertex and the upper vertex of the obstacle; the first right boundary is determined by the upper vertex and the right vertex of the obstacle.

[0022] Optionally, it also includes:

[0023] S7: Vertex SL coordinate correction;

[0024] S71: The obstacle is located on the left side of the center line. The longitudinal coordinates of the previous and next traversal points of the left vertex of the obstacle are equal to the longitudinal coordinate of the left vertex. The longitudinal coordinates of the previous and next traversal points of the right vertex of the obstacle are equal to the longitudinal coordinate of the right vertex. The transverse coordinates of the previous and next traversal points of the lower vertex of the obstacle are equal to the transverse coordinate of the lower vertex.

[0025] S72: The obstacle is located on the right side of the center line. The longitudinal coordinates of the previous and next traversal points of the left vertex of the obstacle are equated with the longitudinal coordinates of the left vertex, the longitudinal coordinates of the previous and next traversal points of the right vertex of the obstacle are equated with the longitudinal coordinates of the right vertex, and the transverse coordinates of the previous and next traversal points of the upper vertex of the obstacle are equated with the transverse coordinates of the upper vertex.

[0026] Optionally, it also includes:

[0027] S8: If there are obstacles A and B;

[0028] S81: Obstacles A and B are both located on the left side of the center line, obstacle A is traversed to determine a first left boundary, obstacle B is traversed to determine a second left boundary, and the maximum driving boundary of the vehicle is determined by combining the first left boundary and the second left boundary;

[0029] S82: Obstacle A is located on the left side of the center line, and obstacle B is located on the right side of the center line. Obstacle A is traversed to determine the left boundary, and obstacle B is traversed to determine the right boundary. The maximum driving boundary of the vehicle is determined by combining the left boundary and the right boundary.

[0030] Optionally, the specific steps of S4 are:

[0031] S41: Acquire the vehicle body boundary of the vehicle and the road boundary of the vehicle; the vehicle body boundary includes the front boundary of the vehicle body, the rear boundary of the vehicle body, the left boundary of the vehicle body and the right boundary of the vehicle body, and the road boundary includes the left road boundary and the right road boundary;

[0032] S42: Calculating a front obstacle detection boundary of the vehicle according to the front boundary of the vehicle and a preset forward detection distance;

[0033] S43: Calculating a rear obstacle detection boundary of the vehicle according to the rear boundary of the vehicle and a preset rearward detection distance;

[0034] S44: Calculating a left obstacle detection boundary of the vehicle according to a left boundary of the vehicle and a preset left detection distance;

[0035] S45: Calculating and obtaining a right obstacle detection boundary of the vehicle according to a right boundary of the vehicle and a preset right detection distance;

[0036] S46: Obtain an obstacle detection boundary of the vehicle by merging the obtained front obstacle detection boundary, rear obstacle detection boundary, left obstacle detection boundary, and right obstacle detection boundary.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] In the present invention, the driving boundary that allows the vehicle to drive when there is no obstacle in the lane is first obtained, and the driving boundary includes a left boundary, a right boundary and a center line, wherein the left boundary is the maximum lateral position for the vehicle to drive to the left when there is no obstacle, and the right boundary is the maximum lateral position for the vehicle to drive to the right when there is no obstacle, so as to determine the driving boundary of the vehicle; an SL coordinate system is established based on the center line of the lane, with the position of the center line in the longitudinal direction as 0, the direction deviating from the left side of the center line as the positive direction, and the direction deviating from the right side of the center line as the negative direction; the data structure expression of the boundary is confirmed to be (S, L, R), and a series of data structures constitute the boundary; an obstacle detection boundary is set, and obstacles within the detection boundary are traversed and classified to identify; the obstacle boundary is iteratively traversed at a certain resolution to obtain the SL coordinate boundary of the obstacle boundary, and the driving boundary under the SL coordinate system is combined with the obstacle boundary to obtain the final maximum boundary. Through the above steps, the drivable boundary of the autonomous driving vehicle is optimized, and the traffic efficiency of the autonomous driving vehicle is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a structural flow chart of the present invention.

[0040] Figure 2 It is a structural flow chart of determining the center point position of an obstacle in the present invention.

