Laser point cloud processing method and device, electronic equipment and storage medium

Abstract

The invention relates to a laser point cloud processing method and device, electronic equipment and a storage medium, and the method comprises the steps: projecting an initial point cloud to each first preset grid, and determining the proportion of low-reflectivity points of each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, the low-reflectivity points are point cloud points of which the reflectivity is smaller than a preset low-reflectivity threshold value; determining a to-be-processed point cloud according to the reflectivity of each point cloud point in the initial point cloud; determining a current weather type according to the distance between adjacent to-be-processed point cloud points in the to-be-processed point cloud, the reflectivity of each to-be-processed point cloud point and the proportion of low-reflectivity points of each first preset grid; and processing the to-be-processed point cloud according to a preset noisy point determination mode corresponding to the current weather type, and determining noisy points in the initial point cloud. The effects of improving the recognition effectiveness of foggy noise points in the laser point cloud and the accuracy of subsequent noise point filtering processing are achieved.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a laser point cloud processing method, device, electronic device and storage medium. Background Art

[0002] Since the light pulses emitted by the lidar sensor in the invisible near-infrared spectrum will not penetrate water particles, once there are foggy water particles in the air, the light pulses emitted by the sensor will be attenuated and backscattered. Therefore, in foggy weather, the lidar point cloud will contain some foggy noise points formed by pseudo echoes. These foggy noise points are easily regarded as point clouds of obstacles in the process of autonomous driving environment perception, causing the vehicle to brake incorrectly.

[0003] At present, the methods for processing foggy noise points are usually based on distance threshold, reflection intensity and deep neural network. Among them, the distance threshold method needs to determine multiple related thresholds for judgment, but all cases are processed by a unified threshold, which is prone to false detection; the reflection intensity is mainly related to the material of the reflector and the distance between the reflector and the sensor. There are some cases where the reflection intensity of the laser point cloud on some conventional obstacles is relatively low, which is also prone to false detection; the method based on deep neural network relies on a large amount of labeled data to train the model. It is difficult to obtain data for foggy scenes and the cost of data labeling is high. Summary of the invention

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the embodiments of the present disclosure provide a laser point cloud processing method, device, electronic device and storage medium, which combine the causes of fog noise, analyze the weather type, and use different methods to identify fog noise for different weather types, thereby achieving the effect of improving the effectiveness of fog noise identification and the accuracy of subsequent noise filtering processing.

[0005] In a first aspect, an embodiment of the present disclosure provides a laser point cloud processing method, the method comprising:

[0006] Projecting the initial point cloud to each first preset grid, and determining the proportion of low reflectivity points in each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, wherein the low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold;

[0007] Determining a point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud;

[0008] Determine the current weather type according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid;

[0009] The point cloud to be processed is processed according to a preset noise point determination method corresponding to the current weather type to determine the noise points in the initial point cloud.

[0010] In a second aspect, the present disclosure also provides a laser point cloud processing device, the device comprising:

[0011] A low reflectivity point ratio determination module is used to project the initial point cloud into each first preset grid, and determine the low reflectivity point ratio of each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, wherein the low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold;

[0012] A point cloud to be processed determining module, used for determining the point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud;

[0013] A current weather type determination module, used to determine the current weather type according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid;

[0014] The noise point determination module is used to process the point cloud to be processed according to a preset noise point determination method corresponding to the current weather type, and determine the noise points in the initial point cloud.

[0015] In a third aspect, an embodiment of the present disclosure further provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the laser point cloud processing method as described above.

[0016] In a fourth aspect, the embodiments of the present disclosure further provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the laser point cloud processing method as described above.

[0017] A laser point cloud processing method provided by an embodiment of the present disclosure projects an initial point cloud onto each first preset grid, determines the proportion of low-reflectivity points in each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, and then determines a point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud, so as to facilitate the subsequent finding of noise points in the point cloud to be processed. Furthermore, the current weather type is determined according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed and the proportion of low-reflectivity points in each first preset grid, so as to facilitate the selection of different methods for determining noise points according to different current weather types. According to the preset noise point determination method corresponding to the current weather type, the point cloud to be processed is processed to determine the noise points in the initial point cloud, thereby achieving the effect of improving the effectiveness of foggy noise point recognition and the accuracy of subsequent noise point filtering processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0019] Figure 1 is a flow chart of a laser point cloud processing method in an embodiment of the present disclosure;

[0020] Figure 2 A schematic diagram of calculating the number of effective connection points in an embodiment of the present disclosure;

[0021] Figure 3 A schematic diagram of calculating the number of another effective connection point in an embodiment of the present disclosure;

[0022] Figure 4 is a schematic diagram of projecting a third point cloud point onto each second preset grid in an embodiment of the present disclosure;

[0023] Figure 5 Schematic diagram of the structure of a laser point cloud device in an embodiment of the present disclosure;

[0024] Figure 6 It is a structural schematic diagram of an electronic device in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0026] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0027] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0028] Figure 1 is a flow chart of a laser point cloud processing method in an embodiment of the present disclosure. The method can be executed by a laser point cloud processing device, which can be implemented in software and / or hardware, and can be configured in an electronic device. Figure 1 As shown, the method may specifically include the following steps:

[0029] S110, projecting the initial point cloud into each first preset grid, and determining the proportion of low reflectivity points in each first preset grid according to the reflectivity of each point cloud point in the initial point cloud.

[0030] The initial point cloud is a point cloud acquired by radar. The first preset grid is a rectangular grid constructed in a rectangular coordinate system according to a preset resolution. The low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold. The low reflectivity point ratio is the ratio of the number of low reflectivity points corresponding to the initial point cloud in the first preset grid to the total number of point cloud points.

[0031] Specifically, each rectangular grid divided according to a preset resolution in a rectangular coordinate system is used as each first preset grid, and the initial point cloud is projected into each first preset grid. According to the low reflectivity threshold, the low reflectivity points in each first preset grid are determined, and the number of these low reflectivity points is determined. For each first preset grid, the ratio of the number of low reflectivity points corresponding to the first preset grid to the total number of point cloud points in the first preset grid is used as the low reflectivity point ratio of the first preset grid.

