A method and device for detecting nearby obstacles based on laser radar

Abstract

A laser radar (LiDAR)-based method and device for detecting nearby obstacles relates to the field of LiDAR sensing. The method comprises: establishing a first coordinate system and dividing the grid based on the installation position and parameters of at least one sensor, and calculating the height threshold corresponding to each grid; the sensor is a LiDAR sensor; establishing original coordinates for each scan point in the original point cloud according to the first coordinate system; projecting each scan point in the original point cloud to obtain projected coordinates; classifying each scan point into a ground point cloud and an obstacle point cloud based on the original coordinates and projected coordinates of each scan point; and clustering, calculating bounding box information, and performing matching tracking on the obstacle point cloud to obtain obstacle information. The present invention effectively reduces the missed detection rate of LiDAR detection of nearby obstacles by adaptively adjusting the classification heights of the ground point cloud and obstacle point cloud at close range based on the installation position and parameters of the LiDAR sensor.

Description

Technical Field

[0001] The present application relates to the field of laser radar perception, and specifically to a method and device for detecting nearby obstacles based on laser radar. Background Art

[0002] With the widespread adoption of autonomous driving technology, automotive LiDAR technology has also developed rapidly. LiDAR can sense surrounding obstacles and provide reliable environmental information for autonomous driving systems.

[0003] In related technologies, target detection using lidar (LiDAR) perception primarily involves two methods: a method based on deep learning model training to identify obstacles of specific labeled categories, and a conventional target detection method based on height-based rule-based point cloud ground removal and clustering. The deep learning-based model training method uses surface point cloud information of all targets as input, and after model processing, directly outputs information such as the category and position of the obstacle. Conventional point cloud clustering methods require first dividing the space into a two-dimensional grid, removing ground points from the three-dimensional point cloud based on a fixed grid height threshold, and then clustering the non-ground point cloud to calculate obstacle information.

[0004] However, model-training-based methods can only identify obstacles within the trained categories. When obstacles are relatively close to the sensor, the vertical viewing angle of the laser sensor is typically around 25°, resulting in the point cloud scanned by the obstacle only showing a localized portion of the obstacle surface, with unclear features. This makes the model unable to classify and identify nearby obstacles. Conventional methods based on fixed-height rules consider point clouds with heights below a threshold to be ground-level. However, when obstacles are relatively close to the sensor, the height of the obstacle point cloud scanned may not exceed the height threshold, resulting in missed obstacle detection. Summary of the Invention

[0005] The present application provides a method and device for detecting nearby obstacles based on laser radar, which can solve the problem in the prior art that laser radar detection is difficult to identify nearby obstacles.

[0006] In a first aspect, the present application provides a method for detecting nearby obstacles based on a laser radar, the method comprising:

[0007] Based on the installation position and parameters of at least one sensor, a first coordinate system is established and divided into grids, and a height threshold corresponding to each grid is calculated; the sensor is a lidar sensor;

[0008] Establishing original coordinates for each scanning point in the original point cloud according to the first coordinate system; projecting each scanning point in the original point cloud to obtain projection coordinates;

[0009] Based on the original coordinates and projection coordinates of each scanning point, each scanning point is classified into ground point cloud and obstacle point cloud;

[0010] The obstacle point cloud is clustered, bounding box information is calculated, and matching tracking is performed to obtain obstacle information.

[0011] In combination with the first aspect, in one embodiment, when there is only one sensor, establishing the first coordinate system includes:

[0012] The projection point of the midpoint of the sensor on the ground is taken as the origin;

[0013] Taking the vector pointing from the origin to the midpoint of the sensor as the Z axis of the first coordinate system, determining the plane containing the origin and perpendicularly intersecting the Z axis as the XY plane;

[0014] On the XY plane, with the origin as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of the sensor is the X axis of the first coordinate system;

[0015] On the XY plane, with the origin as the starting point, a vector obtained by rotating the X axis counterclockwise by 90 degrees is used as the Y axis of the first coordinate system.

