Ground point cloud point determination method and device based on hierarchical statistics

Abstract

The invention belongs to the field of machine vision, and provides a ground point cloud point determination method based on hierarchical statistics, and the method comprises the steps: dividing all to-be-determined point cloud points into a plurality of cylinder grids according to coordinate positions, and enabling the to-be-determined point cloud points to be point cloud points which are collected by a laser radar and are located in a ground point detection range, and all the columnar grids are combined to form the ground point detection range. The columnar grid is divided into a first interval, a second interval and a third interval in the Z-axis direction, the columnar grid corresponds to parameters of five dimensions, and the reserved columnar grid is determined according to the second parameter. And according to the fourth parameter and the fifth parameter, non-target points in the reserved cylinder grids are removed, and all the finally reserved point cloud points are used as the ground point cloud points. According to the scheme, the point cloud points in each interval are screened, layered detection of the ground points is realized, the false detection rate of ground point filtering is effectively reduced, and the filtering effect of non-target points is improved.

Description

Technical Field

[0001] This application belongs to the field of machine vision, and particularly relates to a method and device for determining ground point cloud points based on hierarchical statistics. Background Art

[0002] A lidar measures the distance and angle from the scanner to an object through rapidly emitted laser pulses, accurately and quickly obtaining the three-dimensional data of the object's surface. It is an important means of information acquisition. The data obtained by scanning is discrete three-dimensional points, called point cloud.

[0003] In the vehicle-road collaborative scenario, the ground point filtering algorithm for lidar point cloud is a widespread and important technology. Currently, mainstream object detection algorithms can only set fixed thresholds to filter ground points when processing the original point cloud data. Fixed thresholds are prone to filtering out target points, resulting in a high false detection rate and poor filtering effect for non-target points in this situation. Summary of the Invention

[0004] The purpose of this application is to provide a method and device for determining ground point cloud points based on hierarchical statistics, which can reduce the false detection rate and improve the filtering effect of non-target points.

[0005] In a first aspect, an embodiment of this application provides a method for determining ground point cloud points based on hierarchical statistics. The method includes:

[0006] Dividing all the to-be-determined point cloud points into multiple columnar grids according to their coordinate positions. The to-be-determined point cloud points are the point cloud points collected by the lidar within the ground point detection range, and all the columnar grids together form the ground point detection range;

[0007] Dividing the columnar grid into a first interval, a second interval, and a third interval along the Z-axis direction. The columnar grid corresponds to five-dimensional parameters. The first parameter is used to store the z-axis coordinate value of the lowest point in the corresponding grid. The second parameter is used to store the number of points whose height difference between the points within the first interval and the lowest point in the grid is within the target point height distribution range in the corresponding grid. The third parameter is used to store the height of the lowest non-target point above the ground. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset. The fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset;

[0008] Determining the retained columnar grids according to the second parameter;

[0009] According to the fourth parameter and the fifth parameter, eliminating non-target points within the retained columnar grids, and taking all the finally retained point cloud points as the ground point cloud points.

[0010] In an implementable embodiment, the method further includes:

[0011] Filter out a point cloud point spatial data set within the ground point detection range from the point cloud data collected by the lidar.

[0012] In an implementable embodiment, the determining the retained cylinder grids according to the second parameter includes:

[0013] Sort all the point cloud points in the first interval of each cylinder grid according to the height value, and determine the point cloud point with the lowest height value to be determined;

[0014] According to the height value of the point cloud point with the lowest height value to be determined corresponding to the first interval of each cylinder grid and the height values of all the point cloud points in the first interval of each cylinder grid, determine the retained cylinder grids.

[0015] In an implementable embodiment, the determining the retained cylinder grids according to the second parameter further includes:

[0016] For each first interval of each cylinder grid, repeatedly compare the height values of any two point cloud points, and compare the point cloud point with the lower height value with any other point cloud point until all point cloud points are compared, and use the point cloud point with the lower height value obtained in the last comparison as the point cloud point with the lowest height value to be determined;

[0017] According to the height value of the point cloud point with the lowest height value to be determined corresponding to the first interval of each cylinder grid and the height values of all the point cloud points in the first interval of each cylinder grid, determine whether each cylinder grid is retained.

[0018] In an implementable embodiment, the determining whether each cylinder grid is retained according to the height value of the point cloud point with the lowest height value to be determined corresponding to the first interval of each cylinder grid and the height values of all the point cloud points in the first interval of each cylinder grid includes:

[0019] For each first interval of each cylinder grid, according to the difference between the height value of the point cloud point with the lowest height value to be determined and the height values of each point cloud point to be determined in the first interval of the cylinder grid, and in combination with a preset target height distribution range, determine the number of qualified point cloud points in the first interval of each cylinder grid;

[0020] According to the number of qualified point cloud points in the first interval of each cylinder grid, and in combination with a preset filtering threshold parameter, determine whether each cylinder grid is retained.

