Point cloud data filtering method, device and laser radar

Abstract

The present application is applicable to the field of computer application technology and provides a method, device, and laser radar for filtering point cloud data. The method includes: obtaining point cloud data to be filtered; determining a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and multiple neighboring points corresponding to the target point; determining the radial distance of the target point and the radial distance of each neighboring point; determining a target filtering parameter corresponding to the target point based on the radial distance of the target point, the radial distance of each neighboring point, and the filter strength coefficient corresponding to the target point; and filtering the target point based on the target filtering parameter. Thus, the target filtering parameter can be determined by performing a simple calculation on the radial distance of points in the neighborhood corresponding to the target point, thereby reducing the computational complexity and computational time of point cloud data filtering and improving the efficiency of point cloud data filtering.

Description

Technical Field

[0001] The present application belongs to the field of computer application technology, and in particular relates to a point cloud data filtering method, device, laser radar, and computer-readable storage medium. Background Art

[0002] When acquiring point cloud data, noise points are inevitably present due to factors such as LiDAR accuracy, operator experience, environmental factors, electromagnetic wave diffraction characteristics, surface property variations of the measured object, and the data stitching and registration process. This noise can severely impact subsequent point cloud applications such as point cloud registration, feature extraction, 3D reconstruction, and visualization. Denoising point cloud data is a crucial step in point cloud data processing, serving as the foundation and key to subsequent data processing.

[0003] In related technologies, due to the large amount of point cloud data, filtering processing of massive point cloud data takes a long time and is inefficient. Summary of the Invention

[0004] The embodiments of the present application provide a point cloud data filtering method, device, laser radar and storage medium, which can solve the problem that due to the large amount of point cloud data, filtering processing of massive point cloud data is time-consuming and inefficient.

[0005] In a first aspect, an embodiment of the present application provides a method for filtering point cloud data, comprising: obtaining point cloud data to be filtered, wherein the point cloud data to be filtered includes multiple points; determining a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and multiple neighborhood points corresponding to the target point; determining the radial distance of the target point and the radial distance of each neighborhood point; determining a target filtering parameter corresponding to the target point based on the radial distance of the target point, the radial distance of each neighborhood point and the filtering strength coefficient corresponding to the target point; and filtering the target point based on the target filtering parameter.

[0006] In a possible implementation of the first aspect, before determining the target filtering parameter corresponding to the target point based on the radial distance of the target point, the radial distance of each neighboring point, and the filter strength coefficient corresponding to the target point, the method further includes:

[0007] Obtain the horizontal angle resolution and range divergence of the laser radar corresponding to the point cloud data to be filtered;

[0008] According to the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angle resolution, as well as the sequence number of each neighborhood point, the sequence number of the target point and the distance divergence, the sudden change neighborhood points and the non-sudden change neighborhood points in the neighborhood points are determined;

[0009] Accordingly, the target filter parameters corresponding to the target point are determined according to the radial distance of the target point, the radial distance of each neighboring point, and the filter strength coefficient corresponding to the target point, including:

[0010] The target filtering parameters are determined according to the radial distance of the target point, the radial distance of each non-mutation neighboring point and the filtering strength coefficient.

[0011] Optionally, in another possible implementation of the first aspect, determining the sudden change neighboring points and the non-sudden change neighboring points in the neighboring points based on the radial distance of each neighboring point, the radial distance of the target point, and the horizontal angular resolution, as well as the sequence number of each neighboring point, the sequence number of the target point, and the distance measurement divergence, includes:

[0012] According to the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angle resolution, the distance between each neighborhood point and the target point is determined respectively;

[0013] According to the radial distance of the target point, the difference between the serial number of the target point and the serial number of each neighboring point, and the distance divergence, the mutation distance threshold corresponding to each neighboring point is determined respectively;

[0014] If the distance between any neighboring point and the target point is greater than the mutation distance threshold corresponding to any neighboring point, then any neighboring point is determined as a mutation neighboring point;

[0015] If the distance between any neighborhood point and the target point is less than or equal to the mutation distance threshold corresponding to any neighborhood point, then any neighborhood point is determined as a non-mutation neighborhood point.

[0016] Optionally, in another possible implementation of the first aspect, the target filter parameters include a first target filter parameter and a second target filter parameter; accordingly, determining the target filter parameters based on the radial distance of the target point, the radial distance of each non-mutation neighboring point, and the filter strength coefficient includes:

[0017] According to the radial distance of the target point and the radial distance of each non-mutation neighboring point, the mean radial distance and the mean square of the radial distance corresponding to the target neighborhood are determined;

[0018] Determine a first target filtering parameter according to the radial distance mean, the radial distance square mean and the filtering strength coefficient;

[0019] A second target filtering parameter is determined according to the first target filtering parameter and the radial distance mean.

[0020] Optionally, in yet another possible implementation of the first aspect, the filtering process on the target point according to the target filtering parameter includes:

[0021] determining the product of the radial distance of the target point and the first target filter parameter;

[0022] The product of the radial distance of the target point and the first target filtering parameter and the sum of the second target filtering parameter are determined as the filtered radial distance corresponding to the target point.

[0023] Optionally, in yet another possible implementation of the first aspect, determining a target neighborhood corresponding to a target point in the point cloud data to be filtered includes:

[0024] Determine the number of target lines and the sequence number of the target point corresponding to the target point;

[0025] The (n-1) / 2 points in the target line number that are adjacent to the target point and whose sequence number is smaller than the sequence number of the target point are determined as neighborhood points, and the (n-1) / 2 points in the target line number that are adjacent to the target point and whose sequence number is larger than the sequence number of the target point are determined as neighborhood points, where n is the size of the target neighborhood.

[0026] Optionally, in another possible implementation of the first aspect, before determining the target filtering parameter corresponding to the target point based on the radial distance of the target point, the radial distance of each neighboring point, and the filter strength coefficient corresponding to the target point, the method further includes:

[0027] The filter strength coefficient is determined based on whether the target neighborhood contains corner points.

[0028] Optionally, in another possible implementation of the first aspect, determining the filter strength coefficient according to whether the target neighborhood includes a corner point includes:

[0029] When the target neighborhood contains corner points, the filter strength coefficient is determined based on the distance between each adjacent point in the target neighborhood;

[0030] When the target neighborhood does not contain a corner point, the filter strength coefficient is determined to be a preset value.

[0031] Optionally, in yet another possible implementation of the first aspect, before determining the filter strength coefficient based on whether the target neighborhood includes a corner point, the method further includes:

[0032] According to the radial distance of each point in the target neighborhood, the radial distance difference between adjacent points in the target neighborhood is determined;

[0033] If the radial distance differences between adjacent points in the target neighborhood are all greater than or equal to 0 or less than or equal to 0, it is determined that the target neighborhood does not contain a corner point; otherwise, it is determined that the target neighborhood contains a corner point.

[0034] In the second aspect, an embodiment of the present application provides a filtering device for point cloud data, including: a first acquisition module, used to acquire point cloud data to be filtered, wherein the point cloud data to be filtered includes multiple points; a first determination module, used to determine the target neighborhood corresponding to the target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and multiple neighborhood points corresponding to the target point; a second determination module, used to determine the radial distance of the target point and the radial distance of each neighborhood point; a third determination module, used to determine the target filtering parameters corresponding to the target point based on the radial distance of the target point, the radial distance of each neighborhood point and the filtering strength coefficient corresponding to the target point; a first filtering module, used to filter the target point according to the target filtering parameters.

