Methods, devices, vehicles, and media for handling crosstalk noise in lidar channels.

Abstract

This invention discloses a method, apparatus, vehicle, and medium for processing channel crosstalk noise in lidar systems. The method includes: acquiring point cloud data of obstacles within a region of interest; determining obstacle features based on the point cloud data; determining candidate obstacles among the obstacles based on the obstacle features; for each candidate obstacle, determining the target position symmetrical about the lidar's central axis; determining the target obstacle based on the distance between each obstacle and the target position; determining noise obstacles generated by channel crosstalk among the candidate obstacles based on the target obstacle corresponding to each candidate obstacle; and outputting a list of obstacles to be filtered out. This embodiment, by identifying noise obstacles generated by channel crosstalk among the candidate obstacles, can filter out channel crosstalk noise obstacles present in the candidate obstacles, improving the safety of autonomous vehicles.

Description

Technical Field

[0001] This invention relates to the field of radar technology, and in particular to methods, devices, vehicles, and media for processing crosstalk noise in lidar channels. Background Technology

[0002] As the mass production of intelligent driving vehicles continues to advance, the number of lidar sensors, as one of the main sensors for intelligent driving vehicles, is also increasing.

[0003] In existing technologies, when lidar acquires point cloud data, noise arises due to channel crosstalk. For example... Figure 1 The image shows a point cloud of a certain type of LiDAR. The three targets outlined on the right are three cones with high reflectivity stripes, while the targets outlined on the left are the generated channel crosstalk noise. In real-world scenarios, there are no obstacles on the left side of the autonomous vehicle. However, due to this noise, an obstacle is perceived on the left, interfering with the vehicle's movement and potentially causing emergency braking.

[0004] Therefore, how to filter out this noise has become an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a method, apparatus, vehicle, and medium for processing crosstalk noise in lidar channels, in order to solve the problem that existing technologies cannot filter out noise from autonomous vehicles, thereby improving vehicle driving safety.

[0006] According to one aspect of the present invention, a method for processing crosstalk noise in a lidar channel is provided, comprising:

[0007] Acquire point cloud data of each obstacle within the region of interest;

[0008] Obstacle features are determined based on the point cloud data, and candidate obstacles are determined among the obstacles based on the obstacle features;

[0009] For each candidate obstacle, determine the target position that is symmetrical about the central axis of the lidar, and determine the target obstacle based on the distance between each obstacle and the target position;

[0010] Based on the target obstacle corresponding to each candidate obstacle, identify the noise obstacles generated by channel crosstalk among the candidate obstacles, and output a list of obstacles to filter out the noise obstacles.

[0011] According to another aspect of the present invention, a processing apparatus for crosstalk noise in a lidar channel is provided, comprising:

[0012] The point cloud data acquisition module is used to acquire point cloud data of each obstacle within the region of interest;

[0013] The candidate obstacle determination module is used to determine obstacle features based on the point cloud data, and to determine candidate obstacles among the obstacles based on the obstacle features;

[0014] The target obstacle determination module is used to determine the target position symmetrical about the central axis of the lidar for each candidate obstacle, and to determine the target obstacle based on the distance between each obstacle and the target position;

[0015] The obstacle list output module is used to determine the noise obstacles generated by channel crosstalk among the candidate obstacles based on the target obstacle corresponding to each candidate obstacle, and output an obstacle list to filter out the noise obstacles.

[0016] According to another aspect of the present invention, a vehicle is provided, the vehicle comprising:

[0017] LiDAR is used to collect environmental information during vehicle operation.

[0018] At least one processor; and

[0019] A memory communicatively connected to the at least one processor; wherein,

[0020] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the method for processing crosstalk noise in the lidar channel as described in any embodiment of the present invention.

[0021] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the method for processing crosstalk noise in a lidar channel according to any embodiment of the present invention.

[0022] The technical solution of this invention determines the target obstacle based on the target position symmetrical about the central axis of the lidar and the distance between each obstacle and the target position; and determines the noise obstacle generated by channel crosstalk in the candidate obstacle according to the target obstacle corresponding to each candidate obstacle. This can filter out the channel crosstalk noise obstacle in the candidate obstacle, thereby improving the safety of autonomous vehicles.

[0023] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a diagram illustrating the interaction between existing lidar technology and obstacles.