[0021] In an implementable embodiment, determining the number of qualified point cloud points in the first interval of each column grid by combining the difference between the height value of the to-be-determined point cloud point with the lowest corresponding height value and the height values of each to-be-determined point cloud point in the first interval within the column grid, and in combination with a preset target height distribution range, includes:

[0022] For each to-be-determined point cloud point, determine whether the difference is within the target height distribution range;

[0023] If so, regard the corresponding to-be-determined point cloud point as a qualified point cloud point;

[0024] Calculate the number of qualified point cloud points in the first interval of each column grid.

[0025] In an implementable embodiment, determining whether to retain each column grid by combining the number of qualified point cloud points in each column grid and a preset filtering threshold parameter includes:

[0026] Determine whether the number of qualified point cloud points corresponding to the first interval of each column grid is higher than the filtering threshold parameter. If it is higher, determine to retain the column grid.

[0027] In an implementable embodiment, removing non-target points within the retained column grids according to the fourth parameter and the fifth parameter, and using all the finally retained point cloud points as the ground point cloud points includes:

[0028] Traverse the point cloud points in the fifth parameter, remove non-target points within the fifth parameter of the retained column grids. If there are target points, continue to traverse the point cloud points in the fourth parameter, remove non-target points within the fourth parameter of the retained column grids, and use all the retained point cloud points as the ground point cloud points.

[0029] In a second aspect, an embodiment of the present application provides a ground point cloud point determination device based on hierarchical statistics, including:

[0030] A preprocessing unit that divides all to-be-determined point cloud points into multiple column grids according to coordinate positions. The to-be-determined point cloud points are point cloud points collected by a lidar within the ground point detection range, and all the columnar grids together form the ground point detection range;

[0031] The storage unit divides the columnar grid into a first interval, a second interval, and a third interval along the Z-axis direction. The columnar grid corresponds to five-dimensional parameters. The first parameter is used to store the z-axis coordinate value of the lowest point in the corresponding grid. The second parameter is used to store the number of points in the corresponding grid whose height difference between the points in the first interval and the height of the lowest point in the grid is within the target point height distribution range. The third parameter is used to store the height of the lowest non-target point above the ground. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset. The fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset;

[0032] The processing unit determines the retained columnar grids according to the second parameter;

[0033] The output unit eliminates non-target points within the retained columnar grids according to the fourth parameter and the fifth parameter, and takes all the finally retained point cloud points as the ground point cloud points.

[0034] In a third aspect, an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the electronic device implements the method described in any one of the above.

[0035] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by an electronic device, it implements the method described in any one of the above.

[0036] The beneficial effects of the embodiments of the present application compared with the prior art are:

[0037] By dividing all the point cloud points to be determined into multiple columnar grids according to their coordinate positions, and then dividing multiple intervals along the z-axis, and screening the point cloud points in each interval, hierarchical detection of ground points is achieved, effectively reducing the false detection rate of ground point filtering and improving the filtering effect of non-target points. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0039] Figure 1 It is an application scenario diagram of a method for determining ground point cloud points based on hierarchical statistics provided by the present application;

[0040] Figure 2 Flow schematic diagram of a method for determining ground point cloud points based on hierarchical statistics provided by this application;

[0041] Figure 3 Flow schematic diagram of a method for dividing column grids in an embodiment provided by this application;

[0042] Figure 4 Flow schematic diagram of a method for determining the lowest ground points in an embodiment provided by this application;

[0043] Figure 5 Flow schematic diagram of a method for determining the lowest ground points in another embodiment provided by this application;

[0044] Figure 6 Flow schematic diagram of a method for determining the lowest ground points in another embodiment provided by this application;

[0045] Figure 7 Flow schematic diagram of a method for retaining column grids in an embodiment provided by this application;

[0046] Figure 8 Embodiment flow schematic diagram of a method for determining ground point cloud points based on hierarchical statistics provided by this application;

[0047] Figure 9 Embodiment scenario schematic diagram of a method for determining ground point cloud points based on hierarchical statistics provided by this application;

[0048] Figure 10 Structural schematic diagram of a device for determining ground point cloud points based on hierarchical statistics provided by this application;

[0049] Figure 11 Structural schematic diagram of an electronic device provided by this application. Detailed implementation manners

[0050] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of this application. However, those skilled in the art should clearly understand that this application can also 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 unnecessary details from interfering with the description of this application.

[0051] It should be understood that when used in the specification and appended claims of this application, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0052] It should also be understood that the term "and / or" used in the specification of this application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0053] As used in the specification of this application and the appended claims, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" depending on the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]" depending on the context.

[0054] In addition, in the description of the specification of this application and the appended claims, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0055] Reference to "one embodiment" or "some embodiments" etc. described in the specification of this application means that a specific feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of this application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in another way. The terms "comprising", "including", "having", and their variants all mean "including but not limited to", unless otherwise specifically emphasized in another way.

