High-precision map data generation method, device, electronic device and storage medium

Abstract

The present disclosure provides a map data generation method, device, electronic device and storage medium, which relate to the field of artificial intelligence technology, especially to the field of autonomous driving, intelligent transportation and high-precision maps. The specific implementation scheme is: according to the first point cloud data set and the second point cloud data set, N point cloud data pairs are determined, each of the N point cloud data pairs includes the first point cloud data and the second point cloud data, N≥1; for each point cloud data pair, according to the point coordinates of the first point cloud data, a straight line or plane first geometric figure is fitted; according to the point coordinates of the second point cloud data, a predetermined number of reference points are determined; according to the coordinates of the first geometric figure and the predetermined number of reference points corresponding to each point cloud pair, the relative posture information between the first point cloud data set and the second point cloud data set is determined; and according to the relative posture information, map data is generated based on the first point cloud data set and the second point cloud data set.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to the fields of autonomous driving, intelligent transportation, and high-precision maps. More specifically, the present disclosure provides a map data generation method, device, electronic device, storage medium, computer program product, and autonomous driving vehicle. Background Art

[0002] High-precision maps, also known as high-accuracy maps, are maps used by self-driving cars. High-precision maps have accurate vehicle location information and rich road element data information, which can help cars predict complex road information, such as slope, curvature, and heading, to better avoid potential risks. In the process of determining high-precision maps using sensors such as LiDAR (Light Detection And Ranging, laser radar), point cloud data from different frames need to be spliced. The accuracy of the splicing result of the point cloud data has a direct impact on the accuracy of the high-precision map. Summary of the invention

[0003] The present disclosure provides a map data generation method, device, electronic device, storage medium, computer program product, and autonomous driving vehicle.

[0004] According to one aspect of the present disclosure, a map data generation method is provided, comprising: determining N point cloud data pairs based on a first point cloud data set and a second point cloud data set, each point cloud data pair in the N point cloud data pairs comprising first point cloud data from the first point cloud data set and second point cloud data from the second point cloud data set, each point cloud data pair being used to characterize the same target object, and N being an integer greater than or equal to 1; fitting a first linear or planar geometric figure according to the point coordinates of the first point cloud data for each point cloud data pair; determining a predetermined number of reference points according to the point coordinates of the second point cloud data; determining relative pose information between the first point cloud data set and the second point cloud data set according to the coordinates of the first geometric figure and the predetermined number of reference points corresponding to each point cloud pair; and generating map data based on the first point cloud data set and the second point cloud data set according to the relative pose information.

[0005] According to another aspect of the present disclosure, a map data generating device is provided, comprising: a first determining module, a second determining module, a posture determining module and a generating module. The first determining module is used to determine N point cloud data pairs according to a first point cloud data set and a second point cloud data set, each of the N point cloud data pairs includes first point cloud data from the first point cloud data set and second point cloud data from the second point cloud data set, each point cloud data pair is used to characterize the same target object, and N is an integer greater than or equal to 1. The second determining module is used to fit a linear or planar first geometric figure according to the point coordinates of the first point cloud data for each point cloud data pair; and is also used to determine a predetermined number of reference points according to the point coordinates of the second point cloud data. The posture determining module is used to determine the relative posture information between the first point cloud data set and the second point cloud data set according to the coordinates of the first geometric figure and the predetermined number of reference points corresponding to each point cloud pair. The generating module is used to generate map data based on the first point cloud data set and the second point cloud data set according to the relative posture information.

[0006] According to another aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method provided by the present disclosure.

[0007] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the method provided by the present disclosure.

[0008] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method provided by the present disclosure is implemented.

[0009] According to one aspect of the present disclosure, an autonomous driving vehicle is provided, comprising the above-mentioned electronic device.

[0010] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0012] Figure 1 is a schematic diagram of an application scenario of the map data generating method and device according to an embodiment of the present disclosure;

[0013] Figure 2 is a schematic flow chart of a map data generating method according to an embodiment of the present disclosure;

[0014] Figure 3 is a schematic flow chart of fitting a first geometric figure according to an embodiment of the present disclosure;

[0015] Figure 4 is a schematic flow chart of determining relative posture information according to an embodiment of the present disclosure;

[0016] Figure 5 is a schematic diagram of a map data generating method according to an embodiment of the present disclosure;

[0017] Figure 6 is a schematic structural block diagram of a map data generating device according to an embodiment of the present disclosure; and

[0018] Figure 7 It is a structural block diagram of an electronic device for implementing the map data generating method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0019] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0020] In some technical solutions, when the first frame point cloud data set and the second frame point cloud data set need to be spliced, the first point feature and the second point feature can be determined from the first frame point cloud data set and the second frame point cloud data set respectively, and the first point feature and the second point feature constitute the same point feature, and the same point feature indicates that the first point feature and the second point feature represent the same target object point. Then, based on the first point feature and the second point feature, the first frame point cloud data set and the second frame point cloud data set are spliced ​​to obtain map data.

