4D lane line labeling data quality inspection method and device, electronic equipment and storage medium

Abstract

The application discloses a 4D lane line marking data quality inspection method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring target point cloud data, and determining target 4D lane line marking data according to the target point cloud data; determining a target quality inspection operation result of the target 4D lane line marking data, wherein the target quality inspection operation result is a result obtained after a target quality inspection operation is performed on the target 4D lane line marking data; determining abnormal 4D lane line marking data in the target 4D lane line marking data based on the target quality inspection operation result, wherein the abnormal 4D lane line marking data is target 4D lane line marking data that exceeds the target quality inspection operation and is represented by the target quality inspection operation result. The scheme realizes the target 4D lane line marking data quality inspection efficiency and accuracy by performing the target quality inspection operation on the target 4D lane line marking data.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving annotation technology, and in particular to a 4D lane line annotation data quality inspection method, device, electronic device and storage medium. Background Technology

[0002] With the rapid development of artificial intelligence technology, autonomous driving is being applied at increasingly higher levels in certain scenarios, and the requirements for the accuracy of perception algorithms in vehicle control are also becoming more stringent. Training perception algorithms requires a large amount of labeled data, and the quality of the labeled data plays a crucial role in the training of perception algorithms.

[0003] Currently, lane marking data collection is affected by paint materials and weather conditions, so a time dimension has been introduced to fuse multiple frames of data captured by vehicle sensors into a single scene to improve the collection effect. However, while introducing the time dimension enhances the collection effect, the increase in marking data under large point clouds further reduces the efficiency and accuracy of manual quality inspection. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for quality inspection of 4D lane line marking data, in order to solve the problems of low efficiency and poor accuracy in manual quality inspection of 4D lane line marking data.

[0005] According to one aspect of the present invention, a 4D lane line annotation data quality inspection method is provided, the method comprising:

[0006] Acquire target point cloud data and determine target 4D lane line annotation data based on target point cloud data. Target point cloud data is the three-dimensional data of the target object in the vehicle driving scene. The target object includes at least several target lane lines. Target 4D lane line annotation data is the 4D annotation data of several target lane lines. 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range.

[0007] The target quality inspection results of the target 4D lane line annotation data are determined. The target quality inspection results are the results obtained after performing target quality inspection operations on the target 4D lane line annotation data. The target quality inspection operations include: jump quality inspection operation, type quality inspection operation, and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane line is the lane line represented by the target lane line according to the target 4D lane line annotation data. The type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between every two labeled lane lines. The preset quality inspection operation is used to check the specified annotation task. The annotation task is the task of performing quality inspection on the target 4D lane line annotation data belonging to the same time range.

[0008] Based on the results of the target quality inspection operation, abnormal 4D lane line annotation data in the target 4D lane line annotation data is determined. Abnormal 4D lane line annotation data refers to the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the results of the target quality inspection operation.

[0009] According to another aspect of the present invention, a 4D lane marking data quality inspection device is provided, the device comprising:

[0010] The first determining module is used to acquire target point cloud data and determine target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene. The target object includes at least a number of target lane lines. The target 4D lane line annotation data is the 4D annotation data of a number of target lane lines. The 4D annotation data is obtained by adding annotations after superimposing the three-dimensional data within a preset time range.

[0011] The second determining module is used to determine the target quality inspection operation result of the target 4D lane line annotation data. The target quality inspection operation result is the result obtained after performing the target quality inspection operation on the target 4D lane line annotation data. The target quality inspection operation includes: jump quality inspection operation, type quality inspection operation and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane line is the lane line represented by the target lane line according to the target 4D lane line annotation data. The type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between each pair of labeled lane lines. The preset quality inspection operation is used to check the specified annotation task. The annotation task is the task of performing quality inspection on the target 4D lane line annotation data belonging to the same time range.

[0012] The third determination module is used to determine abnormal 4D lane line annotation data in the target 4D lane line annotation data based on the target quality inspection operation results. Abnormal 4D lane line annotation data refers to the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation results.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] At least one processor; and

[0015] A memory that is communicatively connected to at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to perform the 4D lane line marking data quality inspection method of any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the 4D lane line marking data quality inspection method of any embodiment of the present invention.

[0018] The technical solution of this invention acquires target point cloud data and determines target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene, and the target object includes at least several target lane lines. The target 4D lane line annotation data is the 4D annotation data of several target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range, thus realizing the acquisition of target 4D lane line annotation data and providing actual data support for the subsequent quality inspection process. The target quality inspection operation result of the target 4D lane line annotation data is determined. The target quality inspection operation result is the result obtained after performing target quality inspection operations on the target 4D lane line annotation data. The target quality inspection operations include: jump quality inspection operation, type quality inspection operation, and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane lines are based on... The target 4D lane line annotation data represents the lane lines of the target lane lines. A type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between every two annotated lane lines. A preset quality inspection operation is used to check specified annotation tasks. The annotation task is to perform quality inspection on target 4D lane line annotation data belonging to the same time range. This realizes the execution of target quality inspection operations on target 4D lane line annotation data, improving the efficiency and accuracy of quality inspection and reducing labor costs. Based on the results of the target quality inspection operation, abnormal 4D lane line annotation data is identified within the target 4D lane line annotation data. Abnormal 4D lane line annotation data is target 4D lane line annotation data that exceeds the target quality inspection operation results. This realizes the identification of incorrectly annotated abnormal 4D lane line annotation data based on the target quality inspection operation results, facilitating secondary quality inspection of abnormal 4D lane line annotation data.

