A point cloud data processing method and device

Abstract

Embodiments of the present application provide a point cloud data processing method and device, and relate to the technical field of data processing. The method comprises: determining a target laser point corresponding to a target object in first point cloud data to be processed; obtaining a relative speed of the target object relative to a traveling object at a target acquisition time corresponding to the target laser point; and adjusting position information of the target laser point in a direction opposite to a direction of the relative speed based on an acquisition time difference and the relative speed. The point cloud data processing scheme provided by the embodiments of the present application can improve the accuracy of point cloud data.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a point cloud data processing method and apparatus. Background Technology

[0002] In existing technologies, a driving object can be equipped with a LiDAR (Light Detection and Ranging) system. A scanning LiDAR can periodically collect point cloud data of the environment surrounding the driving object in various directions by rotating. The point cloud data composed of all the laser points collected by the LiDAR within one acquisition cycle can be called a frame of point cloud data. The moment when the LiDAR collects any laser point within one acquisition cycle can be taken as the acquisition time of that frame of point cloud data. After acquiring the point cloud data, the data processing device can process the data to obtain environmental information about the environment surrounding the driving object, and use this environmental information as the environmental information of the environment surrounding the driving object at the time the point cloud data was acquired.

[0003] However, since it takes time for a lidar to acquire a frame of point cloud data, the acquisition time corresponding to each laser point in the same frame may be different. Furthermore, during the movement of the aforementioned vehicle, the environment surrounding the vehicle often changes at different acquisition times. In this case, the environmental information reflected by different laser points in the same frame may be environmental information from different times, resulting in low accuracy of the point cloud data composed of laser points acquired by the lidar at different acquisition times. Summary of the Invention

[0004] The purpose of this invention is to provide a point cloud data processing method and apparatus to improve the accuracy of point cloud data. The specific technical solution is as follows:

[0005] In a first aspect, embodiments of the present invention provide a point cloud data processing method, the method comprising:

[0006] In the first point cloud data to be processed, the target laser point corresponding to the target object is determined, wherein the first point cloud data is the point cloud data collected by the lidar installed on the moving object;

[0007] The relative velocity of the target object to the moving object is obtained at the target acquisition time corresponding to the target laser point;

[0008] Based on the acquisition time difference and the relative velocity, the position information of the target laser point is adjusted in the direction opposite to the direction of the relative velocity. The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data.

[0009] Secondly, embodiments of the present invention provide a point cloud data processing apparatus, the apparatus comprising:

[0010] The target determination module is used to determine the target laser point of the corresponding target object in the first point cloud data to be processed, wherein the first point cloud data is the point cloud data collected by the lidar installed on the moving object;

[0011] A speed acquisition module is used to acquire the relative speed of the target object relative to the moving object at the target acquisition time corresponding to the target laser point;

[0012] The position adjustment module is used to adjust the position information of the target laser point in the opposite direction to the direction of the relative velocity based on the acquisition time difference and the relative velocity. The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data.

[0013] Thirdly, embodiments of the present invention provide an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0014] Memory, used to store computer programs;

[0015] When a processor executes a program stored in a memory, it implements the point cloud data processing method steps described in any of the first aspects above.

[0016] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the point cloud data processing method steps described in any of the first aspects above.

[0017] Fifthly, embodiments of the present invention also provide a computer program product containing instructions that, when run on a computer, cause the computer to execute any of the point cloud data processing methods described in the first aspect above.

[0018] Beneficial effects of the embodiments of the present invention:

[0019] In the point cloud data processing scheme provided in this embodiment of the invention, the target laser point corresponding to the target object is first determined in the first point cloud data, then the relative speed of the target object relative to the moving object is obtained at the target acquisition time corresponding to the target laser point, and finally the position information of the target laser point is adjusted in the direction opposite to the direction of the relative speed based on the acquisition time difference and the relative speed.