[0021] It is understandable that, with the above technical solution, the same-name point features will directly affect the accuracy of the map data. However, in actual applications, errors are prone to occur when selecting point features in two frames of point cloud data sets, resulting in the two selected point features not being the same-name point features, which in turn causes inaccurate map data.

[0022] The embodiment of the present disclosure aims to propose a map data generation method, which selects the same-name line features or the same-name surface features representing the same target object to replace the same-name point features in the above technical solution, thereby reducing the accuracy requirement for selecting the same-name features and ensuring the accuracy of the map data.

[0023] The technical solution provided by the present disclosure will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] Figure 1 It is a schematic diagram of an application scenario of the map data generating method and device according to an embodiment of the present disclosure.

[0025] It should be noted that Figure 1 What is shown is merely an example of a system architecture to which the embodiments of the present disclosure can be applied, in order to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0026] like Figure 1 As shown, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0027] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptops, desktop computers, etc.

[0028] The server 105 may be a server that provides various services, such as a background management server (only an example) that supports websites browsed by users using the terminal devices 101, 102, and 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as map data generated according to the first point cloud data set and the second point cloud data set) to the terminal device.

[0029] It should be noted that the map data generation method provided in the embodiment of the present disclosure can generally be executed by the server 105. Accordingly, the map data generation device provided in the embodiment of the present disclosure can generally be set in the server 105. The map data generation method provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105. Accordingly, the map data generation device provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105.

[0030] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0031] Figure 2 is a schematic flow chart of a map data generating method according to an embodiment of the present disclosure.

[0032] like Figure 2 As shown, the map data generating method 200 may include operations S210 to S240.

[0033] In operation S210, N point cloud data pairs are determined based on the first point cloud data set and the second point cloud data set, each point cloud data pair of the N point cloud data pairs includes first point cloud data from the first point cloud data set and second point cloud data from the second point cloud data set, each point cloud data pair is used to represent the same target object, and N is an integer greater than or equal to 1.

[0034] The first point cloud data and the second point cloud data may both be three-dimensional point cloud data. The first point cloud data and the second point cloud data may be acquired using a laser radar or other types of sensors.

[0035] A plurality of first point cloud data may be determined from the first point cloud data set, a plurality of second point cloud data may be determined from the second point cloud data set, and then the first point cloud data and the second point cloud data representing the same target object may be determined as a point cloud data pair.

[0036] Point cloud data pairs are used to represent the same target object, such as the same vertical pole, horizontal pole, tree, edge and other columnar objects, or the same wall, the same ground, the same road sign and other surface objects.

[0037] The first point cloud data and the second point cloud data may be of the same type, for example, both are linear or planar.

[0038] In some embodiments, in order to constrain the point cloud data pairs in multiple degrees of freedom, multiple point cloud data pairs can be determined. For example, in order to ensure that the relative pose information of 6 degrees of freedom can be determined, at least three point cloud data pairs can be determined. For each point cloud data set in the first point cloud data set and the second point cloud data set, three orthogonal planar point cloud data, or three orthogonal linear point cloud data, or one point cloud data representing the ground, one point cloud data representing the wall and one linear point cloud data, or one point cloud data representing the ground and two mutually perpendicular linear point cloud data can be determined from the point cloud data set.

[0039] In operation S220, for each point cloud data pair, a first geometric figure of a straight line or a plane is fitted according to the point coordinates of the first point cloud data; and a predetermined number of reference points are determined according to the point coordinates of the second point cloud data.

[0040] For example, the first geometric figure is located in a first coordinate system where the first point cloud data is located, and the first geometric figure may correspond to a first parametric equation.

[0041] For example, when the first geometric figure is a straight line, the predetermined number may be an integer greater than or equal to 2, and the line connecting the predetermined number of reference points may be a straight line. When the first geometric figure is a plane, the predetermined number may be an integer greater than or equal to 3, and the predetermined number of reference points may constitute a plane.

[0042] For example, a predetermined number of points may be selected from the second point cloud data as reference points.

[0043] In operation S230 , relative pose information between the first point cloud dataset and the second point cloud dataset is determined based on the first geometric figure corresponding to each point cloud pair and the coordinates of a predetermined number of reference points.

[0044] In operation S240 , map data is generated based on the first point cloud data set and the second point cloud data set according to the relative pose information.

[0045] For example, the first point cloud data set and the second point cloud data set can be deduplicated, spliced, and other operations can be performed to obtain map data. After the point cloud data set is obtained based on the splicing and other operations, deep learning and other methods can be used to perform element recognition and classification on the point cloud data set, and road elements such as traffic lights and signs can be identified to form map data. In the process of building a map, manual or automatic annotation can also be performed to correct the data in order to improve the accuracy of the map.

[0046] According to the technical solution provided by the embodiment of the present disclosure, since a straight line or plane point cloud data pair is used to replace the same-name point features, for a straight line point cloud data pair, the two point cloud data in the point cloud data pair may have a position deviation in the straight line direction. For a plane point cloud data pair, the two point cloud data in the point cloud data pair may have a position deviation in the extension direction of the plane. Therefore, the accuracy requirements for selecting point cloud data pairs can be reduced, thereby ensuring the accuracy of the map.