[0016] In combination with the first aspect, in one embodiment, when there are multiple sensors, establishing the first coordinate system includes:

[0017] Determine the origin based on the projection point of each sensor's midpoint on the ground;

[0018] The vector from the projection point of any sensor on the ground to the midpoint of the sensor is used as the first vector;

[0019] Taking the origin as the starting point and a vector parallel to the first vector as the Z axis of the first coordinate system, determine a plane containing the origin and perpendicularly intersecting the Z axis as the XY plane;

[0020] On the XY plane, with the origin as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of any sensor is used as the X axis of the first coordinate system;

[0021] On the XY plane, with the origin as the starting point, a vector obtained by rotating the X axis counterclockwise by 90 degrees is used as the Y axis of the first coordinate system.

[0022] In conjunction with the first aspect, in one embodiment, dividing the grid includes:

[0023] The XY plane of the first coordinate system is divided into square areas of equal area using a preset grid resolution.

[0024] In combination with the first aspect, in one embodiment, when there is only one sensor, the height threshold corresponding to each grid is calculated as follows:

[0025] Get the minimum distance the sensor can scan the ground;

[0026] The height threshold corresponding to each grid within the horizontal scanning angle of the sensor is calculated based on the horizontal distance between the midpoint of each grid and the origin of the first coordinate system, a preset height ratio threshold, the height of the sensor from the ground, the minimum distance of the sensor scanning the ground, and the vertical scanning angle of the sensor.

[0027] In combination with the first aspect, in one embodiment, when there are multiple sensors, the height threshold corresponding to each grid is calculated as follows:

[0028] Establishing a second coordinate system corresponding to each sensor according to the installation position and parameters of each sensor, and calculating the coordinates of the midpoint of each grid in the first coordinate system;

[0029] According to the parameters of each sensor, the coordinates of the midpoints of each grid are projected into the second coordinate system corresponding to each sensor to obtain the projected coordinates;

[0030] Get the minimum distance for each sensor to scan the ground;

[0031] Calculating a first height threshold corresponding to each grid within a horizontal scanning angle of each sensor based on a horizontal distance between the projected coordinates and the origin of the second coordinate system corresponding to each sensor, a preset height ratio threshold, a height of each sensor from the ground, a minimum distance scanned by each sensor against the ground, and a vertical scanning angle of each sensor;

[0032] The maximum value of all first height thresholds corresponding to each grid is taken as the height threshold of each grid.

[0033] In conjunction with the first aspect, in one embodiment, establishing a second coordinate system corresponding to each sensor according to the installation position and parameters of each sensor includes:

[0034] The projection point of the midpoint of each sensor on the XY plane of the first coordinate system is used as the origin of the second coordinate system corresponding to each sensor;

[0035] On the XY plane, with the origin of the second coordinate system corresponding to each sensor as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of each sensor as the X axis of the second coordinate system corresponding to each sensor;

[0036] On the XY plane, with the origin of the second coordinate system corresponding to each sensor as the starting point, the vector obtained by rotating the X axis of the second coordinate system corresponding to each sensor 90 degrees counterclockwise is used as the Y axis of the second coordinate system corresponding to each sensor.

[0037] In combination with the first aspect, in one embodiment, classifying each scanning point into a ground point cloud and an obstacle point cloud based on the original coordinates and the projected coordinates of each scanning point includes:

[0038] Determine the grid where the projection coordinates of each scanning point are located, and calculate the Z value difference between the original coordinates and the projection coordinates of each scanning point;

[0039] Find the minimum value among all Z value differences, and calculate the first difference between each Z value difference and the minimum value;

[0040] If the first difference is less than the height threshold of the grid where the corresponding projection coordinate is located, the scanning point corresponding to the first difference is classified as a ground point cloud;

[0041] If the first difference is greater than the height threshold of the grid where the corresponding projection coordinate is located, the scanning point corresponding to the first difference is classified as an obstacle point cloud.

[0042] In combination with the first aspect, in one embodiment, the minimum distance that the sensor scans the ground is calculated based on the height of the sensor from the ground and the vertical scanning angle of the sensor.