[0032] S120 , determining a point cloud to be processed according to the reflectivity of each point in the initial point cloud.

[0033] The point cloud to be processed is a set of point clouds that meet screening requirements, and the screening requirements may include reflectivity requirements, range requirements, and the like.

[0034] Specifically, according to the preset requirement of filtering the point cloud to be processed, the reflectivity of each point cloud point in the initial point cloud is screened, and the point cloud to be processed that meets the requirement is determined, so as to be used for subsequently determining the noise points therefrom.

[0035] Based on the above example, the point cloud to be processed can be determined according to the reflectivity of each point in the initial point cloud in the following way:

[0036] The initial point cloud is filtered according to a preset range to determine the process point cloud, and the process point cloud is clustered to determine multiple clusters;

[0037] For each cluster, when the proportion of high reflectivity points in the cluster is greater than the high reflectivity proportion threshold and the proportion of low reflectivity points is less than the low reflectivity proportion threshold, it is determined that the points in the cluster are marked as real object points;

[0038] The point cloud points in the process point cloud that are not marked as real object points constitute the point cloud to be processed.

[0039] Among them, the preset range is a preset area range and / or road network range. The road network range can be understood as the range of the road area. The preset area range is the range within a certain distance (such as 3m, etc.) from the measurement point. The process point cloud is the part of the initial point cloud within the preset range. The clustering clusters are multiple clusters obtained by clustering the process point cloud according to the distance between the point cloud points. The high reflectivity point is a point cloud point whose reflectivity is greater than the preset high reflectivity threshold. The real object point is the point cloud point corresponding to the real object, such as the point cloud point corresponding to the obstacle.

[0040] Specifically, the initial point cloud is filtered using a preset range, and the part within the preset range is used as the process point cloud. The process point cloud is clustered to obtain multiple clusters. For each cluster, the point cloud points in the cluster whose reflectivity is greater than the preset high reflectivity threshold are taken as high reflectivity points, and the point cloud points in the cluster whose reflectivity is less than the preset low reflectivity threshold are taken as low reflectivity points. The number of high reflectivity points and the number of low reflectivity points in the cluster are counted, and the proportion of high reflectivity points and the proportion of low reflectivity points in the cluster are divided by the total number of point cloud points in the cluster, respectively. If the proportion of high reflectivity points in the cluster is greater than the high reflectivity proportion threshold and the proportion of low reflectivity points is less than the low reflectivity proportion threshold, it can be determined that the cluster reflects a real object, and therefore, the points in the cluster are determined to be marked as real object points. Furthermore, the point cloud points in the process point cloud that are not marked as real object points constitute the point cloud to be processed.

[0041] S130, determining the current weather type according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid.

[0042] The current weather type is a weather type with different adaptability to the laser radar, for example, the current weather type may be general foggy weather, dense foggy weather, and non-foggy weather.

[0043] Specifically, the distance between each adjacent point cloud point to be processed in the point cloud to be processed is calculated, and the distance between each adjacent point cloud point to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid can be analyzed and classified according to each pre-set threshold value to obtain the corresponding category, which is the current weather type.

[0044] Based on the above example, the current weather type can be determined according to the distance between adjacent point cloud points to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid in the following manner:

[0045] Determine the number of first point cloud points according to the distance between adjacent point cloud points to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid;

[0046] The current weather type is determined according to the number of points in the first point cloud, each second preset grid, the initial point cloud, and the point cloud to be processed.

[0047] Among them, the first point cloud point is a point cloud point whose distance to each adjacent point cloud point to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid all meet the preset threshold requirements, which can be understood as a point cloud point that may be a noise point, that is, a suspected point cloud point. The second preset grid is each sector or sector ring grid constructed in the polar coordinate system according to the preset angle and preset spacing.

[0048] Specifically, the distance between each adjacent point cloud point to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid are analyzed using each preset threshold, and each point cloud point analyzed is used as the first point cloud point, and the number of first point cloud points is counted. The initial point cloud is processed using each second preset grid to obtain each point cloud point in each second preset grid, and the point cloud to be processed, the number of first point cloud points, and each point cloud point in each second preset grid are analyzed, and numerical analysis and category analysis are performed to obtain the current weather type.

[0049] Based on the above example, the number of first point cloud points can be determined according to the distance between adjacent point cloud points to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid in the following manner:

[0050] Determine the number of valid connection points for each point cloud point to be processed;

[0051] For each point cloud point to be processed, in response to the number of valid connection points of the point cloud point to be processed being less than the preset number of connection points, the reflectivity of the point cloud point to be processed being less than the low reflectivity threshold, and the proportion of low reflectivity points of the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed being greater than the preset first proportion, the point cloud point to be processed is determined to be the first point cloud point;

[0052] Determine the number of points in the first point cloud.

[0053] Among them, the valid connection points of the point cloud points to be processed satisfy: the valid connection points and the corresponding point cloud points to be processed are located on the same laser radar line beam, the point cloud points adjacent to the valid connection points in the direction close to the corresponding point cloud points to be processed are another valid connection point or the corresponding point cloud points to be processed, and the distance between the point cloud points adjacent to the valid connection points in the direction close to the corresponding point cloud points to be processed and the valid connection points is less than the preset distance. The preset number of connection points and the preset first proportion are preset thresholds. The neighborhood grids are the first preset grids adjacent to the first preset grid corresponding to the point cloud points to be processed.

[0054] In the embodiment of the present invention, the effective connection point is not the laser radar origin.

[0055] Specifically, for each point cloud point to be processed, the valid connection points corresponding to the point cloud point to be processed are determined according to the conditions that the valid connection points of the point cloud point to be processed need to meet, and then the number of valid connection points of the point cloud point to be processed can be obtained by counting. For each point cloud point to be processed, if the number of valid connection points of the point cloud point to be processed is less than the preset number of connection points, the reflectivity of the point cloud point to be processed is less than the low reflectivity threshold (such as 5, etc.), and the proportion of low reflectivity points of the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed is greater than the preset first proportion, then the point cloud point to be processed may be a noise point, and therefore, the point cloud point to be processed is determined to be the first point cloud point, and the number of the first point cloud points is obtained by counting.