[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of a 4D lane line annotation data quality inspection method provided in Embodiment 1 of the present invention;

[0022] Figure 2 This is a flowchart of a 4D lane line annotation data quality inspection method provided in Embodiment 2 of the present invention;

[0023] Figure 3 This is a flowchart of a 4D lane line annotation data quality inspection method provided in Embodiment 3 of the present invention;

[0024] Figure 4 This is a schematic diagram of lane marking provided in Embodiment 3 of the present invention;

[0025] Figure 5 This is a schematic diagram of the structure of a 4D lane line marking data quality inspection device provided in Embodiment 4 of the present invention;

[0026] Figure 6 This is a schematic diagram of the structure of an electronic device for implementing a 4D lane line marking data quality inspection method, as provided in Embodiment 5 of the present invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0030] Example 1

[0031] Figure 1 This is a flowchart illustrating a 4D lane line annotation data quality inspection method according to Embodiment 1 of the present invention. This embodiment of the invention is applicable to situations involving quality inspection of 4D lane line annotation data. The method can be executed by a 4D lane line annotation data quality inspection device, which can be implemented in hardware and / or software and configured in an electronic device implementing the 4D lane line annotation data quality inspection method. Figure 1 As shown, the method includes:

[0032] S101. Obtain target point cloud data and determine target 4D lane line annotation data based on target point cloud data. Target point cloud data is the three-dimensional data of the target object in the driving scene. The target object includes at least several target lane lines. Target 4D lane line annotation data is the 4D annotation data of several target lane lines. 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range.

[0033] In this embodiment of the invention, the driving scenario can refer to various situations or environments encountered by the vehicle during autonomous driving. The target object can refer to several objects in the driving scenario, such as the road where the vehicle is located, nearby pedestrians, signs, or buildings. The target lane lines can refer to the lane lines of the road where the vehicle is located in the driving scenario, used to divide the road, guide traffic flow, indicate driving direction, and define lane boundaries. For example, when the vehicle is driving on a road with only one lane, the target lane lines can be the lane lines on both sides of the lane; when the vehicle is driving on a road with multiple lanes, the target lane lines can be multiple lane lines that divide the road.

[0034] Target point cloud data refers to the 3D data of target objects in a driving scene. Statistical filtering can be used to remove discrete points from the target point cloud data, improving the accuracy of subsequent processing. Furthermore, when the amount of target point cloud data exceeds a preset threshold, voxel mesh filtering can be combined to reduce the data volume and improve system performance. 4D labeled data refers to data annotated with a time dimension in 3D space. It is obtained by overlaying multiple 3D data frames within a preset time range and then adding annotations.

[0035] Specifically, when a vehicle is driving in a driving scenario, the 3D data of the target object in the driving scenario can be acquired as target point cloud data. The target point cloud data includes at least the 3D data of several target lane lines. Then, by overlaying the 3D data of several target lane lines within a preset time range and adding annotations, 4D annotation data of several target lane lines can be obtained, which serves as the target 4D lane line annotation data.

[0036] For example, target point cloud data of the target lane lines can be collected by sensors configured in the vehicle at 0.1-second intervals, and several 3D data of the target lane lines can be selected from the target point cloud data. Furthermore, when the preset time range is 1 second, 10 frames of 3D data can be acquired within the preset time range. By superimposing the 10 frames of 3D data, annotations can be added using semantic recognition algorithms or manual annotation to obtain 4D annotation data of the target lane lines.

[0037] S102. Determine the target quality inspection operation result of the target 4D lane line annotation data. The target quality inspection operation result is the result obtained after performing the target quality inspection operation on the target 4D lane line annotation data. The target quality inspection operation includes: jump quality inspection operation, type quality inspection operation and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane line is the lane line represented by the target lane line according to the target 4D lane line annotation data. The type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between each pair of labeled lane lines. The preset quality inspection operation is used to check the specified annotation task. The annotation task is the task of performing quality inspection on the target 4D lane line annotation data belonging to the same time range.

[0038] In this embodiment of the invention, the labeled lane line refers to the lane line represented by sequentially connecting the nodes in the target 4D lane line annotation data according to the labeled target lane line. Correspondingly, the number of labeled lane lines is the same as the number of target lane lines. The jump quality inspection operation refers to checking whether lane line jumps have occurred in the labeled lane lines. When the smoothness of the labeled lane line exceeds a preset smoothness threshold, it can be considered that a lane line jump has occurred. The three dimensions in the target 4D lane line annotation data can be the X-axis, Y-axis, and Z-axis, respectively. Correspondingly, lane line jumps can include X-axis lane line jumps, Y-axis lane line jumps, and Z-axis lane line jumps.