[0047] The technical solution provided by the embodiments of the present disclosure is applicable to scenarios where the intersection rate between the first point cloud data set and the second point cloud data set is low, such as on- and off-bridge scenarios, sunken tunnel entrance scenarios, opposite lanes with hard isolation, dynamic object scenarios with relatively large traffic volume, etc.

[0048] According to another embodiment of the present disclosure, the above operation of determining N point cloud data pairs based on the first point cloud data set and the second point cloud data set may include the following operations: displaying the first point cloud data set and the second point cloud data set. Then, in response to receiving a selection instruction, determining N first point cloud data and N second point cloud data from the first point cloud data set and the second point cloud data set respectively according to the selection instruction.

[0049] For example, the client displays the first point cloud data set and the second point cloud data set, and the user selects a local point cloud in the first point cloud data set and the second point cloud data set using a polygon, a bounding box, etc. The selection operation triggers a selection instruction.

[0050] Then, according to the selection instruction, the selected local point cloud can be segmented from the first point cloud data set and the second point cloud data set to obtain the first point cloud data and the second point cloud data.

[0051] It should be understood that in the first point cloud data set and the second point cloud data set, there are more linear or planar point cloud data, which interferes with the selection of point cloud data pairs from the first point cloud data set and the second point cloud data set, so it is more difficult to select point cloud data pairs. Especially in the low intersection rate scenario, the low intersection rate means that there are fewer features with the same name representing the same target object in the first point cloud data set and the second point cloud data set, which also increases the difficulty of selecting point cloud data pairs. In addition, some linear or planar point cloud data represent dynamic objects, which interferes with the selection of point cloud data pairs.

[0052] The disclosed embodiment selects point cloud data from a point cloud data set by selecting instructions, without selecting point cloud data by algorithm, thereby simplifying the processing flow of selecting point cloud data. In addition, the first point cloud data and the second point cloud data are determined by manual assistance, which can ensure the selection of appropriate first point cloud data and second point cloud data, thereby ensuring the accuracy of map data.

[0053] In another embodiment of the present disclosure, the above operation of determining N point cloud data pairs based on the first point cloud data set and the second point cloud data set may further include the following operation: in response to receiving a matching instruction, determining a matching relationship between the N first point cloud data and the N second point cloud data according to the matching instruction. Then, according to the matching relationship, the N first point cloud data and the N second point cloud data are divided into N point cloud data pairs.

[0054] For example, a user specifies a matching relationship between N first point cloud data and N second point cloud data through a client, and the specifying operation triggers a matching instruction. Then, according to the matching instruction, the matching first point cloud data and second point cloud data can be determined as a point cloud data pair.

[0055] The disclosed embodiment determines the point cloud data pair through matching instructions, without the need to match the first point cloud data and the second point cloud data through an algorithm, thereby simplifying the processing flow of determining the point cloud data pair.

[0056] Figure 3 is a schematic flowchart of fitting a first geometric figure according to an embodiment of the present disclosure.

[0057] like Figure 3 As shown, according to another embodiment of the present disclosure, the method 320 for fitting a straight line or plane first geometric figure according to the point coordinates of the first point cloud data may include operations S321 to S325.

[0058] In operation S321 , a covariance matrix of the first point cloud data is determined according to point coordinates of the first point cloud data.

[0059] In operation S322, the type of the first point cloud data is determined according to the eigenvalues ​​of the covariance matrix.

[0060] For example, a PCA (Principal Components Analysis) method may be used to calculate the covariance matrix of the first point cloud data and perform eigenvalue decomposition to obtain the eigenvalues ​​of the first point cloud data.

[0061] The eigenvalue can represent the expansion of the first point cloud data in three degrees of freedom. For example, for the first point cloud data of a straight line type, the expansion of the first degree of freedom along the straight line direction is larger, and the expansion of the second degree of freedom and the third degree of freedom along the plane perpendicular to the straight line direction is smaller. For the first point cloud data of a planar type, the expansion of the first degree of freedom and the second degree of freedom in the two perpendicular directions along the plane itself is larger, and the expansion of the third degree of freedom along the plane normal is smaller. It can be seen that the eigenvalue can be used to distinguish the type of the first point cloud data. The type of the first point cloud data represents the shape of the first geometric figure fitted by the first point cloud data, for example, whether the first geometric figure is planar or linear.

[0062] For example, when the second largest eigenvalue is greater than or equal to a predetermined value, it can be determined that the first geometric figure fitted by the first point cloud data is a plane type. When the second largest eigenvalue is less than a predetermined value, it can be determined that the first geometric figure fitted by the first point cloud data is a straight line type. The predetermined value can be 0.1.

[0063] For example, when the ratio of the maximum eigenvalue to the second maximum eigenvalue is less than or equal to a predetermined ratio, it can be determined that the first geometric figure fitted by the first point cloud data is a plane type. When the ratio of the maximum eigenvalue to the second maximum eigenvalue is greater than a predetermined value, it can be determined that the first geometric figure fitted by the first point cloud data is a straight line type. The predetermined ratio can be 5.

