Vehicle heading angle calculation method, vehicle posture calculation method, device and equipment

Abstract

The present invention discloses a vehicle heading angle calculation method, a vehicle posture calculation method, an apparatus, and equipment, relating to the field of communication technology, to solve the problem of large computational complexity in existing methods. The vehicle heading angle calculation method comprises: obtaining a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; determining a first point cloud mapped from the target point cloud to the ground, and determining a target rectangular frame surrounding the first point cloud; respectively determining the longitudinal edge region and the transverse edge region in the target rectangular frame; determining G rotation caddies based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region; respectively determining the edge distance corresponding to each of the G rotation caddies, and determining the target rotation caddie based on the edge distance; and determining the heading angle of the target vehicle based on the target rotation caddie. Embodiments of the present invention can reduce algorithm complexity and computational complexity.

Description

Technical Field

[0001] The present invention relates to the field of computer processing technology, and in particular to a vehicle heading angle calculation method, a vehicle posture calculation method, a device and equipment. Background Art

[0002] In the field of autonomous driving, the vehicle's posture is represented by a three-dimensional detection frame. Specifically, the vehicle's posture can be represented by a three-dimensional rectangular frame that includes length, width, height, and heading angle information. The vehicle posture algorithm in related technologies mainly maps the vehicle's three-dimensional point cloud to the ground, and then calculates the heading angle through the mapped ground point cloud. Specifically, the vehicle heading vector, the ground normal vector, and the cross product direction of the heading vector and the ground normal vector can be used as the target coordinate system. The difference between the maximum and minimum points of the ground point cloud in the target coordinate system is used as the target length, width, and height, and the target three-dimensional detection frame is further constructed. Therefore, in the calculation of vehicle posture, the key lies in the calculation of ground information and the vehicle target heading angle.

[0003] The relevant processing method is to calculate the heading angle based on the shape of the vehicle point cloud. This not only requires processing a large amount of point cloud data, but also is prone to incomplete removal of the background (non-target vehicle objects), resulting in an increase in the amount of calculation and affecting the accuracy of the calculation results.

[0004] It can be seen that the existing processing methods have the problem of large amount of calculation. Summary of the Invention

[0005] The embodiments of the present invention provide a vehicle heading angle calculation method, a vehicle posture calculation method, an apparatus and a device to solve the problem of large computational complexity in existing processing methods.

[0006] In a first aspect, an embodiment of the present invention provides a method for calculating a vehicle heading angle, comprising:

[0007] Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0008] Determine a first point cloud mapped from the target point cloud to the ground, and determine a target rectangular box surrounding the first point cloud;

[0009] Determining the longitudinal edge area and the transverse edge area in the target rectangular frame respectively, wherein the longitudinal edge area is the area corresponding to the long side of the target vehicle;

[0010] Determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located in the longitudinal edge region, and the transverse edge point cloud is a point cloud in the first point cloud located in the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1;

[0011] Determine the edge distance corresponding to each of the G rotating calipers respectively, and determine the target rotating caliper based on the edge distances, wherein the edge distance corresponding to the first rotating caliper includes a longitudinal distance and a transverse distance, the longitudinal distance includes the distance between the longitudinal edge point cloud and the longitudinal line segment of the first rotating caliper, and the transverse distance includes the distance between the transverse edge point cloud and the transverse line segment of the first rotating caliper, and the first rotating caliper is any rotating caliper among the G rotating calipers;

[0012] Based on the target rotation jam, a heading angle of the target vehicle is determined.

[0013] Optionally, after respectively determining the longitudinal edge area and the transverse edge area in the target rectangular frame, the method further includes:

[0014] removing noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud;

[0015] The determining G rotation caulking parts based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region comprises:

[0016] G rotation calipers are determined based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge region.

[0017] Optionally, determining G rotation caulking points based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region includes:

[0018] Selecting a first target point closest to the lateral outer line of the lateral edge area from the lateral edge point cloud;

[0019] Selecting a second target point closest to the longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud;

[0020] Selecting G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point;

[0021] Based on the first target point, the second target point and the G third target points, G rotational calipers are determined, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

[0022] Optionally, before determining the first point cloud mapped from the target point cloud to the ground, the method further includes:

[0023] Get ground point cloud;

[0024] Performing rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid;

[0025] Determining a ground expression corresponding to a grid according to each sub-ground point cloud;

[0026] Determining that the target point cloud is mapped to a first point cloud on the ground includes:

[0027] Determine the target point cloud center corresponding to the target point cloud;

[0028] Determine a target ground expression of a grid corresponding to the center of the target point cloud;

[0029] Based on the target ground expression, ground mapping processing is performed on the target point cloud to obtain a mapped first point cloud.

[0030] Optionally, respectively determining the edge distance corresponding to each of the G rotating caulkings includes:

[0031] Determine the sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotating caliper as the longitudinal distance, and determine the sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotating caliper as the transverse distance;

[0032] The sum of the longitudinal distance and the transverse distance is determined as the edge distance corresponding to the first rotating chuck.

[0033] Optionally, determining the target rotation jam based on the edge distance includes:

[0034] The rotating caliper with the smallest corresponding edge distance among the G rotating calipers is determined as the target rotating caliper.

[0035] Optionally, the separately determining the longitudinal edge area and the transverse edge area in the target rectangular frame includes:

[0036] Based on the preset longitudinal edge area width and transverse edge area width, and the initial heading angle of the target vehicle, the longitudinal edge area and the transverse edge area are respectively determined in the target rectangular frame.

[0037] Optionally, the target rectangular box is a minimum rectangular box surrounding the first point cloud.

[0038] In a second aspect, an embodiment of the present invention further provides a vehicle posture calculation method, comprising:

[0039] Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0040] Based on the heading angle of the target vehicle and the target point cloud, a three-dimensional detection frame for representing the position and posture of the target vehicle is determined, wherein the heading angle of the target vehicle is the heading angle determined using the method described in the first aspect above.

[0041] In a third aspect, an embodiment of the present invention further provides a vehicle heading angle calculation device, comprising:

[0042] A first acquisition module is configured to acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0043] A first determining module is configured to determine a first point cloud mapped from the target point cloud to the ground, and to determine a target rectangular frame surrounding the first point cloud;

[0044] A second determining module is configured to respectively determine a longitudinal edge region and a transverse edge region in the target rectangular frame, wherein the longitudinal edge region is a region corresponding to a long side of the target vehicle;

[0045] a third determination module, configured to determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located within the longitudinal edge region, the transverse edge point cloud is a point cloud in the first point cloud located within the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1;

[0046] a fourth determination module, configured to respectively determine an edge distance corresponding to each of the G rotational caulkings, and determine a target rotational caulking based on the edge distances, wherein the edge distance corresponding to a first rotational caulking includes a longitudinal distance and a transverse distance, the longitudinal distance including a distance between the longitudinal edge point cloud and a longitudinal line segment of the first rotational caulking, and the transverse distance including a distance between the transverse edge point cloud and a transverse line segment of the first rotational caulking, and the first rotational caulking is any one of the G rotational caulkings;

[0047] A fifth determination module is configured to determine a heading angle of the target vehicle based on the target rotation jam.