[0056] For example, the schematic diagram of calculating the number of effective connection points is as follows: Figure 2 and Figure 3 As shown. Figure 2In , the number of valid connection points of P1 is 0; the adjacent point cloud point on the left side of P2 is the radar origin, so the number of valid connection points on the left side is 0, and the number of valid connection points on the right side of P2 is 2, so the number of valid connection points of P2 is 2; the number of valid connection points on the left side of P3 is 1, and the number of valid connection points on the right side of P3 is also 1, so the number of valid connection points of P3 is 2; the number of valid connection points on the left side of P4 is 2, and there is no valid connection point on the right side, so the number of valid connection points of P4 is 2. Figure 3 In the figure, the point cloud point adjacent to the left side of P1 is the radar origin, which is not a valid connection point. Therefore, the number of valid connection points on the left side is 0. The distance between the adjacent point cloud point P2 on the right side of P1 and P1 is greater than the preset distance, so it is not a valid connection point either. The number of valid connection points on the right side is 0, so the number of valid connection points of P1 is 0; the distance between the adjacent point cloud point P1 on the left side of P2 and P2 is greater than the preset distance, so it is not a valid connection point. The number of valid connection points on the left side of P2 is 0. The distance between the adjacent point cloud point P3 on the right side of P2 and P2 is The distance between P2 and P3 is less than the preset distance, P3 is the valid connection point of P2, but the distance between P4 and P3 is greater than the preset distance, P4 is not the valid connection point of P2, therefore, the number of valid connection points on the right side of P2 is 1, so the number of valid connection points of P2 is 1; the distance between the adjacent point cloud points P2 and P3 on the left side of P3 is less than the preset distance, P2 is the valid connection point of P3, but the distance between P1 and P2 is greater than the preset distance, P1 is not the valid connection point of P3, therefore, the number of valid connection points on the left side of P3 is 1, and the number of valid connection points on the right side of P3 is 1. The distance between adjacent point cloud points P4 and P3 is greater than the preset distance, so they are not valid connection points. The number of valid connection points on the right side of P3 is 0, so the number of valid connection points of P3 is 1; the distance between adjacent point cloud points P3 and P4 on the left side of P4 is greater than the preset distance, so they are not valid connection points. The number of valid connection points on the left side of P4 is 0. The distance between adjacent point cloud points P5 and P4 on the right side of P4 is less than the preset distance. P5 is a valid connection point of P4. The adjacent point cloud point on the right side of P5 is the radar origin, which is not a valid connection point of P4. Therefore, the number of valid connection points on the right side of P4 is 1, so the number of valid connection points of P4 is 1; the distance between the adjacent point cloud points P4 and P5 on the left side of P5 is less than the preset distance, and P4 is a valid connection point of P5, but the distance between P3 and P4 is greater than the preset distance, and P3 is not a valid connection point of P5, so the number of valid connection points on the left side of P5 is 1, and the adjacent point cloud point on the right side of P5 is the radar origin, which is not a valid connection point, so the number of valid connection points on the right side is 0, so the number of valid connection points of P5 is 1.

[0057] It can be understood that the first point cloud point meets the following conditions: the initial point cloud collected by the lidar is within a certain range nearby, within the road network area, is not the point cloud of the vehicle itself and is not the origin of the lidar; the reflectivity is less than the low reflectivity threshold; the corresponding marking result is a non-real object point; the proportion of low reflectivity points in the corresponding first preset grid and the neighborhood grid is greater than the preset first proportion (such as 0.9); the number of corresponding valid connection points is less than the preset number of connection points.

[0058] Based on the above example, the current weather type can be determined according to the number of points in the first point cloud, each second preset grid, the initial point cloud, and the point cloud to be processed in the following manner:

[0059] For each point cloud point to be processed, if the reflectivity of the point cloud point to be processed is less than the low reflectivity threshold and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed is greater than the preset second proportion, then the point cloud point to be processed is determined to be the second point cloud point;

[0060] determining the number of points in the second point cloud;

[0061] The ratio of the number of points in the first point cloud to the number of points in the second point cloud is used as a reference value;

[0062] In response to the number of the first point cloud points not being equal to zero, the number of the first point cloud points being less than a first number threshold, and the reference value being greater than a preset ratio, determining that the current weather type is general foggy weather;

[0063] In response to the number of the first point cloud points being greater than the second number threshold, determining that the current weather type is dense fog weather;

[0064] In response to the number of the first point cloud points being greater than or equal to the first number threshold and less than or equal to the second number threshold, projecting the third point cloud points into each second preset grid;

[0065] Determining a valid grid from each second preset grid;

[0066] Determine the first target circle with the largest number of valid grids;

[0067] Determine the angle of the largest connected domain formed by valid grids in the first target ring;

[0068] Determine the average point cloud density of valid grids in the first target ring;

[0069] In response to the angle of the maximum connected domain being greater than a preset angle, and the average point cloud density being greater than a preset point cloud density, it is determined that the current weather type is dense fog weather.

[0070] In addition, in one embodiment of the present invention, if it is determined that the current weather type is neither ordinary foggy weather nor dense foggy weather, the current weather type is determined to be non-foggy weather.

[0071] Among them, the second point cloud point is a point cloud point whose reflectivity is less than the low reflectivity threshold and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed is greater than the preset second proportion. The third point cloud point is a point cloud point whose reflectivity is less than the low reflectivity threshold in the initial point cloud, and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the third point cloud point is greater than the preset third proportion. The reference value is the ratio of the number of first point cloud points to the number of second point cloud points. The second preset grid is a polar coordinate grid with the laser radar as the origin, and the second preset grid constitutes a plurality of circular rings. The valid grid is the second preset grid containing at least one third point cloud point. The first target circular ring is a circular ring containing the largest number of valid grids among the circular rings. The average point cloud density is the ratio of the number of point cloud points in each valid grid in the first target circular ring to the number of valid grids in the first target circular ring. The first quantity threshold, the second quantity threshold, the preset second proportion, the preset third proportion, the preset ratio, the preset angle and the preset point cloud density are all preset values, among which the preset first proportion, the preset second proportion and the preset third proportion may be the same or different.