[0041] Figure 3 It is a structural flow chart of correcting SL coordinates in the present invention.

[0042] Figure 4 It is the equivalent diagram of the corrected traversal points in the present invention.

[0043] Figure 5 It is the corrected obstacle boundary map in the present invention.

[0044] The above drawings include the following reference numerals: DETAILED DESCRIPTION

[0045] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0046] The technical solution of the present invention proposes a boundary refinement method for autonomous driving based on the SL coordinate system.

[0047] Reference Figures 1 to 5, in this embodiment, includes:

[0048] S1: Obtain the driving boundaries that the vehicle is allowed to drive when there are no obstacles in the lane. The driving boundaries include the left boundary, the right boundary and the center line;

[0049] S2: The SL coordinate system is established based on the centerline of the lane. S represents the longitudinal position along the centerline, and L represents the lateral position perpendicular to the centerline. The position of the centerline in the longitudinal direction is 0, the direction deviating from the left side of the centerline is the positive direction, and the direction deviating from the right side of the centerline is the negative direction.

[0050] S3: The data structure representation of the confirmed boundary is (S, L, R), where S represents the longitudinal position, L represents the maximum traversable lateral position in the positive direction at the corresponding S position, and R represents the maximum traversable lateral position in the negative direction at the corresponding S position;

[0051] S4: Set the obstacle detection boundary, and traverse and classify obstacles within the detection boundary;

[0052] S5: Iterate the obstacle boundary at a certain resolution to obtain the SL coordinate boundary of the obstacle boundary, and combine the driving boundary in the SL coordinate system with the obstacle boundary to obtain the final maximum boundary.

[0053] Optionally, in this embodiment, the driving boundary that allows the vehicle to travel when there is no obstacle in the lane is first obtained, and the driving boundary includes a left boundary, a right boundary and a center line, wherein the left boundary is the maximum lateral position for the vehicle to travel to the left when there is no obstacle, and the right boundary is the maximum lateral position for the vehicle to travel to the right when there is no obstacle, so as to determine the driving boundary of the vehicle; an SL coordinate system is established based on the center line of the lane, wherein the S coordinate in the SL coordinate system represents the longitudinal position along the center line direction, and the L coordinate represents the lateral position perpendicular to the center line direction, and at the same time, the position of the center line in the longitudinal direction is 0, the direction deviating from the left side of the center line is the positive direction, and the direction deviating from the right side of the center line is the negative direction; the data structure expression of the boundary is confirmed, which is (S, L, R), wherein S represents the longitudinal position, L and R represent the maximum positive drivable lateral position and the maximum negative drivable lateral position of the vehicle at the corresponding S position, respectively, and a series of data structures constitute the boundary, for example (0, 1, -1) represents the maximum positive drivable lateral position and the maximum negative drivable lateral position of the vehicle at the corresponding S position. When S=0m on the road, the vehicle can move 1m to the left and 1m to the right laterally; and (1, 1, -1) means that when the longitudinal position is 1m, the vehicle can move 1m to the left and 1m to the right laterally, that is, the center line is used as the positive and negative dividing line; set the obstacle detection boundary, traverse and classify to identify obstacles within the detection boundary. Specifically, the process of setting the obstacle boundary is: obtain the vehicle's own body boundary and the lane's road boundary, set the vehicle's own detection range by defining the vehicle's obstacle detection boundary, include obstacles within the range in the detection and perform subsequent analysis, and ignore obstacles outside the range, thereby reducing the amount of computation for vehicle obstacle detection; iterate and traverse the obstacle boundary at a certain resolution to obtain the SL coordinate boundary of the obstacle boundary, and combine the driving boundary under the SL coordinate system with the obstacle boundary to obtain the final maximum boundary. Through the above steps, the drivable boundary of the autonomous driving vehicle is optimized and the traffic efficiency of the autonomous driving vehicle is improved.