[0064] In operation S323, a first geometric figure is determined according to the type and point coordinates of the first point cloud data.

[0065] It is understandable that the parametric equation can be determined according to the type of the first point cloud data and the point coordinates, and the parametric equation can be used to characterize the first geometric figure. The embodiment of the present disclosure does not limit the method of determining the parametric equation of the first geometric figure.

[0066] For example, when the type of the first point cloud data is a plane type, the parametric equation of the plane may be solved using the least square method.

[0067] For example, when the type of the first point cloud data is a plane type, the following operations may be repeated until a predetermined number of executions is met: three non-collinear points in the first point cloud data are determined, and a candidate plane is determined based on the coordinates of the three points. Then, when it is determined that the distance between the point in the first point cloud data and the candidate plane is less than or equal to a distance threshold, the point is determined to be an inner point of the candidate plane, and the number of inner points corresponding to the candidate plane is obtained. Then, the candidate plane corresponding to the maximum number of inner points is determined as the first geometric figure fitted by the first point cloud data, and the parametric equation of the first geometric figure may be solved.

[0068] For example, when the type of the first point cloud data is a straight line type, the center point of the first point cloud data may be calculated, and then the eigenvector corresponding to the maximum eigenvalue of the covariance matrix may be used as a direction vector to determine the parametric equation of the first geometric figure.

[0069] It can be seen that in this embodiment, the eigenvalue of the covariance matrix can accurately reflect the type of the first point cloud data, thereby ensuring the accuracy of the first geometric figure.

[0070] In operation S324, external points of the first geometric figure are determined from the first point cloud data according to the point coordinates of the first geometric figure and the first point cloud data.

[0071] It is understandable that after determining the first point cloud data from the first point cloud data set, the first point cloud data may include external points. For example, points in the first point cloud data that are spaced from the first geometric figure by a predetermined distance or more may be determined as external points. The predetermined distance may be 0 meters, 0.2 meters, etc.

[0072] In operation S325, it is determined whether the number of outliers is greater than a number threshold. If not, operation S326 is performed. If yes, operation S327 is performed.

[0073] For example, the quantity threshold may be 0, 1, 2, etc.

[0074] In operation S326, the first geometric figure is output, for example, a parametric equation representing the first geometric figure is output.

[0075] In operation S327 , outliers are filtered out from the first point cloud data.

[0076] For example, after filtering out the outliers, the operation of fitting a straight line or plane first geometric figure according to the point coordinates of the first point cloud data may be returned to, for example, returning to the above operation S321.

[0077] It can be seen that after filtering out the outliers on the first point cloud data, the points included in the first point cloud data will change, that is, the filtering outliers can update the first point cloud data. Therefore, after returning to the operation of fitting the first geometric figure according to the point coordinates of the first point cloud data, the first geometric figure can be refitted using the updated first point cloud data, thereby ensuring the accuracy of the first geometric figure.

[0078] In other embodiments, the above operations S324 to S327 may be omitted.

[0079] In other embodiments, other methods may be used to determine the type of the first point cloud data. For example, when the first point cloud data is segmented from the first point cloud data set using a bounding box, the type of the first point cloud data may be determined using the ratio of the maximum size to the second maximum size of the length, width, height, and height of the bounding box. For example, if the ratio of the maximum size to the second maximum size is greater than or equal to 0.2, the first point cloud data is determined to be a linear type, otherwise it is a planar type. For another example, the type of the first point cloud data may be determined according to an instruction for specifying the type of the first point cloud data.

[0080] According to another embodiment of the present disclosure, the above-mentioned operation of determining a predetermined number of reference points based on the point coordinates of the second point cloud data may include the following operations: fitting a second geometric figure based on the point coordinates of the second point cloud data, and then determining a predetermined number of reference points from the second geometric figure.

[0081] For example, the second geometric figure can be fitted by referring to the operation of fitting a linear or planar first geometric figure according to the point coordinates of the first point cloud data, which will not be described in detail in this embodiment. It can be understood that the second geometric figure can be located in the second coordinate system where the second point cloud data is located, and the second geometric figure can correspond to a second parametric equation.

[0082] It is understandable that, since the first point cloud data and the second point cloud data are features representing the same target object, that is, they are features with the same name, the types of geometric figures fitted by the first point cloud data and the second point cloud data can be the same. When the second geometric figure is a plane, the selected predetermined number of reference points can form a plane.

[0083] In one example, a predetermined number of reference points may be randomly selected from the second geometric figure.

[0084] In another example, the center point of the second point cloud data may be determined first, and then the reference point may be determined according to a parametric equation corresponding to the center point and the second geometric figure.

[0085] For example, when the second geometric figure is a straight line, a point located on the straight line and spaced a first predetermined distance from the center point may be determined as the reference point. The first predetermined distance may be 0.5 meters.

[0086] For another example, when the second geometric figure is a plane, a point obtained by moving the center point in a predetermined direction within the plane by a predetermined distance can be determined as the reference point. The second predetermined distance can be 0.5 meters, and the predetermined direction can be upward, downward, rightward, etc.