[0072] Specifically, if the number of the first point cloud points is not equal to zero and the number of the first point cloud points is less than the first number threshold, it is necessary to continue to determine whether the current weather type is dense fog weather. Therefore, for each point cloud point to be processed, if the reflectivity of the point cloud point to be processed is less than the low reflectivity threshold and the first preset grid corresponding to the point cloud point to be processed and the proportion of low reflectivity points in each neighborhood grid are greater than the preset second proportion, then the point cloud point to be processed is taken as the second point cloud point, and the number of the second point cloud points is obtained by counting. The ratio of the number of the first point cloud points to the number of the second point cloud points is used as a reference value. If the number of the first point cloud points is not equal to zero, the number of the first point cloud points is less than the first number threshold and the reference value is greater than the preset ratio, it is determined that the current weather type is general foggy weather. If the number of the first point cloud points is greater than the second number threshold, it can be directly determined that the current weather type is dense foggy weather. If the number of the first point cloud points is greater than or equal to the first number threshold and less than or equal to the second number threshold, it is necessary to distinguish between dense fog weather and general foggy weather. Accordingly, it is necessary to project the third point cloud point into each second preset grid, such as Figure 4 As shown, Figure 4Each second preset grid is not marked, the sector ring covered by the dotted line is a second preset grid, and each sector ring with the same distance from the center of the circle constitutes a circular ring, and the center angle of each second preset grid is the same. Then, the valid grids are determined from each second preset grid, the number of valid grids in each circular ring is counted, and the circular ring with the largest number of valid grids is taken as the first target circular ring. The first target circular ring is analyzed to determine the angle of the maximum connected domain formed by the valid grids and the average point cloud density of the valid grids in the first target circular ring. If the angle of the maximum connected domain is greater than the preset angle, and the average point cloud density is greater than the preset point cloud density, it is determined that the current weather type is dense fog weather.

[0073] It should be noted that the current weather type that does not belong to any of the above situations is non-foggy day. For non-foggy day, there is no need to determine and filter the noise points corresponding to foggy day. The subsequent steps can be stopped to determine and filter other noise points.

[0074] It can be understood that the second point cloud point meets the following conditions: the initial point cloud collected by the laser radar is within a certain range nearby, within the road network area, is not a self-vehicle point cloud and is not the laser radar origin; the reflectivity is less than the low reflectivity threshold; the corresponding marking result is a non-real object point; the corresponding first preset grid and the proportion of low reflectivity points in the neighborhood grid is greater than the preset second proportion (such as 0.9). The third point cloud point meets the following conditions: the initial point cloud collected by the laser radar is within a certain range nearby, is not a self-vehicle point cloud and is not the laser radar origin; the reflectivity is less than the low reflectivity threshold; the corresponding first preset grid and the proportion of low reflectivity points in the neighborhood grid is greater than the preset third proportion (such as 0.9).

[0075] Exemplarily, the number of the first point cloud points is recorded as N, and the number of the second point cloud points is recorded as M. It can be seen that the reference value is N / M. When N is not 0 and is less than or equal to the first quantity threshold (such as 100, etc.), continue to judge whether it is a general foggy day. If N / M is greater than a preset ratio (such as 0.6, etc.), it is determined that the current weather type is a general foggy day. If N is greater than the first quantity threshold, it starts to judge whether it is dense fog weather or general foggy day. If N is greater than the second quantity threshold (such as 600, etc.), it is directly determined that the current weather type is dense fog weather. If N is greater than or equal to the first quantity threshold and less than or equal to the second quantity threshold. Further judgment is required to project the third point cloud point into each second preset grid to determine each valid grid. Traverse each ring, determine the number of valid grids in each ring, and take the ring with the largest number of valid grids as the first target ring. The maximum connected domain formed by the valid grids in the first target circle is calculated, and the angle of connection of the maximum connected domain is calculated. In addition, the average point cloud density of the valid grids in the first target circle is calculated. If the angle of connection of the maximum connected domain is greater than a preset angle (such as 90°, etc.), and the average point cloud density is greater than a preset point cloud density (such as 8, etc.), the current weather type is determined to be dense fog. Furthermore, the current weather type corresponding to other unmentioned branches can be determined to be non-foggy.

[0076] S140: Process the point cloud to be processed according to a preset noise point determination method corresponding to the current weather type, and determine the noise points in the initial point cloud.

[0077] The preset noise point determination method is a method for determining noise points corresponding to the current weather type, and different current weather types correspond to different preset noise point determination methods.

[0078] Specifically, a correspondence between various current weather types and various preset noise point determination methods can be established in advance, and the correspondence is one-to-one. Therefore, according to the current weather type determined in the previous step, the corresponding preset noise point determination method can be determined from the correspondence. Then, using the determined preset noise point determination method, the foggy noise point identification processing can be performed on the processing point cloud, and the processing point cloud can be processed for subsequent filtering and removal.

[0079] Based on the above example, when it is determined that the current weather type is a general foggy day, the point cloud to be processed can be processed according to the preset noise point determination method corresponding to the current weather type in the following manner to determine the noise points in the initial point cloud:

[0080] Projecting the point cloud points to be processed onto a third preset grid;

[0081] Determine the second target ring with the largest number of non-empty grids;

[0082] The second target circle and its adjacent circles are regarded as noise areas;

[0083] For each point cloud point to be processed in the noise point area, in response to the reflectivity of the point cloud point to be processed being less than or equal to the first threshold, the point cloud point to be processed is determined to be a noise point in the initial point cloud; in response to the reflectivity of the point cloud point to be processed being greater than the first threshold and less than the second threshold, if the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset first number of points, the point cloud point to be processed is determined to be a noise point in the initial point cloud;

[0084] For each point cloud point to be processed in other circular rings outside the noise point area, in response to the reflectivity of the point cloud point to be processed being less than or equal to the low reflectivity threshold, the number of valid connection points corresponding to the point cloud point to be processed being less than or equal to the preset second number of points, and the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed being greater than the preset fifth proportion, the point cloud point to be processed is determined to be a noise point in the initial point cloud; in response to the reflectivity of the point cloud point to be processed being less than or equal to the low reflectivity threshold and the number of valid connection points corresponding to the point cloud point to be processed being less than the preset first number of points, the point cloud point to be processed is determined to be a noise point in the initial point cloud.