[0026] Figure 1a This is a flowchart of a laser radar channel crosstalk noise reduction method provided in Embodiment 1 of the present invention;

[0027] Figure 1b This is a flowchart of a method for processing raw point cloud images according to Embodiment 1 of the present invention;

[0028] Figure 1c This is a structural schematic diagram of the relationship between obstacles and lidar according to Embodiment 1 of the present invention.

[0029] Figure 2 This is a flowchart of a laser radar channel crosstalk noise reduction method provided in Embodiment 2 of the present invention;

[0030] Figure 3 This is a flowchart of another method for reducing crosstalk noise in a lidar channel according to Embodiment 3 of the present invention;

[0031] Figure 4 This is a schematic diagram of the structure of a lidar channel crosstalk noise device provided in Embodiment 4 of the present invention;

[0032] Figure 5 This is a schematic diagram of the structure of a vehicle that implements the lidar channel crosstalk noise reduction method of the present invention. Detailed Implementation

[0033] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0034] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0035] Example 1

[0036] Figure 1a This is a flowchart illustrating a method for processing crosstalk noise in a lidar channel according to Embodiment 1 of the present invention. This embodiment is applicable to scenarios involving processing crosstalk noise present in a lidar channel. The method can be executed by a lidar channel crosstalk noise processing device, which can be implemented in hardware and / or software and can be configured in a vehicle. Figure 1a As shown, the method includes:

[0037] S110. Obtain point cloud data of each obstacle within the region of interest.

[0038] Among them, the Region of Interest (ROI) can be understood as a specific area in an image that is of interest; point cloud data can be a massive set of points representing the surface characteristics of a target, usually obtained through lidar measurement or photogrammetry, used to reflect the real situation of the ground, such as ground conditions and features of ground reflectors; and the near range can be a circular area with a radius of 10 meters centered on the lidar itself.

[0039] Optionally, target objects with high reflectivity can be cones, triangular signs, and vehicle license plates with high reflectivity stripes.

[0040] This embodiment can acquire point cloud data of various highly reflective objects within a region of interest using a lidar system. Specifically, an M1 solid-state lidar can be used to scan the highly reflective objects at close range. This embodiment does not impose any specific limitations on this.

[0041] For example, S1110 may include:

[0042] The original point cloud image is acquired, and then subjected to map filtering, ground point filtering, and point cloud clustering to obtain the point cloud data of obstacles within the region of interest.

[0043] The original point cloud image can be obtained by a LiDAR system using multiple preset laser probes, rotating the probes 360° to capture the surrounding point cloud image. Since most points in the original point cloud image are outside the drivable area, map filtering is needed to select points within the drivable area. However, the drivable area contains multiple ground points, so ground point filtering is also necessary. Subsequently, point cloud clustering is performed on obstacles within the drivable area to obtain point cloud data of obstacles within the region of interest. Specifically, this can be achieved by... Figure 1b As shown in the flowchart for obstacle acquisition, all obstacles within the region of interest can be obtained by performing steps such as map filtering, ground point filtering, and point cloud clustering on the original point cloud image.

[0044] S120. Determine obstacle features based on point cloud data, and determine candidate obstacles among each obstacle based on obstacle features.

[0045] Among them, obstacle features can be the average point cloud intensity of the obstacle (intensity_avg), the length of the obstacle, the width of the obstacle, the height of the obstacle, the center coordinates of the bounding box of the obstacle (center_x, center_y, center_z), and the number of obstacle points (obj_num).

[0046] For example, for each obstacle, the length, width, height, and number of obstacle points are determined based on the corresponding point cloud data. The aspect ratio is determined based on the length and width of the obstacle, and whether the obstacle is a candidate obstacle is determined based on at least one of the aspect ratio, height, and number of obstacle points.

[0047] Furthermore, for each obstacle, the features of the current obstacle are determined based on the point cloud data, and the features of the current obstacle are statistically traversed to obtain the length, width, height, and number of obstacle points (obj_num) of the current obstacle; furthermore, the aspect ratio (length / width) is determined based on the length and width of the current obstacle, and whether the current obstacle is a candidate obstacle is determined based on at least one of the aspect ratio, height, and number of obstacle points.

[0048] For example, determining whether a current obstacle is a candidate obstacle based on at least one of the current obstacle's aspect ratio, height, and obstacle point count may include: determining the current obstacle as a candidate obstacle if the current obstacle's aspect ratio is greater than a set aspect ratio filter value; determining the current obstacle as a candidate obstacle if the current obstacle's height is less than a set height filter value; and determining the current obstacle as a candidate obstacle if the current obstacle's obstacle point count is less than a set point count filter threshold.