[0043] In a second aspect, the present application provides a near obstacle detection device based on a near obstacle detection method using a laser radar, the device comprising:

[0044] a modeling module configured to establish a first coordinate system and divide the system into grids based on the installation position and parameters of at least one sensor, wherein the sensor is a lidar sensor; and calculate a height threshold corresponding to each grid.

[0045] an analysis module, configured to establish original coordinates for each scan point in the original point cloud according to the first coordinate system; project each scan point in the original point cloud to obtain projection coordinates; and classify each scan point into a ground point cloud and an obstacle point cloud based on the original coordinates and the projection coordinates of each scan point;

[0046] The processing module performs clustering, bounding box information calculation and matching tracking on the obstacle point cloud to obtain obstacle information.

[0047] The beneficial effects of the technical solutions provided in the embodiments of the present application include:

[0048] Based on the installation position and parameters of at least one sensor, a first coordinate system is established and grids are divided, and the height threshold corresponding to each grid is calculated; the original coordinates of each scanning point in the original point cloud are established according to the first coordinate system; each scanning point in the original point cloud is projected to obtain projection coordinates; based on the original coordinates and projection coordinates of each scanning point, each scanning point is classified into a ground point cloud and an obstacle point cloud; the obstacle point cloud is clustered, bounding box information is calculated and matched and tracked to obtain obstacle information, thereby realizing customized height thresholds according to the installation position of the laser radar, dividing each scanning point in the original point cloud according to the height threshold, thereby distinguishing between the ground point cloud and the obstacle point cloud, so that the laser radar can flexibly identify nearby obstacles. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flowchart of a method for detecting nearby obstacles based on lidar is provided in an embodiment of the present application.

[0050] Figure 2 A schematic diagram of a first coordinate system for a method for detecting nearby obstacles based on lidar provided in an embodiment of the present application.

[0051] Figure 3 A schematic diagram of single-sensor grid division for a LiDAR-based method for detecting nearby obstacles provided in an embodiment of the present application.

[0052] Figure 4 A schematic diagram of multi-sensor grid division for a LiDAR-based method for detecting nearby obstacles provided in an embodiment of the present application. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] First, some technical terms in this application are explained to facilitate those skilled in the art to understand this application.

[0055] Laser Radar (LIDAR) is a radar system that uses laser beams to detect target characteristics such as position and speed. Its operating principle is to transmit a detection signal (laser beam) toward the target. The received signal reflected from the target (target echo) is then compared with the transmitted signal. After appropriate processing, relevant target information such as range, direction, altitude, speed, attitude, and shape is obtained, enabling detection, tracking, and identification. It consists of a laser transmitter, an optical receiver, a turntable, and an information processing system. The laser converts electrical pulses into light pulses and transmits them. The optical receiver then converts the light pulses reflected from the target back into electrical pulses and sends them to a display.

[0056] In order to make the objectives, technical solutions and advantages of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0057] First, please refer to Figure 1 , Figure 1 This is a flow chart of a method for detecting nearby obstacles based on a laser radar. This embodiment provides a method for detecting nearby obstacles based on a laser radar, which includes the following steps:

[0058] Step S1: Based on the installation position and parameters of at least one sensor, a first coordinate system is established and divided into grids, and a height threshold corresponding to each grid is calculated. The above-mentioned sensor is a lidar sensor.

[0059] In this embodiment, the installation position of the sensor includes the height of the sensor from the ground, the orientation and inclination of the sensor, and the sensor parameters include the horizontal scanning angle of the sensor and the vertical scanning angle of the sensor.

[0060] Step S2: establishing original coordinates for each scanning point in the original point cloud according to the first coordinate system.

[0061] It should be noted that the above-mentioned original point cloud is the point cloud detected by the sensor. If there is only one such sensor, the point cloud detected by the sensor is directly used as the original point cloud; if there are multiple such sensors, the point clouds detected by multiple sensors are spliced, and the spliced ​​point cloud is used as the original point cloud.