[0035] In a possible implementation of the second aspect, the apparatus further includes:

[0036] The second acquisition module is used to obtain the horizontal angle resolution and range divergence of the laser radar corresponding to the point cloud data to be filtered;

[0037] The fourth determination module is used to determine the sudden change neighboring points and the non-sudden change neighboring points in the neighborhood points according to the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angle resolution, as well as the sequence number of each neighborhood point, the sequence number of the target point and the distance measurement divergence;

[0038] Accordingly, the third determining module includes:

[0039] The first determining unit is used to determine the target filtering parameter according to the radial distance of the target point, the radial distance of each non-mutation neighboring point and the filtering strength coefficient.

[0040] Optionally, in another possible implementation of the second aspect, the fourth determining module includes:

[0041] a second determining unit, configured to determine the distance between each neighborhood point and the target point according to the radial distance of each neighborhood point, the radial distance of the target point, and the horizontal angular resolution;

[0042] The third determining unit is used to determine the mutation distance threshold corresponding to each neighboring point according to the radial distance of the target point, the difference between the sequence number of the target point and the sequence number of each neighboring point, and the distance measurement divergence;

[0043] a fourth determining unit, configured to determine any neighborhood point as a sudden change neighborhood point if the distance between any neighborhood point and the target point is greater than a sudden change distance threshold corresponding to the any neighborhood point;

[0044] The fifth determining unit is configured to determine any neighborhood point as a non-mutation neighborhood point if the distance between any neighborhood point and the target point is less than or equal to a mutation distance threshold corresponding to any neighborhood point.

[0045] Optionally, in another possible implementation of the second aspect, the target filtering parameter includes a first target filtering parameter and a second target filtering parameter; accordingly, the first determining unit is specifically configured to:

[0046] According to the radial distance of the target point and the radial distance of each non-mutation neighboring point, the mean radial distance and the mean square of the radial distance corresponding to the target neighborhood are determined;

[0047] Determine a first target filtering parameter according to the radial distance mean, the radial distance square mean and the filtering strength coefficient;

[0048] A second target filtering parameter is determined according to the first target filtering parameter and the radial distance mean.

[0049] Optionally, in yet another possible implementation of the second aspect, the first filtering module includes:

[0050] a sixth determining unit, configured to determine a product of a radial distance of the target point and a first target filtering parameter;

[0051] The seventh determining unit is configured to determine the filtered radial distance corresponding to the target point by taking the sum of the product of the radial distance of the target point and the first target filtering parameter and the second target filtering parameter.

[0052] Optionally, in yet another possible implementation of the second aspect, the first determining module includes:

[0053] An eighth determining unit, configured to determine the number of target lines corresponding to the target point and the sequence number of the target point;

[0054] The ninth determination unit is used to determine the (n-1) / 2 points in the target line number that are adjacent to the target point and whose serial numbers are smaller than the serial number of the target point as neighborhood points, and to determine the (n-1) / 2 points in the target line number that are adjacent to the target point and whose serial numbers are larger than the serial number of the target point as neighborhood points, where n is the size of the target neighborhood.

[0055] Optionally, in another possible implementation of the second aspect, the apparatus further includes:

[0056] The fifth determination module is used to determine the filter strength coefficient according to whether the target neighborhood contains a corner point.

[0057] Optionally, in yet another possible implementation of the second aspect, the fifth determining module includes:

[0058] a tenth determining unit, configured to determine a filter strength coefficient based on distances between adjacent points in the target neighborhood when a corner point is included in the target neighborhood;

[0059] The eleventh determining unit is configured to determine the filter strength coefficient as a preset value when the target neighborhood does not include a corner point.

[0060] Optionally, in yet another possible implementation of the second aspect, the apparatus further includes:

[0061] a sixth determining module, configured to determine a radial distance difference between adjacent points in the target neighborhood based on the radial distance of each point in the target neighborhood;

[0062] The seventh determination module is configured to determine that the target neighborhood does not contain a corner point if the radial distance differences between adjacent points in the target neighborhood are all greater than or equal to 0 or are all less than or equal to 0; otherwise, determine that the target neighborhood contains a corner point.

[0063] In a third aspect, an embodiment of the present application provides a laser radar, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the point cloud data filtering method described above when executing the computer program.

[0064] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the point cloud data filtering method as described above is implemented.

[0065] In a fifth aspect, an embodiment of the present application provides a computer program product, which, when running on a laser radar, enables the laser radar to perform the point cloud data filtering method as described above.

[0066] Compared with the prior art, the beneficial effects of the embodiments of the present application are as follows: by determining the target filtering parameters of each point in the point cloud data according to the radial distance of each point in the neighborhood corresponding to each point in the point cloud data, the point cloud data is filtered to improve the smoothing effect of the point cloud data. Since the target filtering parameters can be determined by performing a simple calculation on the radial distance of the points in the neighborhood, the computational complexity and computational time of the point cloud data filtering are reduced, thereby improving the efficiency of the point cloud data filtering. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. 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 any creative work.

[0068] Figure 1 1 is a flow chart of a method for filtering point cloud data provided by an embodiment of the present application;

[0069] Figure 2 This is a schematic diagram of a partial structure of point cloud data provided by an embodiment of the present application;

[0070] Figure 3 is a flowchart of a method for filtering point cloud data provided by another embodiment of the present application;

[0071] Figure 4 is a schematic diagram of the positional relationship between a target point and neighboring points provided in an embodiment of the present application;

[0072] Figure 5 This is a schematic diagram of another positional relationship between a target point and neighboring points provided by an embodiment of the present application;

[0073] Figure 6 1 is a flow chart of a method for filtering point cloud data provided in yet another embodiment of the present application;

[0074] Figure 7 This is a schematic diagram of a target neighborhood in which no corner points exist, provided by an embodiment of the present application;

[0075] Figure 8 This is a schematic diagram of another target neighborhood provided by an embodiment of the present application in which no corner points exist;

[0076] Figure 9 This is a schematic diagram of a target neighborhood with corner points provided by an embodiment of the present application;

[0077] Figure 10 This is a schematic diagram showing another example of a target neighborhood having corner points provided by an embodiment of the present application;

[0078] Figure 11 This is a schematic diagram of a corner point existing in another target neighborhood provided by an embodiment of the present application;

[0079] Figure 12 Schematic diagram of the structure of the point cloud data filtering device provided in an embodiment of the present application;

[0080] Figure 13 It is a schematic diagram of the structure of the laser radar provided in an embodiment of the present application. DETAILED DESCRIPTION

[0081] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0082] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0083] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0084] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0085] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0086] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0087] The point cloud data filtering method, device, laser radar, storage medium and computer program provided in this application are described in detail below with reference to the accompanying drawings.

[0088] Figure 1 A flow chart of a point cloud data filtering method provided in an embodiment of the present application is shown.

[0089] Step 101: Acquire point cloud data to be filtered, wherein the point cloud data to be filtered includes a plurality of points.

[0090] It should be noted that the point cloud data filtering method of the embodiment of the present application can be performed by the point cloud data filtering device of the embodiment of the present application. The point cloud data filtering device of the embodiment of the present application can be configured in any laser radar to perform the point cloud data filtering method of the embodiment of the present application, filter the point cloud data generated by the laser radar detecting the environment, and improve the smoothing effect of the point cloud data.

[0091] The point cloud data to be filtered may refer to point cloud data generated by any type of laser radar detecting its environment.

[0092] In an embodiment of the present application, the point cloud data to be filtered can be the point cloud data generated in real time by the laser radar during operation. Every time the laser radar generates a new frame of point cloud data, the point cloud data can be used as the point cloud data to be filtered, so as to filter each frame of point cloud data generated by the laser radar and improve the smoothing effect of the point cloud data.