[0049] In this embodiment, the current obstacle is determined to be a candidate obstacle when the obstacle features meet any of the following conditions: the length / width of the current obstacle is greater than the preset length / width filter value (rabbit_size_ratio), or the height of the current obstacle is less than the preset height filter value (rabbit_height_thr), or the number of points of the current obstacle (obj_num) is less than the preset number of points filter threshold (rabbit_pt_num).

[0050] Preferably, the preset aspect ratio filter value can be 2; the preset height filter value can be 0.3; and the preset point count filter threshold can be 100. This embodiment does not impose specific restrictions on these values ​​and can be flexibly set as needed.

[0051] S130. For each candidate obstacle, determine the target position that is symmetrical about the central axis of the lidar, and determine the target obstacle based on the distance between each obstacle and the target position.

[0052] In this embodiment, determining the target position symmetrical about the central axis of the lidar for the current candidate obstacle is specifically as follows: traverse each candidate obstacle, find the obstacle that is symmetrical about the central axis of the lidar, compare the distances of each obstacle to the target position, and determine the obstacle closest to the symmetrical position about the central axis of the lidar as the target obstacle based on the comparison results.

[0053] For example, S130 may include: for each candidate obstacle, determining the center coordinates of the obstacle bounding box of the current obstacle based on the corresponding point cloud data; determining candidate obstacles in the area surrounding the target position symmetrical about the central axis of the lidar based on the center coordinates of the obstacle bounding box, wherein the area surrounding the target position is a circular area with the target position as the center and a set length as the radius; and determining the target obstacle within the candidate obstacles in the area surrounding the target position based on the distance between the candidate obstacles in the area surrounding the target position and the target position.

[0054] Here, the obstacle bounding box can be understood as the border of the obstacle in the point cloud data, for example, the smallest outer rectangle of the obstacle.

[0055] In this embodiment, (center_x_a) can be used to represent the x-coordinate of the center of the current obstacle a. Similarly, (center_x_b), (center_y_a), (center_y_b), and (nearby_obj_dis) are used as the set distance thresholds for judging the symmetrical position. For all candidate obstacles within a certain range of the symmetrical position of the current obstacle a, the distance dis from the candidate obstacle to the symmetrical position of obstacle a is calculated.

[0056] Specifically, the distance from a candidate obstacle to the symmetrical position of obstacle a can be calculated using the following formula:

[0057] dis=sqrt((center_x_a-center_x_b)*(center_x_a-center_x_b)+(center_y_a+center_y_b)*(center_y_a+center_y_b)).

[0058] Furthermore, the obstacle with the smallest dis among all obstacles at symmetrical positions is marked as the nearest obstacle at the symmetrical position of the noise candidate target. Figure 1c A schematic diagram showing the relationship between the lidar and obstacles is provided, such as... Figure 1c The position of obstacle a' is the symmetrical position of obstacle a with respect to the lidar, and obstacle c is the nearest obstacle to the symmetrical position of obstacle a.

[0059] Preferably, the set symmetrical position judgment distance threshold can be 1.0m, but this embodiment does not impose a specific limitation on it.

[0060] For example, determining the candidate obstacles within the area surrounding the target position symmetrical about the central axis of the lidar based on the center coordinates of the obstacle bounding box may include: traversing all candidate obstacles except the current obstacle, and calculating the distances between the current obstacle and the candidate obstacles on the x-axis and y-axis respectively; if the distance between the current obstacle and the candidate obstacle on the x-axis is less than a set symmetry position judgment distance threshold, and the distance between the current obstacle and the candidate obstacle on the y-axis is less than the set symmetry position judgment distance threshold, then it is determined that the candidate obstacle is located within the area surrounding the target position symmetrical about the central axis of the lidar based on the center coordinates of the obstacle bounding box.

[0061] In this embodiment, it is necessary to traverse the candidate obstacles to find the obstacle closest to its symmetrical position about the central axis of the lidar. Specifically, this involves traversing all candidate target obstacles b except for the current obstacle a, such as... Figure 1c As shown, if all of the following conditions are met, then obstacle b is within a certain range of positions symmetrical to obstacle a:

[0062] abs(center_x_a-center_x_b) <nearby_obj_dis;

[0063] abs(center_y_a+center_y_b) <nearby_obj_dis;

[0064] center_y_a*center_y_b<0;

[0065] Here, abs represents the absolute value; center_x_a represents the x-coordinate of the center coordinate of obstacle a, and the same applies to center_x_b, center_y_a, and center_y_b; nearby_obj_dis is the set distance threshold for judging symmetrical positions.