[0085] Among them, the third preset grid is a polar coordinate grid with the laser radar as the origin, and the third preset grid constitutes a plurality of circular rings. The non-empty grid is the third preset grid having a point cloud point to be processed whose reflectivity is less than the low reflectivity threshold. The valid connection points of the point cloud points to be processed meet the conditions of the above example. The second target circular ring is the circular ring with the largest number of non-empty grids obtained by projecting the third preset grid. The noise area is the second target circular ring and its adjacent circular rings. The first threshold, the second threshold, the preset first number of points, the preset second number of points, the preset fourth proportion, and the preset fifth proportion are all preset values, wherein the preset fourth proportion is less than the preset fifth proportion, and the preset first number of points is less than the preset second number of points.

[0086] Specifically, the point cloud points to be processed are projected into the third preset grid, the number of non-empty grids in each third preset grid can be determined, the circular ring with the largest number of non-empty grids is used as the second target circular ring, and the second target circular ring and its adjacent circular rings are used as the noise area. For each point cloud point to be processed in the noise area, if the reflectivity of the point cloud point to be processed is less than or equal to the first threshold, the point cloud point to be processed is directly determined to be a noise point in the initial point cloud; if the reflectivity of the point cloud point to be processed is greater than the first threshold and less than the second threshold, then if the proportion of low reflectivity points in the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset first number of points, the point cloud point to be processed is used as a noise point in the initial point cloud. For each point cloud point to be processed in other circular rings outside the noise point area, if the point cloud point to be processed satisfies the following three conditions: the reflectivity is less than or equal to the low reflectivity threshold, the number of valid connection points corresponding to the point cloud point to be processed is less than or equal to the preset second number of points, and the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fifth proportion, then the point cloud point to be processed can be determined as a noise point in the initial point cloud. Moreover, for each point cloud point to be processed in other circular rings outside the noise point area, if the reflectivity of the point cloud point to be processed is less than or equal to the low reflectivity threshold and the number of valid connection points corresponding to the point cloud point to be processed is less than the preset first number of points, then the point cloud point to be processed can also be determined as a noise point in the initial point cloud.

[0087] Exemplarily, first, the point cloud points to be processed are projected to the third preset grid, and the third preset grid with the point cloud points to be processed whose reflectivity is less than the low reflectivity threshold (such as 5, etc.) is taken as the number of non-empty grids. Then, the position of the ring with the largest number of non-empty grids is counted, which is the second target ring. Secondly, for each point cloud point to be processed in the second target ring and its adjacent rings (noise area) before and after, if the reflectivity is less than or equal to the first threshold (such as 8, etc.), then the point cloud point to be processed is set as a foggy noise point (noise point in the initial point cloud); if the reflectivity is greater than the first threshold and less than the second threshold (such as 13, etc.), the proportion of low reflectivity points in the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion (such as 0.5, etc.), or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset first number of points (such as 3, etc.), then the point cloud point to be processed is set as a foggy noise point (noise point in the initial point cloud). For each point cloud point to be processed in other circular rings outside the noise area, if the reflectivity is less than or equal to the low reflectivity threshold (such as 5) and the number of valid connection points corresponding to the point cloud point to be processed is less than or equal to the preset second number of points (such as 5, etc.) and the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fifth proportion (such as 0.8, etc.), then the point cloud point to be processed is set as a foggy noise point (noise point in the initial point cloud). For each point cloud point to be processed in other circular rings outside the noise area, if the reflectivity is less than or equal to the low reflectivity threshold (such as 5) and the number of valid connection points corresponding to the point cloud point to be processed is less than the preset first number of points (such as 3, etc.), then the point cloud point to be processed is also set as a foggy noise point (noise point in the initial point cloud).

[0088] Based on the above example, when it is determined that the current weather type is dense fog, the point cloud to be processed can be processed according to the preset noise point determination method corresponding to the current weather type in the following manner to determine the noise points in the initial point cloud:

[0089] In response to the reflectivity of the point cloud point to be processed being less than or equal to the first threshold, determining the point cloud point to be processed to be a noise point in the initial point cloud;

[0090] In response to the reflectivity of the point cloud point to be processed being greater than the first threshold and less than the second threshold, if the proportion of low-reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset third number of points, then the point cloud point to be processed is determined to be a noise point in the initial point cloud.

[0091] The valid connection points of the point cloud points to be processed still meet the conditions of the above example. The third number of points is preset to a preset value.

[0092] Specifically, each point cloud point to be processed is analyzed separately, and if the reflectivity of the point cloud point to be processed is less than or equal to the first threshold, the point cloud point to be processed is determined as a noise point in the initial point cloud. If the reflectivity of the point cloud point to be processed is greater than the first threshold and less than the second threshold, further judgment is performed, and if the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset third number of points, the point cloud point to be processed is determined as a noise point in the initial point cloud.

[0093] Exemplarily, for each point cloud point to be processed, if the reflectivity is a first threshold value (such as 8, etc.), it is set as a foggy noise point (noise point in the initial point cloud) for the point cloud point to be processed; or, if the reflectivity is greater than the first threshold value and less than the second threshold value (such as 13, etc.), and the proportion of low-reflectivity points in the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion (such as 0.5, etc.), then the point cloud point to be processed is set as a foggy noise point (noise point in the initial point cloud); or, if the reflectivity is greater than the first threshold value and less than the second threshold value (such as 13, etc.), and the number of valid connection points corresponding to the point cloud point to be processed is less than the preset third number of points (such as 2, etc.), then the point cloud point to be processed is set as a foggy noise point (noise point in the initial point cloud).