[0062] Step S3: Project each scanning point in the original point cloud to obtain projection coordinates.

[0063] It should be noted that, projecting each scanning point in the original point cloud here refers to projecting each scanning point onto the XY plane of the first coordinate system.

[0064] Step S4: Based on the original coordinates and projection coordinates of each scanning point, each scanning point is classified into a ground point cloud and an obstacle point cloud.

[0065] Step S5: clustering, bounding box information calculation and matching tracking are performed on the above obstacle point cloud to obtain obstacle information.

[0066] In some embodiments, in the above step S1, when there is only one sensor, establishing the first coordinate system includes the following steps:

[0067] S111: The projection point of the sensor midpoint on the ground is taken as the origin.

[0068] S112: Taking the vector from the origin to the midpoint of the sensor as the Z axis of the first coordinate system, determine the plane containing the origin and perpendicularly intersecting the Z axis as the XY plane.

[0069] S113: On the XY plane, with the origin as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of the sensor is used as the X axis of the first coordinate system.

[0070] S114: On the XY plane, with the origin as the starting point, the vector obtained by rotating the X axis counterclockwise by 90 degrees is used as the Y axis of the first coordinate system.

[0071] In some embodiments, in the above step S1, when there are multiple sensors, establishing the first coordinate system includes the following steps:

[0072] S121: Determine the origin based on the projection point of the midpoint of each sensor on the ground.

[0073] S122: A vector from the projection point of any sensor on the ground to the midpoint of the sensor is used as a first vector.

[0074] S123: With the origin as the starting point and a vector parallel to the first vector as the Z axis of the first coordinate system, determine that a plane including the origin and perpendicularly intersecting the Z axis is the XY plane.

[0075] S124: On the XY plane, with the origin as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of any sensor is used as the X axis of the first coordinate system.

[0076] S125: On the XY plane, with the origin as the starting point, the vector obtained by rotating the X axis counterclockwise by 90 degrees is used as the Y axis of the first coordinate system.

[0077] In some embodiments, the grid division in step S1 includes:

[0078] The XY plane of the first coordinate system is divided into square areas of equal area using a preset grid resolution.

[0079] In this embodiment, please refer to the above grid division Figure 3 and Figure 4 .

[0080] Figure 3 A schematic diagram of a single sensor grid division of a laser radar-based near obstacle detection method provided in an embodiment of the present application, in which is the horizontal scanning angle of the sensor.

[0081] Figure 4 A schematic diagram of multi-sensor grid division for a near obstacle detection method based on laser radar provided in an embodiment of the present application, in which 、 、 are the horizontal scanning angles of different sensors respectively.

[0082] In some embodiments, when there is only one sensor, the height threshold corresponding to each grid in step S1 is calculated as follows:

[0083] S131: Obtain the minimum distance for the sensor to scan the ground.

[0084] S132: Calculate the height threshold corresponding to each of the above grids within the horizontal scanning angle of the sensor based on the horizontal distance between the midpoint of each of the above grids and the origin of the above first coordinate system, a preset height ratio threshold, the height of the sensor from the ground, the minimum distance of the above sensor scanning the ground, and the vertical scanning angle of the sensor.

[0085] In some embodiments, when there are multiple sensors, the height threshold corresponding to each grid in step S1 is calculated as follows:

[0086] S141: establishing a second coordinate system corresponding to each sensor according to the installation position and parameters of each sensor, and calculating the coordinates of the midpoint of each grid in the first coordinate system.

[0087] S142: Projecting the coordinates of the midpoints of the grids into the second coordinate system corresponding to the sensors according to the parameters of the sensors to obtain projected coordinates.

[0088] S143: Obtain the minimum distance for each sensor to scan the ground.

[0089] S144: Calculate the first height threshold corresponding to each grid within the horizontal scanning angle of each sensor based on the projected coordinates and the horizontal distance between the origin of the second coordinate system corresponding to each sensor, a preset height ratio threshold, the height of each sensor from the ground, the minimum distance of each sensor scanning the ground, and the vertical scanning angle of each sensor.