[0087] It should be noted that after the second geometric figure is determined using the second point cloud data set, some external points outside the second geometric figure may exist in the second point cloud data. Since the relative pose information needs to be determined based on the reference point and the first geometric figure, if some points are randomly selected from the second point cloud data as reference points, and the selected points are external points, the accuracy of the relative pose information will be affected.

[0088] It can be understood that the embodiment of the present disclosure selects the reference point from the second geometric figure, so the second point cloud data may not include the reference point.

[0089] The disclosed embodiment determines reference points from a second geometric figure fitted using second point cloud data. It can be seen that the reference points are all interior points of the second geometric figure. Therefore, in the process of determining relative pose information based on the reference points and the first geometric figure, the accuracy of the relative pose information can be ensured, thereby ensuring the accuracy of the map data.

[0090] Figure 4 It is a schematic flowchart of determining relative posture information according to an embodiment of the present disclosure.

[0091] like Figure 4 As shown, according to another embodiment of the present disclosure, the above method 430 for determining the relative pose information between the first point cloud data set and the second point cloud data set based on the coordinates of the first geometric figure corresponding to each point cloud pair and a predetermined number of reference points may include operations S431 to S434.

[0092] In operation S431, for each point cloud data pair, coordinates of a predetermined number of reference points are transformed based on the candidate transformation matrix to obtain transformed coordinates as pseudo coordinates of the predetermined number of reference points in the first coordinate system where the first geometric figure is located.

[0093] It can be understood that the reference point is determined based on the second point cloud data in the second coordinate system, so the coordinates of the reference point are the coordinates of the reference point in the second coordinate system. By using the candidate transformation matrix to transform the coordinates of the reference point, the reference point can be transformed into other coordinate systems, and the transformed coordinates of the reference point in other coordinate systems can be obtained.

[0094] In operation S432 , a first distance corresponding to each point cloud data pair is determined according to the sum of distances from the pseudo coordinates of a predetermined number of reference points to the first geometric figure.

[0095] For example, in a first coordinate system where the first geometric figure is located, the projection distance from the pseudo coordinate of each reference point to the first geometric figure is calculated to obtain a predetermined number of projection distances. The sum of the predetermined number of projection distances may be determined as the first distance.

[0096] It can be understood that the first distance of 0 may indicate that: in the first coordinate system, the distance from the pseudo-coordinates of the predetermined number of reference points to the first geometric figure is 0, that is, the geometric figure formed by the predetermined number of reference points coincides with the first geometric figure.

[0097] In operation S433 , second distances corresponding to the N point cloud data pairs are determined according to the first distance corresponding to each point cloud data pair.

[0098] For example, the sum of N first distances corresponding to N data pairs may be determined as the second distance.

[0099] In operation S434, relative pose information is determined according to a candidate transformation matrix corresponding to the minimum second distance among the plurality of candidate transformation matrices.

[0100] For example, using multiple candidate transformation matrices, repeating the above operations S431 to S433 for each candidate transformation matrix, thereby performing multiple coordinate transformations on the coordinates of the reference point to obtain multiple second distances corresponding to the multiple candidate transformation matrices. Then, the candidate transformation matrix corresponding to the minimum second distance can be determined as the relative pose information.

[0101] The disclosed embodiment utilizes a candidate transformation matrix to determine the second distances corresponding to N point cloud pairs, and then determines the relative pose information between the first point cloud data set and the second point cloud data set according to the candidate transformation matrix corresponding to the minimum second distance. Thus, the relative pose information can be accurately determined to ensure the accuracy of the map data.

[0102] In addition, since the coordinates of the reference point are transformed using the candidate transformation matrix without the need to transform the first geometric figure, the calculation process is simpler, thereby improving the efficiency of determining relative pose information.

[0103] In some other embodiments, the above operations S431 and S432 may be replaced by the following scheme: for each point cloud data pair, coordinate transformation is performed on the first geometric figure based on the candidate transformation matrix, and the transformed first geometric figure is used as a pseudo-figure of the first geometric figure in the second coordinate system. Then, the first distance is determined according to the sum of the distances from the coordinates of a predetermined number of reference points to the pseudo-figure. Then, the above operations S433 and S434 may be performed to determine the relative pose information.

[0104] In some other embodiments, the following formula may be used to determine the relative posture information.

[0105]

[0106] Among them, T * represents the relative pose information between the first point cloud data set and the second point cloud data set. T is the optimization variable, which represents the candidate transformation matrix. i Indicates the coordinates of the reference point. j Indicates the parameters in the parametric equation corresponding to the first point cloud data of the straight line. j Represents the parameters in the parametric equation corresponding to the first point cloud data of the plane. p2line is a first function for calculating the distance from a point to a first geometric figure of a straight line according to the pseudo coordinates of the reference point and the parameters corresponding to the first point cloud data. p2plane is a second function for calculating the distance from the point to the first geometric figure of the plane according to the pseudo coordinates of the reference point and the parameters corresponding to the first point cloud data. i is an integer greater than or equal to 1, and j is an integer greater than or equal to 1.