[0094] The laser point cloud processing method provided in the present embodiment projects the initial point cloud into each first preset grid, determines the proportion of low-reflectivity points in each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, and then determines the point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud, so as to find noise points in the point cloud to be processed later. Furthermore, the current weather type is determined according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed and the proportion of low-reflectivity points in each first preset grid, so as to select different methods to determine the noise points according to different current weather types. According to the preset noise point determination method corresponding to the current weather type, the point cloud to be processed is processed to determine the noise points in the initial point cloud, thereby achieving the effect of improving the effectiveness of foggy noise point recognition and the accuracy of subsequent noise point filtering processing.

[0095] Figure 5 FIG. 1 is a schematic diagram of the structure of a laser point cloud device in an embodiment of the present disclosure. Figure 5 As shown, the device includes: a low reflectivity point ratio determination module 210, a to-be-processed point cloud determination module 220, a current weather type determination module 230 and a noise point determination module 240.

[0096] Among them, the low reflectivity point ratio determination module 210 is used to project the initial point cloud to each first preset grid, and determine the low reflectivity point ratio of each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, wherein the low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold; the to-be-processed point cloud determination module 220 is used to determine the to-be-processed point cloud according to the reflectivity of each point cloud point in the initial point cloud; the current weather type determination module 230 is used to determine the current weather type according to the distance between each adjacent to-be-processed point cloud point in the to-be-processed point cloud, the reflectivity of each to-be-processed point cloud point and the low reflectivity point ratio of each first preset grid; the noise point determination module 240 is used to process the to-be-processed point cloud according to the preset noise point determination method corresponding to the current weather type, and determine the noise points in the initial point cloud.

[0097] Based on the above example, optionally, the module 220 for determining the point cloud to be processed is also used to filter the initial point cloud according to a preset range, determine the process point cloud, and cluster the process point cloud to determine multiple cluster clusters; for each cluster cluster, when the proportion of high reflectivity points in the cluster cluster is greater than the high reflectivity proportion threshold and the proportion of low reflectivity points is less than the low reflectivity proportion threshold, determine that the points in the cluster cluster are marked as real object points, wherein the high reflectivity points are point cloud points whose reflectivity is greater than the preset high reflectivity threshold; the point cloud points in the process point cloud that are not marked as real object points constitute the point cloud to be processed.

[0098] Based on the above example, optionally, the current weather type determination module 230 is also used to determine the number of first point cloud points based on the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid; determine the current weather type based on the number of the first point cloud points, each second preset grid, the initial point cloud and the point cloud to be processed.

[0099] On the basis of the above example, optionally, the current weather type determination module 230 is further used to determine the number of valid connection points of each of the point cloud points to be processed, wherein the valid connection points of the point cloud points to be processed satisfy: the valid connection point and the corresponding point cloud point to be processed are located on the same laser radar line beam, the point cloud point adjacent to the valid connection point in the direction close to the corresponding point cloud point to be processed is another valid connection point or the corresponding point cloud point to be processed, and the distance between the point cloud point adjacent to the valid connection point in the direction close to the corresponding point cloud point to be processed and the valid connection point is less than a preset distance; for each point cloud point to be processed, in response to the number of valid connection points of the point cloud point to be processed being less than the preset number of connection points, the reflectivity of the point cloud point to be processed being less than the low reflectivity threshold, and the proportion of low reflectivity points of the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed being greater than the preset first proportion, the point cloud point to be processed is determined to be the first point cloud point; wherein the neighborhood grids are each first preset grid adjacent to the first preset grid corresponding to the point cloud point to be processed; and the number of the first point cloud points is determined.

[0100] On the basis of the above example, optionally, the current weather type determination module 230 is also used to, for each point cloud point to be processed, if the reflectivity of the point cloud point to be processed is less than the low reflectivity threshold and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed is greater than the preset second proportion, then determine that the point cloud point to be processed is a second point cloud point; determine the number of the second point cloud points; take the ratio of the number of the first point cloud points to the number of the second point cloud points as a reference value; in response to the number of the first point cloud points not being equal to zero, the number of the first point cloud points being less than the first quantity threshold and the reference value being greater than the preset ratio, determine that the current weather type is general foggy weather; in response to the number of the first point cloud points being greater than the second quantity threshold, determine that the current weather type is dense foggy weather; in response to the number of the first point cloud points being greater than or equal to the first quantity threshold and less than or equal to the a second quantity threshold, the third point cloud point is projected into each second preset grid, wherein the third point cloud point is a point cloud point whose reflectivity in the initial point cloud is less than the low reflectivity threshold, and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the third point cloud point is greater than the preset third proportion, the second preset grid is a polar coordinate grid with the laser radar as the origin, and the second preset grid constitutes a plurality of circular rings; a valid grid is determined from each second preset grid, and the valid grid is a second preset grid containing at least one third point cloud point; a first target circular ring with the largest number of valid grids is determined; an angle of the maximum connected domain formed by the valid grids in the first target circular ring is determined; an average point cloud density of the valid grids in the first target circular ring is determined; in response to the angle of the maximum connected domain being greater than the preset angle, and the average point cloud density being greater than the preset point cloud density, the current weather type is determined to be dense fog.

[0101] Based on the above example, optionally, if the current weather type is general foggy weather, the noise point determination module 240 is further used to project the point cloud points to be processed onto a third preset grid, wherein the third preset grid is a polar coordinate grid with the laser radar as the origin, and the third preset grid constitutes a plurality of circular rings; determine a second target circular ring with the largest number of non-empty grids, wherein the non-empty grid is a third preset grid with point cloud points to be processed having a reflectivity less than a low reflectivity threshold; use the second target circular ring and its adjacent circular rings as noise point areas; for each point cloud point to be processed in the noise point area, Processing point cloud points, in response to the reflectivity of the point cloud points to be processed being less than or equal to the first threshold, determining that the point cloud points to be processed are noise points in the initial point cloud; in response to the reflectivity of the point cloud points to be processed being greater than the first threshold and less than the second threshold, if the proportion of low reflectivity points of the first preset grid corresponding to the point cloud points to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud points to be processed is less than the preset first number of points, determining that the point cloud points to be processed are noise points in the initial point cloud; for each point cloud point to be processed in other circular rings outside the noise point area In response to the reflectivity of the point cloud point to be processed being less than or equal to the low reflectivity threshold, the number of valid connection points corresponding to the point cloud point to be processed being less than or equal to the preset second number of points, and the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed being greater than the preset fifth proportion, it is determined that the point cloud point to be processed is a noise point in the initial point cloud; in response to the reflectivity of the point cloud point to be processed being less than or equal to the low reflectivity threshold and the number of valid connection points corresponding to the point cloud point to be processed being less than the preset first number of points, it is determined that the point cloud point to be processed is a noise point in the initial point cloud. noise points; wherein, the preset fourth proportion is less than the preset fifth proportion; the preset first number of points is less than the preset second number of points; the effective connection point of the point cloud point to be processed satisfies: the effective connection point and the corresponding point cloud point to be processed are located on the same laser radar line beam, the point cloud point adjacent to the effective connection point in the direction close to the corresponding point cloud point to be processed is another effective connection point or the corresponding point cloud point to be processed, and the distance between the point cloud point adjacent to the effective connection point and the effective connection point in the direction close to the corresponding point cloud point to be processed is less than the preset distance.