[0056] It should be understood that the magnitude of the sequence numbers of the steps in this embodiment does not mean the order of execution is prior or posterior, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0057] At present, the intelligentization of automobiles, as a major trend in the development of the times, has always been the focus of attention in many industries such as the Internet and automobile enterprises. Environmental perception, as one of the core technologies in the field of intelligent driving, is the only way for automobiles to achieve intelligentization. Currently, for intelligent automobiles, such as vehicles of the type of intelligent logistics vehicles, one way to obtain external environmental information is to use sensors such as cameras. Although cameras and other sensors are inexpensive, there are problems such as low measurement accuracy of the distance to obstacles and weak perception ability of the three-dimensional information of obstacles. Another relatively better way is to use a low-beam lidar to obtain external environmental information. However, due to the relatively low sampling accuracy of the low-beam lidar, the recognition of obstacles is not accurate enough. Therefore, how to improve the detection and filtering of ground point clouds is an urgent problem to be solved. Figure 1 It is a scene diagram of the prior art. The existing ground point filtering algorithm can only filter point cloud points through a fixed threshold, such as Figure 1 By only screening the point cloud points within the preset threshold area, through Figure 1 It is known that this area not only contains pedestrians, vehicles but also the greening scenes on both sides of the road. Under such conditions, the existing filtering method will inevitably misjudge the ground point cloud points, and the overall filtering effect of the ground point cloud points is not good.

[0058] This application proposes a method for determining ground point clouds based on lidar, including:

[0059] S201, dividing all the point cloud points to be determined into multiple cylindrical grids according to their coordinate positions. The point cloud points to be determined are the point cloud points collected by the lidar within the ground point detection range, and all the cylindrical grids together form the ground point detection range;

[0060] S202, dividing the cylindrical grid into a first interval, a second interval, and a third interval along the Z-axis direction. The cylindrical grid corresponds to five-dimensional parameters. The first parameter is used to store the z-axis coordinate value of the lowest point in the corresponding grid. The second parameter is used to store the number of points in the corresponding grid whose height difference between the points in the first interval and the lowest point in the grid is within the target point height distribution range. The third parameter is used to store the height of the lowest non-target point above the ground. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset. The fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset;

[0061] S203, determining the retained cylindrical grids according to the second parameter;

[0062] S204, removing non-target points within the retained cylindrical grids according to the fourth parameter and the fifth parameter, and taking all the finally retained point cloud points as the ground point clouds.

[0063] Exemplarily, all the point cloud points within a preset spatial range are divided into different columnar grids according to their coordinate positions. For example, Figure 3 as shown, the entire space is divided into the same columnar grids as shown in F1. According to the coordinates of the point cloud points in the space, they are divided into the corresponding columnar grids. The entire space is divided into three intervals on the Z-axis to screen the ground point cloud points. All the grids identical to F1 form the ground point detection range. The F1 grid contains multiple point cloud points. Whether to retain the F1 grid is determined according to the height values of the point cloud points in the first interval. Then, the point cloud points in the remaining third interval and second interval are traversed, and the non-target points in the retained columnar grids are removed. All the finally retained point cloud points are used as the ground point cloud points. The point cloud points to be determined within the F1 grid are the ground point cloud points.

[0064] In this application, by dividing the point cloud points within the spatial range into the corresponding grids, and then determining whether to retain the grid according to the height of the point cloud points, the point cloud points are further filtered by this method, preventing the omission of incompletely filtered point cloud points, effectively reducing the false detection rate of ground point filtering, and improving the filtering effect of non-target points.

[0065] In an implementable embodiment of this application, the method for determining ground point cloud points based on hierarchical statistics further includes:

[0066] Screening out the point cloud point spatial dataset within the ground point detection range from the point cloud data collected by the lidar.

[0067] Exemplarily, assume there is a set of three-dimensional point cloud data collected by a lidar currently. Each point includes three coordinate values (x, y, z), representing the position of the point in the three-dimensional space. By defining a height threshold, the point cloud points within the height detection range can be determined. Assume the preset height threshold is between 0 - 0.5 meters. Then all the point cloud data between 0 - 0.5 meters form the point cloud spatial dataset. Specifically, there is a point cloud point with coordinates (3, 4, 0.3). Determine whether this point is within the ground point detection range. The x coordinate of this point is 3, the y coordinate is 4, and the z coordinate is 0.3. In this case, the z coordinate of this point is within the range of 0 - 0.5 meters. Therefore, this point can be used as a point that meets the ground point detection range.

[0068] In this application, by this method, all the point cloud data is traversed, and the qualified point cloud point spatial dataset is screened out, preprocessing the point cloud points, reducing redundant data, and facilitating further filtering of the point cloud points.

[0069] In an implementable embodiment of this application, determining the retained columnar grids according to the second parameter includes:

[0070] Determine whether each column grid is retained according to the height value of the lowest ground point corresponding to the first interval of each column grid and the height values of all point cloud points within the first interval of each column grid.