[0020] As can be seen from the above, both the target acquisition time and the first acquisition time are the acquisition times of a portion of the laser points in the first point cloud data collected by the lidar. Since the time required for the lidar to acquire one frame of the first point cloud data is usually short, the acquisition time difference between the target acquisition time and the first acquisition time is usually small. The relative velocity mentioned above is the velocity of the target object relative to the moving object at the target acquisition time. Since the acquisition time difference is usually small, the velocity of the target object relative to the moving object between the target acquisition time and the first acquisition time can be approximated as the relative velocity. Therefore, based on the acquisition time difference and the relative velocity, the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time can be obtained. Adjusting the position information of the target laser points in the direction opposite to the direction of the relative velocity according to the relative displacement can offset the offset of the target laser points in the first point cloud data caused by the relative displacement between the target object and the moving object, so that the laser points in the adjusted first point cloud data are closer to the laser points that the lidar can acquire at the first acquisition time. Therefore, the solution provided by the embodiments of the present invention can improve the accuracy of point cloud data. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0022] Figure 1 A flowchart illustrating the first point cloud data processing method provided in an embodiment of the present invention;

[0023] Figure 2 A flowchart illustrating the second point cloud data processing method provided in this embodiment of the invention;

[0024] Figure 3 A flowchart illustrating the third point cloud data processing method provided in this embodiment of the invention;

[0025] Figure 4 A flowchart illustrating the fourth point cloud data processing method provided in this embodiment of the invention;

[0026] Figure 5 A flowchart illustrating the fifth point cloud data processing method provided in this embodiment of the invention;

[0027] Figure 6 A flowchart illustrating the sixth point cloud data processing method provided in this embodiment of the invention;

[0028] Figure 7This is a schematic diagram of the structure of the first point cloud data processing device provided in an embodiment of the present invention;

[0029] Figure 8 This is a schematic diagram of the structure of the second point cloud data processing device provided in an embodiment of the present invention;

[0030] Figure 9 This is a schematic diagram of the structure of the third point cloud data processing device provided in an embodiment of the present invention;

[0031] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 based on the present invention are within the scope of protection of the present invention.

[0033] To address the issue of low accuracy in point cloud data processing in existing technologies, this invention provides a point cloud data processing method and apparatus.

[0034] In one embodiment of the present invention, a point cloud data processing method is provided, the method comprising:

[0035] In the first point cloud data to be processed, the target laser point of the corresponding target object is determined. The first point cloud data is the point cloud data collected by the lidar installed on the moving object.

[0036] Obtain the relative velocity of the target object with respect to the moving object at the target acquisition time corresponding to the target laser point;

[0037] Based on the acquisition time difference and relative velocity, the position information of the target laser point is adjusted in the direction opposite to the direction of the relative velocity. The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data.

[0038] As can be seen from the above, both the target acquisition time and the first acquisition time are the acquisition times of a portion of the laser points in the first point cloud data collected by the lidar. Since the time required for the lidar to acquire one frame of the first point cloud data is usually short, the acquisition time difference between the target acquisition time and the first acquisition time is usually small. The relative velocity mentioned above is the velocity of the target object relative to the moving object at the target acquisition time. Since the acquisition time difference is usually small, the velocity of the target object relative to the moving object between the target acquisition time and the first acquisition time can be approximated as the relative velocity. Therefore, based on the acquisition time difference and the relative velocity, the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time can be obtained. Adjusting the position information of the target laser points in the direction opposite to the direction of the relative velocity according to the relative displacement can offset the offset of the target laser points in the first point cloud data caused by the relative displacement between the target object and the moving object, so that the laser points in the adjusted first point cloud data are closer to the laser points that the lidar can acquire at the first acquisition time. Therefore, the solution provided by the embodiments of the present invention can improve the accuracy of point cloud data.

[0039] First, the subject responsible for implementing the solution provided in the embodiments of the present invention will be described.

[0040] In one scenario, the entity executing the solution provided in this embodiment of the invention can be a data processing device installed in the vehicle. In this case, the lidar and the data processing device can communicate wirelessly or via wired communication. After the lidar collects point cloud data, it sends the point cloud data to the data processing device. Upon receiving the point cloud data, the data processing device can process the point cloud data using the solution provided in this embodiment of the invention, thereby obtaining environmental information about the driving environment of the vehicle based on the processed point cloud data. This allows the device to control the vehicle's movement based on the environmental information.

[0041] In another scenario, the entity executing the solution provided in this embodiment of the invention can also be a remote server. In this case, the lidar and the remote server can communicate wirelessly. After the lidar collects point cloud data, it sends the point cloud data to the remote server via wireless communication. Upon receiving the point cloud data, the remote server can process the point cloud data using the solution provided in this embodiment of the invention, and then obtain environmental information about the driving environment of the driving object based on the processed point cloud data, thereby analyzing the driving process of the driving object based on the environmental information.

[0042] The following detailed description of a point cloud data processing method, apparatus, electronic device, and storage medium provided by the present invention will be provided through specific embodiments.