[0090] S145: Taking the maximum value of all first height thresholds corresponding to each grid as the height threshold of each grid.

[0091] It should be noted that the maximum value among all first height thresholds corresponding to each grid refers to the maximum value among all first height thresholds corresponding to each grid, rather than the maximum value among all first height thresholds of all grids.

[0092] In this embodiment, the method for calculating the height threshold corresponding to each grid within the horizontal scanning angle of the sensor in step S132 and the method for calculating the first height threshold corresponding to each grid within the horizontal scanning angle of each sensor in step S144 are as follows:

[0093] (1)

[0094] Please refer to Figure 2 , Figure 2 This is a schematic diagram of a first coordinate system of a method for detecting nearby obstacles based on a laser radar provided in an embodiment of the present application. In formula (1), is the original height threshold corresponding to each grid within the horizontal scanning angle of the above sensor, is the horizontal distance between the midpoint of each grid and the origin of the first coordinate system, is the preset height ratio threshold, is the height of the sensor from the ground, The minimum distance for the sensor to scan the ground. is the vertical scanning angle of the sensor.

[0095] In the above step S132, if the grid corresponds to If the height threshold is greater than or equal to the fixed height threshold, the height threshold corresponding to the grid is set as the fixed height threshold. If the height threshold is less than the fixed height threshold, the height threshold corresponding to the grid is set to the height threshold corresponding to the grid. .

[0096] In the above step S144, if the grid corresponds to is greater than or equal to the fixed height threshold, the first height threshold corresponding to the grid is set as the fixed height threshold. If the height threshold is less than the fixed height threshold, the first height threshold corresponding to the grid is set to the height threshold corresponding to the grid.

[0097] It should be noted that the height ratio threshold refers to the ratio of the height threshold corresponding to each grid within the horizontal scanning angle of the sensor to the height of the sensor from the ground. In this embodiment, it is preferably set to 0.5. The fixed height threshold refers to the maximum value of the height thresholds corresponding to the grids. In this embodiment, it is preferably set to 0.4.

[0098] In some embodiments, the minimum distance that the sensor scans the ground in steps S131 and S143 is calculated based on the height of the sensor from the ground and the vertical scanning angle of the sensor.

[0099] In this embodiment, the minimum distance for the sensor to scan the ground is calculated as follows:

[0100] (2)

[0101] In formula (2), The minimum distance for the sensor to scan the ground. is the height of the sensor from the ground, is the vertical scanning angle of the sensor.

[0102] In some embodiments, the step S142 above establishes a second coordinate system corresponding to each sensor according to the installation position and parameters of each sensor, including:

[0103] S151: The projection point of the midpoint of each sensor on the XY plane of the first coordinate system is used as the origin of the second coordinate system corresponding to each sensor.

[0104] S152: On the XY plane, with the origin of the second coordinate system corresponding to each sensor as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of each sensor as the X axis of the second coordinate system corresponding to each sensor.

[0105] S153: On the XY plane, with the origin of the second coordinate system corresponding to each sensor as the starting point, the vector obtained by rotating the X axis of the second coordinate system corresponding to each sensor by 90 degrees counterclockwise is used as the Y axis of the second coordinate system corresponding to each sensor.

[0106] In some embodiments, in the above step S4, each scanning point is classified into a ground point cloud and an obstacle point cloud based on the original coordinates and the projected coordinates of each scanning point, specifically including:

[0107] S41: Determine the grid where the projection coordinates corresponding to each scanning point are located, and calculate the Z value difference between the original coordinates and the projection coordinates of each scanning point.

[0108] S42: Find the minimum value among all Z value differences, and calculate the first difference between each Z value difference and the minimum value.

[0109] S43: Determine whether the first difference is less than the height threshold of the grid where the corresponding projection coordinate is located. If yes, proceed to S44; if not, proceed to S45.

[0110] S44: The scanning point corresponding to the first difference is classified as a ground point cloud, and this determination is completed.