[0071] Exemplarily, assume there is a column grid that contains point cloud data, and each point cloud data has corresponding three-dimensional coordinates (x, y, z). Assume the height of the lowest ground point is 0.2 meters. Preset a height range (0.5, 0.8), and determine whether the column grid is retained by judging whether the difference between the point cloud points in the column grid and the lowest ground point is within the preset height range. Specifically, if the point cloud coordinates within the column grid at this time are: D1(6, 5, 0.6), D2(4, 6, 0.9), D3(5, 6, 0.4), D4(3, 4, 0.5), and the z coordinates of D1, D2, D3, and D4 are 0.6, 0.9, 0.4, and 0.5 respectively. Subtract the height of the lowest ground point from the vertical coordinates of these point cloud points, and calculate the differences as: 0.4, 0.7, 0.2, 0.3 respectively. Since the preset height range is (0.5, 0.8), D2 is the point cloud point that meets the conditions, so this column grid can be retained.

[0072] This application determines whether to retain a column grid based on the difference between the height values of all point cloud points in each column grid and the height value of the lowest ground point, and improves the filtering effect of point cloud points through this step.

[0073] In an implementable embodiment of this application, determining the retained column grids according to the second parameter further includes:

[0074] S401, sort all the point cloud points in the first interval of each column grid according to the height value, and determine the point cloud point with the lowest height value to be determined;

[0075] S402, determine the retained column grids according to the height value of the point cloud point with the lowest height value to be determined corresponding to the first interval of each column grid and the height values of all point cloud points in the first interval of each column grid.

[0076] Exemplarily, assume there is a cylindrical grid containing point cloud data, and each point cloud data has corresponding three-dimensional coordinates (x, y, z). First, traverse all the point cloud points in the first interval of the cylindrical grid, and determine the lowest ground point height according to the z-axis coordinate values of all the point cloud points. Specifically, assume that all the point cloud points in the first interval of the cylindrical grid are B1(1, 3, 1), B2(1, 2, 0.5), B3(2, 1, 0.8), B4(3, 1, 0.4), B5(2, 2, 0.3), B6(3, 1, 0.1), B7(4, 3, 0.6). At this time, by comparing the z-axis coordinates of all the point cloud points, it is found that the z-axis coordinate of B6 is the lowest ground point height of 0.1 meter. Preset a height range (0, 1), and determine whether to save the cylindrical grid by judging whether the difference between the point cloud points in the cylindrical grid and the lowest ground point is within the preset height range. Since the z coordinates of B1, B2, B3, B4, B5, B6, B7 are 1, 0.5, 0.8, 0.4, 0.3, 0.1, 0.6 respectively, subtract the z-axis coordinates of these point cloud points from the lowest ground point height, and calculate the differences as 0.9, 0.4, 0.7, 0.3, 0.2, 0, 0.5 respectively. The preset height range is (0, 1), so all the point cloud points are qualified point cloud points, and thus the cylindrical grid can be retained.

[0077] This application first traverses the lowest ground point, and then determines whether to retain the cylindrical grid by the difference between the height values of all the point cloud points in the first interval of each cylindrical grid and the lowest ground point height value, which improves the filtering effect of the ground point cloud points through this step.

[0078] In the implementable embodiments of this application, the determining the retained cylindrical grid according to the second parameter further includes:

[0079] S501, for each first interval of the cylindrical grid, repeatedly compare the height values of any two point cloud points, and compare the point cloud point with the lower height value with any other point cloud point until all the point cloud points are compared. Take the point cloud point with the lower height value obtained in the last comparison as the point cloud point to be determined with the lowest height value;

[0080] S502, determine whether to retain each cylindrical grid according to the height value of the point cloud point to be determined with the lowest height value corresponding to each first interval of the cylindrical grid and the height values of all the point cloud points in the first interval of each cylindrical grid.