[0043] See Figure 1 The flowchart of the first point cloud data processing method is provided. The method includes the following steps S101-S103.

[0044] Step S101: In the first point cloud data to be processed, determine the target laser point corresponding to the target object.

[0045] The first point cloud data is the point cloud data collected by the lidar installed on the moving object.

[0046] In one embodiment of the present invention, before performing the above step S101, the point cloud data collected by the lidar can be segmented to determine the dynamic point cloud data and static point cloud data in the point cloud data. The dynamic point cloud data is used to describe the point cloud data of an object in motion, and the static point cloud data is used to describe the point cloud data of an object in a stationary state.

[0047] Both the dynamic point cloud data and the static point cloud data mentioned above can be used as the first point cloud data.

[0048] In addition, point cloud segmentation can be performed on point cloud data using existing technologies, which will not be detailed here.

[0049] The aforementioned objects that can be driven can be objects with driving functions, such as vehicles, robots, etc.

[0050] The target object mentioned above can be an object in the driving environment where the driving object is located. For example, if the driving object is a vehicle, the target object can be other vehicles around the vehicle, or pedestrians, buildings, etc. around the vehicle.

[0051] The aforementioned lidar mainly refers to lidar that can collect point cloud data in a rotating manner, such as mechanically rotating lidar.

[0052] Since lidar typically collects point cloud data of the surrounding environment by emitting laser beams and receiving the reflected laser beams, each laser beam received by the lidar corresponds to a laser point in the point cloud data. Therefore, the target laser point corresponding to the target object is the laser point in the point cloud data that corresponds to the laser beam reflected back from the target object after it comes into contact with the target object.

[0053] In one implementation, feature extraction can be performed on the first point cloud data. Based on the extracted features, laser points in the first point cloud data that match the object features of the target object can be determined as target laser points. Feature extraction on the first point cloud data can be implemented based on a feature extraction algorithm or based on a deep learning network used for feature extraction. Furthermore, the method of extracting features from the first point cloud data and determining laser points in the first point cloud data that match the object features of the target object based on the extracted features can be implemented using existing technologies, which will not be detailed here.

[0054] In another implementation, point cloud clustering can be performed on the first point cloud data to determine the target laser point. For details on determining the target laser point through point cloud clustering of the first point cloud data, please refer to the following sections. Figure 6 The embodiment shown in step S101A will not be described in detail here.

[0055] Furthermore, there may be one or more target objects. If there are multiple target objects, there will be multiple sets of target laser points corresponding to different target objects in the first point cloud data. For each target object, the target laser point corresponding to that target object can be determined in the first point cloud data through the above step S101.

[0056] Step S102: Obtain the relative speed of the target object relative to the moving object at the target acquisition time corresponding to the target laser point.

[0057] The target acquisition time mentioned above refers to the time when the lidar acquires the target laser point.

[0058] Specifically, in the first point cloud data, there are usually multiple target laser points corresponding to a single target object. The target acquisition times for different laser points may differ, but since the time required to acquire all the target laser points corresponding to a single target object is usually short, the difference between the target acquisition times of different laser points is small, and the change in relative velocity between different target acquisition times is also small. Therefore, it can be assumed that the relative velocities corresponding to the target acquisition times of different laser points are the same. Based on this, the relative velocity of the target object relative to the moving object at the target acquisition time of any laser point can be taken as the relative velocity of the target object relative to the moving object at the target acquisition times of all laser points in the target laser system.

[0059] In addition, the aforementioned relative speed is the speed of the target object relative to the moving object, with the moving object as the reference frame.

[0060] In one embodiment of the present invention, it can be achieved through subsequent... Figure 3The relative velocity is obtained in steps S102A-S102D of the embodiment shown, which will not be described in detail here.

[0061] In another embodiment of the present invention, the first point cloud data can be subjected to point cloud clustering processing, and the laser points belonging to the same cluster obtained by clustering can be determined as the laser points of the corresponding target object. Then, based on the determined laser points, the relative speed of the target object relative to the moving object can be obtained using the Kalman filter method or particle filter method in the prior art.

[0062] The specific details of point cloud clustering processing for the first point cloud data will be discussed later. Figure 6 Step S101A in the illustrated embodiment will be explained, but will not be detailed here.