[0111] S45: Classify the scanning point corresponding to the first difference as an obstacle point cloud.

[0112] In a second aspect, the present application provides a near obstacle detection device based on a near obstacle detection method using a laser radar, the device comprising:

[0113] A modeling module is used to establish a first coordinate system and divide the grid based on the installation position and parameters of at least one sensor, and calculate the height threshold corresponding to each grid; the above-mentioned sensor is a lidar sensor.

[0114] An analysis module is configured to establish original coordinates for each scanning point in the original point cloud according to the first coordinate system; project each scanning point in the original point cloud to obtain projection coordinates; and classify each scanning point into a ground point cloud and an obstacle point cloud based on the original coordinates and projection coordinates of each scanning point.

[0115] The processing module clusters the obstacle point cloud, calculates bounding box information, and performs matching tracking to obtain obstacle information.

[0116] Through the methods and devices in the above embodiments, the laser radar can adaptively adjust the height threshold for distinguishing nearby obstacles, distinguish the ground point cloud and the obstacle point cloud in the original point cloud, and then cluster the obstacle point cloud, calculate the bounding box information and match and track it, and finally obtain the obstacle information, which effectively reduces the problem that the laser radar has difficulty in identifying nearby obstacles.

[0117] It should be noted that the serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0118] The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit the "first", "second" and "third" to different types.

[0119] In the description of the embodiments of this application, the words "exemplary," "for example," or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary," "for example," or "for example" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary," "for example," or "for example" is intended to present the relevant concepts in a concrete manner.

[0120] In the description of the embodiments of the present application, unless otherwise specified, “ / ” means or, for example, A / B can mean A or B; “and / or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, “multiple” refers to two or more than two.

[0121] In some processes described in the embodiments of the present application, multiple operations or steps are included that appear in a specific order. However, it should be understood that these operations or steps may not be performed in the order in which they appear in the embodiments of the present application or may be performed in parallel. The sequence numbers of the operations are only used to distinguish between different operations, and the sequence numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations or steps may be performed in sequence or in parallel, and these operations or steps may be combined.

[0122] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, or the part that contributes to the existing technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above and includes a number of instructions for enabling a terminal device to execute the methods described in each embodiment of this application.

[0123] The above are only preferred embodiments of the present application and do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for detecting nearby obstacles based on laser radar, characterized in that: The method comprises: Based on the installation position and parameters of at least one sensor, a first coordinate system is established and a grid is divided on the XY plane, and a height threshold corresponding to each grid is calculated; the sensor is a lidar sensor; Establishing original coordinates for each scanning point in the original point cloud according to the first coordinate system; projecting each scanning point in the original point cloud to obtain projection coordinates; Based on the original coordinates and projection coordinates of each scanning point, each scanning point is classified into ground point cloud and obstacle point cloud; Clustering, bounding box information calculation and matching tracking are performed on the obstacle point cloud to obtain obstacle information; When there is only one sensor, the height threshold corresponding to each grid is calculated as follows: Get the minimum distance the sensor can scan the ground; Calculating a height threshold corresponding to each grid within a horizontal scanning angle of the sensor based on a horizontal distance between a midpoint of each grid and the origin of the first coordinate system, a preset height ratio threshold, a height of the sensor from the ground, a minimum distance for the sensor to scan the ground, and a vertical scanning angle of the sensor; The method of classifying each scanning point into a ground point cloud and an obstacle point cloud based on the original coordinates and the projected coordinates of each scanning point includes: Determine the grid where the projection coordinates of each scanning point are located, and calculate the Z value difference between the original coordinates and the projection coordinates of each scanning point; Find the minimum value among all Z value differences, and calculate the first difference between each Z value difference and the minimum value; If the first difference is less than the height threshold of the grid where the corresponding projection coordinate is located, the scanning point corresponding to the first difference is classified as a ground point cloud; If the first difference is greater than the height threshold of the grid where the corresponding projection coordinate is located, the scanning point corresponding to the first difference is classified as an obstacle point cloud.