[0081] Exemplarily, assume there is a cylindrical grid that is divided into three intervals along the Z-axis, and each interval contains point cloud data. Each point cloud data in the first interval has corresponding three-dimensional coordinates (x, y, z). First, compare the height values of any two point cloud points in the first interval, and compare the point cloud point with the lower height value with any other point cloud point until each point cloud point is compared. The point cloud point with the lower height value obtained from the last comparison is used as the point cloud point with the lowest height to be determined. Specifically, assume that all the point cloud points in the cylindrical grid are A1(2, 3, 0.5), A2(1, 3, 0.5), A3(2, 4, 0.8), A4(1, 1, 0.4), A5(4, 2, 0.7), A6(3, 2, 0.1), A7(2, 1, 0.6). First, compare the z-axis coordinates of A1 and A2. The z-axis coordinates of A1 and A2 are both 0.5. Therefore, 0.5 is used as the height of the lowest point cloud point to be determined. Continue to compare the z-axis coordinate of point A1 with the z-axis coordinate of point A3. 0.8 is greater than 0.5, and 0.5 is used as the height of the lowest point cloud point to be determined. Continue to compare the z-axis coordinate of point A1 with the z-axis coordinate of point A4. 0.4 is less than 0.5. At this time, 0.4 is used as the height of the lowest point cloud point to be determined. Continue to compare the z-axis coordinate of point A4 with the z-axis coordinate of point A5. 0.4 is less than 0.7, and 0.4 is used as the height of the lowest point cloud point to be determined. Continue to compare the z-axis coordinate of point A4 with the z-axis coordinate of point A6. 0.1 is less than 0.4, and 0.1 is used as the height of the lowest point cloud point to be determined. Continue to compare the z-axis coordinate of point A6 with the z-axis coordinate of point A7. 0.1 is less than 0.6, and finally 0.1 is used as the height of the lowest point cloud point. Preset a height range (0, 1), and determine whether to save the cylindrical grid by judging whether the difference between the point cloud points in the cylindrical grid and the lowest ground point is within the preset height range. Since the z coordinates of A1, A2, A3, A4, A5, A6, A7 are: 0.5, 0.5, 0.8, 0.4, 0.7, 0.1, 0.6, calculate the differences between the z-axis coordinates of these point cloud points and the lowest ground point height respectively, and the differences are: 0.4, 0.4, 0.7, 0.3, 0.6, 0, 0.5. The preset height range is (0, 1). Therefore, all point cloud points are qualified point cloud points, and thus the cylindrical grid can be retained.

[0082] This step determines the height of the lowest ground point in the first interval of the grid through real-time comparison, and then determines whether to retain the cylindrical grid by the difference between the height values of all point cloud points in each cylindrical grid and the height value of the lowest ground point. Through this step, the filtering effect of the ground point cloud points is further improved.

[0083] In an implementable embodiment of the present application, determining whether each column grid is retained based on the height value of the lowest point cloud point to be determined corresponding to each first interval of the column grid and the height values of all point cloud points in the first interval within each column grid includes:

[0084] S601. For each first interval of the column grid, determine the number of qualified point cloud points in the first interval of each column grid according to the difference between the height value of the lowest point cloud point to be determined and the height values of each point cloud point to be determined in the first interval within the column grid, in combination with a preset target height distribution range;

[0085] S602. Determine whether each column grid is retained according to the number of qualified point cloud points in the first interval of each column grid, in combination with a preset filtering threshold parameter.

[0086] Exemplarily, assume that in column grid F1, a target height distribution range is preset as (0, 0.5), and the preset filtering threshold parameter is 2. Now, there are point cloud points C1(5, 5, 0.8), C2(6, 4, 0.5), C3(2, 5, 0.2), C4(3, 4, 0.6), C5(1, 2, 0.3), C6(4, 4, 0.7), C7(4, 3, 0.9) in grid F1. The height of the lowest point cloud point to be determined determined by the above method is 0.2. At this time, the height difference between C1 and the lowest point cloud point to be determined is 0.6, which does not conform to the target height distribution range; the height difference between C2 and the lowest point cloud point to be determined is 0.3, which conforms to the target height distribution range; the height difference between C3 and the lowest point cloud point to be determined is 0, which conforms to the target height distribution range; the height difference between C4 and the lowest point cloud point to be determined is 0.4, which conforms to the target height distribution range; the height difference between C5 and the lowest point cloud point to be determined is 0.1, which conforms to the target height distribution range; the height difference between C6 and the lowest point cloud point to be determined is 0.5, which conforms to the target height distribution range; the height difference between C7 and the lowest point cloud point to be determined is 0.7, which does not conform to the target height distribution range. In this embodiment, C2 - C6 all conform to the target height distribution range, and the number of point cloud points conforming to the target height distribution range exceeds the preset filtering threshold parameter 2. Therefore, this column grid F1 is determined to be retained.

[0087] Through this step, the present application further refines the filtering of point cloud points, effectively reducing the false detection rate of ground point filtering and improving the filtering effect of non-target points.

[0088] In an implementable embodiment of the present application, the determining the number of qualified point cloud points in the first interval of each column grid according to the difference between the height value of the lowest point cloud point to be determined and the height values of each point cloud point to be determined in the first interval within the column grid, in combination with a preset target height distribution range, includes:

[0089] S701. For each point cloud point to be determined, determine whether the difference is within the target height distribution range;

[0090] S702. If so, use the corresponding point cloud point to be determined as a qualified point cloud point;

[0091] S703. Calculate the number of qualified point cloud points in the first interval of each column grid.