[0102] On the basis of the above example, optionally, if the current weather type is dense fog weather, the noise point determination module 240 is also used to determine that the point cloud point to be processed is a noise point in the initial point cloud in response to the reflectivity of the point cloud point to be processed being less than or equal to the first threshold; in response to the reflectivity of the point cloud point to be processed being greater than the first threshold and less than the second threshold, if the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset third number of points, then the point cloud point to be processed is determined to be a noise point in the initial point cloud; wherein the valid connection point of the point cloud point to be processed satisfies: the valid connection point and the corresponding point cloud point to be processed are located on the same laser radar line beam, the point cloud point adjacent to the valid connection point in the direction close to the corresponding point cloud point to be processed is another valid connection point or the corresponding point cloud point to be processed, and the distance between the point cloud point adjacent to the valid connection point in the direction close to the corresponding point cloud point to be processed and the valid connection point is less than the preset distance.

[0103] The laser point cloud device provided in the embodiment of the present disclosure can execute the steps in the laser point cloud method provided in the method embodiment of the present disclosure, and the execution steps and beneficial effects are no longer repeated here.

[0104] Figure 6 Schematic diagram of the structure of an electronic device in the embodiment of the present disclosure. Figure 6 , which shows a structural schematic diagram of an electronic device 300 suitable for implementing the embodiments of the present disclosure. Figure 6 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0105] like Figure 6 As shown, the electronic device 300 may include a processing device 301, a read-only memory (ROM) 302, a random access memory (RAM) 303, a bus 304, an input / output (I / O) interface 305, an input device 306, an output device 307, a storage device 308, and a communication device 309. The processing device (e.g., a central processing unit, a graphics processor, etc.) 301 can perform various appropriate actions and processes to implement the method of the embodiment described in the present disclosure according to the program in the ROM 302 or the program loaded from the storage device 308 to the RAM 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through the bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0106] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart, thereby implementing the laser point cloud processing method as described above. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above functions defined in the method of the embodiment of the present disclosure are executed.

[0107] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0108] The computer-readable medium may be included in the electronic device, or may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device:

[0109] Projecting the initial point cloud to each first preset grid, and determining the proportion of low reflectivity points in each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, wherein the low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold;

[0110] Determining a point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud;

[0111] Determine the current weather type according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid;

[0112] The point cloud to be processed is processed according to a preset noise point determination method corresponding to the current weather type to determine the noise points in the initial point cloud.

[0113] Optionally, when the above one or more programs are executed by the electronic device, the electronic device may also execute other steps described in the above embodiments.

[0114] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0115] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

Claims

1. A laser point cloud processing method, characterized in that: The method comprises: Projecting the initial point cloud to each first preset grid, and determining the proportion of low reflectivity points in each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, wherein the low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold; Determining a point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud; Determine the current weather type according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid; The point cloud to be processed is processed according to a preset noise point determination method corresponding to the current weather type to determine the noise points in the initial point cloud.

2. The method according to claim 1, characterized in that The step of determining the point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud comprises: Filtering the initial point cloud according to a preset range to determine a process point cloud, and clustering the process point cloud to determine a plurality of clusters; For each cluster, when the proportion of high reflectivity points in the cluster is greater than the high reflectivity proportion threshold and the proportion of low reflectivity points is less than the low reflectivity proportion threshold, it is determined that the points in the cluster are marked as real object points, wherein the high reflectivity points are point cloud points whose reflectivity is greater than a preset high reflectivity threshold; Point cloud points in the process point cloud that are not marked as real object points constitute the point cloud to be processed.

3. The method according to claim 1, characterized in that The determining the current weather type according to the distance between adjacent to-be-processed point cloud points in the to-be-processed point cloud, the reflectivity of each to-be-processed point cloud point, and the proportion of low-reflectivity points in each first preset grid includes: Determine the number of first point cloud points according to the distance between adjacent point cloud points to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid; The current weather type is determined according to the number of points in the first point cloud, each second preset grid, the initial point cloud and the point cloud to be processed.

4. The method according to claim 3, characterized in that The determining the number of first point cloud points according to the distance between adjacent point cloud points to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid includes: Determine the number of valid connection points of each of the to-be-processed point cloud points, wherein the valid connection points of the to-be-processed point cloud points satisfy: the valid connection point and the corresponding to-be-processed point cloud point are located on the same laser radar line beam, the point cloud point adjacent to the valid connection point in the direction close to the corresponding to-be-processed point cloud point is another valid connection point or the corresponding to-be-processed point cloud point, and the distance between the point cloud point adjacent to the valid connection point in the direction close to the corresponding to-be-processed point cloud point and the valid connection point is less than a preset distance; For each point cloud point to be processed, in response to the number of valid connection points of the point cloud point to be processed being less than the preset number of connection points, the reflectivity of the point cloud point to be processed being less than the low reflectivity threshold, and the proportion of low reflectivity points of the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed being greater than the preset first proportion, the point cloud point to be processed is determined to be the first point cloud point; wherein the neighborhood grids are each first preset grid adjacent to the first preset grid corresponding to the point cloud point to be processed; The number of points in the first point cloud is determined.