[0039] In this embodiment of the invention, the type quality inspection operation can refer to inspecting the type of the target 4D lane line annotation data when two intersecting lane lines exist among several lane line annotations. For example, when two lane lines intersect, the type of the corresponding target 4D lane line annotation data cannot be the same. The preset quality inspection operation can refer to performing a customized inspection operation on a specified annotation task to achieve flexible automatic quality inspection.

[0040] Specifically, the jump quality inspection operation, type quality inspection operation, and preset quality inspection operation performed on the target 4D lane line annotation data can be used as the target quality inspection operation performed on the target 4D lane line annotation data. Correspondingly, the jump quality inspection operation results, type quality inspection operation results, and preset quality inspection operation results obtained after performing the jump quality inspection operation, type quality inspection operation, and preset quality inspection operation on the target 4D lane line annotation data can be used as the target quality inspection operation results.

[0041] As an optional step, the target quality inspection results for determining the target 4D lane line annotation data include the following steps A1-A2:

[0042] Step A1: Determine the intersection status between every two lane lines in a number of lane marking lines.

[0043] Step A2: Determine the type of the target 4D lane line annotation data based on the intersection status and perform quality inspection.

[0044] Specifically, each lane marking line can be discretized to obtain a discrete point sequence for each lane marking line. This allows for pairwise determination of whether the discrete point sequences of each lane marking line intersect. If there is an intersection between the discrete point sequences, the corresponding two lane marking lines are considered to be intersecting. Furthermore, if intersecting lane marking lines are detected, the type inspection result of the target 4D lane marking data corresponding to the intersecting lane marking lines is used as the type quality inspection result.

[0045] As an optional step, the target quality inspection results for determining the target 4D lane line annotation data include the following steps B1-B3:

[0046] Step B1: Determine the target quality inspection object in the annotation task. The target quality inspection object is the object to be inspected when performing quality inspection on the target 4D lane line annotation data.

[0047] Step B2: Determine the reference configuration operation for the target quality inspection object. The reference configuration operation includes: target configuration, action configuration, condition configuration, and logical relationship configuration.

[0048] Step B3: Determine the preset quality inspection operation results of the target 4D lane line annotation data based on the target quality inspection object and reference configuration operation.

[0049] In this embodiment of the invention, the target quality inspection object can refer to the object pre-set for inspecting target 4D lane line annotation data, such as global annotations or 3D point cloud lane lines. The target configuration can refer to the pre-set quality inspection operations performed when inspecting the target quality inspection object, such as inspecting lane line type or lane line transitions. Action configuration can include equal to, not equal to, greater than, less than, empty, or illegal. Condition configuration can refer to the pre-set specific detection values ​​when inspecting the target quality inspection object. Logical relationship configuration can include AND and OR.

[0050] Specifically, verification operations for the target quality inspection object can be generated by configuring the target, action, and conditions in the annotation task. After configuring the logical relationships between multiple verification operations, a reference configuration operation for the target quality inspection object can be generated. Then, the target 4D lane line annotation data is compared and verified with the reference configuration operation of the target quality inspection object to obtain the preset quality inspection operation result for the target 4D lane line annotation data. For example, the target quality inspection object can be specified as a 3D point cloud lane line, the target configuration can be set to check lane line jumps, the action configuration can be set to "greater than", and the condition configuration can be set to a jump detection threshold of 0.9. Thus, the operation verification operation for the 3D point cloud lane line is to check if the lane line jump is greater than the jump detection threshold of 0.9.

[0051] S103. Based on the target quality inspection operation results, determine the abnormal 4D lane line annotation data in the target 4D lane line annotation data. The abnormal 4D lane line annotation data is the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation results.

[0052] Specifically, the results of the target quality inspection operation can identify abnormal 4D lane line annotation data with errors in the target 4D lane line annotation data, so as to conduct a second quality inspection on the abnormal 4D lane line annotation data and thus ensure the accuracy of the lane line annotation data.

[0053] The technical solution of this invention acquires target point cloud data and determines target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene, and the target object includes at least several target lane lines. The target 4D lane line annotation data is the 4D annotation data of several target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range, thus realizing the acquisition of target 4D lane line annotation data and providing actual data support for the subsequent quality inspection process. The target quality inspection operation result of the target 4D lane line annotation data is determined. The target quality inspection operation result is the result obtained after performing target quality inspection operations on the target 4D lane line annotation data. The target quality inspection operations include: jump quality inspection operation, type quality inspection operation, and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane lines are based on... The target 4D lane line annotation data represents the lane lines of the target lane lines. A type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between every two annotated lane lines. A preset quality inspection operation is used to check specified annotation tasks. The annotation task is to perform quality inspection on target 4D lane line annotation data belonging to the same time range. This realizes the execution of target quality inspection operations on target 4D lane line annotation data, improving the efficiency and accuracy of quality inspection and reducing labor costs. Based on the results of the target quality inspection operation, abnormal 4D lane line annotation data is identified within the target 4D lane line annotation data. Abnormal 4D lane line annotation data is target 4D lane line annotation data that exceeds the target quality inspection operation results. This realizes the identification of incorrectly annotated abnormal 4D lane line annotation data based on the target quality inspection operation results, facilitating secondary quality inspection of abnormal 4D lane line annotation data.