[0092] Exemplarily, if in grid F2, the target height distribution range is set to (0.1, 0.2), and there are three point cloud points to be determined in grid F2, D1(1, 1, 0.2), D2(2, 2, 0.1), D3(3, 2, 0.5). At this time, the lowest height value of the point cloud points to be determined is 0.1. Subtract the height value of each point cloud point to be determined from the lowest height value to be determined to determine the difference. At this time, the difference of D1 is 0.1, which meets the target height distribution range, and D1 is used as a qualified point cloud point. The difference of D2 is 0, which does not meet the target height distribution range, so D2 is an unqualified point cloud point. The difference of D3 is 0.4, which does not meet the target height distribution range, so D3 is an unqualified point cloud point. In summary, in grid F2, only D1 meets the target height distribution range, so the number of qualified point cloud points is 1.

[0093] This application effectively reduces the false detection rate of ground point filtering and improves the filtering effect of non-target points through this method.

[0094] In an implementable embodiment of this application, determining whether to retain each column grid according to the number of qualified point cloud points in each column grid and in combination with a preset filtering threshold parameter includes:

[0095] Determine whether the number of qualified point cloud points corresponding to the first interval of each column grid is higher than the filtering threshold parameter. If it is higher, determine to retain the column grid.

[0096] Exemplarily, the preset filtering value parameter is 10. Assume that the number of qualified point cloud points in column grid A is 15, and the number of qualified point cloud points in column grid B is 8. According to the set filtering threshold parameter of 10 points, determine to retain column grid A because the number of its qualified point cloud points is higher than 10, and exclude column grid B because the number of its qualified point cloud points is lower than or equal to 10.

[0097] This application determines whether to retain each column grid according to the number of qualified point cloud points in each column grid and the preset filtering threshold parameter, so as to realize the screening and processing of point cloud data.

[0098] In an implementable embodiment of the present application, the step of removing non-target points in the retained columnar grid according to the fourth parameter and the fifth parameter, and taking all the finally retained point cloud points as the ground point cloud points includes:

[0099] Traverse the point cloud points in the fifth parameter, and remove non-target points in the retained columnar grid of the fifth parameter. If there are target points, continue to traverse the point cloud points in the fourth parameter, and remove non-target points in the retained columnar grid of the fourth parameter. All the retained point cloud points are used as the ground point cloud points.

[0100] Exemplarily, the entire road space is divided into three partitions on the Z axis, and the first interval, the second interval, and the third interval are respectively corresponding to the ground small target area, the ground medium target area, and the ground large area. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset, and the fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset. After the above method filters the point cloud points in the first interval and retains the qualified columnar grids, traverse all the point cloud points in the third interval, and remove non-target points in the retained columnar grid. When there are target ground point cloud points in the third interval, continue to traverse all the point cloud points in the second interval, and remove non-target points in the retained columnar grid again. When there are target ground point cloud points in the second interval, retain all the point cloud points as the ground point cloud points.

[0101] Figure 8 It is a schematic flow chart of a complete embodiment of a method for determining ground point cloud points based on hierarchical statistics in the present application. Exemplarily, for a road scene Figure 9 First, collect all the point cloud data of the entire scene. The dashed box indicated by z1 is the preset point cloud data range, and grid division is performed on the part within this space range. z2 is the first interval on the Z axis, that is, the first effective detection range, z3 is the second interval on the Z axis, and z4 is the third interval on the Z axis. Then set all the required range parameters. Set the point cloud data range as: [Xmin, Ymin, Zmin, Xmax, Ymax, Zmax]. Points exceeding the point cloud data range will be excluded first. Since the point cloud has the characteristics of being dense at close range and sparse at long range, set the required point cloud range: [Xfilater, Yfilater]. The setting of this parameter only filters the ground points at close range. According to the coordinates of all points in the point cloud data, divide the three-dimensional point cloud data space into m*n*5 columnar grids, and the size of each grid is equal. In addition, it is also necessary to set the target distribution range in the first interval on the Z axis as [Emin, Emax] and the filtering threshold parameter T = 1. Specifically, traverse all the point cloud points in z1, and count the number of points in each grid that meet E min <Zp -Z grid ≤E max The number of points that meet the conditions. For each columnar grid, filter out the points that do not meet the condition of E min <Z p -Z grid ≤E max Compare the number threshold T of the conditions that meet the conditions. When the number of points that meet the conditions is greater than T, it is considered that there is a target in this grid, and all points of this grid are retained. Subsequently, traverse the point cloud points in the third interval, filter out the non-target points in the third interval of the retained columnar grid. If there are target points, continue to traverse the point cloud points in the second interval, filter out the non-target points in the second interval of the retained columnar grid, and all the retained point cloud points are used as the ground point cloud points. If there are no target point clouds in the second interval, retain the target point clouds in the second interval and all the point cloud points in the first interval. Put the points of all the retained grids together to complete the filtering of the ground point cloud points.

[0102] This application effectively reduces the false detection rate of ground point filtering and improves the filtering effect of non-target points through this method.