[0093] For example, assuming that the laser radar is installed in a vehicle, the point cloud data to be filtered can be point cloud data generated in real time by the laser radar scanning environmental information while the vehicle is driving.

[0094] In an embodiment of the present application, the point cloud data to be filtered may include a large number of points, and the data corresponding to each point may include one or more of the information such as the distance between the spatial position represented by the point and the laser radar, three-dimensional coordinates, color, light intensity, reflectivity, etc. The embodiment of the present application does not limit this.

[0095] Step 102 : determining a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and a plurality of neighboring points corresponding to the target point.

[0096] The target neighborhood may refer to the neighborhood corresponding to the target point in the point cloud data to be filtered. It should be noted that the center point of the target neighborhood is the target point.

[0097] It should be noted that when filtering the point cloud data to be filtered, each point in the point cloud data to be filtered can be determined as a target point in turn, so that each point in the point cloud data to be filtered can be filtered in turn to achieve filtering of the entire point cloud data to be filtered.

[0098] In the embodiment of the present application, since a point in the point cloud data is generally very close to the points in its neighborhood, the points in the neighborhood generally have a high similarity. Therefore, if there are points with large differences in the neighborhood, it means that there may be noise points here. Therefore, by filtering the corresponding points through the points in the neighborhood corresponding to the point, the noise points can be removed, and the smoothing effect and quality of the point cloud data can be improved. Therefore, when filtering the point cloud data to be filtered, for a target point, the target point can be filtered based on the data of the surrounding points of the target point, so that the point cloud data can transition smoothly and remove noise points in the point cloud data. Therefore, when filtering a target point, the target neighborhood corresponding to the target point can be determined, so that the target point can be filtered based on the information of the points in the target neighborhood.

[0099] It should be noted that in actual use, the target neighborhood corresponding to the target point can be determined through a binary tree, KD tree, octree and other structures, or the target neighborhood corresponding to the target point can be determined through a k-nearest neighbor search, an R radius search or a hybrid search method. The embodiments of the present application do not limit this.

[0100] Furthermore, since point cloud data is usually an ordered point cloud, for example, point cloud data generated by a laser radar can be arranged by line number and point sequence, the target neighborhood can be directly determined from the point cloud data to be filtered based on the line number and sequence number of the target point, thereby further reducing computational complexity. That is, in one possible implementation of the embodiment of the present application, the above step 102 may include:

[0101] Determine the number of target lines and the sequence number of the target point corresponding to the target point;

[0102] The (n-1) / 2 points in the target line number that are adjacent to the target point and whose sequence number is smaller than the sequence number of the target point are determined as neighborhood points, and the (n-1) / 2 points in the target line number that are adjacent to the target point and whose sequence number is larger than the sequence number of the target point are determined as neighborhood points, where n is the size of the target neighborhood.

[0103] The target line number may refer to the line number where the target point is located in the point cloud data to be filtered; and the sequence number of the target point may refer to the sequence number corresponding to the target point in the target line number.

[0104] The size of the target neighborhood may refer to the number of points contained in the target neighborhood.

[0105] As a possible implementation, take a 16-line mechanical laser radar as an example. This laser radar can emit 16 laser lines each time, and the data detected by each laser line can correspond to a point in the point cloud data. Assuming that the laser radar needs to emit lasers N times to generate a frame of point cloud data, the point cloud data generated by the laser radar can include 16 lines, each line includes N points, and the points in each line are arranged in the order of acquisition time. Therefore, when determining the target neighborhood, based on the pre-set target neighborhood size n, n-1 points before and after the target point in the target line are selected as the neighborhood points of the target point to generate the target neighborhood.

[0106] For example, if Figure 2 As shown, it is a partial structural diagram of a point cloud data provided by an embodiment of the present application. Assuming that the size of the target neighborhood is 5, if the target point is the point with sequence number 3 in line number 1, the points with sequence numbers 1, 2, 4, and 5 in line number 1 can be determined as neighborhood points, that is, the target neighborhood includes the points with sequence numbers 1, 2, 3, 4, and 5 in line number 1, among which the point with sequence number 3 is the center point; if the target point is the point with sequence number 4 in line number 2, the points with sequence numbers 2, 3, 5, and 6 in line number 2 can be determined as neighborhood points, that is, the target neighborhood includes the points with sequence numbers 2, 3, 4, 5, and 6 in line number 2, among which the point with sequence number 4 is the center point.

[0107] Step 103: Determine the radial distance of the target point and the radial distance of each neighboring point.

[0108] The radial distance may refer to the distance between the spatial position corresponding to the point in the point cloud data to be filtered measured by the laser radar and the laser radar.

[0109] In an embodiment of the present application, since the point cloud data to be filtered can include data corresponding to each point, and the data corresponding to each point can include the radial distance of the point, after determining any point as a target point and determining the target neighborhood corresponding to the target point, the radial distance corresponding to the target point can be obtained from the data corresponding to the target point, and the radial distance corresponding to the neighborhood point can be obtained from the data corresponding to the neighborhood point.

[0110] Step 104 : determining a target filtering parameter corresponding to the target point according to the radial distance of the target point, the radial distance of each neighboring point, and the filtering strength coefficient corresponding to the target point.

[0111] As a possible implementation method, the point cloud data to be filtered can be filtered by a guided graph filtering method, so that the target filtering parameters can include a first filtering parameter and a second filtering parameter, and the first filtering parameter and the second filtering parameter can be determined respectively by the following formulas:

[0112]

[0113]

[0114] Among them, a i is the first filter parameter corresponding to the target point, i is the serial number of the target point, n is the size of the target neighborhood, p j is the radial distance of the jth point in the target neighborhood, j is the sequence number of the point in the target neighborhood, is the mean radial distance of each point in the target neighborhood, and ε is the filter strength coefficient corresponding to the target point.

[0115]

[0116] Among them, b i is the second filtering parameter corresponding to the target point.

[0117] It should be noted that the filter strength coefficient corresponding to a target point can indicate the degree of filtering applied to that target point. In practice, the specific value of the preset filter strength coefficient can be determined based on actual needs and specific application scenarios. Alternatively, the filter strength coefficient of the target point can be determined in real time based on its characteristics to further improve the accuracy and reliability of filtering and enhance the quality of the point cloud.

[0118] Step 105: Filter the target point according to the target filtering parameters.

[0119] In an embodiment of the present application, the radial distance of the target point may be filtered according to the target filtering parameter to generate a filtered radial distance corresponding to the target point.

[0120] As a possible implementation method, when filtering the point cloud data to be filtered using the guided graph filtering method, the radial distance after filtering corresponding to the target point can be determined by the following formula based on the determined first target filtering parameters and the second target filtering parameters:

[0121] p′ i =a i ·p i +b i

[0122] Among them, p′ i is the filtered radial distance corresponding to the target point, p i is the radial distance of the target point, a i is the first filter parameter corresponding to the target point, b i is the second filtering parameter corresponding to the target point.

[0123] It should be noted that, in actual use, each point in the point cloud data to be filtered is determined as a target point in turn, and each target point is filtered in turn according to the above steps to generate a filtering result for the point cloud data to be filtered.

[0124] The point cloud data filtering method provided in the embodiment of the present application determines the target neighborhood corresponding to the target point in the point cloud data to be filtered, and determines the radial distance of the target point and the radial distance of each neighborhood point. Then, based on the radial distance of the target point, the radial distance of each neighborhood point, and the filter strength coefficient corresponding to the target point, the target filter parameter corresponding to the target point is determined, and then the target point is filtered according to the target filter parameter. Thus, by determining the target filter parameter for each point in the point cloud data based on the radial distance of each point in the neighborhood corresponding to each point in the point cloud data, the point cloud data is filtered, thereby improving the smoothing effect of the point cloud data. Since the target filter parameter can be determined by performing a simple calculation on the radial distance of the points in the neighborhood, the computational complexity and computational time of the point cloud data filtering are reduced, thereby improving the efficiency of the point cloud data filtering.

