Point cloud data processing method, device, equipment and storage medium
By acquiring the lidar point cloud data set and dividing the point cloud data based on the height value and the extreme difference of the ground height, the problem of low processing accuracy of the lidar point cloud data is solved, and more accurate obstacle recognition and visualization of intelligent driving scenarios are achieved. , improving the reliability of autonomous driving.
Patent Information
- Application Number
- CN202111177379.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-10-09
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-10-09
AI Technical Summary
In the prior art, the laser radar point cloud data processing accuracy is low, resulting in inaccurate obstacle identification and affecting the reliability of autonomous driving.
By obtaining the point cloud data set collected by lidar, the height value of each point and the extreme difference radian and limit radian of the ground height are determined. Based on these parameters, the point cloud data is divided into ground points and non-ground points, and further processed using density clustering algorithm Target obstacle point cloud data to improve data accuracy.
It improves the accuracy of point cloud data processing, can more accurately divide ground points and non-ground points, form a visual vehicle driving scenario, and enhances the reliability of intelligent driving.
Smart Images

Figure CN113989289B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of laser radar data processing technology, and in particular to a point cloud data processing method, device, equipment and storage medium. Background Art
[0002] In the field of autonomous driving environmental perception, environmental perception is the key and difficulty in realizing autonomous driving. It provides the basis for decision-making planning and control, and is the premise and necessary condition for realizing autonomous driving.
[0003] The main sensors currently used are LiDAR, millimeter-wave radar, and cameras. LiDAR has become the main sensor for high-level autonomous driving due to its wide field of view, weak light dependence, and high ranging accuracy.
[0004] However, in the prior art, when processing laser radar data, the point cloud data of the obstacle is determined by calculating the slope of each laser point, so that the calculated result has the disadvantage of low accuracy. Summary of the invention
[0005] The present invention aims to solve the technical problem of low point cloud data processing accuracy of laser radar in the prior art.
[0006] In order to solve the above technical problems, the present application discloses a point cloud data processing method, which comprises the following steps:
[0007] Obtain a point cloud data set; the point cloud data set includes three-dimensional coordinates of multiple points; the point cloud data is collected by a laser radar;
[0008] Determine the height value of each point from the three-dimensional coordinates of the multiple points;
[0009] Get the ground height extreme difference arc and ground height limit arc;
[0010] Determine the attribute of each point based on the height value of each point, the ground height extreme difference arc and the ground height limit arc;
[0011] A point cloud dataset of the target obstacle is determined based on the attribute of each point and the point cloud dataset.
[0012] Optionally, this attribute includes ground points and non-ground points;
[0013] The attribute of each point is determined based on the height value of each point, the ground height range radian and the ground height limit radian, including:
[0014] Determine the three-dimensional coordinates of the target point from the point cloud data set;
[0015] Determine the ground height limit value of the target point according to the three-dimensional coordinates of the target point and the ground height limit arc;
[0016] Determine the attribute of the target point according to the height value of the target point and the ground height limit value of the target point;
[0017] The attributes of the remaining target points are determined based on the attributes of the target point, the height values of the remaining target points, the ground height extreme difference arc and the ground height limit arc.
[0018] Optionally, the determining the attributes of the remaining target points based on the attributes of the target point, the height values of the remaining target points, the ground height range radian and the ground height limit radian includes:
[0019] determining the target point as the previous target point;
[0020] Determine the current target point from the remaining target points; the remaining target points are target points whose attributes have not been determined;
[0021] Determine the ground height extreme difference value of the current target point according to the three-dimensional coordinates of the current target point, the three-dimensional coordinates of the previous target point and the ground height extreme difference arc;
[0022] Determine the ground height limit value of the current target point according to the three-dimensional coordinates of the current target point and the ground height limit arc;
[0023] Determining the distance between the current target point and the previous target point;
[0024] If the distance between the current target point and the previous target point is less than or equal to a preset threshold, a first result of the current target point is determined according to the attribute of the previous target point, the height value of the current target point, the height value of the previous target point and the ground height extreme difference value of the current target point;
[0025] Determine the attribute of the current target point based on the first result;
[0026] Determine the current target point as the updated previous target point;
[0027] Re-determine the updated current target point from the remaining target points; repeat the above steps of determining the attributes of the current target point to determine the attributes of the updated current target point; repeat the above steps of determining the updated previous target point, the updated current target point and determining the attributes of the updated current target point until the attributes of all remaining target points are determined.
[0028] Optionally, if the distance between the current target point and the previous target point is less than or equal to a preset threshold, then determine the first result of the current target point according to the attribute of the previous target point, the height value of the current target point, the height value of the previous target point, and the ground height extreme difference value of the current target point; determine the attribute of the current target point based on the first result, including:
[0029] If the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a ground point, and the height value of the current target point satisfies the first preset condition, then determine that the first result of the current target point is directly confirmable attribute;
[0030] Determine that the attribute of the current target point is a ground point based on the first result;
[0031] If the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a non-ground point, and the height value of the current target point satisfies the first preset condition, then determine that the first result of the current target point is not directly confirmable attribute;
[0032] Process the height value of the current target and the second preset condition based on the first result to obtain a second result;
[0033] If the second result is that the height value of the current target satisfies the second preset condition, then determine that the attribute of the current target point is a ground point; otherwise, determine that the attribute of the current target point is a non-ground point.
[0034] Optionally, the first preset condition is:
[0035] Z0 - a < Z1 < Z0 + a, where Z0 represents the height value of the previous target point; Z1 represents the height value of the current target point; a represents the ground height extreme difference value of the current target point;
[0036] The second preset condition is:
[0037] Z0 - b < Z1 < Z0 + b, where b represents the ground height limit value of the current target point.