[0048] Optionally, the vehicle posture calculation device further includes:

[0049] a first processing module, configured to remove noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud;

[0050] The third determination module is used to determine G rotation caulking points based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge area.

[0051] Optionally, the third determining module includes:

[0052] a first selection unit, configured to select a first target point closest to a lateral outer line of the lateral edge region from the lateral edge point cloud;

[0053] a second selection unit, configured to select a second target point closest to a longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud;

[0054] a third selection unit, configured to select G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point;

[0055] The first determining unit is configured to determine G rotational calipers based on the first target point, the second target point, and the G third target points, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

[0056] Optionally, the vehicle posture calculation device further includes:

[0057] The second acquisition module is used to obtain ground point cloud;

[0058] a sixth determining module, configured to perform rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid;

[0059] a seventh determining module, configured to determine a ground expression corresponding to a grid based on each sub-ground point cloud;

[0060] The first determining module includes:

[0061] A second determining unit is used to determine the target point cloud center corresponding to the target point cloud;

[0062] A third determining unit is used to determine a target ground expression of a grid corresponding to the center of the target point cloud;

[0063] A processing unit is used to perform ground mapping processing on the target point cloud based on the target ground expression to obtain a mapped first point cloud.

[0064] Optionally, the fourth determining module includes:

[0065] a fourth determining unit, configured to determine a sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotation caliper as the longitudinal distance, and to determine a sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotation caliper as the transverse distance;

[0066] The fifth determining unit is configured to determine the sum of the longitudinal distance and the transverse distance as the edge distance corresponding to the first rotating housing.

[0067] Optionally, the fourth determining module further includes:

[0068] The sixth determining unit is configured to determine a rotation chuck having the smallest corresponding edge distance among the G rotation chucks as a target rotation chuck.

[0069] Optionally, the second determining module is used to determine the longitudinal edge area and the transverse edge area in the target rectangular frame respectively based on a preset longitudinal edge area width and transverse edge area width, and an initial heading angle of the target vehicle.

[0070] Optionally, the target rectangular box is a minimum rectangular box surrounding the first point cloud.

[0071] In a fourth aspect, an embodiment of the present invention further provides a vehicle posture calculation device, comprising:

[0072] An acquisition module, configured to acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0073] A determination module is used to determine a three-dimensional detection frame for representing the position and posture of the target vehicle based on the heading angle of the target vehicle and the target point cloud, wherein the heading angle of the target vehicle is the heading angle determined using the method described in the first aspect above.

[0074] In the fifth aspect, an embodiment of the present invention also provides a vehicle heading angle calculation device, comprising: a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the computer program, the steps in the vehicle heading angle calculation method described in the first aspect above are implemented.

[0075] In the sixth aspect, an embodiment of the present invention also provides a vehicle posture calculation device, comprising: a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the computer program, the steps in the vehicle posture calculation method described in the second aspect above are implemented.

[0076] In the seventh aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the vehicle heading angle calculation method as described in the first aspect above are implemented; or the steps in the vehicle posture calculation method as described in the second aspect above are implemented.

[0077] In an embodiment of the present invention, a target point cloud is obtained, wherein the target point cloud includes a point cloud of a target vehicle; a first point cloud mapped to the ground by the target point cloud is determined, and a target rectangular frame surrounding the first point cloud is determined; a longitudinal edge region and a transverse edge region in the target rectangular frame are respectively determined, wherein the longitudinal edge region is a region corresponding to a long side of the target vehicle; based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, G rotation cassettes are determined, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located in the longitudinal edge region, the transverse edge point cloud is a point cloud in the first point cloud located in the transverse edge region, and the rotation cassette is composed of mutually perpendicular longitudinal line segments and transverse line segments. The longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1; the edge distance corresponding to each of the G rotating caddies is determined respectively, and based on the edge distance, the target rotating caddy is determined, wherein the edge distance corresponding to the first rotating caddy includes the longitudinal distance and the transverse distance, the longitudinal distance includes the distance from the longitudinal edge point cloud to the longitudinal line segment of the first rotating caddy, and the transverse distance includes the distance from the transverse edge point cloud to the transverse line segment of the first rotating caddy, and the first rotating caddy is any rotating caddy among the G rotating caddies; based on the target rotating caddy, the heading angle of the target vehicle is determined.

[0078] In this way, since it is only necessary to calculate the heading angle of the target vehicle based on the edge point clouds in the longitudinal edge area and the lateral edge area in the target rectangular box corresponding to the target point cloud, and adopt the rotating camber method, the algorithm complexity can be reduced to a certain extent and the amount of calculation can be reduced compared with the calculation method based on a large number of target point clouds in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in describing the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0080] Figure 1 is a flow chart of a method for calculating a vehicle heading angle provided by an embodiment of the present invention;

[0081] Figure 2 Schematic diagram of mapping a target point cloud onto the ground provided by an embodiment of the present invention;

[0082] Figure 3is a schematic diagram of selecting a region of interest in a target rectangular frame provided by an embodiment of the present invention;

[0083] Figure 4 is a schematic diagram of determining a rotation caulking based on a region of interest provided by an embodiment of the present invention;

[0084] Figure 5 is a flow chart of a vehicle posture method provided by an embodiment of the present invention;

[0085] Figure 6 is a structural diagram of a vehicle heading angle calculation device provided by an embodiment of the present invention;

[0086] Figure 7 is a structural diagram of a vehicle posture calculation device provided by an embodiment of the present invention;

[0087] Figure 8 is a structural diagram of a vehicle heading angle calculation device provided by an embodiment of the present invention;

[0088] Figure 9 It is a structural diagram of a vehicle posture calculation device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0089] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0090] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for calculating a vehicle heading angle according to an embodiment of the present invention. Figure 1 As shown, the following steps are included:

[0091] Step 101: Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle.

[0092] The embodiment of the present invention is a vehicle heading angle calculation method proposed for a roadside perception scenario, that is, the above-mentioned target point cloud is obtained through roadside perception.

[0093] Specifically, when the target vehicle is in motion, a laser radar on the road surface can be used to perform real-time perception of the roadside to obtain a roadside point cloud. Background point clouds, such as those representing static objects such as buildings, scenery, and obstacles, can then be removed from the roadside point cloud. Finally, a clustering algorithm can be used to cluster the remaining roadside point clouds after removing the background point clouds to obtain the target vehicle's point cloud, also known as the target point cloud. This can specifically include target point cloud clusters and target point cloud centers. The target vehicle is the vehicle for which the heading angle or vehicle posture needs to be calculated.