[0054] In this embodiment, the step S6 is also included: determining whether the obstacle is inclined; thereby calculating the influence of the obstacle's orientation angle on the boundary;

[0055] S61: If the angle between the boundary of the obstacle and the road is 0 or approaches 0, the obstacle is considered to be not tilted, and the boundary of the obstacle is the maximum driving boundary of the vehicle; the obstacle is considered to be regular, that is, the angle between the side of the obstacle closest to the center line and the center line is 0° or approaches 0°, and the obstacle is set to be parallel to the center line of the road, that is, the boundary of the obstacle closest to the center line is the driving boundary of the vehicle;

[0056] S62: If the angle between the boundary of the obstacle and the road is greater than a certain threshold, the obstacle is set to be inclined, and the longitudinal starting point S is obtained according to the four vertices of the obstacle. S, longitudinal end point E S , horizontal starting point S L , lateral end point E L , according to the SL coordinates of the four vertices, the corresponding left and right boundaries at the boundary S are obtained; if the angle between the boundary of the obstacle and the road is greater than the set threshold, the obstacle is considered to be inclined, and the left and right boundaries at the boundary S are obtained through the SL coordinates of the four vertices of the obstacle.

[0057] In this embodiment, the step further includes: if (S L +E L ) / 2>0, the center point of the obstacle is located on the left side of the lane centerline, and the left boundary corresponding to the boundary S includes the first boundary on the left and the second boundary on the left; where S L is the starting point of the obstacle in the longitudinal direction of the road, E L is the end point of the obstacle in the transverse direction of the road, that is, the obstacle is located on the left side of the center line, and the left boundary corresponding to the boundary S includes the first left boundary and the second left boundary;

[0058] S64: The first boundary on the left side is determined by the left vertex and the lower vertex of the obstacle; the second boundary on the left side is determined by the lower vertex and the right vertex of the obstacle; since the obstacle is located on the left side of the center line, the first boundary on the left side is the boundary formed by the left vertex and the lower vertex of the obstacle; the second boundary on the left side is the boundary formed by the lower vertex and the right vertex of the obstacle.

[0059] Furthermore, in this embodiment, the process further includes S65: if (S L +E L ) / 2<0, the center point of the obstacle is located on the right side of the lane centerline, and the right boundary corresponding to boundary S includes the first boundary on the right side and the second boundary on the right side; where S L is the starting point of the obstacle in the longitudinal direction of the road, E L The end point of the obstacle in the transverse direction of the road, that is, the obstacle is located on the right side of the center line. The right boundary corresponding to the boundary S includes the first boundary on the right side and the second boundary on the right side.

[0060] S66: The first boundary on the right side is determined by the left vertex and the upper vertex of the obstacle; the first boundary on the right side is determined by the upper vertex and the right vertex of the obstacle; since the obstacle is located on the right side of the center line, the first boundary on the right side is the boundary formed by the left vertex and the upper vertex of the obstacle; the second boundary on the right side is the boundary formed by the upper vertex and the right vertex of the obstacle.

[0061] In this embodiment, S7 is also included: correcting the SL coordinates of the vertex; when setting the resolution, the vertex may be sandwiched between two traversal points, so that the boundary of the vertex is cut off by the two traversal points, resulting in boundary distortion, thereby causing the vehicle to collide with the obstacle, so it is necessary to correct and process the SL coordinates of the vertex to prevent the boundary distortion;

[0062] S71: The obstacle is located on the left side of the center line. The longitudinal coordinates of the previous and next traversal points of the left vertex of the obstacle are equal to the longitudinal coordinates of the left vertex. The longitudinal coordinates of the previous and next traversal points of the right vertex of the obstacle are equal to the longitudinal coordinates of the right vertex. The transverse coordinates of the previous and next traversal points of the lower vertex of the obstacle are equal to the transverse coordinates of the lower vertex. Figure 4 , Figure 4 There are six points a, b, c, d, e, and f in the figure, that is, the two boundary sampling points before and after the obstacle boundary vertex will not be interpolated, that is, the longitudinal coordinates of points a and b are forced to be equal to the longitudinal coordinates of the left vertex, the transverse coordinates of points c and d are forced to be equal to the transverse coordinates of the lower vertex, and the longitudinal coordinates of points e and f are forced to be equal to the longitudinal coordinates of the right vertex.