[0125] In a possible implementation form of the present application, when there are mutated noise points in the target neighborhood of the target point, if the mutated noise points are used for filtering, points that did not originally exist will be generated in the filtered point cloud data, thereby affecting the accuracy and reliability of the filtering. Therefore, when filtering the target point, the mutated neighborhood points in the target neighborhood can be removed to further improve the accuracy and reliability of the filtering, and further improve the quality of the point cloud.

[0126] The following combination Figure 3 , the filtering method of point cloud data provided in the embodiment of the present application is further explained.

[0127] Figure 3 A flow chart of another method for filtering point cloud data provided in an embodiment of the present application is shown.

[0128] like Figure 3 As shown, the filtering method of the point cloud data includes the following steps:

[0129] Step 201 : obtaining point cloud data to be filtered, wherein the point cloud data to be filtered includes a plurality of points.

[0130] Step 202 : determining a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and a plurality of neighboring points corresponding to the target point.

[0131] Step 203: Determine the radial distance of the target point and the radial distance of each neighboring point.

[0132] The specific implementation process and principles of the above steps 201-203 can be referred to the detailed description of the above embodiment and will not be repeated here.

[0133] Step 204 : Obtain the horizontal angular resolution and range divergence of the laser radar corresponding to the point cloud data to be filtered.

[0134] The horizontal angular resolution of the laser radar may refer to the angle between two adjacent points belonging to the same line number in the point cloud data to be filtered.

[0135] Among them, the ranging divergence of the lidar can be used to indicate the degree of discreteness of multiple measured distances obtained by the lidar for multiple detections of the object at the same distance; the ranging divergence can measure the ranging accuracy of the lidar.

[0136] In an embodiment of the present application, the horizontal angular resolution and ranging divergence of the laser radar are inherent performance parameters of the laser radar itself. Therefore, the horizontal angular resolution and ranging divergence of the laser radar can be pre-stored in a storage device of the laser radar, or can be pre-stored in a device that filters the point cloud data to be filtered generated by the laser radar, so that when the horizontal angular resolution and ranging divergence of the laser radar are needed, the pre-stored horizontal angular resolution and ranging divergence can be directly read.

[0137] Step 205 , determining the sudden change neighborhood points and the non-sudden change neighborhood points in the neighborhood points according to the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angle resolution, as well as the sequence number of each neighborhood point, the sequence number of the target point and the distance measurement divergence.

[0138] Among them, the mutation neighborhood point can refer to the neighborhood point in the target neighborhood whose distance to the target point is greater than a certain threshold; the non-mutation neighborhood point can refer to the neighborhood point in the target neighborhood whose distance to the target point is less than or equal to a certain threshold.

[0139] In an embodiment of the present application, for a neighborhood point, the distance between the neighborhood point and the target point can be determined based on the radial distance of the neighborhood point, the radial distance of the target point, and the horizontal angular resolution of the laser radar. Then, based on the sequence number of the neighborhood point and the target point, the radial distance of the target point, and the distance divergence of the laser radar, the mutation distance threshold corresponding to the neighborhood point is determined. Then, based on the relationship between the distance between the neighborhood point and the target point and the mutation distance threshold, it is determined whether the neighborhood point is a mutation neighborhood point. That is, in a possible implementation of the embodiment of the present application, the above step 205 may include:

[0140] According to the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angle resolution, the distance between each neighborhood point and the target point is determined respectively;

[0141] According to the radial distance of the target point, the difference between the serial number of the target point and the serial number of each neighboring point, and the distance divergence, the mutation distance threshold corresponding to each neighboring point is determined respectively;

[0142] If the distance between any neighboring point and the target point is greater than the mutation distance threshold corresponding to any neighboring point, then any neighboring point is determined as a mutation neighboring point;

[0143] If the distance between any neighborhood point and the target point is less than or equal to the mutation distance threshold corresponding to any neighborhood point, then any neighborhood point is determined as a non-mutation neighborhood point.

[0144] As a possible implementation, Figure 4 As shown in FIG. 1 , a schematic diagram of the positional relationship between a target point and neighboring points provided by an embodiment of the present application is shown, where it is assumed that point O is the coordinate origin (i.e., the location of the laser radar), point B is the target point, the serial number is i, and the radial distance OB of the target point B is p. i , points A, C, and D are some neighborhood points of the target point B. The serial numbers of points A, C, and D are i-1, i+1, and i+2 respectively, and the radial distances OA, OC, and OD are p respectively. i-1 、p i+1 、p i+2 , θ is the horizontal angular resolution of the laser radar, that is, the angle between the radial distances of adjacent points. Since the radial distances between the target point B and each neighboring point are known, the angles between the target point B and each neighboring point are also known. Therefore, the distances AB, BC, and BD between the target point B and each neighboring point can be determined according to the law of cosines, that is:

[0145]

[0146] Where Δr ij is the distance between the target point B and the jth neighboring point, i is the serial number of the target point B, j is the serial number of the neighboring point, p i is the radial distance to the target point B, p j is the radial distance of the jth neighborhood point, and θ is the horizontal angular resolution.

[0147] As a possible implementation method, since each point in the target neighborhood is a plurality of points continuously scanned by the lidar, and in order to achieve a certain detection accuracy, the horizontal angle resolution is usually a very small value, such as 0.01, 0.001, etc., therefore, each point in the target neighborhood is usually a very close point in the actual space. Therefore, under the premise that no abnormal situation such as mutation occurs, it can be considered that each point in the target neighborhood is in the same plane, and then the conditions that need to be met for each point in the target neighborhood to be in the same plane can be used to deduce the mutation distance threshold corresponding to each neighborhood point.

[0148] like Figure 5 As shown in FIG. 1 , another schematic diagram of the positional relationship between a target point and a neighboring point provided by an embodiment of the present application is shown, where it is assumed that point O is the coordinate origin (i.e., the location of the laser radar), point B is the target point, the serial number is i, and the radial distance OB of the target point B is p. i , points A, C, and D are some neighborhood points of the target point B. The serial numbers of points A, C, and D are i-1, i+1, and i+2 respectively, and the radial distances OA, OC, and OD are p respectively. i-1 、p i+1 、p i+2 , θ is the horizontal angle resolution of the laser radar, OP is perpendicular to AC, α is the angle between OP and OB, and if the length of OP is L, then according to Figure 5 The position shown and the cosine theorem give:

[0149] Δr ii =L·tan(α+|ij|·θ)-L·tanα

[0150] L=p i cosα, α∈[0,θ]

[0151] Right now:

[0152] Δr ij =p i ·cosα[tan(α+|ij|·θ)-tanα]

[0153] Where Δr ij is the distance between the target point B and the jth neighboring point, i is the serial number of the target point B, j is the serial number of the neighboring point, p i is the radial distance to the target point B.

[0154] Thus, when the jth neighborhood point is a non-mutation neighborhood point, Δr can be determined ij The following conditions must be met:

[0155] Δr ij ≤m·p i ·|ij|

[0156] That is to say, the mutation distance threshold corresponding to the jth neighborhood point can be determined as m·p i ·|ij|, where m is a constant. m can be determined based on the range of cosα[tan(α+|ij|·θ)-tanα] when the condition that each neighborhood point and the target point are in the same plane is satisfied.