[0094] Step 102: Determine a first point cloud mapped from the target point cloud to the ground, and determine a target rectangular box surrounding the first point cloud.

[0095] After obtaining the target point cloud, a mapping process can be performed on the target point cloud, specifically mapping the target point cloud to the ground, thereby obtaining a first point cloud. Specifically, a mapping relationship between the target point cloud and the ground point cloud can be determined based on a previously obtained ground point cloud, that is, a mapping formula between the target point cloud coordinates and the ground point cloud coordinates is established, and then, based on the mapping formula, a first point cloud mapped from the target point cloud to the ground is determined.

[0096] After obtaining the first point cloud, a target rectangular box can be further determined on the ground surface corresponding to the first point cloud to encompass the first point cloud. Specifically, the target rectangular box can be the smallest rectangular box encompassing the first point cloud. In other words, the target rectangular box can be the smallest rectangular box encompassing the first point cloud. For example, based on the distribution of the first point cloud on the ground and the edge points in the first point cloud, a rectangular box encompassing the first point cloud with the smallest area can be determined. This ensures minimal computational effort and accuracy.

[0097] Optionally, before step 102, the method further includes:

[0098] Get ground point cloud;

[0099] Performing rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid;

[0100] Determining a ground expression corresponding to a grid according to each sub-ground point cloud;

[0101] Determining that the target point cloud is mapped to a first point cloud on the ground includes:

[0102] Determine the target point cloud center corresponding to the target point cloud;

[0103] Determine a target ground expression of a grid corresponding to the center of the target point cloud;

[0104] Based on the target ground expression, ground mapping processing is performed on the target point cloud to obtain a mapped first point cloud.

[0105] That is, in one embodiment, the ground point cloud can be obtained in advance and rasterized, and then a mapping formula for mapping the target point cloud to the ground can be established. Then, the target point cloud can be mapped using the mapping formula to obtain the first point cloud.

[0106] Specifically, before performing real-time roadside perception to obtain the target point cloud, point cloud perception data can be collected by using a lidar on the road surface, and the collected point cloud perception data can be segmented using a region growing algorithm to obtain a ground point cloud.

[0107] Then, the ground point cloud can be rasterized, specifically, the ground corresponding to the ground point cloud is divided into multiple grids, and the ground point cloud is dispersed in multiple grids, which is equivalent to the ground point cloud being divided into multiple sub-ground point clouds by the multiple grids, and each grid includes a sub-ground point cloud.

[0108] Next, the ground expression of each grid corresponding to each sub-ground point cloud can be determined. Specifically, the RANdom Sampling Consensus (RANSAC) algorithm can be used to fit the sub-ground corresponding to the sub-ground point cloud in each grid to obtain the ground equation representing the curvature of each sub-ground, and thus the ground expression of each grid can be determined. For example, the ground expression of the i-th grid is determined as: A i x+B i y+C i z+D i =0.

[0109] In this way, after determining the ground expression of each grid, the grid corresponding to the target point cloud can be determined, and then the target point cloud can be ground mapped based on the ground expression of the grid corresponding to the target point cloud. Specifically, when determining the target point cloud, the target point cloud center corresponding to the target point cloud can be determined by a clustering algorithm, and then the target ground expression of the grid corresponding to the target point cloud center can be determined. Finally, based on the target ground expression, the corresponding target mapping formula can be determined to obtain the first point cloud mapped to the ground by the target point cloud.

[0110] For example, if the ground expression of grid i is A i x+B i y+C i z+D i =0, then the corresponding mapping formula can be determined as Among them, (x o ,y o ,z o ) is the target point cloud coordinate, (x p ,y p ,z p ) is the first point cloud coordinate of the ground. The mapping diagram of the 3D target point cloud to the ground can be seen in Figure 2 , P A ′、P B ′ and P C ′ is the target point cloud P A 、P B and P C The point cloud after mapping to the ground, is the ground normal vector.

[0111] In this way, through this embodiment, the first point cloud mapped from the target point cloud to the ground can be determined more accurately and quickly.

[0112] Step 103 : Determine the longitudinal edge region and the transverse edge region in the target rectangular frame respectively, wherein the longitudinal edge region is the region corresponding to the long side of the target vehicle.

[0113] After determining the target rectangular frame, in order to reduce the amount of calculation and ensure the calculation accuracy, the points in the region of interest can be used as the calculation points of the vehicle heading angle. Specifically, considering that the point cloud distribution is denser on both sides of the vehicle close to the laser radar, it has greater reference significance for describing the target vehicle. Therefore, the areas on both sides of the vehicle can be used as the region of interest. The higher quality areas in the target point cloud can be divided into two parts, namely the edge area close to the laser radar ( Figure 3 The lateral edge area 32 in the longitudinal direction close to the laser radar ( Figure 3 The longitudinal edge region 31 in FIG.

[0114] That is, a longitudinal edge region and a transverse edge region can be respectively determined within the target rectangular frame, and used to calculate a rotation caliper capable of fitting the vehicle heading angle based on points within these two edge regions. The longitudinal edge region can be an edge region within the target rectangular frame that corresponds to the long side of the target vehicle and is close to the road surface laser radar, and the transverse edge region can be a side within the target rectangular frame that corresponds to the short side of the target vehicle and is laterally close to the road surface laser radar.

[0115] Specifically, the initial heading angle of the target vehicle can be calculated based on the principal component analysis (PCA) algorithm, and the heading angle of the vehicle is used as the direction of the rectangular frame, and the corresponding longitudinal edge area and transverse edge area are determined in the target rectangular frame. For example, see Figure 3 , L pca It represents the initial heading angle calculated by the PCA algorithm, which can be used as the direction of the target rectangular box 30. When the target rectangular box is located on the left or right side of the laser radar 34, the longitudinal edge area 31 and the transverse edge area 32 as shown in the figure can be determined in the target rectangular box 30. The area 33 in the figure represents the corresponding inner area of ​​the vehicle.

[0116] It should be noted that the lengths of the longitudinal edge region and the transverse edge region may be consistent with the length of the target rectangular frame, and the widths of the longitudinal edge region and the transverse edge region may be selected empirically or preset.

[0117] Optionally, step 103 includes:

[0118] Based on the preset longitudinal edge area width and transverse edge area width, and the initial heading angle of the target vehicle, the longitudinal edge area and the transverse edge area are respectively determined in the target rectangular frame.

[0119] That is, in one method, the width of the longitudinal edge region and the width of the transverse edge region can be pre-set, that is, the widths of the two edge regions can be initialized separately, such as initializing them to width Δw. Thus, after obtaining the target rectangular frame, the longitudinal edge region and the transverse edge region in the target rectangular frame can be determined based on the preset longitudinal edge region width and the transverse edge region width, and in combination with the initial heading angle of the target vehicle. Specifically, the longitudinal edge region and the transverse edge region can be determined on the specific side of the target rectangular frame based on the initial heading angle of the target vehicle, and then based on the preset longitudinal edge region width and the transverse edge region width, the specific longitudinal edge region and the transverse edge region can be determined, wherein the initial heading angle of the target vehicle can be calculated using the PCA algorithm on the target point cloud.