[0038] Optionally, before determining the three-dimensional coordinates of the target point from the point cloud dataset, it further includes:
[0039] Obtain the horizontal angular resolution and the horizontally rotatable angle of the lidar;
[0040] Determine the number of sub-point cloud datasets according to the horizontal angular resolution and the horizontally rotatable angle;
[0041] Determine the horizontal deflection angle and the radius of each point; the radius is the distance between each point and the lidar;
[0042] Determine the sequence number of the sub-point cloud data set to which each point belongs according to the horizontal deflection angle of each point and the horizontal angular resolution; the point cloud data set includes multiple sub-point cloud data sets;
[0043] Determining the three-dimensional coordinates of the target point from the point cloud data set includes:
[0044] The following processing is performed on each of the multiple sub-point cloud data sets, and each of the multiple sub-point cloud data sets is processed in parallel:
[0045] The points in the sub-point cloud data set are sorted from small to large according to the size of the radius to generate a first data set to be processed;
[0046] Determine the point with the smallest radius in the first data set to be processed as the target point;
[0047] After determining the attributes of the remaining target points based on the attributes of the target point, the height values of the remaining target points, the ground height range radian and the ground height limit radian, the method further includes:
[0048] The points in the sub-point cloud data set are sorted from large to small according to the radius to generate a second data set to be processed;
[0049] Determine the point with the largest radius in the second data set to be processed as the target point; repeat the above steps of determining the attributes of the target point and the attributes of the remaining target points;
[0050] The point cloud data of the target obstacle is determined based on the attribute of each point and the point cloud data set, including:
[0051] A point cloud data set of the target obstacle is determined according to the attribute of each point in the first data set to be processed, the attribute of each point in the second data set to be processed, and the point cloud data set.
[0052] Optionally, determining the point cloud data of the target obstacle based on the attribute of each point and the point cloud data set includes:
[0053] Determine a point cloud data set of the target obstacle to be processed according to the attribute of each point and the point cloud data set;
[0054] The point cloud dataset of the target obstacle to be processed is clustered using a density clustering algorithm to obtain a point cloud dataset of the target obstacle.
[0055] In another aspect, the present application further discloses a point cloud data processing device, which comprises:
[0056] An acquisition module is used to acquire a point cloud data set, a ground height extreme difference arc, and a ground height limit arc; the point cloud data set includes three-dimensional coordinates of multiple points; the point cloud data is collected by a laser radar;
[0057] A height value determination module, used to determine the height value of each point from the three-dimensional coordinates of the multiple points;
[0058] An attribute determination module is used to determine the attribute of each point based on the height value of each point, the ground height extreme difference arc and the ground height limit arc;
[0059] The point cloud data set determination module is used to determine the point cloud data set of the target obstacle based on the attribute of each point and the point cloud data set.
[0060] On the other hand, the present application also discloses a computer device, which includes a processor and a memory, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the above-mentioned point cloud data processing method.
[0061] On the other hand, the present application also discloses a computer storage medium, in which at least one instruction or at least one program is stored. The at least one instruction or at least one program is loaded and executed by a processor to implement the above-mentioned point cloud data processing method.
[0062] By adopting the above technical solution, the point cloud data processing method provided by this application has the following beneficial effects:
[0063] The point cloud data processing method provided by the present application first obtains a point cloud data set; the point cloud data set includes the three-dimensional coordinates of multiple points; the point cloud data is collected by a laser radar; the height value of each point is determined from the three-dimensional coordinates of the multiple points; the ground height extreme difference arc and the ground height limit arc are obtained; so that the attributes of each point can be determined based on the height value of each point, the ground height extreme difference arc and the ground height limit arc; the point cloud data set of the target obstacle is determined based on the attributes of each point and the point cloud data set. In this way, the point cloud data in the above point cloud data set can be more accurately divided into ground points and non-ground points, and then the point cloud data of the ground points are processed, so that the data of the target obstacle can be obtained, which is convenient for forming a visualized vehicle driving scene. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0065] Figure 1 An optional application scenario diagram for this application;
[0066] Figure 2 This is a flowchart of the first optional point cloud data processing method of this application;
[0067] Figure 3 A schematic diagram of an optional point cloud data for this application;
[0068] Figure 4 This is a flowchart of a second optional point cloud data processing method of this application;
[0069] Figure 5 A schematic diagram of an optional model for determining the ground height limit value of a target point in this application;
[0070] Figure 6 This is a flowchart of a third optional point cloud data processing method of this application;
[0071] Figure 7 A schematic diagram of an optional model for determining the extreme difference value of the ground height of a target point in this application;
[0072] Figure 8 This is a schematic diagram of the structure of an optional point cloud data processing device of the present application;
[0073] Fig. 9 This is a hardware structure block diagram of a server of an optional point cloud data processing method of the present application. DETAILED DESCRIPTION
[0074] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0075] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0076] like Figure 1 As shown, Figure 1 This is an optional application scenario diagram for this application. The scenario includes a vehicle 10, a data processing unit 30, and a laser radar 20 arranged on the vehicle; the laser radar 20 is used to collect point cloud data, thereby forming a point cloud data set, and sending it to the data processing unit 30; the data processing unit 30 is used to receive a point cloud data set, the point cloud data set including the three-dimensional coordinates of multiple points; and determine the height value of each point from the three-dimensional coordinates of the multiple points; and then obtain the ground height extreme radian and the ground height limit radian; so that the attributes of each point can be determined based on the height value of each point, the ground height extreme radian and the ground height limit radian, and the point cloud data set of the target obstacle can be determined based on the attributes of each point and the point cloud data set, so that the point cloud data of the target obstacle processed by the data processing unit 30 can be more accurate, the reliability of intelligent driving is improved, and the technical development in the field of intelligent driving is promoted.