[0063] In another embodiment of the present invention, the target speed of the target object and the driving speed of the driving object can be obtained at the target acquisition time, and the relative speed of the target object with respect to the driving object at the target acquisition time can be obtained by subtracting the driving speed of the driving object from the target speed of the target object.

[0064] For example, if the moving object is vehicle M and the target object is vehicle N, and the speed of vehicle M at the target acquisition time is 80 km / h, and the speed of vehicle N at the target acquisition time is 100 km / h, then the relative speed of the target object relative to the moving object at the target acquisition time is 100 - 80 = 20 km / h.

[0065] Specifically, the target velocity of the target object at the time of target acquisition can be obtained in any of the following methods (i)-(ii).

[0066] Method (1): If a speed measuring device for detecting speed is installed in the driving environment where the target object is located, the target speed of the target object can be obtained by the aforementioned speed measuring device.

[0067] Method (2): When the target object is a vehicle, robot or other object with driving function, a speed sensor can be installed in the target object, and the target speed of the target object can be collected by the speed sensor.

[0068] The speed of the moving object at the target acquisition time can be obtained by any of the following methods (iii)-(iv).

[0069] Method (3): If a speed measuring device for detecting speed is installed in the driving environment where the driving object is located, the driving speed of the driving object can be obtained by the aforementioned speed measuring device.

[0070] Method (4): When a speed sensor is installed on the vehicle, the vehicle's speed can be the speed collected by the speed sensor.

[0071] In addition, if speed measuring devices are installed in the driving environment where the target object and the driving object are located, the speed measuring devices can obtain the target speed of the target object and the driving speed of the driving object at the time of target acquisition, and subtract the driving speed from the target speed to obtain the relative speed of the target object relative to the driving object.

[0072] Step S103: Based on the acquisition time difference and relative velocity, adjust the position information of the target laser point in the direction opposite to the direction of the relative velocity.

[0073] The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data.

[0074] The aforementioned first preset laser point can be the laser point with the earliest acquisition time in the first point cloud data, the laser point with the latest acquisition time in the first point cloud data, or any other laser point in the first point cloud data.

[0075] The aforementioned acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The target acquisition time is the acquisition time corresponding to the target laser point, and the first acquisition time is the acquisition time corresponding to the first preset laser point. Therefore, the aforementioned acquisition time difference is also the time difference between the acquisition time corresponding to the target laser point and the acquisition time of the first preset laser point in the first point cloud data.

[0076] The position information of the target laser point can be used to characterize the relative position of the target object with respect to the moving object. Based on the position information of the target laser point, the relative position of the target object with respect to the moving object can be determined.

[0077] For example, when the above location information is represented in the form of three-dimensional coordinates in a spatial coordinate system, the origin of this spatial coordinate system can be the location of the moving object. The X-axis in this spatial coordinate system is defined as due east, the Y-axis as due south, and the Z-axis as a direction perpendicular to the ground upwards. If the above location information is (100, 20, 5), it means that the target object is located 100 meters due east, 20 meters due south, and 5 meters perpendicular to the ground upwards from the moving object.

[0078] The relative velocity described above is a vector with magnitude and direction. Based on the direction of the relative velocity, the direction opposite to the direction of the relative velocity can be determined. For example, if the direction of the relative velocity is due east, then the direction opposite to the direction of the relative velocity is due west; if the direction of the relative velocity is southeast, then the direction opposite to the direction of the relative velocity is northwest.

[0079] In one embodiment of the present invention, the above-mentioned acquisition time difference can be multiplied by the above-mentioned relative speed to obtain the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time.

[0080] When the aforementioned positional information is expressed in the form of three-dimensional coordinates in a spatial coordinate system with the location of the moving object as the origin, the aforementioned relative displacement can be converted into three-dimensional coordinates in the aforementioned spatial coordinate system. Based on the direction of the aforementioned relative displacement and the positive directions of the three coordinate axes in the aforementioned spatial coordinate system, the sign of each coordinate in the three-dimensional coordinate system representing the aforementioned relative displacement can be determined.

[0081] For example, if the three-dimensional coordinates of the relative displacement in the spatial coordinate system are (10, 4, 2), and the X-axis is defined as due east, the Y-axis as due south, and the Z-axis as the direction perpendicular to the ground upward, then it means that the target object moved 10 meters due east, 4 meters due south, and 2 meters perpendicular to the ground upward relative to the moving object within the acquisition time difference.