[0157] As a possible implementation method, since there is usually a certain error in the distance measurement of the laser radar, the distance measurement divergence of the laser radar can be introduced into the mutation distance threshold to further improve the accuracy of the determination of the mutation neighborhood point. Thus, the mutation distance threshold corresponding to the jth neighborhood point can be determined as m·p i |ij| + Δd, where Δd is the range divergence of the lidar.

[0158] That is, the distance Δr between the jth neighboring point and the target point can be ij ≤m·p i When |ij|+Δd, the jth neighboring point is determined to be a non-mutation neighboring point; when the distance Δr between the jth neighboring point and the target point is ij >m·p i When |ij|+Δd, the jth neighborhood point is determined as the mutation neighborhood point.

[0159] Step 206: Determine target filtering parameters based on the radial distance of the target point, the radial distance of each non-mutation neighboring point, and the filtering strength coefficient.

[0160] In an embodiment of the present application, after the sudden change neighborhood points in the target neighborhood are determined, the sudden change neighborhood points can be removed, and the target filtering parameters corresponding to the target point are determined only based on the non-sudden change neighborhood points in the target neighborhood.

[0161] Furthermore, when filtering the point cloud data to be filtered using the guided graph filtering method, the target filtering parameters may include a first filtering parameter and a second filtering parameter, so that the first filtering parameter and the second filtering parameter can be determined based on the radial distance of the target point, the radial distance of each non-mutation neighboring point, and the filtering strength coefficient, respectively, to further improve the accuracy of the filtering parameter determination, thereby improving the smoothing effect of the point cloud data. That is, in a possible implementation of the embodiment of the present application, the above step 206 may include:

[0162] According to the radial distance of the target point and the radial distance of each non-mutation neighboring point, the mean radial distance and the mean square of the radial distance corresponding to the target neighborhood are determined;

[0163] Determine a first target filtering parameter according to the radial distance mean, the radial distance square mean and the filtering strength coefficient;

[0164] A second target filtering parameter is determined according to the first target filtering parameter and the radial distance mean.

[0165] As a possible implementation, the first filtering parameter and the second filtering parameter can be determined by the following formulas:

[0166]

[0167]

[0168] Among them, a i is the first filter parameter corresponding to the target point, i is the serial number of the target point, k is the sum of the number of the target point and the non-mutation neighboring points, p j is the radial distance of the jth point in the target neighborhood, j is the sequence number of the target point and the non-mutation neighborhood point in the target neighborhood, is the mean radial distance between the target point and each non-mutation neighborhood point in the target neighborhood, and ε is the filter strength coefficient corresponding to the target point.

[0169]

[0170] Among them, b i is the second filtering parameter corresponding to the target point.

[0171] It should be noted that, for other implementation processes and principles of step 206, reference can be made to the detailed description of the above embodiment, which will not be repeated here.

[0172] Step 207: Filter the target point according to the target filtering parameters.

[0173] As a possible implementation, when filtering the point cloud data to be filtered using the guided graph filtering method, the radial distance of the target point can be filtered based on the first and second filtering parameters determined for the guided graph filtering of the target point. That is, in a possible implementation of the embodiment of the present application, the above step 207 can include:

[0174] The product of the radial distance of the target point and the first target filter parameter is determined.

[0175] The product of the radial distance of the target point and the first target filtering parameter and the sum of the second target filtering parameter are determined as the filtered radial distance corresponding to the target point.

[0176] As a possible implementation method, after determining the first filtering parameter and the second filtering parameter for performing the guide map filtering on the target point, the filtered radial distance corresponding to the target point can be determined by the following formula:

[0177] p′ i =a i ·p i +b i

[0178] Among them, p′ i is the filtered radial distance corresponding to the target point, p i is the radial distance of the target point, a iis the first filter parameter corresponding to the target point, b i is the second filtering parameter corresponding to the target point.

[0179] It should be noted that, for other specific implementation processes and principles of step 207, reference can be made to the detailed description of the above embodiment, which will not be repeated here.

[0180] The filtering method for point cloud data provided in an embodiment of the present application determines the target neighborhood corresponding to the target point in the point cloud data to be filtered, and the radial distance of the target point in the target neighborhood and the radial distance of each neighborhood point, and then obtains the horizontal angular resolution and ranging divergence of the laser radar corresponding to the point cloud data to be filtered, and determines the sudden change neighborhood points and non-sudden change neighborhood points in the neighborhood points based on the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angular resolution, as well as the serial number of each neighborhood point, the serial number of the target point and the ranging divergence, and then determines the target filtering parameters based on the radial distance of the target point, the radial distance of each non-sudden change neighborhood point and the filtering strength coefficient, so as to filter the target point according to the target filtering parameters. Therefore, when filtering the point cloud data, when there are mutation neighborhood points in the target neighborhood corresponding to the target point, the data of the mutation neighborhood points are not used to filter the target point, thereby reducing the impact of mutation noise points on point cloud filtering, and reducing the probability of originally existing points in the filtered point cloud data. This not only reduces the computational complexity and computational time of point cloud data filtering, improves the efficiency of point cloud data filtering, but also further improves the accuracy and reliability of point cloud data filtering, and further improves the quality and smoothing effect of the filtered point cloud.

[0181] In one possible implementation of the present application, if the same filter strength is used for each point when filtering point cloud data, it is easy to cause corner points in the point cloud data to be over-smoothed, resulting in the inability to fully preserve the angular characteristics of the point cloud data, thereby affecting the smoothing effect of the point cloud data. Therefore, in an embodiment of the present application, the filter strength coefficient can be determined based on whether there are corner points in the neighborhood to further improve the accuracy and reliability of point cloud data filtering and further improve the smoothing effect of the point cloud.

[0182] The following combination Figure 6 , the filtering method of point cloud data provided in the embodiment of the present application is further explained.

[0183] Figure 6 A flow chart of another method for filtering point cloud data provided in an embodiment of the present application is shown.

[0184] like Figure 6 As shown, the filtering method of the point cloud data includes the following steps:

[0185] Step 301: Acquire point cloud data to be filtered, wherein the point cloud data to be filtered includes a plurality of points.

[0186] Step 302 : determining a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and a plurality of neighboring points corresponding to the target point.

[0187] Step 303: Determine the radial distance of the target point and the radial distance of each neighboring point.

[0188] The specific implementation process and principles of the above steps 301-303 can be referred to the detailed description of the above embodiment and will not be repeated here.

[0189] Step 304: Determine the filter strength coefficient based on whether the target neighborhood contains a corner point.

[0190] In the embodiment of the present application, when the point cloud data to be filtered is subjected to the guide map filtering, the first target filtering parameter a is calculated according to the description in the above embodiment. i , the second target filter parameter b i And the filtered radial distance p′ corresponding to the target point i From the formula, we can see that the smaller the filter strength coefficient corresponding to the target point, the i The closer to 1, the better i The closer it is to 0, and the radial distance p′ after filtering i The closer to the target point the radial distance p i That is to say, the smaller the filter strength coefficient corresponding to the target point, the smaller the filter strength for the target point, and the target point can retain its original features to a large extent. Therefore, when it is determined that there is a corner point in the target neighborhood corresponding to the target point, the filter strength coefficient corresponding to the target point can be determined to be a smaller value to retain the corner point features of the point cloud data; when it is determined that there is no corner point in the target neighborhood corresponding to the target point, the filter strength coefficient corresponding to the target point can be determined to be a larger value to achieve a better smoothing effect for the target point.

[0191] Furthermore, the filter strength coefficient corresponding to the target point can be determined based on the distance between each adjacent point in the target neighborhood, so as to adaptively adjust the filter strength coefficient according to the characteristics of each point itself to further improve the accuracy of point cloud filtering. That is, in a possible implementation of the embodiment of the present application, the above step 304 may include:

[0192] When the target neighborhood contains corner points, the filter strength coefficient is determined based on the distance between each adjacent point in the target neighborhood;

[0193] When the target neighborhood does not contain a corner point, the filter strength coefficient is determined to be a preset value.