[0077] Optionally, the data processing unit may be arranged on the vehicle, or may be placed in a server or other terminal outside the vehicle.
[0078] Optionally, the terminal may be a physical device of the type of a desktop computer, a laptop computer, a mobile phone, a tablet computer, a digital assistant, a smart wearable device, etc.; among which, the smart wearable device may include a smart bracelet, a smart watch, smart glasses, a smart helmet, etc.
[0079] The terminal may include a display screen, a storage device and a processor connected via a data bus. The display screen is used for displaying the virtual image of the device to be monitored and the connection relationship between the various sub-devices in the device to be monitored. The display screen may be a touch screen of a mobile phone or a tablet computer. The storage device is used to store the program code and data of the shooting device. The storage device may be a memory of the terminal, or a storage device such as a smart media card, a secure digital card, or a flash card. The processor may be a single-core or multi-core processor.
[0080] The following describes a specific embodiment of a point cloud data processing method of the present application. Figure 2 This is a flowchart of the first optional point cloud data processing method of the present application. This specification provides method operation steps such as the embodiments or flowcharts, but may include more or fewer operation steps based on conventional or non-creative labor. The order of steps listed in the embodiments is only one way of executing the steps among many orders, and does not represent the only order of execution. When the actual system or server product is executed, it can be executed in the order shown in the embodiments or the drawings or in parallel (for example, in a parallel processor or multi-threaded processing environment). Specifically, Figure 2 As shown, the method may include:
[0081] S201: Obtain a point cloud data set; the point cloud data set includes three-dimensional coordinates of multiple points; the point cloud data is collected by a laser radar.
[0082] Optionally, the three-dimensional coordinates of the points in the point cloud data set can be the three-dimensional coordinates in the world coordinate system, or the three-dimensional coordinates corresponding to the spatial coordinate system established with the laser radar as the origin, and can be subsequently converted accordingly as needed.
[0083] Optionally, the point cloud data set in the embodiment of the present application can be directly based on the point cloud data collected by the laser radar; it can also be the point cloud data after preliminary screening and processing. Optionally, step S201 can be specifically explained as: using the laser radar to collect environmental information around the vehicle to form a preliminary point cloud data set, and using the point cloud voxel downsampling method to process the preliminary point cloud data set to obtain the point cloud data set; thereby effectively reducing the number of point cloud data and reducing the burden of subsequent data processing; while improving the speed of program processing, maintaining the overall characteristics of the point cloud and ensuring the accuracy of subsequent algorithms.
[0084] Optionally, the above-mentioned point cloud voxel downsampling method can be the following data processing process: create a three-dimensional voxel grid, and in each voxel, use the centroid point of all the points in the voxel to represent the characteristics of all the points in the voxel, so that all the points in the voxel are finally represented by a centroid point, and a filtered point cloud is obtained after processing all voxels; compared with the method of approximating with the voxel center, this method is more accurate in representing the surface corresponding to the sampling point.
[0085] Optionally, the laser radar in the present application needs to be calibrated before collecting data to ensure the accuracy of the subsequently collected data; the specific calibration method can be any one of the following: Method 1, calibrating it by auxiliary calibration, it is necessary to provide a flat calibration site, and complete the calibration of the laser radar by setting the reference object position information and size information; Method 2, calibrating it by automatic calibration, it only needs to drive the vehicle to the specified area, and the calibration can be completed based on the debugging tool.
[0086] Optionally, in the present embodiment, the point cloud dataset in the above step S201 may be a dataset formed by point cloud data collected by multiple lidars; optionally, the present application performs the above and subsequent processing on the above point cloud dataset as a whole, thereby improving data integrity and processing efficiency.
[0087] In order to improve processing efficiency, the above-mentioned point cloud dataset can be optionally divided into multiple sub-point cloud datasets, so that each sub-point cloud dataset can be processed in parallel subsequently; after step S201, the point cloud data processing method also includes: obtaining the horizontal angular resolution and the horizontal rotatable angle of the laser radar, determining the number of sub-point cloud datasets according to the horizontal angular resolution and the horizontal rotatable angle, determining the horizontal deflection angle and radius of each point, the radius being the distance of each point from the laser radar, determining the serial number of the sub-point cloud dataset to which each point belongs according to the horizontal deflection angle and the horizontal angular resolution of each point, and the point cloud dataset includes multiple sub-point cloud datasets.
[0088] Optional, such as Figure 3 As shown, Figure 3 This is a schematic diagram of an optional point cloud data of this application. In order to better describe the solution of this application and reflect its beneficial effects, the horizontal plane of this application is Figure 3 The plane formed by the xy axes in the image, and the z axis is perpendicular to the plane formed by the xy axes; the horizontal angular resolution is δ, and the horizontal rotatable angle is the angle that the laser radar can rotate around the z axis, for example, 360 degrees, δ = 2, then the number of sub-point cloud data sets n = 360 / 2 = 180; for point m (x1, y1, z1); its horizontal deflection angle is θ, and the following formula can be used to determine θ of m:
[0089] θ=artg(y1 / x1)……Formula (1)
[0090] In order to quickly classify the points, the final θ can be determined by rounding down to ten. For example, when the actual θ of m is 33 degrees, based on the above rounding down principle, the θ of m determined here is 30 degrees, that is, θ is in the range of 30-40 degrees (excluding 40), and θ is 30 degrees.