[0066] S140. Based on the target obstacle corresponding to each candidate obstacle, determine the noise obstacles generated by channel crosstalk among the candidate obstacles, and output the list of obstacles to filter out noise obstacles.

[0067] In this context, the target obstacle can be understood as the obstacle closest to the candidate obstacle at a position symmetrical about the central axis of the lidar. Furthermore, since channel crosstalk is generally generated in the symmetrical channel of the obstacle, an obstacle with high reflectivity and significantly higher point cloud count can be found within a certain range of the channel crosstalk noise relative to the symmetrical position of the lidar.

[0068] For example, determining the noise obstacle generated by channel crosstalk among the candidate obstacles based on the target obstacle corresponding to each candidate obstacle may include: for each target obstacle corresponding to a candidate obstacle, comparing the average intensity of the target obstacle with a set point cloud intensity threshold, and comparing the number of obstacle points of the target obstacle with the current candidate obstacle; and determining whether the current candidate obstacle is a noise obstacle generated by channel crosstalk based on the comparison result.

[0069] For example, determining whether the current candidate obstacle is a noise obstacle caused by channel crosstalk based on the comparison result may include: if the average intensity of the target obstacle is greater than a set point cloud intensity threshold and the number of obstacle points of the target obstacle is greater than the number of obstacle points of the current candidate obstacle, then the current candidate obstacle is determined to be a noise obstacle caused by channel crosstalk.

[0070] In this embodiment Figure 1c Whether a current candidate obstacle a is a noise obstacle can be determined based on the nearest obstacle c at the symmetrical position of the candidate obstacle. The current candidate obstacle a is considered a noise obstacle caused by channel crosstalk when the following condition is met:

[0071] intensity_avg_c>intensity_thr;

[0072] obj_num_c / obj_num_a>num_ratio;

[0073] Where intensity_thr is the preset point cloud intensity threshold, and num_ratio is the preset point ratio threshold.

[0074] Preferably, the preset point cloud intensity threshold can be 70; the preset point ratio threshold can be 20.m, and this embodiment does not impose specific restrictions on this.

[0075] In this embodiment, once noise obstacles caused by channel crosstalk are identified among the candidate obstacles, a list of obstacles to be filtered out (i.e., the list of truly existing obstacles within the range of interest) is output. Obstacles are then filtered out sequentially according to this list.

[0076] This embodiment determines the target obstacle based on the target position symmetrical about the central axis of the lidar and the distance between each obstacle and the target position. Based on the target obstacle corresponding to each candidate obstacle, noise obstacles caused by channel crosstalk are identified in the candidate obstacles. This can filter out the channel crosstalk noise obstacles in the candidate obstacles, thereby improving the safety of autonomous vehicles.

[0077] Example 2

[0078] Figure 2 This is a flowchart of a method for processing crosstalk noise in a lidar channel according to Embodiment 2 of the present invention. After obtaining the point cloud data of each obstacle within the region of interest, this embodiment further includes: identifying suspended obstacles and filtering them out. Terms identical to those used in the above steps are not repeated here. Figure 2 As shown, the method includes:

[0079] S210. Obtain point cloud data of each obstacle within the region of interest.

[0080] S220. Obtain the first feature point of each obstacle and the second feature point of the ground height corresponding to the obstacle based on the point cloud data.

[0081] The first feature point can be the lowest point of the obstacle; the second feature point can be the highest point of the obstacle at the corresponding ground height.

[0082] S230. Based on the first feature point and the second feature point, determine the suspended obstacles among the obstacles, treat the suspended obstacles as noise obstacles generated by channel crosstalk, and filter out the suspended obstacles.

[0083] In this embodiment, for obstacles within a circular area with a radius of 10 meters centered on the LiDAR itself, it is necessary to calculate the lowest point (obj_lowest height) of each obstacle from the ground and the highest point (ground_highest_height) of each obstacle from the ground. Based on the lowest point of the obstacle from the ground and the highest point of each obstacle from the ground, suspended obstacles are determined.