[0054] Example 2

[0055] Figure 2 This is a flowchart illustrating a 4D lane line annotation data quality inspection method provided in Embodiment 2 of the present invention. The technical solution of this embodiment further optimizes the target quality inspection operation results of the aforementioned embodiments based on the technical solutions of the previous embodiments. Solutions not described in detail in this embodiment are found in the above embodiments. This embodiment can be combined with various optional solutions from one or more of the above embodiments. Figure 2 As shown, the method includes:

[0056] S201. Obtain target point cloud data and determine target 4D lane line annotation data based on target point cloud data. Target point cloud data is the three-dimensional data of the target object in the driving scene. The target object includes at least several target lane lines. Target 4D lane line annotation data is the 4D annotation data of several target lane lines. 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range.

[0057] S202. Determine the target surface equation of the target 4D lane line annotation data. The target surface equation is used to fit the vehicle drivable trajectory corresponding to the target 4D lane line annotation data.

[0058] Specifically, the target surface equation can be obtained by performing surface fitting on the vehicle's drivable trajectory corresponding to the target 4D lane line annotation data. Here, the vehicle's drivable trajectory refers to the road on which the vehicle travels in the driving scenario, as represented by the target 4D lane line annotation data.

[0059] As an option, determining the target surface equation for the target 4D lane line annotation data includes the following steps C1-C4:

[0060] Step C1: Determine the vehicle's drivable trajectory corresponding to the target 4D lane line annotation data based on the target point cloud data.

[0061] Step C2: Divide the vehicle's drivable trajectory into segments according to the preset segmentation distance.

[0062] Step C3: Determine the target plane equation corresponding to each segment of the vehicle's drivable trajectory.

[0063] Step C4: The target plane equation of each segment of the vehicle's drivable trajectory is weighted and summed to obtain the target surface equation.

[0064] Specifically, based on vehicle motion characteristics, the drivable trajectory of the vehicle can be segmented from the target point cloud data, and then segmented according to a preset distance. The vehicle's drivable trajectory can be divided into The target plane equation is obtained by performing planar fitting on each segment of the vehicle's drivable trajectory. Furthermore, the target surface equation is obtained by weighted summation of each target plane equation based on the vehicle's drivable trajectory.

[0065] For example, the PANSAC algorithm can be used to perform planar fitting on each segment of the vehicle's drivable trajectory. Each segment of the vehicle's drivable trajectory can be used as... and from Three non-collinear points are randomly selected to fit the drivable trajectory of the vehicle, resulting in the fitted plane equation. .calculate The distances from other points in the plane to the fitted plane equation are calculated and compared with a preset distance threshold. Points with distances less than the preset threshold are designated as inliers, and the proportion of inliers is calculated. ,in, Indicates the number of interior points. express The quantity of all points. Iterate the above process until... After all increments are less than a preset threshold, the final fitted plane equation is taken as the target plane equation. Then, by weighted summation of each target plane equation, the target surface equation can be obtained. ,in, Represents the weights of each objective plane equation and is based on the interior point scale. Sure.

[0066] S203. Determine the Z-axis jitter tolerance value of the target 4D lane line annotation data. The three dimensions of the target 4D lane line annotation data are the X-axis, Y-axis and Z-axis. The Z-axis jitter tolerance value is used to check whether the smoothness of each labeled lane line on the Z-axis meets the preset smoothness threshold.

[0067] Specifically, the allowable jitter error value for the Z-axis coordinates of each node in the target 4D lane line annotation data can be preset as the Z-axis jitter tolerance value. If the Z-axis coordinates of each node in the target 4D lane line annotation data meet the Z-axis jitter tolerance value, then the smoothness of the labeled lane line corresponding to the target 4D lane line annotation data on the Z-axis can be considered to meet the preset smoothness threshold.

[0068] S204. Determine the jump quality inspection results of the target 4D lane line annotation data based on the target surface equation and Z-axis jitter tolerance value.

[0069] Specifically, the coordinates of each node in the target 4D lane line annotation data can be substituted into the target surface equation to obtain the Z-axis jitter value of each node coordinate relative to the target surface equation. Then, the absolute value of the Z-axis jitter value of each node coordinate relative to the target surface equation can be compared with the Z-axis jitter tolerance value. If the absolute value of the Z-axis jitter value of a node is greater than the Z-axis jitter tolerance value, the lane line represented by the target 4D lane line annotation data is considered to have a Z-axis lane line jump, and the node is marked. If the absolute value of the Z-axis jitter value of no node is greater than the Z-axis jitter tolerance value, the lane line represented by the target 4D lane line annotation data is considered not to have a Z-axis lane line jump. Finally, the presence or absence of a Z-axis lane line jump in the lane line represented by the target 4D lane line annotation data can be used as the result of the jump quality inspection operation.