[0107] Figure 5 is a schematic diagram of a map data generating method according to an embodiment of the present disclosure.

[0108] like Figure 5 As shown, N first point cloud data 504 and N second point cloud data 505 can be selected from the first point cloud data set 501 and the second point cloud data set 502 respectively, and the N first point cloud data 504 and the N second point cloud data 505 can be divided into N point cloud data pairs 503, and the N point cloud data pairs 503 include the first point cloud data pair 5031, the second point cloud data pair 5032,..., the Nth point cloud data pair 5033.

[0109] Then, based on the first candidate transformation matrix 509 , predetermined processing may be performed on each of the N point cloud data pairs 503 to obtain a first second distance 5121 corresponding to the first candidate transformation matrix 509 .

[0110] This embodiment is described by taking the processing of the first point cloud data pair 5031 as an example. For example, the first point cloud data pair 5031 includes the first point cloud data 504 and the second point cloud data 505, and the first point cloud data 504 is fitted into the first geometric figure 506. The second point cloud data 505 is fitted into the second geometric figure 507, and a predetermined number of reference points 508 are determined from the second geometric figure 507. Then, the coordinates of the reference points 508 are transformed according to the first candidate transformation matrix 509 to obtain pseudo coordinates 510 of the reference points. According to the pseudo coordinates 510 of the reference points and the first geometric figure 506, the first first distance 5111 corresponding to the first point cloud data pair 5031 is determined.

[0111] Similarly, the second point cloud data pair 5032 may be processed based on the first candidate transformation matrix 509 to obtain a second first distance 5112 corresponding to the second point cloud data pair 5032. The Nth point cloud data pair 5033 may be processed to obtain an Nth first distance 5113 corresponding to the Nth point cloud data pair 5033.

[0112] It can be seen that by performing the above predetermined processing on the N point cloud data pairs 503 , N first distances 511 can be obtained, and the first second distance 5121 corresponding to the first candidate transformation matrix 509 can be determined using the N first distances 511 .

[0113] Then, the candidate transformation matrix can be adjusted to re-determine the second distance. For example, the predetermined processing can be performed on each of the N point cloud data pairs 503 based on the second candidate transformation matrix to obtain the second second distance 5122 corresponding to the second candidate transformation matrix. Similarly, the predetermined processing can be performed on each of the N point cloud data pairs 503 based on the third candidate transformation matrix to obtain the third second distance 5123 corresponding to the third candidate transformation matrix.

[0114] It can be seen that by adjusting the candidate transformation matrix multiple times, multiple second distances 512 can be obtained, and the relative posture information 513 can be determined according to the candidate transformation matrix corresponding to the minimum second distance.

[0115] Map data 514 may be generated according to the relative position information 513 , the first point cloud data set 501 , and the second point cloud data set 502 .

[0116] Figure 6 is a schematic structural block diagram of a map data generating device according to an embodiment of the present disclosure.

[0117] like Figure 6As shown, the map data generating device 600 may include a first determining module 610 , a second determining module 620 , a posture determining module 630 and a generating module 640 .

[0118] The first determination module 610 is used to determine N point cloud data pairs based on the first point cloud data set and the second point cloud data set, each point cloud data pair in the N point cloud data pairs includes first point cloud data from the first point cloud data set and second point cloud data from the second point cloud data set, each point cloud data pair is used to represent the same target object, and N is an integer greater than or equal to 1.

[0119] The second determination module 620 is used to fit a straight line or plane first geometric figure according to the point coordinates of the first point cloud data for each point cloud data pair; and is also used to determine a predetermined number of reference points according to the point coordinates of the second point cloud data.

[0120] The pose determination module 630 is used to determine the relative pose information between the first point cloud data set and the second point cloud data set based on the coordinates of the first geometric figure corresponding to each point cloud pair and a predetermined number of reference points.

[0121] The generation module 640 is used to generate map data based on the first point cloud data set and the second point cloud data set according to the relative posture information.

[0122] According to another embodiment of the present disclosure, the second determination module includes: a fitting submodule and a reference point determination submodule. The fitting submodule is used to fit the second geometric figure according to the point coordinates of the second point cloud data. The reference point determination submodule is used to determine a predetermined number of reference points from the second geometric figure.

[0123] According to another embodiment of the present disclosure, the second determination module includes: a matrix determination submodule, a type determination submodule and a graphic determination submodule. The matrix determination submodule is used to determine the covariance matrix of the first point cloud data according to the point coordinates of the first point cloud data. The type determination submodule is used to determine the type of the first point cloud data according to the eigenvalues ​​of the covariance matrix. The graphic determination submodule is used to determine the first geometric figure according to the type and point coordinates of the first point cloud data.

[0124] According to another embodiment of the present disclosure, the map data generating device further includes: an external point determination module and a filtering module. The external point determination module is used to determine the external points of the first geometric figure from the first point cloud data according to the first geometric figure and the point coordinates of the first point cloud data after determining the first geometric figure according to the type and point coordinates of the first point cloud data. The filtering module is used to filter out the external points from the first point cloud data when it is determined that the number of external points is greater than a number threshold, and return the operation of fitting the first geometric figure of a straight line or a plane type according to the point coordinates of the first point cloud data.