[0084] In this embodiment, it is necessary to filter near-range suspended obstacles and obtain point cloud data of the remaining obstacles after filtering out the suspended obstacles. Specifically, for obstacles within a circular area with a radius of 10 meters centered on the LiDAR itself, the lowest point (obj_lowest height) and the highest point (ground_highest height) of each obstacle are calculated. When the difference between the lowest point and the highest point of an obstacle is less than a preset suspended obstacle filtering threshold (rabbit_float_height), that is:

[0085] When (obj_lowest_height–ground_highest_height>rabbit_float_height), the current obstacle can be determined to be a nearby suspended obstacle that needs to be filtered. Therefore, the suspended obstacle is treated as a noise obstacle generated by channel crosstalk and is filtered out.

[0086] Preferably, the preset threshold for filtering suspended objects can be 0.1m. This embodiment does not impose a specific limitation on this and can be flexibly set as needed. S240: Determine obstacle features based on point cloud data, and determine candidate obstacles among each obstacle based on obstacle features.

[0087] In this embodiment, it is necessary to obtain the point cloud data of the remaining obstacles after filtering out suspended obstacles; for each remaining obstacle, the length, width, height and number of obstacle points of the current obstacle are determined according to the corresponding point cloud data; for each remaining obstacle, the aspect ratio is determined according to the length and width of the current obstacle, and whether the current obstacle is a candidate obstacle is determined according to at least one of the aspect ratio, height and number of obstacle points of the current obstacle.

[0088] In this embodiment, for each remaining obstacle, the features of the current obstacle can be determined based on the point cloud data, and the features of the current obstacle can be statistically traversed to obtain the length, width, height, and number of obstacle points. Furthermore, the aspect ratio (length / width) is determined based on the length and width of the current obstacle, and whether the current obstacle is a candidate obstacle is determined based on at least one of the aspect ratio, height, and number of obstacle points.

[0089] Specifically, candidate targets for channel crosstalk noise within the current range are identified. An obstacle is considered a candidate target if its characteristics satisfy one of the following conditions:

[0090] length / width>rabbit_size_ratio;

[0091] height <rabbit_height_thr;

[0092] obj_num <rabbit_pt_num;

[0093] Where rabbit_size_ratio is the set aspect ratio filter value, rabbit_height_thr is the set height filter value, and rabbit_pt_num is the set point count filter threshold.

[0094] S250. For each candidate obstacle, determine the target position that is symmetrical about the central axis of the lidar, and determine the target obstacle based on the distance between each obstacle and the target position.

[0095] For example, find the nearest obstacle to the symmetrical position of all noise candidate targets. Traverse each noise candidate target and find the nearest obstacle to its symmetrical position about the lidar's central axis. Specifically, traverse all candidate target obstacles b except the currently visited obstacle a. If all of the following conditions are met, then obstacle b is within a certain range of the symmetrical position of obstacle a:

[0096] abs(center_x_a-center_x_b) <nearby_obj_dis;

[0097] abs(center_y_a+center_y_b) <nearby_obj_dis;

[0098] center_y_a*center_y_b<0;

[0099] Where center_x_a represents the x-coordinate of the center of obstacle a, and the same applies to center_x_b, center_y_a, and center_y_b. Nearby_obj_dis is the set distance threshold for determining the symmetrical position. For all obstacles within a certain range of the symmetrical position of obstacle a, the distance dis from the obstacle to the symmetrical position of obstacle a is calculated.

[0100] dis=sqrt((center_x_a-center_x_b)*(center_x_a-center_x_b)+(center_y_a+center_y_b)*(center_y_a+center_y_b));

[0101] The obstacle with the smallest dis among all obstacles at symmetrical positions is marked as the nearest obstacle at the symmetrical position of the noise candidate target, i.e., the target obstacle.

[0102] S260. Based on the target obstacle corresponding to each candidate obstacle, determine the noise obstacles generated by channel crosstalk among the candidate obstacles, and output the list of obstacles to filter out noise obstacles.

[0103] This embodiment improves the safety of autonomous vehicles by filtering out channel crosstalk noise obstacles in candidate obstacles by filtering out suspended obstacles.

[0104] Example 3

[0105] Figure 3 This is a flowchart of another method for processing crosstalk noise in a lidar channel according to Embodiment 2 of the present invention. Terms used in the same steps as described above will not be repeated here. Figure 3 As shown, the method includes:

[0106] S310, Filter out suspended obstacles.