[0070] S205. Based on the target quality inspection operation results, determine the abnormal 4D lane line annotation data in the target 4D lane line annotation data. The abnormal 4D lane line annotation data is the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation results.

[0071] The technical solution of this invention acquires target point cloud data and determines target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene, and the target object includes at least several target lane lines. The target 4D lane line annotation data is the 4D annotation data of several target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range, thus realizing the acquisition of target 4D lane line annotation data and providing practical data support for the subsequent quality inspection process. The invention also determines the target surface equation of the target 4D lane line annotation data, which is used to fit the vehicle's drivable trajectory corresponding to the target 4D lane line annotation data. Finally, the invention determines the Z-axis jitter tolerance value of the target 4D lane line annotation data. The three dimensions of the target 4D lane line annotation data are the X-axis, Y-axis, and Z-axis, and the Z-axis jitter tolerance value is used to verify each... The system checks whether the smoothness of the lane markings on the Z-axis meets a preset smoothness threshold; it determines the jump quality inspection results of the target 4D lane marking data based on the target surface equation and Z-axis jitter tolerance value, thereby obtaining the target surface equation of the vehicle's drivable trajectory corresponding to the target 4D lane marking data, and determining whether there are Z-axis lane marking jumps based on the target surface equation and Z-axis jitter tolerance value, in order to improve the accuracy of the jump quality inspection operation on the target 4D lane marking data; it identifies abnormal 4D lane marking data in the target 4D lane marking data based on the target quality inspection operation results. Abnormal 4D lane marking data are target 4D lane marking data that exceed the target quality inspection operation as indicated by the target quality inspection operation results, thereby identifying erroneously labeled abnormal 4D lane marking data based on the target quality inspection operation results, so as to perform secondary quality inspection on the abnormal 4D lane marking data.

[0072] Example 3

[0073] Figure 3 This is a flowchart illustrating a 4D lane line annotation data quality inspection method provided in Embodiment 3 of the present invention. The technical solution of this embodiment further optimizes the target quality inspection operation results of the aforementioned embodiments based on the technical solutions of the above embodiments. Solutions not described in detail in this embodiment can be found in the above embodiments. This embodiment can be combined with various optional solutions in one or more of the above embodiments. Figure 3 As shown, the method includes:

[0074] S301. Obtain target point cloud data and determine target 4D lane line annotation data based on target point cloud data. Target point cloud data is the three-dimensional data of the target object in the driving scene. The target object includes at least several target lane lines. Target 4D lane line annotation data is the 4D annotation data of several target lane lines. 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range.

[0075] S302. Based on the target 4D lane line annotation data, several annotated lane lines are segmented to obtain several reference lane lines.

[0076] Specifically, lane markings are formed by sequentially connecting the nodes in the target 4D lane marking data. Therefore, several lane markings can be segmented based on the nodes in the target 4D lane marking data to obtain several reference lane lines. For example, lane markings can be segmented based on nodes at different positions within the lane markings, with a preset number of nodes between these nodes. The preset number of nodes is not less than 0, and a smaller preset number indicates a better segmentation effect. If the preset number of nodes is 0, it means the lane markings are segmented based on two adjacent nodes.

[0077] S303. Determine the tangent vectors of several reference lane lines. The tangent vectors are the vectors formed by the two endpoints of the reference lane lines mapped onto the horizontal plane.

[0078] Specifically, the tangent vector of the reference lane line can refer to the tangent vector of the reference lane line on the X-axis or Y-axis. Therefore, each segment of the reference lane line can be mapped onto the horizontal plane, that is, only the X-axis and Y-axis coordinates are retained for the two endpoints of the reference lane line, so that the tangent vector of the reference lane line can be determined based on the X-axis and Y-axis coordinates of the two endpoints.

[0079] For example, for the two endpoints of the reference lane line and Tangent vector of reference lane line It can be represented as:

[0080] ;

[0081] For tangent vectors After normalization, we can obtain:

[0082] ;

[0083] .

[0084] S304. Determine the jump quality inspection results of the target 4D lane line annotation data based on the tangent vector.

[0085] Specifically, the presence of lane line jumps on the X-axis or Y-axis can be determined based on the tangent vectors of each segment of the reference lane line in each lane line. The presence of lane line jumps on the X-axis or Y-axis represented by the target 4D lane line annotation data is used as the result of the jump quality inspection operation.

[0086] As an optional step, the quality inspection results of the target 4D lane line annotation data are determined based on the tangent vector, including the following steps D1-D3:

[0087] Step D1: Determine the vector angle between the tangent vectors of every two adjacent reference lane lines on each marked lane line.

[0088] Step D2: Determine the preset vector angle threshold. The preset vector angle threshold is used to check whether the smoothness of each marked lane line on the X-axis or Y-axis meets the preset smoothness threshold.

[0089] Step D3: Determine the jump quality inspection results of the target 4D lane line annotation data based on the vector angle and the preset vector angle threshold.