[0091] The serial number W of the sub-point cloud dataset to which each point belongs can be calculated based on the following formula:
[0092] W=θ / δ……Formula (2)
[0093] Continuing with the above example, the sequence number of the sub-point cloud dataset to which point m belongs is 30 / 2=15.
[0094] The above method can be used to determine the serial number of each sub-point cloud data set and the points included in each sub-point cloud data set, so as to segment all points. Subsequently, multiple sub-point cloud data sets can be processed simultaneously using parallel processing, thereby greatly improving processing efficiency.
[0095] It should be noted that the points in the same sub-point cloud dataset are on the same ray (for example, m and n are on the same ray), and the point cloud dataset includes multiple rays, each of which corresponds to a sub-point cloud dataset.
[0096] Of course, as needed, this application can also process the data without classifying it, which simplifies the calculation method.
[0097] Optionally, in order to further improve the efficiency of subsequent processing of the point cloud dataset and reduce the amount of data in the point cloud dataset, step S201 can also be specifically explained as: using a high-precision map to obtain a preliminary point cloud dataset within a preset range of the vehicle; obtaining the position coordinates of the lidar, and determining the point cloud dataset of the area of interest based on the position coordinates of the lidar, the high-precision map and the preliminary point cloud dataset; and determining the point cloud dataset of the area of interest as the point cloud dataset.
[0098] Optionally, by confirming the high-precision map, the point cloud data outside the elevated road around the vehicle and the point cloud data within a preset range around the vehicle as the origin can be determined as non-areas of interest, filtered out, and then the point cloud data set of the above-mentioned area of interest can be obtained.
[0099] S202: Determine the height value of each point from the three-dimensional coordinates of the multiple points.
[0100] Optionally, in this embodiment, the height value of each point may be a z value.
[0101] S203: Obtain the ground height extreme difference arc and the ground height limit arc.
[0102] Optionally, the ground height extreme difference arc is recorded as k, and the ground height limit arc is recorded as l. The above k and l can be determined based on experience. The angle corresponding to k is recorded as α, and the angle corresponding to l is recorded as β.
[0103] S204: Determine the attribute of each point based on the height value of each point, the ground height range radian and the ground height limit radian.
[0104] In some possible implementations, the attributes include ground points and non-ground points.
[0105] Optional, see Figure 4 , Figure 4 This is a flow chart of a second optional point cloud data processing method of this application. Step S204 can be specifically described as follows:
[0106] S2041: Determine the three-dimensional coordinates of the target point from the point cloud data set.
[0107] In order to improve processing efficiency and processing accuracy, step S2041 can be specifically described as follows: for each of the multiple sub-point cloud data sets, the following processing is performed, and each of the multiple sub-point cloud data sets is processed in parallel: each point in the sub-point cloud data set is sorted from small to large according to the size of its radius to generate a first data set to be processed; the point with the smallest radius in the first data set to be processed is determined as the target point. For example, Figure 3 , point m is the target point. Subsequently, the data processing efficiency can be improved by first judging the point with the smallest radius and processing each point in order from small to large radius.
[0108] It should be noted that the radius involved in the above description is a value within the plane where the xy axis is located, that is, a two-dimensional value.
[0109] In order to avoid subsequent processing based only on the first data set to be processed, that is, to determine whether the current point is a ground point, and only judge it together with the previous point, thereby causing the ground attributes and height of the previous point to have a great influence on the result, as well as misjudgment, the accuracy of data processing is improved; optionally, after determining the attributes of each point in the first data set to be processed, each point in the sub-point cloud data set needs to be sorted from large to small according to the size of its radius to generate a second data set to be processed, and the point with the largest radius in the second data set to be processed is determined as the target point; repeat the above steps of determining the attributes of the target point and the attributes of the remaining target points; optionally, in order to further improve processing efficiency, the above steps of generating and processing the first data set to be processed and subsequently generating and processing the second data set to be processed can be processed in parallel.
[0110] It should be noted that, when the first data set to be processed and the second data set to be processed are processed in the order of radius arrangement, the first target point processed for the first data set to be processed is the one with the smallest radius, and the last target point is the one with the largest radius; for the second data set to be processed, the point with the largest radius is taken as the first target point, and the last target point is the point with the smallest radius; thus, the points in the above point cloud data set can be screened twice, that is, from near to far, and then from far to near, so as to effectively improve the accuracy of data processing while improving the data processing efficiency.
[0111] S2042: Determine the ground height limit value of the target point according to the three-dimensional coordinates of the target point and the ground height limit arc.
[0112] Optional, Participation Figure 5 , Figure 5 This is a schematic diagram of an optional model for determining the ground height limit value of a target point in this application. When the target point is point m and its corresponding radius is r1, the ground height limit value b of the target point is determined based on the following formula:
[0113] b=r1×tang[(180 / π)l]……Formula (3)
[0114] Among them, π is the ratio of a circle to a circle, and r1 is the distance from point m to the origin 0.
[0115] S2043: Determine the attribute of the target point according to the height value of the target point and the ground height limit value of the target point.
[0116] Optionally, z1 may be directly compared with the above b. When z1 is greater than b, the target point is determined to be a non-ground point; otherwise, it is determined to be a ground point.
[0117] S2044: Determine the attributes of the remaining target points based on the attributes of the target point, the height values of the remaining target points, the ground height extreme radian and the ground height limit radian.