[0107] S320. Obtain the point cloud data of the remaining obstacles after filtering out suspended obstacles; for each remaining obstacle, determine the length, width, height and number of obstacle points based on the corresponding point cloud data; for each remaining obstacle, determine the aspect ratio based on the length and width of the current obstacle, and determine whether the current obstacle is a candidate obstacle based on at least one of the aspect ratio, height and number of obstacle points of the current obstacle.

[0108] S330. If the aspect ratio of the current obstacle is greater than the set aspect ratio filter value, determine the current obstacle as a candidate obstacle; if the height of the current obstacle is less than the set height filter value, determine the current obstacle as a candidate obstacle; if the number of obstacle points of the current obstacle is less than the set number of points filter threshold, determine the current obstacle as a candidate obstacle.

[0109] S340. For each candidate obstacle, determine the center coordinates of the obstacle bounding box of the current obstacle based on the corresponding point cloud data. Based on the center coordinates of the obstacle bounding box, determine the candidate obstacles in the area surrounding the target position that is symmetrical about the central axis of the lidar. The area surrounding the target position is a circular area with the target position as the center and a set length as the radius. Based on the distance between the candidate obstacles in the area surrounding the target position and the target position, determine the target obstacle within the candidate obstacles in the area surrounding the target position.

[0110] S350. For each candidate obstacle corresponding to a target obstacle, compare the average intensity of the target obstacle with the set point cloud intensity threshold, and compare the number of obstacle points of the target obstacle with the current candidate obstacle; determine whether the current candidate obstacle is a noise obstacle caused by channel crosstalk based on the comparison result.

[0111] This embodiment filters out suspended obstacles and statistically determines candidate obstacles by identifying the remaining obstacles. It also determines the closest distance between the candidate obstacle and the target obstacle, thereby filtering out channel crosstalk noise obstacles in the candidate obstacles and improving the safety of autonomous vehicles.

[0112] Example 4

[0113] Figure 4 This is a schematic diagram of a device for processing crosstalk noise in a lidar channel according to Embodiment 4 of the present invention. Figure 4 As shown, the device includes:

[0114] The point cloud data acquisition module 401 is used to acquire point cloud data of each obstacle in the region of interest.

[0115] The candidate obstacle determination module 402 is used to determine obstacle features based on the point cloud data, and to determine candidate obstacles among the obstacles based on the obstacle features;

[0116] The target obstacle determination module 403 is used to determine the target position symmetrical about the central axis of the lidar for each candidate obstacle, and to determine the target obstacle based on the distance between each obstacle and the target position;

[0117] The obstacle list output module 404 is used to determine the noise obstacles generated by channel crosstalk among the candidate obstacles based on the target obstacle corresponding to each candidate obstacle, and output an obstacle list to filter out the noise obstacles.

[0118] Optionally, the point cloud data acquisition module 401 is also used for:

[0119] Based on the point cloud data, obtain the first feature points of each obstacle and the second feature points of the ground height corresponding to the obstacle;

[0120] Based on the first feature point and the second feature point, the suspended obstacles among the obstacles are determined, and the suspended obstacles are treated as noise obstacles generated by channel crosstalk and filtered out.

[0121] Optional, the point cloud data acquisition module 401 is specifically used for:

[0122] Obtain point cloud data of the remaining obstacles after filtering out suspended obstacles;

[0123] For each remaining obstacle, determine the length, width, height, and number of obstacle points based on the corresponding point cloud data;

[0124] For each remaining obstacle, the aspect ratio is determined based on the length and width of the current obstacle, and whether the current obstacle is a candidate obstacle is determined based on at least one of the aspect ratio, height, and number of obstacle points.

[0125] Optional, the point cloud data acquisition module 401 is specifically used for:

[0126] If the aspect ratio of the current obstacle is greater than the set aspect ratio filter value, the current obstacle is determined to be a candidate obstacle;

[0127] If the height of the current obstacle is less than the set height filter value, the current obstacle is determined to be a candidate obstacle;

[0128] If the number of obstacle points of the current obstacle is less than the set number of points filtering threshold, the current obstacle is determined to be a candidate obstacle.

[0129] Optionally, the target obstacle determination module 403 is specifically used for:

[0130] For each candidate obstacle, the center coordinates of the obstacle bounding box of the current obstacle are determined based on the corresponding point cloud data. Based on the center coordinates of the obstacle bounding box, the candidate obstacles in the surrounding area of ​​the target position symmetrical about the central axis of the lidar are determined. The surrounding area of ​​the target position is a circular area with the target position as the center and a set length as the radius.