[0120] In this way, through this embodiment, the two regions of interest, namely the longitudinal edge region and the transverse edge region in the target rectangular frame, can be determined quickly and accurately.

[0121] Step 104: Determine G rotation chucks based on the longitudinal edge point cloud in the longitudinal edge area and the transverse edge point cloud in the transverse edge area, wherein the longitudinal edge point cloud is the point cloud in the first point cloud located in the longitudinal edge area, and the transverse edge point cloud is the point cloud in the first point cloud located in the transverse edge area. The rotation chucks are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1.

[0122] In an embodiment of the present invention, the principle of the rotation caddy can be used to fit the final heading angle. Therefore, in this step, multiple rotation caddies can be determined based on the longitudinal edge point cloud in the longitudinal edge area and the transverse edge point cloud in the transverse edge area, so as to determine a target rotation caddy that meets the requirements from the multiple rotation caddies; wherein the longitudinal edge point cloud is the portion of the point cloud in the first point cloud distributed in the longitudinal edge area, and the transverse edge point cloud is the portion of the point cloud in the first point cloud distributed in the transverse edge area.

[0123] Specifically, some target points can be selected from the longitudinal edge point cloud and the transverse edge point cloud respectively to form mutually perpendicular rotation caulkings, that is, right-angle caulkings. In order to ensure that the rotation caulking can be as close as possible to the actual heading angle, a target point located on the outer side, that is, closer to the outer transverse edge of the transverse edge area, can be selected from the transverse edge point cloud. A target point located on the outer side, that is, closer to the outer longitudinal edge of the longitudinal edge area can also be selected from the longitudinal edge point cloud. Another target point can also be randomly selected from the longitudinal edge point cloud. That is, a rotation caulking can be formed by selecting two, three or more target points.

[0124] Optionally, step 104 includes:

[0125] Selecting a first target point closest to the lateral outer line of the lateral edge area from the lateral edge point cloud;

[0126] Selecting a second target point closest to the longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud;

[0127] Selecting G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point;

[0128] Based on the first target point, the second target point and the G third target points, G rotational calipers are determined, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

[0129] That is, in one embodiment, the desired target points can be selected in a relatively regular manner, and a plurality of pending rotation caulking members can be formed by selecting different target points.

[0130] Specifically, a first target point closest to the lateral outer line of the lateral edge area can be selected from the lateral edge point cloud, and a second target point closest to the longitudinal outer line of the longitudinal edge area can be selected from the longitudinal edge point cloud, and G third target points can be selected from the remaining points, that is, the points in the longitudinal edge point cloud other than the second target point. In order to obtain a more comprehensive result, all the remaining points can be used as third target points.

[0131] That is, for the lateral edge region, only one point located at the outermost side (corresponding to the point closest to the lateral edge of the target vehicle) can be selected as the first target point. For the longitudinal edge region, one point located at the outermost side (corresponding to the point closest to the longitudinal edge of the target vehicle) can be selected as the second target point, and one point can be selected from the remaining points in sequence as the third target point. The first target point, the second target point, and any one of the third target points each constitute a rotational chuck. In determining the rotational chuck, a longitudinal line segment can be obtained by connecting the second target point and any one of the third target points. A transverse line segment perpendicular to the longitudinal line segment can then be drawn through the first target point to obtain a rotational chuck.

[0132] Among them, for the convenience of calculation, each time a third target point is selected, the currently selected third target point can be numbered, and the constructed rotation caulking can be numbered accordingly, so that for the G third target points, the first rotation caulking, the second rotation caulking,... and the Gth rotation caulking can be obtained in sequence.

[0133] For example, see Figure 4 , we can first determine the horizontal edge area 41 and the outer line L b The closest point P0 is used as one of the fulcrums of the rotating chuck, and then a point in the longitudinal edge area 42 that is adjacent to the outer line L is selected. r The closest point P1 and any other point P2 in the longitudinal edge region 42 form a line segment L l , then determine a line that passes through P0 and intersects with line segment L l The vertical line segment L w , so line segment L l and line segment L wThe direction of the right-angled caulking is the line segment L marked in the figure. l The direction indicated by the arrow can be used as the corresponding heading angle.

[0134] In this way, by selecting the outermost target points from the longitudinal edge area and the transverse edge area respectively, and selecting any other point from the longitudinal edge area respectively, a plurality of rotating caulkings can be formed, which can not only reduce unnecessary calculation amount but also ensure that the calculated results are more accurate and reliable.

[0135] Optionally, after step 103, the method further includes:

[0136] removing noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud;

[0137] The step 104 includes:

[0138] G rotation calipers are determined based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge region.

[0139] That is, in one embodiment, the longitudinal edge region may be first subjected to denoising processing, that is, noise points in the longitudinal edge point cloud may be removed. Specifically, the RANSAC algorithm may be used to perform straight line fitting on the longitudinal edge point cloud in the longitudinal edge region to obtain a fitted straight line. Then, points in the longitudinal edge point cloud that are farther away from the fitted straight line are removed. For example, points whose distance from the fitted straight line is greater than a preset distance threshold will be removed as noise points. Other points in the longitudinal edge point cloud, that is, points that are not removed but retained, will be used as longitudinal edge points to participate in the rotation caulking calculation. In other words, when determining the rotation caulking, G rotation caulkings may be determined based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge region. The specific determination method is similar to that in step 104 and will not be repeated here.

[0140] In this way, by removing the noise points in the longitudinal edge point cloud, the amount of calculation when determining the rotation camber can be further reduced, the operation speed can be improved, thereby making the real-time performance higher, and the ability to resist noise interference can be improved, ensuring that the heading angle finally obtained by fitting the rotation camber is more accurate.

[0141] Step 105: Determine the edge distance corresponding to each of the G rotational caulkings respectively, and determine the target rotational caulking based on the edge distances, wherein the edge distance corresponding to the first rotational caulking includes a longitudinal distance and a transverse distance, the longitudinal distance includes the distance between the longitudinal edge point cloud and the longitudinal line segment of the first rotational caulking, and the transverse distance includes the distance between the transverse edge point cloud and the transverse line segment of the first rotational caulking, and the first rotational caulking is any rotational caulking among the G rotational caulkings.

[0142] After determining the G rotational caddies, in order to obtain a target rotational caddie that can more realistically reflect the heading angle of the target vehicle by fitting, the target rotational caddie can be determined by respectively calculating the edge distance corresponding to each of the G rotational caddies and based on the edge distance corresponding to each rotational caddie.