[0118] See also Figure 6 , Figure 6 This is a flowchart of a third optional point cloud data processing method of the present application. In some feasible implementations, step S2044 can be specifically described as follows:
[0119] S601: Determine the target point as the previous target point, and determine the current target point from the remaining target points; the remaining target points are target points whose attributes have not been determined;
[0120] S602: Determine the ground height extreme difference value of the current target point according to the three-dimensional coordinates of the current target point, the three-dimensional coordinates of the previous target point and the ground height extreme difference arc;
[0121] S603: Determine the ground height limit value of the current target point according to the three-dimensional coordinates of the current target point and the ground height limit arc;
[0122] S604: Determine the distance between the current target point and the previous target point;
[0123] S605: If the distance between the current target point and the previous target point is less than or equal to a preset threshold, determining a first result of the current target point according to the attribute of the previous target point, the height value of the current target point, the height value of the previous target point and the ground height extreme difference value of the current target point;
[0124] S606: Determine the attribute of the current target point based on the first result;
[0125] S607: Determine the current target point as the updated previous target point;
[0126] S608: Re-determine the updated current target point from the remaining target points, repeat the above steps of determining the attributes of the current target point, determine the attributes of the updated current target point, repeat the above steps of determining the updated previous target point, the updated current target point and determining the attributes of the updated current target point, until the attributes of all remaining target points are determined.
[0127] Optional, Participation Figure 7 , Figure 7 This is a schematic diagram of an optional model for determining the ground height extreme value of a target point in this application. m is the previous target point, n is the current target point, and when determining the attributes of the current target point n, if the height extreme value a of the current target point is to be determined, and its corresponding radius is r2, a can be determined based on the following formula:
[0128] a=(r2-r1)×tang[(180 / π)k]……Formula (4)
[0129] Optionally, the preset thresholds include but are not limited to being set to 1, 2, 3, 4, etc., which are used to reflect that the distance between the two points is small, that is, the two points are relatively close, so that the attributes can be determined more accurately based on the relationship between the two points.
[0130] It should be noted that the radius and the distance between any two points involved in the above description are all values in the plane where the xy axis is located, that is, two-dimensional values. For example, for point m, the radius r1 satisfies the following relationship: (r1) 2 =(x1) 2 +(y1) 2 .
[0131] To ensure the accuracy of data processing and improve the flexibility of data processing; in some feasible embodiments, the above steps S605 - S606 may include the following processes: First, if the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a ground point, and the height value of the current target point meets the first preset condition, then determine that the first result of the current target point is directly confirmable for the attribute; based on the first result, determine that the attribute of the current target point is a ground point; Second, if the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a non - ground point, and the height value of the current target point meets the first preset condition, then determine that the first result of the current target point is not directly confirmable for the attribute; process the height value of the current target and the second preset condition based on the first result to obtain a second result; if the second result is that the height value of the current target meets the second preset condition, then determine that the attribute of the current target point is a ground point, otherwise, determine that the attribute of the current target point is a non - ground point; Third, if the distance between the current target point and the previous target point is greater than the preset threshold, and the attribute of the previous target point is a ground point, and the height value of the current target point does not meet the first preset condition, then determine that the first result of the current target point is not directly confirmable for the attribute; process the height value of the current target and the second preset condition based on the first result to obtain a second result; if the second result is that the height value of the current target meets the second preset condition, then determine that the attribute of the current target point is a ground point, otherwise, determine that the attribute of the current target point is a non - ground point; Fourth, if the distance between the current target point and the previous target point is greater than the preset threshold, and the attribute of the previous target point is a non - ground point, and the height value of the current target point does not meet the first preset condition, then determine that the first result of the current target point is not directly confirmable for the attribute; process the height value of the current target and the second preset condition based on the first result to obtain a second result; if the second result is that the height value of the current target meets the second preset condition, then determine that the attribute of the current target point is a ground point, otherwise, determine that the attribute of the current target point is a non - ground.
[0132] In some feasible embodiments, the first preset condition is: Z0 - a < Z1 < Z0 + a, where Z0 represents the height value of the previous target point; Z1 represents the height value of the current target point; a represents the ground height extreme difference of the current target point; the second preset condition is: Z0 - b < Z1 < Z0 + b, where b represents the ground height limit value of the current target point.
[0133] Optionally, from the above description, it can be seen that when the distance between two points is relatively close, the attribute of the current point can be directly determined as a ground point based on the attribute of the previous target point as a ground point and the first preset condition. Otherwise, when it is other types of situations, it is also necessary to judge whether the height value of the current point meets the second preset condition. This processing method can improve the processing accuracy while ensuring processing efficiency.
[0134] S205: Determine a point cloud dataset of the target obstacle based on the attributes of each point and the point cloud dataset.
[0135] In order to improve the accuracy of point cloud data processing, optionally, step S205 can be specifically explained as: determining the point cloud data set of the target obstacle according to the attributes of each point in the first data set to be processed, the attributes of each point in the second data set to be processed and the point cloud data set.
[0136] Optionally, the attributes of each point in the first data set to be processed can be compared with the attributes of each point in the second data set to be processed, and the points with the same attributes in the two data sets can be merged into one data set. Based on the attributes of the points in the data set, the point cloud data set of the target obstacle can be determined from the point cloud data set, and the attributes of the points of the target obstacle are non-ground points.
[0137] In order to further improve the accuracy of the point cloud data; in some feasible implementations, the above step S205 can be specifically explained as: determining the point cloud data set of the target obstacle to be processed according to the attributes of each point and the point cloud data set; using a density clustering algorithm to cluster the point cloud data set of the target obstacle to be processed to obtain the point cloud data set of the target obstacle.