[0063] S72: The obstacle is located on the right side of the center line. The longitudinal coordinates of the previous and next traversal points of the left vertex of the obstacle are equated with the longitudinal coordinates of the left vertex, the longitudinal coordinates of the previous and next traversal points of the right vertex of the obstacle are equated with the longitudinal coordinates of the right vertex, and the transverse coordinates of the previous and next traversal points of the upper vertex of the obstacle are equated with the transverse coordinates of the upper vertex.

[0064] In this embodiment, S8 is also included: if there are obstacles A and B; in other embodiments, there may be more than two obstacles, and the processing method is the same;

[0065] S81: Obstacles A and B are both located on the left side of the center line, obstacle A is traversed to determine the first left boundary, obstacle B is traversed to determine the second left boundary, and the first left boundary and the second left boundary are combined to determine the maximum driving boundary of the vehicle; Obstacles A and B are both located on the left side of the center line, that is, obstacle A determines the first left boundary, obstacle B determines the second left boundary, and the boundaries formed by the two obstacles are combined to obtain the final drivable boundary;

[0066] S82: Obstacle A is located on the left side of the center line, and obstacle B is located on the right side of the center line. Obstacle A is traversed to determine the left boundary, and obstacle B is traversed to determine the right boundary. The left boundary and the right boundary are combined to determine the maximum driving boundary of the vehicle. Obstacle A and obstacle B are located on both sides of the center line. Obstacle A and obstacle B are respectively determined to have a left boundary and a right boundary, and the left boundary and the right boundary are combined to obtain the final drivable boundary.

[0067] In this embodiment, the specific step S4 includes S41: obtaining the vehicle body boundary of the vehicle and the road boundary of the vehicle, the vehicle body boundary includes the front boundary of the vehicle body, the rear boundary of the vehicle body, the left boundary of the vehicle body and the right boundary of the vehicle body, and the road boundary includes the left road boundary and the right road boundary;

[0068] S42: Calculating a front obstacle detection boundary of the vehicle according to the front boundary of the vehicle and a preset forward detection distance;

[0069] S43: Calculating a rear obstacle detection boundary of the vehicle according to the rear boundary of the vehicle and a preset rearward detection distance;

[0070] S44: Calculating a left obstacle detection boundary of the vehicle according to a left boundary of the vehicle and a preset left detection distance;

[0071] S45: Calculating and obtaining a right obstacle detection boundary of the vehicle according to a right boundary of the vehicle and a preset right detection distance;

[0072] S46: Obtain the obstacle detection boundary of the vehicle based on the obtained front obstacle detection boundary, rear obstacle detection boundary, left obstacle detection boundary and right obstacle detection boundary; by defining the obstacle detection boundary of the vehicle and setting the vehicle's own detection range, obstacles within the range are included in the detection and subsequent analysis is performed, while obstacles outside the range are ignored, thereby reducing the amount of calculation for vehicle obstacle detection and improving planning efficiency.

[0073] The above-mentioned embodiments only express the implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.

Claims

1. A boundary refinement method for autonomous driving based on SL coordinate system, characterized in that: The following steps are involved: S1: Obtaining the driving boundary within which the vehicle is allowed to drive when there is no obstacle in the lane, wherein the driving boundary includes a left boundary, a right boundary and a center line; S2: The SL coordinate system is established based on the centerline of the lane. S represents the longitudinal position along the centerline, and L represents the lateral position perpendicular to the centerline. The position of the centerline in the longitudinal direction is 0, the direction deviating from the left side of the centerline is the positive direction, and the direction deviating from the right side of the centerline is the negative direction. S3: The data structure representation of the confirmed boundary is (S, L, R), where S represents the longitudinal position, L represents the maximum traversable lateral position in the positive direction at the corresponding S position, and R represents the maximum traversable lateral position in the negative direction at the corresponding S position; S4: Set the obstacle detection boundary, and traverse and classify obstacles within the detection boundary; S5: Iterate the obstacle boundary at a certain resolution to obtain the SL coordinate boundary of the obstacle boundary, and combine the driving boundary in the SL coordinate system with the obstacle boundary to obtain the final maximum boundary.