[0131] The target obstacle is determined based on the distance between the target location and the candidate obstacles in the area surrounding the target location.

[0132] Optionally, the target obstacle determination module 403 is specifically used for:

[0133] Traverse the candidate obstacles other than the current obstacle, and calculate the distances between the current obstacle and the candidate obstacles on the x-axis and y-axis respectively;

[0134] If the distance between the current obstacle and the candidate obstacle on the x-axis is less than a set symmetry position judgment distance threshold, and the distance between the current obstacle and the candidate obstacle on the y-axis is less than a set symmetry position judgment distance threshold, then it is determined that the candidate obstacle is located in the area surrounding the target position symmetrical about the central axis of the lidar of the current candidate obstacle.

[0135] Optional, obstacle list output module 404, specifically used for:

[0136] For each candidate obstacle corresponding to a target obstacle, the average intensity of the target obstacle is compared with a set point cloud intensity threshold, and the number of obstacle points of the target obstacle is compared with that of the current candidate obstacle;

[0137] Based on the comparison results, determine whether the current candidate obstacle is a noise obstacle caused by channel crosstalk.

[0138] Optional, obstacle list output module 404, specifically used for:

[0139] If the average intensity of the target obstacle is greater than a set point cloud intensity threshold, and the number of obstacle points of the target obstacle is greater than the number of obstacle points of the current candidate obstacle, then the current candidate obstacle is determined to be a noise obstacle caused by channel crosstalk.

[0140] Optional, the point cloud data acquisition module 401 is specifically used for:

[0141] The original point cloud image is acquired, and map filtering, ground point filtering, and point cloud clustering are performed on the original point cloud image to obtain the point cloud data of obstacles in the region of interest.

[0142] The laser radar channel crosstalk noise processing device provided in this embodiment of the invention can execute the laser radar channel crosstalk noise processing method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0143] Example 5

[0144] Figure 5 A schematic diagram of the structure of a vehicle 10 that can be used to implement an embodiment of the present invention is shown. For example... Figure 5 As shown, vehicle 10 includes at least one processor 11 and a memory, such as read-only memory (ROM) 12 and random access memory (RAM) 13, communicatively connected to at least one processor 11. The memory stores computer programs executable by at least one processor. Processor 11 can perform various appropriate actions and processes based on the computer program stored in ROM 12 or loaded from storage unit 18 into RAM 13. RAM 13 can also store various programs and data required for the operation of vehicle 10. Processor 11, ROM 12, and RAM 13 are interconnected via bus 14. Input / output (I / O) interface 15 is also connected to bus 14.

[0145] Multiple components in vehicle 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows vehicle 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0146] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a method for processing crosstalk noise in a lidar channel.

[0147] In some embodiments, a method for processing lidar channel crosstalk noise can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on vehicle 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the lidar channel crosstalk noise processing method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform a lidar channel crosstalk noise processing method by any other suitable means (e.g., by means of firmware).

[0148] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0149] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0150] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0151] To provide interaction with the user, the systems and technologies described herein can be implemented in a vehicle having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the vehicle. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0152] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0153] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0154] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0155] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for processing crosstalk noise in a lidar channel, characterized in that, include: Acquire point cloud data of each obstacle within the region of interest; Obstacle features are determined based on the point cloud data, and candidate obstacles are determined among the obstacles based on the obstacle features; For each candidate obstacle, determine the target position that is symmetrical about the central axis of the lidar, and determine the target obstacle based on the distance between each obstacle and the target position; Based on the target obstacle corresponding to each candidate obstacle, identify the noise obstacles generated by channel crosstalk among the candidate obstacles, and output a list of obstacles to filter out the noise obstacles.

2. The method according to claim 1, characterized in that, After acquiring the point cloud data of each obstacle within the region of interest, the process also includes: The first feature point of each obstacle and the second feature point of the ground height corresponding to the obstacle are obtained based on the point cloud data; wherein, the first feature point is the lowest point of the obstacle; and the second feature point is the highest point of the ground height corresponding to the obstacle. Based on the first feature point and the second feature point, the suspended obstacles among the obstacles are determined, and the suspended obstacles are treated as noise obstacles generated by channel crosstalk and filtered out.