[0103] The above mainly introduced the method of the embodiment of the present application in combination with the drawings. It should be noted that all the numerical values mentioned above are only for examples and do not constitute specific limitations to the present application. At the same time, it should be understood that although the steps in the flowcharts involved in the above-mentioned embodiments are shown in sequence, these steps are not necessarily executed in the order shown in the figures. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or steps in other steps. The device of the embodiment of the present application will be introduced below in combination with the drawings. For the sake of brevity, appropriate omissions will be made when introducing the device below, and the relevant content can be referred to the relevant descriptions in the above method introduction and will not be repeated.

[0104] Figure 10 It is a schematic structural diagram of a ground point cloud point determination device 900 based on lidar provided by an embodiment of the present application. For the convenience of description, only the parts related to the embodiment of the present application are shown.

[0105] A ground point cloud point determination device based on lidar includes:

[0106] The preprocessing unit 1001 divides all the point cloud points to be determined into multiple cylindrical grids according to their coordinate positions. The point cloud points to be determined are the point cloud points collected by the lidar within the ground point detection range, and all the cylindrical grids together form the ground point detection range.

[0107] The storage unit 1002 divides the cylindrical grid into a first interval, a second interval, and a third interval along the Z-axis direction. The cylindrical grid corresponds to five-dimensional parameters. The first parameter is used to store the z-axis coordinate value of the lowest point in the corresponding grid. The second parameter is used to store the number of points in the corresponding grid whose height difference between the points in the first interval and the lowest point in the grid is within the target point height distribution range. The third parameter is used to store the height of the lowest non-target point above the ground. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset. The fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset.

[0108] The processing unit 1003 determines the cylindrical grids to be retained according to the second parameter.

[0109] The output unit 1004 eliminates the non-target points within the retained cylindrical grids according to the fourth parameter and the fifth parameter, and uses all the finally retained point cloud points as the ground point cloud points.

[0110] Figure 11 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As Figure 11 shown, the electronic device 1100 in this embodiment includes: at least one processor 1101 ( Figure 11 only one is shown), a memory 1102, and a computer program 1103 stored in the memory 1102 and executable on at least one processor 1101. When the processor 1101 executes the computer program 1103, the steps in the above embodiment are implemented.

[0111] The processor 1101 may be a central processing unit (CPU). The processor 1101 may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0112] The memory 1102 may be an internal storage unit of the electronic device 1100 in some embodiments, such as the hard disk or memory of the electronic device 1100. The memory 1102 may also be an external storage device of the electronic device 1100 in other embodiments, such as a plug-in hard disk equipped on the electronic device 1100, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory 1102 may also include both the internal storage unit of the electronic device 1100 and external storage devices. The memory 1102 is used to store an operating system, application programs, Boot Loader data, and other programs, such as the program code of a computer program. The memory 1102 may also be used to temporarily store data that has been output or will be output.

[0113] It should be noted that for the information interaction, execution process, etc. between the above units, since they are based on the same concept as the method embodiments of the present application, their specific functions and the technical effects brought can be specifically referred to the method embodiment part, and will not be elaborated here.

[0114] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the above division of each functional unit is used as an example for illustration. In practical applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit in the embodiment can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit are only for the convenience of mutual distinction and do not limit the protection scope of the present application. The specific working process of the units in the above system can refer to the corresponding process in the foregoing method embodiment and will not be elaborated here.

[0115] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments can be implemented.

[0116] The embodiment of the present application provides a computer program product. When the computer program product runs on a computer, it can implement the above various methods.

[0117] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of this application, a computer program can be used to instruct relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the photographing device / electronic device, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk, or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0118] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do 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 to the implementation process of the embodiments of this application. In the description, specific details such as a specific system structure and technology are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of this application.

[0119] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0120] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0121] In the embodiments provided in the present application, it should be understood that the disclosed device / network device and method can be implemented in other ways. For example, the device / network device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, 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 displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.

[0122] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0123] 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 foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for determining ground point cloud points based on hierarchical statistics, characterized in that, Including: Dividing all the to-be-determined point cloud points into multiple cylindrical grids according to their coordinate positions, where the to-be-determined point cloud points are the point cloud points collected by the lidar within the ground point detection range, and all the cylindrical grids together form the ground point detection range; Dividing the cylindrical grid into a first interval, a second interval, and a third interval along the Z-axis direction. The cylindrical grid corresponds to five-dimensional parameters. The first parameter is used to store the z-axis coordinate value of the lowest point in the corresponding grid. The second parameter is used to store the number of points in the corresponding grid whose height difference between the points in the first interval and the lowest point in the grid is within the target point height distribution range. The third parameter is used to store the height of the lowest non-target point above the ground. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset. The fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset; Determining the retained cylindrical grids according to the second parameter; According to the fourth parameter and the fifth parameter, removing non-target points within the retained cylindrical grids, and taking all the finally retained point cloud points as the ground point cloud points.

2. The method for determining ground point cloud points based on hierarchical statistics according to claim 1, wherein The method further includes: Filtering out the point cloud point spatial dataset within the ground point detection range from the point cloud data collected by the lidar.