5. The method according to claim 3, characterized in that: The determining the current weather type according to the number of points in the first point cloud, each second preset grid, the initial point cloud, and the point cloud to be processed includes: For each point cloud point to be processed, if the reflectivity of the point cloud point to be processed is less than the low reflectivity threshold and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the point cloud point to be processed is greater than the preset second proportion, then the point cloud point to be processed is determined to be a second point cloud point; determining the number of points in the second point cloud; taking a ratio of the number of points in the first point cloud to the number of points in the second point cloud as a reference value; In response to the number of the first point cloud points not being equal to zero, the number of the first point cloud points being less than a first number threshold, and the reference value being greater than a preset ratio, determining that the current weather type is general foggy weather; In response to the number of the first point cloud points being greater than a second number threshold, determining that the current weather type is dense fog; In response to the number of the first point cloud points being greater than or equal to the first number threshold and less than or equal to the second number threshold, projecting the third point cloud points into each second preset grid, wherein the third point cloud points are point cloud points whose reflectivity in the initial point cloud is less than the low reflectivity threshold, and the proportion of low reflectivity points in the first preset grid and each neighborhood grid corresponding to the third point cloud points is greater than the preset third proportion, the second preset grid is a polar coordinate grid with the laser radar as the origin, and the second preset grid forms a plurality of circular rings; Determine a valid grid from each second preset grid, wherein the valid grid is a second preset grid including at least one third point cloud point; Determine the first target circle with the largest number of valid grids; Determine the angle of the largest connected domain formed by valid grids in the first target ring; Determine the average point cloud density of valid grids in the first target ring; In response to the angle of the maximum connected domain being greater than a preset angle, and the average point cloud density being greater than a preset point cloud density, it is determined that the current weather type is dense fog.

6. The method according to claim 1, characterized in that If the current weather type is a general foggy day, the step of processing the point cloud to be processed according to the preset noise point determination method corresponding to the current weather type to determine the noise points in the initial point cloud includes: Projecting the point cloud points to be processed onto a third preset grid, wherein the third preset grid is a polar coordinate grid with the laser radar as the origin, and the third preset grid forms a plurality of circular rings; Determine a second target circular ring having the largest number of non-empty grids, wherein the non-empty grids are third preset grids having point cloud points to be processed having reflectivity less than a low reflectivity threshold; The second target circle and its adjacent circles are regarded as noise areas; For each point cloud point to be processed in the noise point area, in response to the reflectivity of the point cloud point to be processed being less than or equal to the first threshold, the point cloud point to be processed is determined to be a noise point in the initial point cloud; in response to the reflectivity of the point cloud point to be processed being greater than the first threshold and less than the second threshold, if the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than the preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than the preset first number of points, the point cloud point to be processed is determined to be a noise point in the initial point cloud; For each point cloud point to be processed in other circular rings outside the noise point area, in response to the reflectivity of the point cloud point to be processed being less than or equal to the low reflectivity threshold, the number of valid connection points corresponding to the point cloud point to be processed being less than or equal to the preset second number of points, and the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed being greater than the preset fifth proportion, the point cloud point to be processed is determined to be a noise point in the initial point cloud; in response to the reflectivity of the point cloud point to be processed being less than or equal to the low reflectivity threshold and the number of valid connection points corresponding to the point cloud point to be processed being less than the preset first number of points, the point cloud point to be processed is determined to be a noise point in the initial point cloud; Among them, the preset fourth proportion is smaller than the preset fifth proportion; the preset first number of points is smaller than the preset second number of points; the effective connection point of the point cloud point to be processed satisfies: the effective connection point and the corresponding point cloud point to be processed are located on the same laser radar beam, the point cloud point adjacent to the effective connection point in the direction close to the corresponding point cloud point to be processed is another effective connection point or the corresponding point cloud point to be processed, and the distance between the point cloud point adjacent to the effective connection point in the direction close to the corresponding point cloud point to be processed and the effective connection point is less than the preset distance.

7. The method according to claim 1, characterized in that If the current weather type is dense fog, the method of determining the noise points in the initial point cloud by processing the point cloud to be processed according to the preset noise point determination method corresponding to the current weather type includes: In response to the reflectivity of the point cloud point to be processed being less than or equal to a first threshold, determining that the point cloud point to be processed is a noise point in the initial point cloud; In response to the reflectivity of the point cloud point to be processed being greater than the first threshold and less than the second threshold, if the proportion of low reflectivity points of the first preset grid corresponding to the point cloud point to be processed is greater than a preset fourth proportion or the number of valid connection points corresponding to the point cloud point to be processed is less than a preset third number of points, it is determined that the point cloud point to be processed is a noise point in the initial point cloud; Among them, the effective connection point of the point cloud point to be processed satisfies: the effective connection point and the corresponding point cloud point to be processed are located on the same laser radar beam, the point cloud point adjacent to the effective connection point in the direction close to the corresponding point cloud point to be processed is another effective connection point or the corresponding point cloud point to be processed, and the distance between the point cloud point adjacent to the effective connection point in the direction close to the corresponding point cloud point to be processed and the effective connection point is less than a preset distance.

8. A laser point cloud processing device, characterized in that: include: A low reflectivity point ratio determination module is used to project the initial point cloud into each first preset grid, and determine the low reflectivity point ratio of each first preset grid according to the reflectivity of each point cloud point in the initial point cloud, wherein the low reflectivity point is a point cloud point whose reflectivity is less than a preset low reflectivity threshold; A point cloud to be processed determining module, used for determining the point cloud to be processed according to the reflectivity of each point cloud point in the initial point cloud; A current weather type determination module, used to determine the current weather type according to the distance between each adjacent point cloud point to be processed in the point cloud to be processed, the reflectivity of each point cloud point to be processed, and the proportion of low reflectivity points in each first preset grid; The noise point determination module is used to process the point cloud to be processed according to a preset noise point determination method corresponding to the current weather type, and determine the noise points in the initial point cloud.

9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the laser point cloud processing method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the laser point cloud processing method as described in any one of claims 1 to 7 is implemented.

Images