3. The method according to claim 1, characterized in that, The step of determining obstacle features based on the point cloud data, and determining candidate obstacles among the obstacles based on the obstacle features, includes: Obtain point cloud data of the remaining obstacles after filtering out suspended obstacles; For each remaining obstacle, determine the length, width, height, and number of obstacle points based on the corresponding point cloud data; For each remaining obstacle, the aspect ratio is determined based on the length and width of the current obstacle, and whether the current obstacle is a candidate obstacle is determined based on at least one of the aspect ratio, height, and number of obstacle points.

4. The method according to claim 3, characterized in that, The step of determining whether the current obstacle is a candidate obstacle based on at least one of the aspect ratio, height, and number of obstacle points includes: If the aspect ratio of the current obstacle is greater than the set aspect ratio filter value, the current obstacle is determined to be a candidate obstacle; If the height of the current obstacle is less than the set height filter value, the current obstacle is determined to be a candidate obstacle; If the number of obstacle points of the current obstacle is less than the set number of points filtering threshold, the current obstacle is determined to be a candidate obstacle.

5. The method according to claim 1, characterized in that, For each candidate obstacle, the process of determining the target position symmetrical about the central axis of the lidar and the distance between each obstacle and the target position includes: For each candidate obstacle, the center coordinates of the obstacle bounding box of the current obstacle are determined according to the corresponding point cloud data. Based on the center coordinates of the obstacle bounding box, the candidate obstacles in the surrounding area of ​​the target position symmetrical about the central axis of the lidar are determined. The surrounding area of ​​the target position is a circular area with the target position as the center and a set length as the radius. The target obstacle is determined based on the distance between the target location and the candidate obstacles in the area surrounding the target location.

6. The method according to claim 5, characterized in that, The step of determining candidate obstacles within the area surrounding the target position symmetrical about the central axis of the lidar based on the center coordinates of the obstacle bounding box includes: Traverse the candidate obstacles other than the current obstacle, and calculate the distances between the current obstacle and the candidate obstacles on the x-axis and y-axis respectively; If the distance between the current obstacle and the candidate obstacle on the x-axis is less than a set symmetry position judgment distance threshold, and the distance between the current obstacle and the candidate obstacle on the y-axis is less than a set symmetry position judgment distance threshold, then it is determined that the candidate obstacle is located in the area surrounding the target position symmetrical about the central axis of the lidar of the current candidate obstacle.

7. The method according to claim 1, characterized in that, The step of determining the noise obstacles generated by channel crosstalk among the candidate obstacles based on the target obstacle corresponding to each candidate obstacle includes: For each candidate obstacle corresponding to a target obstacle, the average intensity of the target obstacle is compared with a set point cloud intensity threshold, and the number of obstacle points of the target obstacle is compared with that of the current candidate obstacle; Based on the comparison results, determine whether the current candidate obstacle is a noise obstacle caused by channel crosstalk.

8. The method according to claim 7, characterized in that, The step of determining whether the current candidate obstacle is a noise obstacle caused by channel crosstalk based on the comparison result includes: If the average intensity of the target obstacle is greater than a set point cloud intensity threshold, and the number of obstacle points of the target obstacle is greater than the number of obstacle points of the current candidate obstacle, then the current candidate obstacle is determined to be a noise obstacle caused by channel crosstalk.

9. The method according to claim 1, characterized in that, The acquisition of point cloud data for each obstacle within the region of interest includes: The original point cloud image is acquired, and map filtering, ground point filtering, and point cloud clustering are performed on the original point cloud image to obtain the point cloud data of obstacles in the region of interest.

10. A device for processing crosstalk noise in a lidar channel, characterized in that, include: The point cloud data acquisition module is used to acquire point cloud data of each obstacle within the region of interest; The candidate obstacle determination module is used to determine obstacle features based on the point cloud data, and to determine candidate obstacles among the obstacles based on the obstacle features; The target obstacle determination module is used to determine the target position symmetrical about the central axis of the lidar for each candidate obstacle, and to determine the target obstacle based on the distance between each obstacle and the target position; The obstacle list output module is used to determine the noise obstacles generated by channel crosstalk among the candidate obstacles based on the target obstacle corresponding to each candidate obstacle, and output an obstacle list to filter out the noise obstacles.

11. A vehicle, characterized in that, The vehicles include: LiDAR is used to collect environmental information during vehicle operation. At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for processing lidar channel crosstalk noise according to any one of claims 1-9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for processing crosstalk noise in the lidar channel as described in any one of claims 1-9.

Images