2. The boundary refinement method based on the SL coordinate system for autonomous driving according to claim 1, characterized in that: Also includes: S6: Determine whether the obstacle is tilted; S61: If the angle between the boundary of the obstacle and the road is 0 or tends to 0, the obstacle is deemed not to be inclined, and the boundary of the obstacle is the maximum driving boundary of the vehicle; S62: If the angle between the boundary of the obstacle and the road is greater than a certain threshold, the obstacle is set to be inclined, and the longitudinal starting point S is obtained according to the four vertices of the obstacle. S , longitudinal end point E S , horizontal starting point S L , lateral end point E L , according to the SL coordinates of the four vertices, the corresponding left and right boundaries at the boundary S are obtained.

3. The boundary refinement method based on SL coordinate system for autonomous driving according to claim 2, characterized in that: Also includes: S63: If (S L +E L ) / 2>0, the center point of the obstacle is located on the left side of the lane centerline, and the left boundary corresponding to the boundary S includes the first left boundary and the second left boundary; S64: The first left boundary is determined by the left vertex and the lower vertex of the obstacle; the second left boundary is determined by the lower vertex and the right vertex of the obstacle.

4. The boundary refinement method based on the SL coordinate system for autonomous driving according to claim 3, characterized in that: Also includes: S65: If (S L +E L ) / 2<0, the center point of the obstacle is located on the right side of the lane centerline, and the right boundary corresponding to the boundary S includes the first right boundary and the second right boundary; S66: The first right boundary is determined by the left vertex and the upper vertex of the obstacle; the first right boundary is determined by the upper vertex and the right vertex of the obstacle.

5. The boundary refinement method based on SL coordinate system for autonomous driving according to claim 2, characterized in that: Also includes: S7: Vertex SL coordinate correction; S71: The obstacle is located on the left side of the center line. The longitudinal coordinates of the previous and next traversal points of the left vertex of the obstacle are equal to the longitudinal coordinate of the left vertex. The longitudinal coordinates of the previous and next traversal points of the right vertex of the obstacle are equal to the longitudinal coordinate of the right vertex. The transverse coordinates of the previous and next traversal points of the lower vertex of the obstacle are equal to the transverse coordinate of the lower vertex. S72: The obstacle is located on the right side of the center line. The longitudinal coordinates of the previous and next traversal points of the left vertex of the obstacle are equated with the longitudinal coordinates of the left vertex, the longitudinal coordinates of the previous and next traversal points of the right vertex of the obstacle are equated with the longitudinal coordinates of the right vertex, and the transverse coordinates of the previous and next traversal points of the upper vertex of the obstacle are equated with the transverse coordinates of the upper vertex.

6. The boundary refinement method based on SL coordinate system for autonomous driving according to claim 1, characterized in that: Also includes: S8: If there are obstacles A and B; S81: Obstacles A and B are both located on the left side of the center line, obstacle A is traversed to determine a first left boundary, obstacle B is traversed to determine a second left boundary, and the maximum driving boundary of the vehicle is determined by combining the first left boundary and the second left boundary; S82: Obstacle A is located on the left side of the center line, and obstacle B is located on the right side of the center line. Obstacle A is traversed to determine the left boundary, and obstacle B is traversed to determine the right boundary. The maximum driving boundary of the vehicle is determined by combining the left boundary and the right boundary.

7. The boundary refinement method based on SL coordinate system for autonomous driving according to claim 6, characterized in that: The specific steps of S4 are: S41: Acquire the vehicle body boundary of the vehicle and the road boundary of the vehicle; the vehicle body boundary includes the front boundary of the vehicle body, the rear boundary of the vehicle body, the left boundary of the vehicle body and the right boundary of the vehicle body, and the road boundary includes the left road boundary and the right road boundary; S42: Calculating a front obstacle detection boundary of the vehicle according to the front boundary of the vehicle and a preset forward detection distance; S43: Calculating a rear obstacle detection boundary of the vehicle according to the rear boundary of the vehicle and a preset rearward detection distance; S44: Calculating a left obstacle detection boundary of the vehicle according to a left boundary of the vehicle and a preset left detection distance; S45: Calculating and obtaining a right obstacle detection boundary of the vehicle according to a right boundary of the vehicle and a preset right detection distance; S46: Obtain an obstacle detection boundary of the vehicle by merging the obtained front obstacle detection boundary, rear obstacle detection boundary, left obstacle detection boundary, and right obstacle detection boundary.

Images