2. The laser radar-based method for detecting nearby obstacles according to claim 1, wherein: When there is only one sensor, establishing the first coordinate system includes: The projection point of the midpoint of the sensor on the ground is taken as the origin; Taking the vector pointing from the origin to the midpoint of the sensor as the Z axis of the first coordinate system, determining the plane containing the origin and perpendicularly intersecting the Z axis as the XY plane; On the XY plane, with the origin as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of the sensor is the X axis of the first coordinate system; On the XY plane, with the origin as the starting point, a vector obtained by rotating the X axis counterclockwise by 90 degrees is used as the Y axis of the first coordinate system.

3. The laser radar-based method for detecting nearby obstacles according to claim 1, wherein: When there are multiple sensors, establishing the first coordinate system includes: Determine the origin based on the projection point of each sensor's midpoint on the ground; The vector from the projection point of any sensor on the ground to the midpoint of the sensor is used as the first vector; Taking the origin as the starting point and a vector parallel to the first vector as the Z axis of the first coordinate system, determine a plane containing the origin and perpendicularly intersecting the Z axis as the XY plane; On the XY plane, with the origin as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of any sensor is used as the X axis of the first coordinate system; On the XY plane, with the origin as the starting point, a vector obtained by rotating the X axis counterclockwise by 90 degrees is used as the Y axis of the first coordinate system.

4. The laser radar-based method for detecting nearby obstacles according to claim 1, wherein: When there are multiple sensors, the height threshold corresponding to each grid is calculated as follows: Establishing a second coordinate system corresponding to each sensor according to the installation position and parameters of each sensor, and calculating the coordinates of the midpoint of each grid in the first coordinate system; According to the parameters of each sensor, the coordinates of the midpoints of each grid are projected into the second coordinate system corresponding to each sensor to obtain the projected coordinates; Get the minimum distance for each sensor to scan the ground; Calculating a first height threshold corresponding to each grid within a horizontal scanning angle of each sensor based on a horizontal distance between the projected coordinates and the origin of the second coordinate system corresponding to each sensor, a preset height ratio threshold, a height of each sensor from the ground, a minimum distance scanned by each sensor against the ground, and a vertical scanning angle of each sensor; The maximum value of all first height thresholds corresponding to each grid is taken as the height threshold of each grid.

5. The method for detecting nearby obstacles based on laser radar according to claim 4, wherein: The second coordinate system corresponding to each sensor is established according to the installation position and parameters of each sensor, including: The projection point of the midpoint of each sensor on the XY plane of the first coordinate system is used as the origin of the second coordinate system corresponding to each sensor; On the XY plane, with the origin of the second coordinate system corresponding to each sensor as the starting point, a vector parallel to the angle bisector of the horizontal scanning angle of each sensor as the X axis of the second coordinate system corresponding to each sensor; On the XY plane, with the origin of the second coordinate system corresponding to each sensor as the starting point, the vector obtained by rotating the X axis of the second coordinate system corresponding to each sensor 90 degrees counterclockwise is used as the Y axis of the second coordinate system corresponding to each sensor.

6. The method for detecting nearby obstacles based on laser radar according to claim 1 or 4, wherein: The minimum distance that the sensor scans the ground is calculated based on the height of the sensor from the ground and the vertical scanning angle of the sensor.

7. A near obstacle detection device based on the near obstacle detection method based on laser radar according to any one of claims 1 to 6, characterized in that: The device comprises: a modeling module configured to establish a first coordinate system and divide the system into grids based on the installation position and parameters of at least one sensor, wherein the sensor is a lidar sensor; and calculate a height threshold corresponding to each grid. an analysis module, configured to establish original coordinates for each scan point in the original point cloud according to the first coordinate system; project each scan point in the original point cloud to obtain projection coordinates; and classify each scan point into a ground point cloud and an obstacle point cloud based on the original coordinates and the projection coordinates of each scan point; The processing module performs clustering, bounding box information calculation and matching tracking on the obstacle point cloud to obtain obstacle information.

Images