[0138] The process of processing point cloud data by the density clustering algorithm can be as follows: randomly select a point f in the point cloud data set, set the neighborhood Eps and the minimum number of points in the neighborhood MinPts; thereby determine whether the point is a core point. If it is a core point, for each point h in the neighborhood of point f, if h is a core point, merge the data set of f and the data set of h. If h is a boundary point and h is not added to other clusters as a boundary point, then add h to the cluster belonging to x. If h has been added to other clusters, the algorithm continues until the neighborhoods of all points are judged, the algorithm ends, and a point cloud data set of the target obstacle is formed. Based on this density algorithm, the point cloud data set is processed without restrictions on the shape of the obstacle, that is, the point cloud data of amorphous obstacles can be clustered, thereby improving the accuracy and flexibility of clustering processing.
[0139] Optionally, based on the above processing method, in order to subsequently output a visualized vehicle driving scene that is easy for users to watch, the point cloud data set can also be attributed and clustered to extract elevated point cloud data from the point cloud data set to fit the road boundary, so that users can understand the road conditions more intuitively.
[0140] It should be noted that the present application only describes the point cloud data processing of a single acquisition device (i.e., lidar). In fact, the image data of the camera can also be fused as needed, that is, a more accurate vehicle driving scene animation can be formed based on a multi-sensory acquisition device.
[0141] In summary, since the thresholds are all empirical values, on some rugged roads (such as unstructured roads in mining areas, etc.), a single screening method cannot completely separate ground points from non-ground points; this application performs attribute judgment on the point cloud data set, that is, in the process of dividing the points into a data set of ground points and a data set of non-ground points, two screenings are performed, that is, after a screening from small to large radius, the ground point with the largest radius is used as the first point of the ray judgment, and in the order of radius from large to small, according to the rules described above, it is judged again whether each point belongs to a ground point. In this way, each point is compared with adjacent points with a smaller radius than itself, and with adjacent points with a larger radius than itself, which greatly reduces the probability of error and improves the effect of point cloud ground segmentation.
[0142] See also Figure 8 , Figure 8 The present application also discloses a point cloud data processing device, which includes:
[0143] An acquisition module 801 is used to acquire a point cloud data set, a ground height extreme difference arc, and a ground height limit arc, wherein the point cloud data set includes three-dimensional coordinates of a plurality of points, and the point cloud data is collected by a laser radar;
[0144] A height value determination module 802 is used to determine the height value of each point from the three-dimensional coordinates of the multiple points;
[0145] The attribute determination module 803 is used to determine the attribute of each point based on the height value of each point, the ground height range radian and the ground height limit radian;
[0146] The point cloud data set determining module 804 is used to determine the point cloud data set of the target obstacle based on the attribute of each point and the point cloud data set.
[0147] In some possible implementations, the attribute includes a ground point and a non-ground point;
[0148] The attribute determination module is used to determine the three-dimensional coordinates of the target point from the point cloud data set, determine the ground height limit value of the target point according to the three-dimensional coordinates of the target point and the ground height limit radian, determine the attribute of the target point according to the height value of the target point and the ground height limit value of the target point, and determine the attributes of the remaining target points based on the attribute of the target point, the height values of the remaining target points, the ground height extreme radian and the ground height limit radian.
[0149] In some feasible implementation manners, the attribute determination module is used to determine the target point as the previous target point, determine the current target point from the remaining target points; the remaining target points are target points whose attributes have not been determined, determine the ground height extreme difference value of the current target point according to the three-dimensional coordinates of the current target point, the three-dimensional coordinates of the previous target point and the ground height extreme difference radian, determine the ground height limit value of the current target point according to the three-dimensional coordinates of the current target point and the ground height limit radian, determine the distance between the current target point and the previous target point, and if the distance between the current target point and the previous target point is less than or equal to the preset threshold value, then determine the first result of the current target point according to the attribute of the previous target point, the height value of the current target point, the height value of the previous target point and the ground height extreme difference value of the current target point, determine the attribute of the current target point based on the first result, determine the current target point as the updated previous target point, and re-determine the updated current target point from the remaining target points; repeat the above steps of determining the attribute of the current target point to determine the attribute of the updated current target point; repeat the above steps of determining the updated previous target point, the updated current target point and determining the attribute of the updated current target point until the attributes of all the remaining target points are determined.
[0150] In some feasible implementations, the attribute determination module is used to determine that the first result of the current target point is a directly confirmable attribute if the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a ground point, and the height value of the current target point satisfies the first preset condition. Based on the first result, the attribute of the current target point is determined to be a ground point. If the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a non-ground point, and the height value of the current target point satisfies the first preset condition, then the first result of the current target point is determined to be an attribute that cannot be directly confirmed. Based on the first result, the height value of the current target and the second preset condition are processed to obtain a second result. If the second result is that the height value of the current target satisfies the second preset condition, then the attribute of the current target point is determined to be a ground point; otherwise, the attribute of the current target point is determined to be a non-ground point.
[0151] In some feasible implementations, the first preset condition is:
[0152] Z0 - a < Z1 < Z0 + a, where Z0 represents the height value of the previous target point; Z1 represents the height value of the current target point; a represents the ground height extreme difference value of the current target point;
[0153] The second preset condition is:
[0154] Z0 - b < Z1 < Z0 + b, where b represents the ground height limit value of the current target point.