[0082] Specifically, by subtracting the coordinates of the target laser point from the coordinates of the same axis in the three-dimensional coordinates of the relative displacement, the position information of the target laser point can be adjusted in the direction opposite to the direction of the relative velocity.

[0083] For example, if the three-dimensional coordinates of the target laser point are (100, 20, 5) and the three-dimensional coordinates of the relative displacement are (10, 4, 2), then for the X-axis, Y-axis, and Z-axis respectively, subtract the coordinates of the relative displacement on the corresponding axes from the coordinates of the target laser point on those axes. In the resulting three-dimensional coordinates, the X-axis coordinate is 100-10=90, the Y-axis coordinate is 20-4=16, and the Z-axis coordinate is 5-2=3. Therefore, the adjusted three-dimensional coordinates of the target laser point are (90, 16, 3).

[0084] In addition, it can also be based on subsequent Figure 2 In the embodiment shown, steps S103A-S103C adjust the position information of the target laser point, which will not be described in detail here.

[0085] In addition, since there are often multiple target laser points corresponding to a target object in the first point cloud data, for each target laser point determined in step S101 above, the acquisition time difference corresponding to the target laser point can be calculated. Then, based on the calculated acquisition time difference and the relative velocity, the position information of the target laser point is adjusted in the direction opposite to the direction of the relative velocity.

[0086] As can be seen from the above, in the solution provided by the embodiments of the present invention, both the target acquisition time and the first acquisition time are the acquisition times of a portion of the laser points in the first point cloud data acquired by the lidar. Since the time required for the lidar to acquire one frame of the first point cloud data is often short, the acquisition time difference between the target acquisition time and the first acquisition time is often small. The relative speed mentioned above is the speed of the target object relative to the moving object at the target acquisition time. Since the acquisition time difference is usually small, the speed of the target object relative to the moving object between the target acquisition time and the first acquisition time can be approximated as the relative speed. Therefore, based on the acquisition time difference and the relative speed, the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time can be obtained. Adjusting the position information of the target laser points in the direction opposite to the direction of the relative speed according to the relative displacement can offset the offset of the target laser points in the first point cloud data caused by the relative displacement between the target object and the moving object, so that the laser points in the adjusted first point cloud data are closer to the laser points that the lidar can acquire at the first acquisition time. Therefore, the accuracy of point cloud data can be improved through the solution provided by the embodiments of the present invention.

[0087] In one embodiment of the present invention, after adjusting the position information of the target laser point, adjusted first point cloud data can be obtained, which can be referred to as adjusted point cloud data for convenience. Target recognition is performed on the adjusted point cloud data to determine the laser point corresponding to the target object in the adjusted point cloud data. Based on the determined position information of the laser point, object information of the target object can be obtained. For example, the object information may include: the object's position, size, shape, or type, etc. Then, based on the object information of the target object, the movement of the vehicle can be controlled, for example, controlling the vehicle to avoid obstacles, or planning the vehicle's route based on the object information and controlling the vehicle to move according to the planned route, etc.

[0088] Since the point cloud data processing scheme provided by the embodiments of the present invention can improve the accuracy of point cloud data, the accuracy of the adjusted point cloud data is higher than that of the first point cloud data. The accuracy of the object information of the target object obtained based on the adjusted point cloud data is also higher. Therefore, the driving of the driving object can be controlled more accurately, and the driving route of the driving object can be planned more accurately, thus ensuring the safety of the driving object during the driving process.

[0089] In one embodiment of the present invention, see Figure 2 It provides a flowchart of the second point cloud data processing method, which is consistent with the aforementioned Figure 1 Compared to the illustrated embodiment, in this embodiment, step S103 can be achieved through the following steps S103A-S103C.

[0090] Step S103A: Decompose the relative velocity into longitudinal velocity and lateral velocity.

[0091] The longitudinal velocity is in the same direction as the direction of travel of the object, while the lateral velocity is perpendicular to the direction of travel of the object.

[0092] The direction of the longitudinal velocity is the same as the direction of travel of the object, and the direction of the lateral velocity is perpendicular to the direction of travel of the object. Therefore, the direction of the longitudinal velocity is perpendicular to the direction of the lateral velocity.

[0093] Specifically, the above decomposes the relative velocity into longitudinal velocity and lateral velocity, which is to say, to perform an orthogonal decomposition of the relative velocity, decomposing the velocity in one direction into two velocities that are perpendicular to each other.