[0194] As a possible implementation method, for point cloud areas where there are no corner points, a unified filter strength coefficient can be used for filtering so that the point cloud data as a whole can present a more uniform smoothing effect. Therefore, when it is determined that the target neighborhood corresponding to the target point does not contain a corner point, the filter strength coefficient can be determined as a preset value to use the default filter strength parameter to filter the target point.

[0195] Accordingly, for point cloud regions containing corner points, the filter strength coefficients corresponding to the points in the region can be adaptively determined based on the corner point characteristics in the region, thereby further ensuring the smoothing effect of the corner points while preserving the corner point characteristics. Therefore, when it is determined that the target neighborhood corresponding to the target point contains a corner point, the filter strength coefficient corresponding to the target point can be determined based on the radial distance difference between each adjacent point in the target neighborhood.

[0196] As an example, the filter strength coefficient corresponding to the target point can be determined by the following formula:

[0197]

[0198] Δp jj+1 =p j -p j+1

[0199] Among them, ε is the filter strength coefficient corresponding to the target point, 2l+1 is the size of the target neighborhood corresponding to the target point, Δp jj+1 is the radial distance difference between the jth point and the j+1th point in the target neighborhood, p j is the radial distance of the jth point in the target neighborhood, p j+1 is the radial distance to the j+1th point in the target neighborhood, i is the serial number of the target point, and j is the serial number of each point in the target neighborhood.

[0200] It should be noted that the radial distance difference Δp between any two adjacent points in the target neighborhood is jj+1 , refers to the difference in radial distances between adjacent points.

[0201] Furthermore, for a point in the target neighborhood, whether the point is a corner point can be determined based on whether the radial distance differences between the adjacent points on both sides of the point and the point have different signs. That is, in a possible implementation of the embodiment of the present application, before the above step 304, the following steps may also be included:

[0202] Based on the radial distance of each point in the target neighborhood, determine the radial distance difference between each adjacent point in the target neighborhood

[0203] If the radial distance differences between adjacent points in the target neighborhood are all greater than or equal to 0 or less than or equal to, it is determined that the target neighborhood does not contain a corner point; otherwise, it is determined that the target neighborhood contains a corner point.

[0204] As a possible implementation method, if the radial distance differences between adjacent points in the target neighborhood are greater than or equal to 0, it can be said that the radial distance between each point in the target neighborhood and the laser radar is gradually decreasing, such as Figure 7 As shown; if the radial distance differences between adjacent points in the target neighborhood are less than or equal to 0, it can be said that the radial distance between each point in the target neighborhood and the laser radar is gradually increasing, as shown in Figure 8 As shown; Therefore, in these two cases, it can be determined that there is no corner point in the target neighborhood; if the target neighborhood has both adjacent points with radial distance differences greater than 0 and adjacent points with radial distance differences less than 0, it can be said that there is at least one point in the target neighborhood with a radial distance difference greater than 0 between it and one of its adjacent points, and a radial distance difference less than 0 between it and another adjacent point, so it can be determined that there is a corner point in the target neighborhood, as shown in Figure 9 shown.

[0205] For example, if Figure 7 As shown, 701 is the laser radar, 704 is the target point, the serial number is i, and the radial distance between the target point i and the laser radar is p i , 702, 703, 705, 706 are the neighboring points of target point i, the serial numbers are i-2, i-1, i+1, i+2, and the radial distances are p i-2 、p i-1 、p i+1 、p i+2 ; As can be seen from the figure, p i-2 >p i-1 >p i >p i+1 >p i+2 , that is, p i-2 -p i-1 >0, p i-1 -p i >0, p i -p i+1 >0, p i+1 -p i+2 >0, so we can determine Figure 7 There are no corner points within the target neighborhood shown.

[0206] like Figure 8 As shown, 801 is the laser radar, 804 is the target point, the serial number is i, and the radial distance between the target point i and the laser radar is p i, 802, 803, 805, and 806 are the neighboring points of target point i, and their serial numbers are i-2, i-1, i+1, and i+2, respectively, and their radial distances are p i-2 、p i-1 、p i+1 、p i+2 ; As can be seen from the figure, p i-2 <p i-1 <p i <p i+1 <p i+2 , that is, p i-2 -p i-1 <0, p i-1 -p i <0, p i -p i+1 <0, p i+1 -p i+2 <0, so we can determine Figure 8 There are no corner points within the target neighborhood shown.

[0207] like Figure 9 As shown, 901 is the laser radar, 904 is the target point, the serial number is i, and the radial distance between the target point i and the laser radar is p i , 902, 903, 905, and 906 are the neighboring points of target point i, and their serial numbers are i-2, i-1, i+1, and i+2, respectively, and their radial distances are p i-2 、p i-1 、p i+1 、p i+2 ; As can be seen from the figure, p i-2 <p i-1 <p i , p i >p i+1 >p i+2 , that is, p i-2 -p i-1 <0, p i-1 -p i <0, p i -p i+1 >0, p i+1 -p i+2 >0, so we can determine Figure 9 There is a corner point in the target neighborhood shown, and the corner point is the target point i.

[0208] like Figure 10 As shown, 1001 is the laser radar, 1004 is the target point, the serial number is i, and the radial distance between the target point i and the laser radar is p i, 1002, 1003, 1005, 1006 are the neighboring points of target point i, the serial numbers are i-2, i-1, i+1, i+2, and the radial distances are p i-2 、p i-1 、p i+1 、p i+2 ; As can be seen from the figure, p i-2 >p i-1 >p i , p i <p i+1 <p i+2 , that is, p i-2 -p i-1 >0, p i-1 -p i >0, p i -p i+1 <0, p i+1 -p i+2 <0, so we can determine Figure 10 There is a corner point in the target neighborhood shown, and the corner point is the target point i.

[0209] like Figure 11 As shown, 1101 is the laser radar, 1104 is the target point, the serial number is i, and the radial distance between the target point i and the laser radar is p i , 1102, 1103, 1105, 1106 are the neighboring points of target point i, the serial numbers are i-2, i-1, i+1, i+2, and the radial distances are p i-2 、p i-1 、p i+1 、p i+2 ; As can be seen from the figure, p i-2 <p i-1 >p i , p i <p i+1 <p i+2 , that is, p i-2 -p i-1 <0, p i-1 -p i >0, p i -p i+1 <0, p i+1 -p i+2 <0, so we can determine Figure 11 There is a corner point in the target neighborhood shown, and the corner point is the neighborhood point i-1 (1103).

[0210] Step 305 : determining a target filtering parameter corresponding to the target point according to the radial distance of the target point, the radial distance of each neighboring point, and the filtering strength coefficient corresponding to the target point.

[0211] Step 306: Filter the target point according to the target filtering parameters.

[0212] The specific implementation process and principle of the above steps 305-306 can be referred to the detailed description of the above embodiment and will not be repeated here.

[0213] The filtering method for point cloud data provided in the embodiment of the present application determines the target neighborhood corresponding to the target point in the point cloud data to be filtered, the radial distance of the target point in the target neighborhood, and the radial distance of each neighboring point, and determines the filter strength coefficient according to whether the target neighborhood contains a corner point. Then, the target filter parameter corresponding to the target point is determined according to the radial distance of the target point, the radial distance of each neighboring point, and the filter strength coefficient corresponding to the target point, and then the target point is filtered according to the target filter parameter. Thus, by determining the filter strength parameter of each target point according to whether there is a corner point in the target neighborhood of the target point, different filter strengths are used for filtering in the corner point area and the non-corner point area, thereby not only improving the smoothing effect of the non-corner point area, but also reducing the filter strength of the corner point area, effectively retaining the corner point features, and can also adaptively adjust the filter strength parameter according to the corner point features of the corner point area, thereby not only reducing the computational complexity and computational time of point cloud data filtering, improving the efficiency of point cloud data filtering, but also further improving the accuracy and reliability of point cloud data filtering, and further improving the smoothing effect of the point cloud.