[0155] In some feasible embodiments, the acquisition module is configured to acquire the horizontal angular resolution and the horizontally rotatable angle of the lidar;
[0156] The device further includes a serial number determination module for sub - point cloud data sets, configured to determine the number of sub - point cloud data sets according to the horizontal angular resolution and the horizontally rotatable angle, determine the horizontal deflection angle and radius of each point; the radius is the distance of each point from the lidar, and determine the serial number of the sub - point cloud data set to which each point belongs according to the horizontal deflection angle and the horizontal angular resolution of each point, and the point cloud data set includes a plurality of sub - point cloud data sets;
[0157] The height value determination module is configured to perform the following processing on each sub - point cloud data set in the plurality of sub - point cloud data sets, and process each sub - point cloud data set in the plurality of sub - point cloud data sets in parallel: sort the points in the sub - point cloud data set from smallest to largest according to the size of the radius to generate a first data set to be processed, and determine the point with the smallest radius in the first data set to be processed as the target point;
[0158] The attribute determination module is configured to sort the points in the sub - point cloud data set from largest to smallest according to the size of the radius to generate a second data set to be processed, and determine the point with the largest radius in the second data set to be processed as the target point; repeat the above steps of determining the attributes of the target point and the attributes of the remaining target points;
[0159] The point cloud data set determination module is configured to determine the point cloud data set of the target obstacle according to the attributes of each point in the first data set to be processed, the attributes of each point in the second data set to be processed, and the point cloud data set.
[0160] In some feasible embodiments, the point cloud data set determination module is configured to determine the point cloud data set of the target obstacle to be processed according to the attributes of each point and the point cloud data set, and perform clustering processing on the point cloud data set of the target obstacle to be processed by using a density clustering algorithm to obtain the point cloud data set of the target obstacle.
[0161] The method embodiments provided in the embodiments of the present application can be executed in a computer terminal, a server or a similar computing device. Taking running on a server as an example, Fig. 9 This is a hardware structure diagram of a server of an optional point cloud data processing method of the present application. Fig. 9 As shown, the server 600 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 910 (the central processing unit 910 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 930 for storing data, and one or more storage media 920 (such as one or more mass storage devices) for storing application programs 923 or data 922. Among them, the memory 930 and the storage medium 920 can be short-term storage or permanent storage. The program stored in the storage medium 920 may include one or more modules, each of which may include a series of instruction operations on the server. Furthermore, the central processing unit 910 can be configured to communicate with the storage medium 920 and execute a series of instruction operations in the storage medium 920 on the server 900. The server 900 may also include one or more power supplies 960, one or more wired or wireless network interfaces 950, one or more input and output interfaces 940, and / or one or more operating systems 921, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.
[0162] The input / output interface 940 may be used to receive or send data via a network. The specific example of the network may include a wireless network provided by a communication provider of the server 900. In one example, the input / output interface 940 includes a network adapter (Network Interface Controller, NIC), which may be connected to other network devices via a base station so as to communicate with the Internet. In one example, the input / output interface 940 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.
[0163] It can be understood by those skilled in the art that Fig. 9 The structure shown is only for illustration and does not limit the structure of the above electronic device. Fig. 9 More or fewer components as shown, or with Fig. 9 Different configurations shown.
[0164] An embodiment of the present application also provides an electronic device, which includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the point cloud data processing method as described above.
[0165] An embodiment of the present application also provides a storage medium, which can be set in a server to store at least one instruction, at least one program, a code set or an instruction set related to a point cloud data processing method in an implementation method embodiment. The at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the above-mentioned point cloud data processing method.
[0166] Optionally, in this embodiment, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.
[0167] It can be seen from the embodiments of the point cloud data processing method, device, electronic device and storage medium provided by the present application that the present application receives payment order information; determines the point cloud data processing scenario type according to the payment order information; determines the target profit sharing rule based on the point cloud data processing scenario type; performs clearing and liquidation on the first clearing entity according to the target profit sharing rule and the payment order information; the first clearing entity is the entity that needs to be cleared and liquidated in real time. In this way, the efficiency and accuracy of identification and cleaning can be improved.
[0168] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of this specification are described. Other embodiments are within the scope of the attached claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0169] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0170] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.
[0171] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
Claims
1. A point cloud data processing method, characterized in that: The following steps are involved: Acquire a point cloud data set; the point cloud data set includes three-dimensional coordinates of multiple points; the point cloud data is collected by a laser radar; Determine the height value of each point from the three-dimensional coordinates of the plurality of points; Get the ground height extreme difference arc and ground height limit arc; Determine the attribute of each point based on the height value of each point, the ground height extreme difference arc and the ground height limit arc; The attributes include ground points and non-ground points; Determine a point cloud dataset of the target obstacle based on the attribute of each point and the point cloud dataset; Wherein, determining the attribute of each point based on the height value of each point, the ground height range radian and the ground height limit radian includes: Determine the three-dimensional coordinates of the target point from the point cloud data set; Determine the ground height limit value of the target point according to the three-dimensional coordinates of the target point and the ground height limit arc; Determine the attribute of the target point according to the height value of the target point and the ground height limit value of the target point; Determine the attributes of the remaining target points based on the attributes of the target points, the height values of the remaining target points, the ground height extreme difference radian and the ground height limit radian; Wherein, determining the point cloud data of the target obstacle based on the attribute of each point and the point cloud data set includes: Determine a point cloud data set of the target obstacle to be processed according to the attribute of each point and the point cloud data set; A density clustering algorithm is used to perform clustering processing on the point cloud data set of the target obstacle to be processed to obtain a point cloud data set of the target obstacle.