[0090] Specifically, the angle between the tangent vectors of every two adjacent reference lane lines on each labeled lane line can be calculated and compared sequentially with a preset vector angle threshold. If the angle between the tangent vectors of two adjacent reference lane lines is greater than the preset threshold, the common endpoint of the two adjacent reference lane lines is marked, indicating that the labeled lane line to which the two adjacent reference lane lines belong has an X-axis or Y-axis jump at that endpoint, causing the smoothness of the labeled lane line to which the two adjacent reference lane lines belong to not meet the preset smoothness threshold on the X-axis or Y-axis. By analogy, it is possible to determine whether each labeled lane line has an X-axis or Y-axis lane line jump, which serves as the result of the jump quality inspection operation.

[0091] For example, such as Figure 4 The tangent vectors for each segment of the reference lane line, as shown in Table 1, are calculated from the marked lane lines.

[0092] Table 1. Schematic diagram of tangent vectors for reference lane lines

[0093]

[0094] Furthermore, based on the vector angle between the tangent vectors of every two adjacent reference lane lines, it can be determined that the marked lane line changes at point F.

[0095] S305. Based on the target quality inspection operation results, determine the abnormal 4D lane line annotation data in the target 4D lane line annotation data. The abnormal 4D lane line annotation data is the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation results.

[0096] The technical solution of this invention acquires target point cloud data and determines target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene, and the target object includes at least several target lane lines. The target 4D lane line annotation data is the 4D annotation data of several target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range, thus realizing the acquisition of target 4D lane line annotation data and providing practical data support for the subsequent quality inspection process. Based on the target 4D lane line annotation data, the several annotated lane lines are segmented to obtain several reference lane lines. The tangent vectors of the several reference lane lines are determined, and the tangent vectors are the reference lane lines mapped onto the horizontal plane. The vector formed by the two endpoints; the jump quality inspection operation result of the target 4D lane line annotation data is determined based on the tangent vector, realizing the determination of whether there is an X-axis or Y-axis lane line jump in the annotation lane line based on the tangent vector of several reference lane lines in the annotation lane line, so as to further improve the accuracy of the jump quality inspection operation of the target 4D lane line annotation data; based on the target quality inspection operation result, abnormal 4D lane line annotation data in the target 4D lane line annotation data is determined. Abnormal 4D lane line annotation data is the target 4D lane line annotation data that exceeds the target quality inspection operation as represented by the target quality inspection operation result, realizing the determination of the erroneous 4D lane line annotation data based on the target quality inspection operation result, so as to perform secondary quality inspection on the abnormal 4D lane line annotation data.

[0097] Example 4

[0098] Figure 5 This is a schematic diagram of a 4D lane line marking data quality inspection device provided in Embodiment 4 of the present invention. This embodiment of the invention is applicable to situations involving quality inspection of 4D lane line marking data. The device can be implemented in hardware and / or software and can be configured in an electronic device that implements the 4D lane line marking data quality inspection method. Figure 5 As shown, the device includes:

[0099] The first determining module 401 is used to acquire target point cloud data and determine target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene. The target object includes at least a number of target lane lines. The target 4D lane line annotation data is the 4D annotation data of a number of target lane lines. The 4D annotation data is obtained by adding annotations after superimposing the three-dimensional data within a preset time range.

[0100] The second determining module 402 is used to determine the target quality inspection operation result of the target 4D lane line annotation data. The target quality inspection operation result is the result obtained after performing the target quality inspection operation on the target 4D lane line annotation data. The target quality inspection operation includes: jump quality inspection operation, type quality inspection operation and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane line is the lane line represented by the target lane line according to the target 4D lane line annotation data. The type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between each pair of labeled lane lines. The preset quality inspection operation is used to check the specified annotation task. The annotation task is the task of performing quality inspection on the target 4D lane line annotation data belonging to the same time range.

[0101] The third determining module 403 is used to determine abnormal 4D lane line annotation data in the target 4D lane line annotation data based on the target quality inspection operation results. Abnormal 4D lane line annotation data refers to target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation results.

[0102] Optionally, determine the target quality inspection results of the target 4D lane line annotation data, including:

[0103] Determine the target surface equation of the target 4D lane line annotation data. The target surface equation is used to fit the vehicle drivable trajectory corresponding to the target 4D lane line annotation data.

[0104] Determine the Z-axis jitter tolerance value of the target 4D lane line annotation data. The three dimensions of the target 4D lane line annotation data are the X-axis, Y-axis and Z-axis. The Z-axis jitter tolerance value is used to check whether the smoothness of each labeled lane line on the Z-axis meets the preset smoothness threshold.

[0105] The jump quality inspection results of the target 4D lane line annotation data are determined based on the target surface equation and the Z-axis jitter tolerance value.