[0214] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0215] Corresponding to the filtering method of point cloud data described in the above embodiment, Figure 7 A structural block diagram of a point cloud data filtering device provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0216] Reference Figure 12 , the device 40, comprises:

[0217] A first acquisition module 41 is configured to acquire point cloud data to be filtered, wherein the point cloud data to be filtered includes a plurality of points;

[0218] A first determining module 42 is configured to determine a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and a plurality of neighboring points corresponding to the target point;

[0219] A second determining module 43 is used to determine the radial distance of the target point and the radial distance of each neighboring point;

[0220] A third determining module 44 is configured to determine a target filtering parameter corresponding to the target point based on the radial distance of the target point, the radial distance of each neighboring point, and the filter strength coefficient corresponding to the target point;

[0221] The first filtering module 45 is used to perform filtering processing on the target point according to the target filtering parameters.

[0222] In actual use, the point cloud data filtering device provided in the embodiment of the present application can be configured in any laser radar to execute the aforementioned point cloud data filtering method.

[0223] The point cloud data filtering device provided in the embodiment of the present application determines the target neighborhood corresponding to the target point in the point cloud data to be filtered, and determines the radial distance of the target point and the radial distance of each neighborhood point. Then, based on the radial distance of the target point, the radial distance of each neighborhood point, and the filter strength coefficient corresponding to the target point, the target filter parameter corresponding to the target point is determined, and then the target point is filtered according to the target filter parameter. Thus, by determining the target filter parameter for each point in the point cloud data based on the radial distance of each point in the neighborhood corresponding to each point in the point cloud data, the point cloud data is filtered, thereby improving the smoothing effect of the point cloud data. Since the target filter parameter can be determined by performing a simple calculation on the radial distance of the points in the neighborhood, the computational complexity and computational time of the point cloud data filtering are reduced, thereby improving the efficiency of the point cloud data filtering.

[0224] In a possible implementation form of the present application, the apparatus 40 further includes:

[0225] The second acquisition module is used to obtain the horizontal angle resolution and range divergence of the laser radar corresponding to the point cloud data to be filtered;

[0226] The fourth determination module is used to determine the sudden change neighboring points and the non-sudden change neighboring points in the neighborhood points according to the radial distance of each neighborhood point, the radial distance of the target point and the horizontal angle resolution, as well as the sequence number of each neighborhood point, the sequence number of the target point and the distance measurement divergence;

[0227] Accordingly, the third determining module 44 includes:

[0228] The first determining unit is used to determine the target filtering parameter according to the radial distance of the target point, the radial distance of each non-mutation neighborhood point and the filtering strength coefficient.

[0229] Furthermore, in another possible implementation form of the present application, the fourth determining module includes:

[0230] a second determining unit, configured to determine the distance between each neighborhood point and the target point according to the radial distance of each neighborhood point, the radial distance of the target point, and the horizontal angular resolution;

[0231] The third determining unit is used to determine the mutation distance threshold corresponding to each neighboring point according to the radial distance of the target point, the difference between the sequence number of the target point and the sequence number of each neighboring point, and the distance measurement divergence;

[0232] a fourth determining unit, configured to determine any neighborhood point as a sudden change neighborhood point if the distance between any neighborhood point and the target point is greater than a sudden change distance threshold corresponding to the any neighborhood point;

[0233] The fifth determining unit is configured to determine any neighborhood point as a non-mutation neighborhood point if the distance between any neighborhood point and the target point is less than or equal to a mutation distance threshold corresponding to any neighborhood point.

[0234] Furthermore, in another possible implementation form of the present application, the target filtering parameters include a first target filtering parameter and a second target filtering parameter; accordingly, the first determining unit is specifically configured to:

[0235] According to the radial distance of the target point and the radial distance of each non-mutation neighboring point, the mean radial distance and the mean square of the radial distance corresponding to the target neighborhood are determined;

[0236] Determine a first target filtering parameter according to the radial distance mean, the radial distance square mean and the filtering strength coefficient;

[0237] A second target filtering parameter is determined according to the first target filtering parameter and the radial distance mean.

[0238] Furthermore, in another possible implementation of the present application, the first filtering module 45 includes:

[0239] a sixth determining unit, configured to determine a product of a radial distance of the target point and a first target filtering parameter;

[0240] The seventh determining unit is configured to determine the filtered radial distance corresponding to the target point by taking the sum of the product of the radial distance of the target point and the first target filtering parameter and the second target filtering parameter.

[0241] Furthermore, in another possible implementation of the present application, the first determining module 42 includes:

[0242] An eighth determining unit, configured to determine the number of target lines corresponding to the target point and the sequence number of the target point;

[0243] The ninth determination unit is used to determine the (n-1) / 2 points in the target line number that are adjacent to the target point and whose serial numbers are smaller than the serial number of the target point as neighborhood points, and to determine the (n-1) / 2 points in the target line number that are adjacent to the target point and whose serial numbers are larger than the serial number of the target point as neighborhood points, where n is the size of the target neighborhood.

[0244] Furthermore, in another possible implementation form of the present application, the apparatus 40 further includes:

[0245] The fifth determination module is used to determine the filter strength coefficient according to whether the target neighborhood contains a corner point.

[0246] Furthermore, in another possible implementation form of the present application, the fifth determining module includes:

[0247] a tenth determining unit, configured to determine a filter strength coefficient based on distances between adjacent points in the target neighborhood when a corner point is included in the target neighborhood;

[0248] The eleventh determining unit is configured to determine the filter strength coefficient as a preset value when the target neighborhood does not include a corner point.

[0249] Furthermore, in another possible implementation form of the present application, the apparatus 40 further includes:

[0250] a sixth determining module, configured to determine a radial distance difference between adjacent points in the target neighborhood based on the radial distance of each point in the target neighborhood;

[0251] The seventh determination module is configured to determine that the target neighborhood does not contain a corner point if the radial distance differences between adjacent points in the target neighborhood are all greater than or equal to 0 or are all less than or equal to 0; otherwise, determine that the target neighborhood contains a corner point.

[0252] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment section and will not be repeated here.

[0253] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0254] In order to implement the above embodiments, the present application also proposes a laser radar.

[0255] Figure 13 This is a schematic structural diagram of a laser radar according to an embodiment of the present application.

[0256] like Figure 13 As shown, the laser radar 200 includes:

[0257] A memory 210 and at least one processor 220, a bus 230 connecting different components (including the memory 210 and the processor 220), the memory 210 stores a computer program, and when the processor 220 executes the program, the point cloud data filtering method described in the embodiment of the present application is implemented.

[0258] Bus 230 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0259] The laser radar 200 typically includes a variety of electronic device readable media. These media can be any available media that can be accessed by the laser radar 200, including volatile and non-volatile media, removable and non-removable media.

[0260] The memory 210 may also include computer system readable media in the form of volatile memory, such as random access memory (RAM) 240 and / or cache memory 250. The laser radar 200 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 260 may be used to read and write non-removable, non-volatile magnetic media ( Figure 13 Not shown, often called a "hard drive"). Although Figure 13 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 230 via one or more data medium interfaces. Memory 210 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of various embodiments of the present application.