3. The method for determining ground point cloud points based on hierarchical statistics according to claim 1, wherein The determining the retained cylindrical grids according to the second parameter includes: Sorting all the point cloud points in the first interval of each cylindrical grid according to the height value, and determining the to-be-determined point cloud point with the lowest height value; Determining the retained cylindrical grids according to the height value of the to-be-determined point cloud point with the lowest height value corresponding to the first interval of each cylindrical grid and the height values of all the point cloud points in the first interval of each cylindrical grid.

4. The method for determining ground point cloud points based on hierarchical statistics according to claim 1, wherein, The determining the retained cylindrical grids according to the second parameter further includes: For each first interval of each cylindrical grid, repeatedly comparing the height values of any two point cloud points, and comparing the point cloud point with the lower height value with any other point cloud point until all the point cloud points are compared. Taking the point cloud point with the lower height value obtained in the last comparison as the to-be-determined point cloud point with the lowest height value; Determining whether each cylindrical grid is retained according to the height value of the to-be-determined point cloud point with the lowest height value corresponding to the first interval of each cylindrical grid and the height values of all the point cloud points in the first interval of each cylindrical grid.

5. Any method for determining ground point cloud points based on hierarchical statistics according to any one of claims 1-4, characterized in that The determining whether each cylindrical grid is retained according to the height value of the to-be-determined point cloud point with the lowest height value corresponding to the first interval of each cylindrical grid and the height values of all the point cloud points in the first interval of each cylindrical grid includes: For each first interval of each cylindrical grid, determining the number of qualified point cloud points in the first interval of each cylindrical grid according to the difference between the height value of the to-be-determined point cloud point with the lowest height value and the height values of each to-be-determined point cloud point in the first interval of the cylindrical grid, in combination with the preset target height distribution range; Determining whether each cylindrical grid is retained according to the number of qualified point cloud points in the first interval of each cylindrical grid, in combination with the preset filtering threshold parameter.

6. The method for determining ground point cloud points based on hierarchical statistics according to claim 5, wherein Determining the number of qualified point cloud points in the first interval of each cylinder grid by combining the difference between the height value of the to-be-determined point cloud point with the lowest corresponding height value and the height values of each to-be-determined point cloud point in the first interval of the cylinder grid, and in combination with a preset target height distribution range, includes: For each to-be-determined point cloud point, determine whether the difference is within the target height distribution range; If so, regard the corresponding to-be-determined point cloud point as a qualified point cloud point; Calculate the number of qualified point cloud points in the first interval of each cylinder grid.

7. The method for determining ground point cloud points based on hierarchical statistics according to claim 5, wherein Determining whether to retain each cylinder grid by combining the number of qualified point cloud points in the first interval of each cylinder grid and a preset filtering threshold parameter, includes: Determine whether the number of qualified point cloud points corresponding to the first interval of each cylinder grid is higher than the filtering threshold parameter. If it is higher, determine to retain the cylinder grid.

8. The method for determining ground point cloud points based on hierarchical statistics according to claim 1, wherein, Removing non-target points within the retained cylinder grids according to the fourth parameter and the fifth parameter, and regarding all the finally retained point cloud points as the ground point cloud points includes: Traverse the point cloud points in the fifth parameter, remove non-target points within the fifth parameter of the retained cylinder grid. If there are target points, continue to traverse the point cloud points in the fourth parameter, remove non-target points within the fourth parameter of the retained cylinder grid, and regard all the retained point cloud points as the ground point cloud points.

9. A ground point cloud point determination device based on hierarchical statistics, characterized in that Includes: A preprocessing unit divides all to-be-determined point cloud points into multiple cylinder grids according to coordinate positions. The to-be-determined point cloud points are point cloud points collected by a lidar within the ground point detection range, and all the columnar grids form the ground point detection range; A storage unit divides the columnar grid into a first interval, a second interval, and a third interval along the Z-axis direction. The columnar grid corresponds to five-dimensional parameters. The first parameter is used to store the z-axis coordinate value of the lowest point in the corresponding grid. The second parameter is used to store the number of points whose height difference between the points in the first interval of the corresponding grid and the lowest point in the grid is within the target point height distribution range. The third parameter is used to store the height of the lowest point of non-target points above the ground. The fourth parameter is used to store the number of points in the second interval that are higher than the average height of small targets in the urban road dataset. The fifth parameter is used to store the number of points in the third interval that are higher than the average height of large targets in the urban road dataset; A processing unit determines the retained cylinder grids according to the second parameter; An output unit removes non-target points within the retained cylinder grids according to the fourth parameter and the fifth parameter, and regards all the finally retained point cloud points as the ground point cloud points.

10. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the electronic device implements the method described in any one of claims 1 to 8.

11. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by an electronic device, it implements the method described in any one of claims 1 to 8.

Images