[0106] Optionally, determine the target surface equation for the target 4D lane line annotation data, including:

[0107] Determine the vehicle's drivable trajectory corresponding to the target's 4D lane line annotation data based on the target point cloud data;

[0108] The vehicle's drivable trajectory is segmented according to a preset segmentation distance;

[0109] Determine the target plane equation corresponding to each segment of the vehicle's drivable trajectory;

[0110] The target surface equation is obtained by weighted summation of the target plane equation for each segment of the vehicle's drivable trajectory.

[0111] Optionally, determine the target quality inspection results of the target 4D lane line annotation data, including:

[0112] Based on the target 4D lane line annotation data, several annotated lane lines are segmented to obtain several reference lane lines;

[0113] Determine the tangent vectors of several reference lane lines. The tangent vectors are the vectors formed by the two endpoints of the reference lane lines mapped onto the horizontal plane.

[0114] The jump quality inspection results of the target 4D lane line annotation data are determined based on the tangent vector.

[0115] Optionally, the jump quality inspection results of the target 4D lane line annotation data are determined based on the tangent vector, including:

[0116] Determine the vector angle between the tangent vectors of every two adjacent reference lane lines on each marked lane line;

[0117] Determine a preset vector angle threshold. The preset vector angle threshold is used to check whether the smoothness of each marked lane line on the X-axis or Y-axis meets the preset smoothness threshold.

[0118] The jump quality inspection results of the target 4D lane line annotation data are determined based on the vector angle and the preset vector angle threshold.

[0119] Optionally, determine the target quality inspection results of the target 4D lane line annotation data, including:

[0120] Determine the intersection status between any two lane lines in a given set of lane markings;

[0121] The quality inspection results are obtained by determining the type of target 4D lane line annotation data based on the intersection status.

[0122] Optionally, determine the target quality inspection results of the target 4D lane line annotation data, including:

[0123] Determine the target quality inspection object in the annotation task. The target quality inspection object is the object when performing quality inspection on the target 4D lane line annotation data.

[0124] The reference configuration operations for determining the target quality inspection object include: target configuration, action configuration, condition configuration, and logical relationship configuration.

[0125] The preset quality inspection operation results of the target 4D lane line annotation data are determined based on the target quality inspection object and the reference configuration operation.

[0126] The technical solution of this invention acquires target point cloud data and determines target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the driving scene, and the target object includes at least several target lane lines. The target 4D lane line annotation data is the 4D annotation data of several target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range, thus realizing the acquisition of target 4D lane line annotation data and providing actual data support for the subsequent quality inspection process. The target quality inspection operation result of the target 4D lane line annotation data is determined. The target quality inspection operation result is the result obtained after performing target quality inspection operations on the target 4D lane line annotation data. The target quality inspection operations include: jump quality inspection operation, type quality inspection operation, and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane lines are based on... The target 4D lane line annotation data represents the lane lines of the target lane lines. A type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between every two annotated lane lines. A preset quality inspection operation is used to check specified annotation tasks. The annotation task is to perform quality inspection on target 4D lane line annotation data belonging to the same time range. This realizes the execution of target quality inspection operations on target 4D lane line annotation data, improving the efficiency and accuracy of quality inspection and reducing labor costs. Based on the results of the target quality inspection operation, abnormal 4D lane line annotation data is identified within the target 4D lane line annotation data. Abnormal 4D lane line annotation data is target 4D lane line annotation data that exceeds the target quality inspection operation results. This realizes the identification of incorrectly annotated abnormal 4D lane line annotation data based on the target quality inspection operation results, facilitating secondary quality inspection of abnormal 4D lane line annotation data.

[0127] The 4D lane line marking data quality inspection device provided in this embodiment of the invention can execute the 4D lane line marking data quality inspection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0128] Example 5

[0129] Figure 6This is a schematic diagram of an electronic device for implementing a 4D lane line marking data quality inspection method, provided in Embodiment 5 of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, 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 processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0130] like Figure 6 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0131] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0132] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the 4D lane line annotation data quality inspection method.

[0133] In some embodiments, the 4D lane line marking data quality inspection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the 4D lane line marking data quality inspection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the 4D lane line marking data quality inspection method by any other suitable means (e.g., by means of firmware).

[0134] Various embodiments 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-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0135] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0136] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0137] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device 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 provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, 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 sound input, voice input, or tactile input).

[0138] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0139] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0140] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0141] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for quality inspection of 4D lane line annotation data, characterized in that, The method includes: Acquire target point cloud data and determine target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the vehicle driving scene. The target object includes at least a number of target lane lines. The target 4D lane line annotation data is the 4D annotation data of the target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range. The target quality inspection result of the target 4D lane line annotation data is determined. The target quality inspection result is the result obtained after performing a target quality inspection operation on the target 4D lane line annotation data. The target quality inspection operation includes: jump quality inspection operation, type quality inspection operation, and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets the preset smoothness threshold. The labeled lane line is the lane line represented by the target lane line according to the target 4D lane line annotation data. The type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between every two labeled lane lines. The preset quality inspection operation is used to check a specified annotation task. The annotation task is a task that performs quality inspection on the target 4D lane line annotation data belonging to the same time range. Based on the target quality inspection operation results, abnormal 4D lane line annotation data in the target 4D lane line annotation data is determined. The abnormal 4D lane line annotation data is the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation results.