[0125] According to another embodiment of the present disclosure, the first determination module includes: a display submodule, a selection submodule, a relationship determination submodule and a division submodule. The display submodule is used to display the first point cloud data set and the second point cloud data set. The selection submodule is used to respond to receiving a selection instruction and, according to the selection instruction, determine N first point cloud data and N second point cloud data from the first point cloud data set and the second point cloud data set, respectively. The relationship determination submodule is used to respond to receiving a matching instruction and, according to the matching instruction, determine the matching relationship between the N first point cloud data and the N second point cloud data. The division submodule is used to divide the N first point cloud data and the N second point cloud data into N point cloud data pairs according to the matching relationship.

[0126] According to another embodiment of the present disclosure, a posture determination module includes: a position determination submodule and a posture determination submodule. The position determination submodule is used to perform coordinate transformation on the coordinates of a predetermined number of reference points based on the candidate transformation matrix for each point cloud data pair, and obtain the transformed coordinates as pseudo-coordinates of the predetermined number of reference points in the coordinate system where the first geometric figure is located. It is also used to determine the first distance corresponding to each point cloud data pair based on the sum of the distances from the pseudo-coordinates of the predetermined number of reference points to the first geometric figure. It is also used to determine the second distance corresponding to N point cloud data pairs based on the first distance corresponding to each point cloud data pair. The posture determination submodule is used to determine the relative posture information based on the candidate transformation matrix corresponding to the minimum second distance among multiple candidate transformation matrices.

[0127] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0128] In the technical solution of the present disclosure, the user's authorization or consent is obtained before obtaining or collecting the user's personal information.

[0129] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, including at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned map data generation method.

[0130] According to an embodiment of the present disclosure, the present disclosure further provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the above-mentioned map data generating method.

[0131] According to an embodiment of the present disclosure, the present disclosure further provides a computer program product, including a computer program, and the computer program implements the above-mentioned map data generating method when executed by a processor.

[0132] According to an embodiment of the present disclosure, the present disclosure also provides an autonomous driving vehicle, comprising the above-mentioned electronic device.

[0133] Figure 7 A schematic block diagram of an example electronic device 700 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0134] like Figure 7 As shown, the device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0135] A number of components in the device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0136] The computing unit 701 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 701 performs the various methods and processes described above, such as the map data generation method. For example, in some embodiments, the map data generation method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the map data generation method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to perform the map data generation method in any other appropriate manner (e.g., by means of firmware).

[0137] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0138] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0139] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0140] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0141] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0142] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0143] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0144] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for generating map data, comprising: Determine N point cloud data pairs according to the first point cloud data set and the second point cloud data set, each point cloud data pair in the N point cloud data pairs includes first point cloud data from the first point cloud data set and second point cloud data from the second point cloud data set, each point cloud data pair is used to represent the same target object, and N is an integer greater than or equal to 1; For each point cloud data pair, Fitting a first geometric figure of a straight line or a plane according to the point coordinates of the first point cloud data; Determining a predetermined number of reference points according to the point coordinates of the second point cloud data; Determining the relative pose information between the first point cloud data set and the second point cloud data set according to the coordinates of the first geometric figure and the predetermined number of reference points corresponding to each point cloud data pair, including: for each point cloud data pair, based on multiple candidate transformation matrices, performing coordinate transformation on the first geometric figure or the predetermined number of reference points, determining the first distance according to the sum of distances from the predetermined number of reference points to the transformed first geometric figure, or determining the first distance according to the sum of distances from the transformed coordinates of the predetermined number of reference points to the first geometric figure; determining the second distance corresponding to the N point cloud data pairs according to the first distance corresponding to each point cloud data pair; determining the relative pose information according to the candidate transformation matrix corresponding to the minimum second distance among the multiple candidate transformation matrices; and According to the relative posture information, map data is generated based on the first point cloud data set and the second point cloud data set.

2. The method according to claim 1, wherein: The determining a predetermined number of reference points according to the point coordinates of the second point cloud data comprises: Fitting a second geometric figure according to the point coordinates of the second point cloud data; and The predetermined number of reference points is determined from the second geometric figure.

3. The method according to claim 1, wherein: The fitting of a straight-line or plane-shaped first geometric figure according to the point coordinates of the first point cloud data comprises: determining a covariance matrix of the first point cloud data according to the point coordinates of the first point cloud data; and Determining the type of the first point cloud data according to the eigenvalues ​​of the covariance matrix; The first geometric figure is determined according to the type and point coordinates of the first point cloud data.

4. The method according to claim 3, further comprising: After determining the first geometric figure according to the type and point coordinates of the first point cloud data, Determining an outer point of the first geometric figure from the first point cloud data according to the point coordinates of the first geometric figure and the first point cloud data; as well as When it is determined that the number of the external points is greater than the number threshold, the external points are filtered out from the first point cloud data, and the operation of fitting a straight line or plane first geometric figure based on the point coordinates of the first point cloud data is returned.