[0143] Among them, for any rotating caulking of the G rotating caulkings, its corresponding edge distance can be determined by calculating the distance from the point in each edge area to the corresponding edge line segment, that is, the longitudinal distance from the longitudinal edge point cloud in the longitudinal edge area to the longitudinal line segment in the rotating caulking, and the transverse distance from the transverse edge point cloud in the transverse edge area to the transverse line segment in the rotating caulking are calculated respectively, and then the corresponding edge distance is determined based on the longitudinal distance and the transverse distance.

[0144] After determining the edge distance corresponding to each rotating chuck, the target rotating chuck that meets the requirements can be determined based on the edge distance corresponding to each rotating chuck. Since the size of the edge distance can reflect the degree of fit of the rotating chuck to the longitudinal edge point cloud and the transverse edge point cloud to a certain extent, for example, the smaller the edge distance, the closer the corresponding rotating chuck is to each point, and the smaller the degree of deviation from each point. Therefore, a rotating chuck with a relatively small edge distance can be selected as the target rotating chuck.

[0145] Optionally, respectively determining the edge distance corresponding to each of the G rotating caulkings includes:

[0146] Determine the sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotating caliper as the longitudinal distance, and determine the sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotating caliper as the transverse distance;

[0147] The sum of the longitudinal distance and the transverse distance is determined as the edge distance corresponding to the first rotating chuck.

[0148] That is, in one embodiment, for any one of the G rotating caulks, referred to as the first rotating caulk, the distance from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotating caulk can be calculated separately, and the sum of these distances is the longitudinal distance corresponding to the first rotating caulk, and the distance from each point in the transverse edge point cloud to the transverse line segment of the first rotating caulk can be calculated separately, and the sum of these distances is the transverse distance corresponding to the first rotating caulk. Finally, the longitudinal distance is added to the transverse distance to obtain the edge distance corresponding to the first rotating caulk.

[0149] That is, for the i-th rotating caulking among the G rotating caulkings, the edge calculation formula can be used To calculate the corresponding edge distance, where D i Indicates the edge distance corresponding to the i-th rotating caliper, dr j Represents the distance from each point in the longitudinal edge point cloud to the longitudinal line segment of the i-th rotation caliper, n is the number of longitudinal edge point clouds, db k Represents the distance from each point in the lateral edge point cloud to the lateral line segment of the i-th rotation caliper, m is the number of lateral edge point clouds, and the value of i ranges from 1 to G.

[0150] In this way, through this embodiment, the edge distance that can truly reflect the degree of fit of each rotating chuck to the longitudinal edge point cloud and the transverse edge point cloud can be quickly calculated, which helps to accurately determine the target rotating chuck.

[0151] Optionally, determining the target rotation jam based on the edge distance includes:

[0152] The rotating caliper with the smallest corresponding edge distance among the G rotating calipers is determined as the target rotating caliper.

[0153] That is, in one embodiment, in order to ensure the calculation accuracy of the heading angle of the target vehicle, the rotating caddy with the smallest corresponding edge distance among the G rotating caddies can be directly used as the target rotating caddy, such as when D1, D2, ..., D are calculated respectively. G After that, the minimum value can be determined, and the rotation caulking corresponding to the minimum value is the target rotation caulking.

[0154] Step 106: Determine the heading angle of the target vehicle based on the target rotation jam.

[0155] After the target rotation jam is determined, the direction of the target rotation jam can be determined based on the target rotation jam. The angle between the direction and the reference direction is the heading angle of the target vehicle. For example, Figure 4 The rotating caliper shown can be based on the longitudinal line segment L lThe direction indicated by the arrow determines the heading angle of the target vehicle.

[0156] The vehicle heading angle calculation method of the embodiment of the present invention obtains a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; determines a first point cloud mapped to the ground by the target point cloud, and determines a target rectangular frame surrounding the first point cloud; determines a longitudinal edge area and a transverse edge area in the target rectangular frame respectively, wherein the longitudinal edge area is an area corresponding to the long side of the target vehicle; based on the longitudinal edge point cloud in the longitudinal edge area and the transverse edge point cloud in the transverse edge area, determines G rotation cassettes, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located in the longitudinal edge area, and the transverse edge point cloud is a point cloud in the first point cloud located in the transverse edge area, and the rotation cassette is composed of mutually perpendicular longitudinal point clouds. The method comprises the following steps: a first rotating caddy is composed of a longitudinal line segment and a transverse line segment, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1; the edge distance corresponding to each rotating caddy in the G rotating caddy is determined respectively, and based on the edge distance, the target rotating caddy is determined, wherein the edge distance corresponding to the first rotating caddy includes a longitudinal distance and a transverse distance, the longitudinal distance includes the distance from the longitudinal edge point cloud to the longitudinal line segment of the first rotating caddy, and the transverse distance includes the distance from the transverse edge point cloud to the transverse line segment of the first rotating caddy, and the first rotating caddy is any rotating caddy in the G rotating caddy; based on the target rotating caddy, the heading angle of the target vehicle is determined.

[0157] In this way, since it is only necessary to calculate the heading angle of the target vehicle based on the edge point clouds in the longitudinal edge area and the lateral edge area in the target rectangular box corresponding to the target point cloud, and adopt the rotating camber method, the algorithm complexity can be reduced to a certain extent and the amount of calculation can be reduced compared with the calculation method based on a large number of target point clouds in the existing technology.

[0158] See also Figure 5 , Figure 5 is a flow chart of a vehicle posture calculation method provided by an embodiment of the present invention, such as Figure 5 As shown, the following steps are included:

[0159] Step 501: Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle.

[0160] The implementation of this step can be found in the description of step 101, and will not be described again here to avoid repetition.

[0161] Step 502: Based on the heading angle of the target vehicle and the target point cloud, a three-dimensional detection frame is determined for representing the position of the target vehicle, wherein the heading angle of the target vehicle is Figure 1 The vehicle heading angle calculation method in the illustrated embodiment determines the obtained heading angle.

[0162] In this embodiment, based on Figure 1 The heading angle of the target vehicle determined in the embodiment of the method shown and the target point cloud are used to construct a three-dimensional detection box surrounding the target point cloud. Figure 1 The relevant introduction of the embodiment of the method shown will not be repeated here.

[0163] The vehicle posture calculation method of the embodiment of the present invention obtains a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; based on the heading angle of the target vehicle and the target point cloud, determines a three-dimensional detection frame for representing the posture of the target vehicle, wherein the heading angle of the target vehicle is Figure 1 The vehicle heading angle calculation method in the illustrated embodiment determines the obtained heading angle. In this way, since the amount of calculation of the heading angle is reduced, the amount of calculation of the vehicle posture calculation based on the heading angle can also be reduced compared to the existing technology.

[0164] The embodiment of the present invention also provides a vehicle heading angle calculation device. Figure 6 , Figure 6 is a structural diagram of a vehicle heading angle calculation device provided by an embodiment of the present invention. Since the principles of the vehicle heading angle calculation device and the vehicle heading angle calculation method in the embodiment of the present invention are similar, the implementation of the vehicle heading angle calculation device can be referred to in the implementation of the method, and the repeated parts will not be repeated here.