2. The point cloud data processing method according to claim 1, characterized in that: The determining the attributes of the remaining target points based on the attributes of the target points, the height values of the remaining target points, the ground height range radian and the ground height limit radian includes: determining the target point as a previous target point; Determine the current target point from the remaining target points; the remaining target points are target points whose attributes have not been determined; Determine the ground height extreme difference value of the current target point according to the three-dimensional coordinates of the current target point, the three-dimensional coordinates of the previous target point and the ground height extreme difference arc; Determine the ground height limit value of the current target point according to the three-dimensional coordinates of the current target point and the ground height limit arc; Determining the distance between the current target point and the previous target point; If the distance between the current target point and the previous target point is less than or equal to a preset threshold, a first result of the current target point is determined according to the attribute of the previous target point, the height value of the current target point, the height value of the previous target point and the ground height extreme difference value of the current target point; Determine the attribute of the current target point based on the first result; Determine the current target point as the updated previous target point; Redetermine an updated current target point from the remaining target points; repeat the above steps of determining the attributes of the current target point to determine the attributes of the updated current target point; repeat the above steps of determining the updated previous target point, the updated current target point, and determining the attributes of the updated current target point until the attributes of all remaining target points are determined.
3. The point cloud data processing method according to claim 2, characterized in that: If the distance between the current target point and the previous target point is less than or equal to a preset threshold, then determine a first result of the current target point according to the attributes of the previous target point, the height value of the current target point, the height value of the previous target point, and the ground height extreme difference value of the current target point; Determining the attributes of the current target point based on the first result includes: If the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a ground point, and the height value of the current target point satisfies a first preset condition, then determine that the first result of the current target point is directly confirmable attribute; Determine the attribute of the current target point as a ground point based on the first result; If the distance between the current target point and the previous target point is less than or equal to the preset threshold, and the attribute of the previous target point is a non-ground point, and the height value of the current target point satisfies the first preset condition, then determine that the first result of the current target point is not directly confirmable attribute; Process the height value of the current target and a second preset condition based on the first result to obtain a second result; If the second result is that the height value of the current target satisfies the second preset condition, then determine the attribute of the current target point as a ground point, otherwise, determine the attribute of the current target point as a non-ground point.
4. The point cloud data processing method according to claim 3, characterized in that: The first preset condition is: Z0 - a < Z1 < Z0 + a, where Z0 represents the height value of the previous target point; Z1 represents the height value of the current target point; a represents the ground height extreme difference value of the current target point; The second preset condition is: Z0 - b < Z1 < Z0 + b, where b represents the ground height limit value of the current target point.
5. The point cloud data processing method according to claim 1, characterized in that: Before determining the three-dimensional coordinates of the target point from the point cloud dataset, it further includes: Obtain the horizontal angular resolution and the horizontally rotatable angle of the lidar; Determine the number of sub-point cloud datasets according to the horizontal angular resolution and the horizontally rotatable angle; Determine the horizontal deflection angle and radius of each point; the radius is the distance of each point from the lidar; Determine the serial number of the sub-point cloud dataset to which each point belongs according to the horizontal deflection angle of each point and the horizontal angular resolution; the point cloud dataset includes multiple sub-point cloud datasets; Determining the three-dimensional coordinates of the target point from the point cloud dataset includes: Perform the following processing on each sub-point cloud dataset in the multiple sub-point cloud datasets, and process each sub-point cloud dataset in the multiple sub-point cloud datasets in parallel: Sort the points in the sub-point cloud dataset in ascending order according to the size of the radius to generate a first dataset to be processed; Determine the point with the smallest radius in the first data set to be processed as the target point; After determining the attributes of the remaining target points based on the attributes of the target points, the height values of the remaining target points, the ground height range radian and the ground height limit radian, the method further includes: sorting the points in the sub-point cloud data set from large to small according to the radius to generate a second data set to be processed; Determine the point with the largest radius in the second data set to be processed as the target point; repeat the above steps of determining the attributes of the target point and the attributes of the remaining target points; The step of determining the point cloud data of the target obstacle based on the attribute of each point and the point cloud data set includes: A point cloud dataset of the target obstacle is determined according to the attribute of each point in the first dataset to be processed, the attribute of each point in the second dataset to be processed, and the point cloud dataset.
6. A point cloud data processing device, characterized in that: include: An acquisition module is used to acquire point cloud data sets, ground height extreme radians and ground height limit radians; The point cloud data set includes three-dimensional coordinates of a plurality of points; The point cloud data is collected by laser radar; A height value determination module, used to determine the height value of each point from the three-dimensional coordinates of the multiple points; An attribute determination module, for determining the attribute of each point based on the height value of each point, the ground height extreme difference radian and the ground height limit radian; the attribute includes ground points and non-ground points; the attribute of each point determined based on the height value of each point, the ground height extreme difference radian and the ground height limit radian, comprises: determining the three-dimensional coordinates of the target point from the point cloud data set; determining the ground height limit value of the target point according to the three-dimensional coordinates of the target point and the ground height limit radian; determining the attribute of the target point according to the height value of the target point and the ground height limit value of the target point; determining the attribute of the remaining target points based on the attribute of the target point, the height values of the remaining target points, the ground height extreme difference radian and the ground height limit radian; A point cloud data set determination module is used to determine a point cloud data set of a target obstacle based on the attributes of each point and the point cloud data set; the point cloud data set of the target obstacle determined based on the attributes of each point and the point cloud data set includes: determining a point cloud data set of the target obstacle to be processed according to the attributes of each point and the point cloud data set; clustering the point cloud data set of the target obstacle to be processed using a density clustering algorithm to obtain a point cloud data set of the target obstacle.
7. A computer device, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the point cloud data processing method as described in any one of claims 1 to 5.
8. A computer storage medium, characterized in that: The computer storage medium stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement the point cloud data processing method according to any one of claims 1 to 5.