[0261] A program / utility 280 having a set (at least one) of program modules 270 may be stored, for example, in memory 210. Such program modules 270 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 270 generally implement the functions and / or methods of the embodiments described herein.

[0262] LiDAR 200 can also communicate with one or more external devices 290 (e.g., a keyboard, pointing device, display 291, etc.), one or more devices that enable a user to interact with LiDAR 200, and / or any device that enables LiDAR 200 to communicate with one or more other computing devices (e.g., a network card, modem, etc.). This communication can occur via input / output (I / O) interface 292. Furthermore, LiDAR 200 can communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet) via network adapter 293. As shown, network adapter 293 communicates with other modules of LiDAR 200 via bus 230. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with LiDAR 200, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0263] The processor 220 executes various functional applications and data processing by running programs stored in the memory 210 .

[0264] It should be noted that the implementation process and technical principles of the laser radar of this embodiment can be found in the aforementioned explanation of the filtering method of point cloud data in the embodiment of the present application, and will not be repeated here.

[0265] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned various method embodiments can be implemented.

[0266] An embodiment of the present application provides a computer program product. When the computer program product runs on a laser radar, the laser radar can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0267] If the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process in the above-mentioned method embodiment by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can at least include: any entity or device capable of carrying computer program code to a camera / terminal device, a recording medium, computer memory, read-only memory (ROM), random access memory (RAM), an electric carrier signal, a telecommunications signal, and a software distribution medium. Examples include a USB flash drive, a mobile hard drive, a magnetic disk, or an optical disk. In some jurisdictions, based on legislation and patent practice, computer-readable media cannot be electric carrier signals or telecommunications signals.

[0268] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0269] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0270] In the embodiments provided in this application, it should be understood that the disclosed devices / lidars and methods can be implemented in other ways. For example, the device / lidar embodiments described above are merely schematic. For example, the division of the modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0271] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0272] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for filtering point cloud data, characterized in that: include: Acquire point cloud data to be filtered, wherein the point cloud data to be filtered includes a plurality of points; Determine a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and a plurality of neighboring points corresponding to the target point; Determining the radial distance of the target point and the radial distance of each of the neighboring points; Obtaining the horizontal angular resolution and range divergence of the laser radar corresponding to the point cloud data to be filtered; Determine, according to the radial distance of each of the neighborhood points, the radial distance of the target point and the horizontal angle resolution, as well as the sequence number of each of the neighborhood points, the sequence number of the target point and the distance measurement divergence, a sudden change neighborhood point and a non-sudden change neighborhood point in the neighborhood points; Determine a target filter parameter corresponding to the target point according to the radial distance of the target point, the radial distance of each of the non-mutation neighboring points, and the filter strength coefficient corresponding to the target point; The target point is filtered according to the target filtering parameters.

2. The method according to claim 1, wherein Determining the sudden change neighboring points and the non-sudden change neighboring points in the neighborhood points according to the radial distance of each of the neighborhood points, the radial distance of the target point, and the horizontal angle resolution, as well as the sequence number of each of the neighborhood points, the sequence number of the target point, and the distance measurement divergence, includes: Determining the distance between each of the neighborhood points and the target point according to the radial distance of each of the neighborhood points, the radial distance of the target point, and the horizontal angular resolution; Determine the mutation distance threshold corresponding to each of the neighborhood points according to the radial distance of the target point, the difference between the sequence number of the target point and the sequence number of each of the neighborhood points, and the distance measurement divergence; If the distance between any neighborhood point and the target point is greater than the mutation distance threshold corresponding to the any neighborhood point, the any neighborhood point is determined as the mutation neighborhood point; If the distance between any of the neighborhood points and the target point is less than or equal to the mutation distance threshold corresponding to the any of the neighborhood points, the any of the neighborhood points is determined as the non-mutation neighborhood point.

3. The method according to claim 1, wherein The target filtering parameters include a first target filtering parameter and a second target filtering parameter, and determining the target filtering parameters corresponding to the target point according to the radial distance of the target point, the radial distance of each of the non-mutation neighboring points, and the filtering strength coefficient corresponding to the target point includes: Determine the mean of the radial distances and the mean of the squares of the radial distances corresponding to the target neighborhood according to the radial distance of the target point and the radial distances of each of the non-mutation neighborhood points; determining the first target filtering parameter according to the radial distance mean, the radial distance square mean, and the filtering strength coefficient; The second target filtering parameter is determined according to the first target filtering parameter and the radial distance mean.

4. The method according to claim 3, wherein The filtering process of the target point according to the target filtering parameter includes: determining a product of a radial distance of the target point and the first target filter parameter; The product of the radial distance of the target point and the first target filtering parameter and the sum of the second target filtering parameter are determined as the filtered radial distance corresponding to the target point.

5. The method according to claim 1, wherein The determining of a target neighborhood corresponding to a target point in the point cloud data to be filtered includes: Determine the number of target lines corresponding to the target point and the sequence number of the target point; The (n-1) / 2 points in the target line number that are adjacent to the target point and whose serial numbers are smaller than the serial number of the target point are determined as the neighborhood points, and the (n-1) / 2 points in the target line number that are adjacent to the target point and whose serial numbers are larger than the serial number of the target point are determined as the neighborhood points, where n is the size of the target neighborhood.

6. The method according to any one of claims 1 to 5, characterized in that: Before determining the target filtering parameter corresponding to the target point according to the radial distance of the target point, the radial distance of each of the non-mutation neighboring points, and the filter strength coefficient corresponding to the target point, the method further includes: The filter strength coefficient is determined according to whether the target neighborhood contains a corner point.

7. The method according to claim 6, wherein The determining of the filter strength coefficient according to whether the target neighborhood includes a corner point includes: When the corner point is included in the target neighborhood, determining the filter strength coefficient according to the distance between adjacent points in the target neighborhood; When the corner point is not included in the target neighborhood, the filtering strength coefficient is determined to be a preset value.

8. The method according to claim 6, wherein Before determining the filter strength coefficient according to whether the target neighborhood includes a corner point, the method further includes: Determining radial distance differences between adjacent points in the target neighborhood based on the radial distance of each point in the target neighborhood; If the radial distance differences between the adjacent points in the target neighborhood are all greater than or equal to 0 or less than or equal to 0, it is determined that the target neighborhood does not include the corner point; otherwise, it is determined that the target neighborhood includes the corner point.

9. A filtering device for point cloud data, characterized in that: include: A first acquisition module is configured to acquire point cloud data to be filtered, wherein the point cloud data to be filtered includes a plurality of points; A first determining module is configured to determine a target neighborhood corresponding to a target point in the point cloud data to be filtered, wherein the target point is any point in the point cloud data to be filtered, and the target neighborhood includes the target point and a plurality of neighboring points corresponding to the target point; A second determining module is used to determine the radial distance of the target point and the radial distance of each of the neighborhood points; A second acquisition module is used to obtain the horizontal angle resolution and range divergence of the laser radar corresponding to the point cloud data to be filtered; a fourth determining module, configured to determine a sudden change neighboring point and a non-sudden change neighboring point in the neighboring points according to the radial distance of each of the neighboring points, the radial distance of the target point, and the horizontal angular resolution, as well as the sequence number of each of the neighboring points, the sequence number of the target point, and the distance measurement divergence; A first determining unit is configured to determine a target filtering parameter corresponding to the target point according to the radial distance of the target point, the radial distance of each of the non-mutation neighboring points, and the filtering strength coefficient corresponding to the target point; The first filtering module is used to perform filtering processing on the target point according to the target filtering parameters.

10. A laser radar comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 8 is implemented.

11. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images