[0165] like Figure 6 As shown, the vehicle heading angle calculation device 600 includes:

[0166] A first acquisition module 601 is configured to acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0167] A first determining module 602 is configured to determine a first point cloud mapped from the target point cloud to the ground, and determine a target rectangular frame surrounding the first point cloud;

[0168] A second determining module 603 is configured to respectively determine a longitudinal edge region and a transverse edge region in the target rectangular frame, wherein the longitudinal edge region is a region corresponding to a long side of the target vehicle;

[0169] a third determining module 604, configured to determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located within the longitudinal edge region, the transverse edge point cloud is a point cloud in the first point cloud located within the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1;

[0170] A fourth determination module 605 is configured to respectively determine an edge distance corresponding to each of the G rotational caulkings, and determine a target rotational caulking based on the edge distances, wherein the edge distance corresponding to a first rotational caulking includes a longitudinal distance and a transverse distance, the longitudinal distance including a distance between the longitudinal edge point cloud and a longitudinal line segment of the first rotational caulking, and the transverse distance including a distance between the transverse edge point cloud and a transverse line segment of the first rotational caulking, and the first rotational caulking is any one of the G rotational caulkings;

[0171] The fifth determining module 606 is configured to determine the heading angle of the target vehicle based on the target rotation stuck.

[0172] Optionally, the vehicle posture calculation device 600 further includes:

[0173] a first processing module, configured to remove noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud;

[0174] The third determination module is used to determine G rotation caulking points based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge area.

[0175] Optionally, the third determining module 604 includes:

[0176] a first selection unit, configured to select a first target point closest to a lateral outer line of the lateral edge region from the lateral edge point cloud;

[0177] a second selection unit, configured to select a second target point closest to a longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud;

[0178] a third selection unit, configured to select G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point;

[0179] The first determining unit is configured to determine G rotational calipers based on the first target point, the second target point, and the G third target points, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

[0180] Optionally, the vehicle posture calculation device 600 further includes:

[0181] The second acquisition module is used to obtain ground point cloud;

[0182] a sixth determining module, configured to perform rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid;

[0183] a seventh determining module, configured to determine a ground expression corresponding to a grid based on each sub-ground point cloud;

[0184] The first determining module 602 includes:

[0185] A second determining unit is used to determine the target point cloud center corresponding to the target point cloud;

[0186] A third determining unit is used to determine a target ground expression of a grid corresponding to the center of the target point cloud;

[0187] A processing unit is used to perform ground mapping processing on the target point cloud based on the target ground expression to obtain a mapped first point cloud.

[0188] Optionally, the fourth determining module 605 includes:

[0189] a fourth determining unit, configured to determine a sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotation caliper as the longitudinal distance, and to determine a sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotation caliper as the transverse distance;

[0190] The fifth determining unit is configured to determine the sum of the longitudinal distance and the transverse distance as the edge distance corresponding to the first rotating housing.

[0191] Optionally, the fourth determining module 605 further includes:

[0192] The sixth determining unit is configured to determine a rotation chuck having the smallest corresponding edge distance among the G rotation chucks as a target rotation chuck.

[0193] Optionally, the second determining module 603 is configured to respectively determine the longitudinal edge area and the transverse edge area in the target rectangular frame based on a preset longitudinal edge area width and transverse edge area width, and an initial heading angle of the target vehicle.

[0194] Optionally, the target rectangular box is a minimum rectangular box surrounding the first point cloud.

[0195] The vehicle heading angle calculation device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effects are similar, which will not be repeated here in this embodiment.

[0196] The embodiment of the present invention also provides a vehicle posture calculation device. Figure 7 , Figure 7 is a structural diagram of a vehicle posture calculation device provided by an embodiment of the present invention. Since the principles of the vehicle posture calculation device for solving the problem are similar to those of the vehicle posture calculation method in the embodiment of the present invention, the implementation of the vehicle posture calculation device can be referred to in the implementation of the method, and the repeated parts will not be repeated here.

[0197] like Figure 7 As shown, the vehicle posture calculation device 700 includes:

[0198] An acquisition module 701 is configured to acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0199] The determination module 702 is used to determine a three-dimensional detection frame for representing the position of the target vehicle based on the heading angle of the target vehicle and the target point cloud, wherein the heading angle of the target vehicle is Figure 1 The vehicle heading angle calculation method in the illustrated embodiment determines the obtained heading angle.

[0200] The vehicle posture calculation device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effects are similar, which will not be repeated here in this embodiment.

[0201] The embodiment of the present invention also provides a vehicle heading angle calculation device. Since the principle of the vehicle heading angle calculation device to solve the problem is similar to the vehicle heading angle calculation method in the embodiment of the present invention, the implementation of the vehicle heading angle calculation device can refer to the implementation of the method, and the repeated parts will not be repeated. Figure 8 As shown, the vehicle heading angle calculation device according to an embodiment of the present invention includes:

[0202] The processor 800 is configured to read the program in the memory 820 and execute the following process:

[0203] Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0204] Determine a first point cloud mapped from the target point cloud to the ground, and determine a target rectangular box surrounding the first point cloud;

[0205] Determining the longitudinal edge area and the transverse edge area in the target rectangular frame respectively, wherein the longitudinal edge area is the area corresponding to the long side of the target vehicle;

[0206] Determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located in the longitudinal edge region, and the transverse edge point cloud is a point cloud in the first point cloud located in the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1;

[0207] Determine the edge distance corresponding to each of the G rotating calipers respectively, and determine the target rotating caliper based on the edge distances, wherein the edge distance corresponding to the first rotating caliper includes a longitudinal distance and a transverse distance, the longitudinal distance includes the distance between the longitudinal edge point cloud and the longitudinal line segment of the first rotating caliper, and the transverse distance includes the distance between the transverse edge point cloud and the transverse line segment of the first rotating caliper, and the first rotating caliper is any rotating caliper among the G rotating calipers;

[0208] Based on the target rotation jam, a heading angle of the target vehicle is determined.

[0209] Among them, Figure 8 In the present disclosure, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits such as one or more processors represented by processor 800 and memory represented by memory 820. The bus architecture may also link various other circuits such as peripherals, voltage regulators, and power management circuits, all of which are well known in the art and are not further described herein. The bus interface provides an interface. Processor 800 is responsible for managing the bus architecture and general processing, while memory 820 may store data used by processor 800 when performing operations.

[0210] Optionally, the processor 800 is further configured to read a program in the memory 820 and execute the following steps:

[0211] removing noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud;

[0212] G rotation calipers are determined based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge region.