2. The method according to claim 1, characterized in that, Determining the target quality inspection results of the target 4D lane line annotation data includes: Determine the target surface equation of the target 4D lane line annotation data, and the target surface equation is used to fit the vehicle drivable trajectory corresponding to the target 4D lane line annotation data. The Z-axis jitter tolerance value of the target 4D lane line annotation data is determined. The three dimensions of the target 4D lane line annotation data are the X-axis, Y-axis and Z-axis. The Z-axis jitter tolerance value is used to check whether the smoothness of each labeled lane line on the Z-axis meets the preset smoothness threshold. The Z-axis jitter tolerance value includes the jitter error value allowed for the Z-axis coordinate values ​​of each node in the target 4D lane line annotation data. The Z-axis jitter value of each node coordinate relative to the target surface equation is obtained by substituting the coordinates of each node in the target 4D lane line annotation data into the target surface equation. Based on the Z-axis jitter tolerance value and the absolute value of the Z-axis jitter value of each node coordinate relative to the target surface equation, the jump quality inspection operation result of the target 4D lane line annotation data is determined.

3. The method according to claim 2, characterized in that, Determining the target surface equation of the target 4D lane line annotation data includes: The vehicle's drivable trajectory corresponding to the target 4D lane line annotation data is determined based on the target point cloud data. The vehicle's drivable trajectory is segmented according to a preset segmentation distance; Determine the target plane equation corresponding to each segment of the vehicle's drivable trajectory; The target surface equation is obtained by weighted summation of the target plane equation for each segment of the vehicle's drivable trajectory.

4. The method according to claim 1, characterized in that, Determining the target quality inspection results of the target 4D lane line annotation data includes: Based on the target 4D lane line annotation data, several annotated lane lines are segmented to obtain several reference lane lines. Determine a number of tangent vectors for the reference lane lines, wherein the tangent vectors are vectors formed by the two endpoints of the reference lane lines mapped onto the horizontal plane; The jump quality inspection result of the target 4D lane line annotation data is determined based on the tangent vector.

5. The method according to claim 4, characterized in that, The jump quality inspection results of the target 4D lane line annotation data are determined based on the tangent vector, including: Determine the vector angle between the tangent vectors of every two adjacent reference lane lines on each of the marked lane lines; A preset vector angle threshold is determined, which is used to check whether the smoothness of each marked lane line on the X-axis or Y-axis meets the preset smoothness threshold. The jump quality inspection result of the target 4D lane line annotation data is determined based on the vector angle and the preset vector angle threshold.

6. The method according to claim 1, characterized in that, Determining the target quality inspection results of the target 4D lane line annotation data includes: Determine the intersection state between every two lane marking lines in a plurality of lane marking lines; The type of the target 4D lane line annotation data is determined based on the intersection state, and the quality inspection results are obtained.

7. The method according to claim 1, characterized in that, Determining the target quality inspection results of the target 4D lane line annotation data includes: Determine the target quality inspection object in the annotation task, wherein the target quality inspection object is the object when performing quality inspection on the target 4D lane line annotation data; The reference configuration operation for determining the target quality inspection object includes: target configuration, action configuration, condition configuration, and logical relationship configuration. The preset quality inspection operation result of the target 4D lane line annotation data is determined based on the target quality inspection object and the reference configuration operation.

8. A 4D lane marking data quality inspection device, characterized in that, The device includes: The first determining module is used to acquire target point cloud data and determine target 4D lane line annotation data based on the target point cloud data. The target point cloud data is the three-dimensional data of the target object in the vehicle driving scene. The target object includes at least a number of target lane lines. The target 4D lane line annotation data is the 4D annotation data of the target lane lines. The 4D annotation data is obtained by adding annotations after superimposing three-dimensional data within a preset time range. The second determining module is used to determine the target quality inspection operation result of the target 4D lane line annotation data. The target quality inspection operation result is the result obtained after performing a target quality inspection operation on the target 4D lane line annotation data. The target quality inspection operation includes: jump quality inspection operation, type quality inspection operation, and preset quality inspection operation. The jump quality inspection operation is used to check whether the smoothness of each labeled lane line meets a preset smoothness threshold. The labeled lane line is a lane line represented by the target lane line according to the target 4D lane line annotation data. The type quality inspection operation is used to check the annotation type of the target 4D lane line annotation data based on the intersection state between every two labeled lane lines. The preset quality inspection operation is used to check a specified annotation task. The annotation task is a task that performs quality inspection on the target 4D lane line annotation data belonging to the same time range. The third determining module is used to determine abnormal 4D lane line annotation data in the target 4D lane line annotation data based on the target quality inspection operation result. The abnormal 4D lane line annotation data is the target 4D lane line annotation data that exceeds the target quality inspection operation as indicated by the target quality inspection operation result.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the 4D lane line annotation data quality inspection method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the 4D lane line annotation data quality inspection method according to any one of claims 1-7.

Images