5. The method according to claim 1, wherein: The determining of N point cloud data pairs according to the first point cloud data set and the second point cloud data set comprises: Displaying the first point cloud data set and the second point cloud data set; In response to receiving a selection instruction, determining N first point cloud data and N second point cloud data from the first point cloud data set and the second point cloud data set respectively according to the selection instruction; In response to receiving a matching instruction, determining a matching relationship between the N first point cloud data and the N second point cloud data according to the matching instruction; and According to the matching relationship, the N first point cloud data and the N second point cloud data are divided into the N point cloud data pairs.

6. The method according to any one of claims 1 to 5, wherein: The step of performing coordinate transformation on the first geometric figure or the predetermined number of reference points based on a plurality of candidate transformation matrices, and determining the first distance according to the sum of distances from the predetermined number of reference points to the transformed first geometric figure, or determining the first distance according to the sum of distances from the transformed coordinates of the predetermined number of reference points to the first geometric figure includes: Based on the candidate transformation matrix, coordinate transformation is performed on the coordinates of the predetermined number of reference points to obtain transformed coordinates as pseudo coordinates of the predetermined number of reference points in the coordinate system of the first geometric figure; and The first distance corresponding to each point cloud data pair is determined according to the sum of the distances from the pseudo coordinates of the predetermined number of reference points to the first geometric figure.

7. A map data generating device, comprising: A first determination module is used to determine N point cloud data pairs according to the first point cloud data set and the second point cloud data set, each point cloud data pair in the N point cloud data pairs includes first point cloud data from the first point cloud data set and second point cloud data from the second point cloud data set, each point cloud data pair is used to represent the same target object, and N is an integer greater than or equal to 1; The second determination module is used for determining, for each point cloud data pair, Fitting a first geometric figure of a straight line or a plane according to the point coordinates of the first point cloud data; Determining a predetermined number of reference points according to the point coordinates of the second point cloud data; A posture determination module, for determining the relative posture information between the first point cloud data set and the second point cloud data set according to the coordinates of the first geometric figure and the predetermined number of reference points corresponding to each point cloud pair, comprising: for each point cloud data pair, based on multiple candidate transformation matrices, performing coordinate transformation on the first geometric figure or the predetermined number of reference points, determining the first distance according to the sum of distances from the predetermined number of reference points to the transformed first geometric figure, or determining the first distance according to the sum of distances from the transformed coordinates of the predetermined number of reference points to the first geometric figure; determining the second distance corresponding to the N point cloud data pairs according to the first distance corresponding to each point cloud data pair; determining the relative posture information according to the candidate transformation matrix corresponding to the minimum second distance among the multiple candidate transformation matrices; and A generation module is used to generate map data based on the relative posture information, the first point cloud data set and the second point cloud data set.

8. The device according to claim 7, wherein: The second determining module comprises: a fitting submodule, configured to fit a second geometric figure according to the point coordinates of the second point cloud data; and The reference point determination submodule is used to determine the predetermined number of reference points from the second geometric figure.

9. The device according to claim 7, wherein: The second determining module comprises: a matrix determination submodule, configured to determine a covariance matrix of the first point cloud data according to point coordinates of the first point cloud data; and A type determination submodule, used to determine the type of the first point cloud data according to the eigenvalue of the covariance matrix; The graphic determination submodule is used to determine the first geometric figure according to the type and point coordinates of the first point cloud data.

10. The apparatus according to claim 9, further comprising: an external point determination module, configured to determine external points of the first geometric figure from the first point cloud data according to the first geometric figure and the point coordinates of the first point cloud data after determining the first geometric figure according to the type and point coordinates of the first point cloud data; as well as A filtering module is used to filter out the external points from the first point cloud data when it is determined that the number of the external points is greater than a number threshold, and return the operation of fitting a straight line or plane first geometric figure based on the point coordinates of the first point cloud data.

11. The device according to claim 7, wherein: The first determining module comprises: A display submodule, used for displaying the first point cloud data set and the second point cloud data set; A selection submodule, configured to, in response to receiving a selection instruction, determine N first point cloud data and N second point cloud data from the first point cloud data set and the second point cloud data set respectively according to the selection instruction; a relationship determination submodule, configured to, in response to receiving a matching instruction, determine a matching relationship between the N first point cloud data and the N second point cloud data according to the matching instruction; and A division submodule is used to divide the N first point cloud data and the N second point cloud data into the N point cloud data pairs according to the matching relationship.

12. The device according to any one of claims 7 to 11, wherein: The posture determination module comprises: A position determination submodule is used to perform coordinate transformation on the coordinates of the predetermined number of reference points based on the candidate transformation matrix to obtain the transformed coordinates as the pseudo-coordinates of the predetermined number of reference points in the coordinate system of the first geometric figure; and determine the first distance corresponding to each point cloud data pair based on the sum of the distances from the pseudo-coordinates of the predetermined number of reference points to the first geometric figure.

13. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

14. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 6.

15. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 6.

16. An autonomous driving vehicle comprising the electronic device as claimed in claim 13.

Images