[0213] Optionally, the processor 800 is further configured to read a program in the memory 820 and execute the following steps:

[0214] Selecting a first target point closest to the lateral outer line of the lateral edge area from the lateral edge point cloud;

[0215] Selecting a second target point closest to the longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud;

[0216] Selecting G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point;

[0217] Based on the first target point, the second target point and the G third target points, G rotational calipers are determined, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

[0218] Optionally, the processor 800 is further configured to read a program in the memory 820 and execute the following steps:

[0219] Get ground point cloud;

[0220] Performing rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid;

[0221] Determining a ground expression corresponding to a grid according to each sub-ground point cloud;

[0222] Determine the target point cloud center corresponding to the target point cloud;

[0223] Determine a target ground expression of a grid corresponding to the center of the target point cloud;

[0224] Based on the target ground expression, ground mapping processing is performed on the target point cloud to obtain a mapped first point cloud.

[0225] Optionally, the processor 800 is further configured to read a program in the memory 820 and execute the following steps:

[0226] Determine the sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotating caliper as the longitudinal distance, and determine the sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotating caliper as the transverse distance;

[0227] The sum of the longitudinal distance and the transverse distance is determined as the edge distance corresponding to the first rotating chuck.

[0228] Optionally, the processor 800 is further configured to read a program in the memory 820 and execute the following steps:

[0229] The rotating caliper with the smallest corresponding edge distance among the G rotating calipers is determined as the target rotating caliper.

[0230] Optionally, the processor 800 is further configured to read a program in the memory 820 and execute the following steps:

[0231] Based on the preset longitudinal edge area width and transverse edge area width, and the initial heading angle of the target vehicle, the longitudinal edge area and the transverse edge area are respectively determined in the target rectangular frame.

[0232] Optionally, the target rectangular box is a minimum rectangular box surrounding the first point cloud.

[0233] The vehicle heading angle calculation device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effects are similar, which will not be repeated here in this embodiment.

[0234] The embodiment of the present invention also provides a vehicle posture calculation device. Since the principle of the vehicle posture calculation device to solve the problem is similar to the vehicle posture calculation method in the embodiment of the present invention, the implementation of the vehicle posture calculation device can refer to the implementation of the method, and the repeated parts will not be repeated. Figure 9 As shown, the vehicle posture calculation device according to an embodiment of the present invention includes:

[0235] The processor 900 is configured to read the program in the memory 920 and execute the following process:

[0236] Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0237] Based on the heading angle of the target vehicle and the target point cloud, a three-dimensional detection frame for representing the position of the target vehicle is determined, wherein the heading angle of the target vehicle is Figure 1 The vehicle heading angle calculation method in the illustrated embodiment determines the obtained heading angle.

[0238] Among them, Figure 9In the present disclosure, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits such as one or more processors represented by processor 900 and memory represented by memory 920. The bus architecture may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, all of which are well known in the art and are not further described herein. The bus interface provides an interface. Processor 900 is responsible for managing the bus architecture and general processing, while memory 920 may store data used by processor 900 when performing operations.

[0239] The vehicle posture calculation device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effects are similar, which will not be repeated here in this embodiment.

[0240] In addition, a computer-readable storage medium according to an embodiment of the present invention is used to store a computer program, and the computer program can be executed by a processor to implement the following steps:

[0241] Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0242] Determine a first point cloud mapped from the target point cloud to the ground, and determine a target rectangular box surrounding the first point cloud;

[0243] Determining the longitudinal edge area and the transverse edge area in the target rectangular frame respectively, wherein the longitudinal edge area is the area corresponding to the long side of the target vehicle;

[0244] Determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located in the longitudinal edge region, and the transverse edge point cloud is a point cloud in the first point cloud located in the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1;

[0245] Determine the edge distance corresponding to each of the G rotating calipers respectively, and determine the target rotating caliper based on the edge distances, wherein the edge distance corresponding to the first rotating caliper includes a longitudinal distance and a transverse distance, the longitudinal distance includes the distance between the longitudinal edge point cloud and the longitudinal line segment of the first rotating caliper, and the transverse distance includes the distance between the transverse edge point cloud and the transverse line segment of the first rotating caliper, and the first rotating caliper is any rotating caliper among the G rotating calipers;

[0246] Based on the target rotation jam, a heading angle of the target vehicle is determined.

[0247] Optionally, after respectively determining the longitudinal edge area and the transverse edge area in the target rectangular frame, the method further includes:

[0248] removing noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud;

[0249] The determining G rotation caulking parts based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region comprises:

[0250] G rotation calipers are determined based on the longitudinal edge point cloud after the noise is removed and the lateral edge point cloud in the lateral edge region.

[0251] Optionally, determining G rotation caulking points based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region includes:

[0252] Selecting a first target point closest to the lateral outer line of the lateral edge area from the lateral edge point cloud;

[0253] Selecting a second target point closest to the longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud;

[0254] Selecting G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point;

[0255] Based on the first target point, the second target point and the G third target points, G rotational calipers are determined, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

[0256] Optionally, before determining the first point cloud mapped from the target point cloud to the ground, the method further includes:

[0257] Get ground point cloud;

[0258] Performing rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid;

[0259] Determining a ground expression corresponding to a grid according to each sub-ground point cloud;

[0260] Determining that the target point cloud is mapped to a first point cloud on the ground includes:

[0261] Determine the target point cloud center corresponding to the target point cloud;

[0262] Determine a target ground expression of a grid corresponding to the center of the target point cloud;

[0263] Based on the target ground expression, ground mapping processing is performed on the target point cloud to obtain a mapped first point cloud.

[0264] Optionally, respectively determining the edge distance corresponding to each of the G rotating caulkings includes:

[0265] Determine the sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotating caliper as the longitudinal distance, and determine the sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotating caliper as the transverse distance;

[0266] The sum of the longitudinal distance and the transverse distance is determined as the edge distance corresponding to the first rotating chuck.

[0267] Optionally, determining the target rotation jam based on the edge distance includes:

[0268] The rotating caliper with the smallest corresponding edge distance among the G rotating calipers is determined as the target rotating caliper.

[0269] Optionally, the separately determining the longitudinal edge area and the transverse edge area in the target rectangular frame includes:

[0270] Based on the preset longitudinal edge area width and transverse edge area width, and the initial heading angle of the target vehicle, the longitudinal edge area and the transverse edge area are respectively determined in the target rectangular frame.

[0271] Optionally, the target rectangular box is a minimum rectangular box surrounding the first point cloud.

[0272] The computer program can also be executed by the processor to implement the following steps:

[0273] Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle;

[0274] Based on the heading angle of the target vehicle and the target point cloud, a three-dimensional detection frame for representing the position and posture of the target vehicle is determined.

[0275] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection of some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0276] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may be physically included separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional units.

[0277] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to perform some steps of the sending and receiving methods described in various embodiments of the present invention. The aforementioned storage medium includes: a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc., various media that can store program code.

[0278] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for calculating a vehicle heading angle, characterized in that: include: Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; Determine a first point cloud mapped from the target point cloud to the ground, and determine a target rectangular box surrounding the first point cloud; Determining the longitudinal edge area and the transverse edge area in the target rectangular frame respectively, wherein the longitudinal edge area is the area corresponding to the long side of the target vehicle; Determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located in the longitudinal edge region, and the transverse edge point cloud is a point cloud in the first point cloud located in the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1; Determine the edge distance corresponding to each of the G rotating calipers respectively, and determine the target rotating caliper based on the edge distances, wherein the edge distance corresponding to the first rotating caliper includes a longitudinal distance and a transverse distance, the longitudinal distance includes the distance between the longitudinal edge point cloud and the longitudinal line segment of the first rotating caliper, and the transverse distance includes the distance between the transverse edge point cloud and the transverse line segment of the first rotating caliper, and the first rotating caliper is any rotating caliper among the G rotating calipers; Determining a heading angle of the target vehicle based on the target rotation jam; The determining G rotation caulking parts based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region comprises: Selecting a first target point closest to the lateral outer line of the lateral edge area from the lateral edge point cloud; Selecting a second target point closest to the longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud; Selecting G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point; Based on the first target point, the second target point and the G third target points, G rotational calipers are determined, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

2. The method according to claim 1, characterized in that After respectively determining the longitudinal edge area and the transverse edge area in the target rectangular frame, the method further includes: removing noise points in the longitudinal edge point cloud, wherein the noise points are points in the longitudinal edge point cloud whose distance from a fitting line is greater than a preset distance threshold, and the fitting line is a line obtained by fitting the longitudinal edge point cloud; The determining G rotation caulking parts based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region comprises: G rotation calipers are determined based on the longitudinal edge point cloud after removing the noise points and the transverse edge point cloud in the transverse edge region.

3. The method according to claim 1, characterized in that Before determining that the target point cloud is mapped to a first point cloud on the ground, the method further includes: Get ground point cloud; Performing rasterization processing on the ground point cloud to divide the ground point cloud into a plurality of sub-ground point clouds, each sub-ground point cloud corresponding to a grid; Determining a ground expression corresponding to a grid according to each sub-ground point cloud; Determining that the target point cloud is mapped to a first point cloud on the ground includes: Determine the target point cloud center corresponding to the target point cloud; Determine a target ground expression of a grid corresponding to the center of the target point cloud; Based on the target ground expression, ground mapping processing is performed on the target point cloud to obtain a mapped first point cloud.

4. The method according to claim 1, wherein The step of respectively determining the edge distance corresponding to each of the G rotating caulking elements comprises: Determine the sum of distances from each point in the longitudinal edge point cloud to the longitudinal line segment of the first rotating caliper as the longitudinal distance, and determine the sum of distances from each point in the transverse edge point cloud to the transverse line segment of the first rotating caliper as the transverse distance; The sum of the longitudinal distance and the transverse distance is determined as the edge distance corresponding to the first rotating chuck.

5. The method according to claim 1, characterized in that The step of determining the target rotation jam based on the edge distance includes: The rotating caliper with the smallest corresponding edge distance among the G rotating calipers is determined as the target rotating caliper.

6. The method according to claim 1, characterized in that The determining of the longitudinal edge area and the transverse edge area in the target rectangular frame respectively includes: Based on the preset longitudinal edge area width and transverse edge area width, and the initial heading angle of the target vehicle, the longitudinal edge area and the transverse edge area are respectively determined in the target rectangular frame.

7. The method according to claim 1, characterized in that The target rectangular box is the minimum rectangular box surrounding the first point cloud.

8. A vehicle posture calculation method, characterized in that: include: Acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; Based on the heading angle of the target vehicle and the target point cloud, a three-dimensional detection frame for representing the position and posture of the target vehicle is determined, wherein the heading angle of the target vehicle is a heading angle determined using the method described in any one of claims 1 to 7.

9. A vehicle heading angle calculation device, characterized in that: include: A first acquisition module is configured to acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; A first determining module is configured to determine a first point cloud mapped from the target point cloud to the ground, and to determine a target rectangular frame surrounding the first point cloud; A second determining module is configured to respectively determine a longitudinal edge region and a transverse edge region in the target rectangular frame, wherein the longitudinal edge region is a region corresponding to a long side of the target vehicle; a third determination module, configured to determine G rotation calipers based on the longitudinal edge point cloud in the longitudinal edge region and the transverse edge point cloud in the transverse edge region, wherein the longitudinal edge point cloud is a point cloud in the first point cloud located within the longitudinal edge region, the transverse edge point cloud is a point cloud in the first point cloud located within the transverse edge region, the rotation calipers are composed of mutually perpendicular longitudinal line segments and transverse line segments, the longitudinal line segment passes through at least two points in the longitudinal edge point cloud, and the transverse line segment passes through at least one point in the transverse edge point cloud, and G is an integer greater than 1; a fourth determination module, configured to respectively determine an edge distance corresponding to each of the G rotational caulkings, and determine a target rotational caulking based on the edge distances, wherein the edge distance corresponding to a first rotational caulking includes a longitudinal distance and a transverse distance, the longitudinal distance including a distance between the longitudinal edge point cloud and a longitudinal line segment of the first rotational caulking, and the transverse distance including a distance between the transverse edge point cloud and a transverse line segment of the first rotational caulking, and the first rotational caulking is any one of the G rotational caulkings; a fifth determining module, configured to determine a heading angle of the target vehicle based on the target rotation jam; The third determining module includes: a first selection unit, configured to select a first target point closest to a lateral outer line of the lateral edge region from the lateral edge point cloud; a second selection unit, configured to select a second target point closest to a longitudinal outer line of the longitudinal edge region from the longitudinal edge point cloud; a third selection unit, configured to select G third target points from the longitudinal edge point cloud, wherein the third target point is any point other than the second target point; The first determining unit is configured to determine G rotational calipers based on the first target point, the second target point, and the G third target points, wherein a longitudinal line segment of each rotational caliper passes through the second target point and one of the G third target points, and a transverse line segment of each rotational caliper passes through the first target point.

10. A vehicle posture calculation device, characterized in that: include: An acquisition module, configured to acquire a target point cloud, wherein the target point cloud includes a point cloud of a target vehicle; A determination module is used to determine a three-dimensional detection frame for representing the position and posture of the target vehicle based on the heading angle of the target vehicle and the target point cloud, wherein the heading angle of the target vehicle is a heading angle determined using the method according to any one of claims 1 to 7.

11. A vehicle heading angle calculation device, comprising: A transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor; wherein the processor is configured to read the program in the memory to implement the steps of the method according to any one of claims 1 to 7.

12. A vehicle posture calculation device, comprising: A transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor; wherein the processor is configured to read the program in the memory to implement the steps of the method according to claim 8.

13. A computer-readable storage medium for storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented; or the steps of the method according to claim 8 are implemented.

Images