[0094] Furthermore, the directions of the longitudinal and lateral velocities mentioned above are determined based on the target object's direction of travel. In addition, the directions of the longitudinal and lateral velocities can also be determined based on geographical location. For example, the longitudinal velocity can be determined by the direction due east, and the lateral velocity by the direction due south.

[0095] Step S103B: Based on the acquisition time difference and longitudinal velocity, obtain the longitudinal compensation amount, and adjust the position information of the target laser point based on the longitudinal compensation amount in the direction opposite to the longitudinal velocity.

[0096] Specifically, the longitudinal compensation amount is obtained by multiplying the above-mentioned acquisition time difference by the above-mentioned longitudinal velocity.

[0097] For example, if the acquisition time difference is 50 milliseconds, or 0.05 seconds, and the longitudinal velocity is 10 m / s, then the longitudinal compensation is 0.05 * 10 = 0.5 m.

[0098] Similar to step S103 above, the position information of the target laser point can be represented in the form of two-dimensional coordinates in a two-dimensional coordinate system. In the above two-dimensional coordinates, the X-axis coordinate represents the distance between the target object and the moving object in the direction of longitudinal velocity, and the Y-axis coordinate represents the distance between the target object and the moving object in the direction of lateral velocity. By subtracting the above longitudinal compensation amount from the X-axis coordinate of the target laser point in the two-dimensional coordinates, the position information of the target laser point can be adjusted based on the longitudinal compensation amount in the direction opposite to the direction of longitudinal velocity.

[0099] Step S103C: Based on the acquisition time difference and lateral velocity, obtain the lateral compensation amount, and adjust the position information of the target laser point based on the lateral compensation amount in the direction opposite to the direction of the lateral velocity.

[0100] Similar to step S103B above, the lateral compensation amount can be obtained by multiplying the acquisition time difference and the lateral velocity. By subtracting the lateral compensation amount from the Y-axis coordinate of the target laser point in the two-dimensional coordinate system, the position information of the target laser point can be adjusted based on the lateral compensation amount in the direction opposite to the direction of the lateral velocity.

[0101] In addition, in the solution provided in this embodiment, step S103B can be executed first, and then step S103C can be executed. That is, in the process of adjusting the position information of the target laser point, the position information of the target laser point can be adjusted first in the direction opposite to the longitudinal velocity, and then the position information of the target laser point can be adjusted in the direction opposite to the transverse velocity.

[0102] In addition, step S103C can be executed first, followed by step S103B. That is, during the process of adjusting the position information of the target laser point, the position information of the target laser point can be adjusted first in the direction opposite to the horizontal velocity, and then the position information of the target laser point can be adjusted in the direction opposite to the vertical velocity.

[0103] As can be seen from the above, in the solution provided by the embodiments of the present invention, the relative velocity is decomposed into longitudinal velocity and lateral velocity, and the position information of the target laser point is adjusted twice in the directions opposite to the longitudinal velocity and the directions opposite to the lateral velocity, respectively. This offsets the offset of the target laser point in the first point cloud data caused by the relative displacement between the target object and the moving object in the two directions, thereby improving the accuracy of the point cloud data.

[0104] In one embodiment of the present invention, see Figure 3 It provides a flowchart of a third point cloud data processing method, which is consistent with the aforementioned Figure 1 Compared to the illustrated embodiment, in this embodiment, step S102 can be achieved through the following steps S102A-S102D.

[0105] Step S102A: Obtain the first position of the target object at the target acquisition time.

[0106] Among them, the first position of the target object is the absolute position of the target object, that is, the position of the target object in the driving environment in which the target object is located.

[0107] Specifically, this step can be implemented in any of the following ways (v) to (viii).

[0108] Method (5): The first position of the target object at the time of target acquisition can be determined by obtaining the relative position of the target object with respect to the moving object and the position of the moving object itself at the time of target acquisition.

[0109] For example, at the time of target acquisition, if the relative position of the target object to the moving object is that the target object is located 100 meters east of the moving object, and the position information of the moving object itself is that the moving object is located 200 meters west of a certain building, then it can be determined that the target object is located 100 meters west of the aforementioned building.

[0110] In one embodiment of the present invention, if a positioning sensor for determining location information is installed in the vehicle, the positioning sensor can collect the position of the vehicle at the target collection time.

[0111] In another embodiment of the present invention, if a positioning device is installed in the driving environment where the driving object is located, the positioning device can collect the position of the driving object at the target collection time.

[0112] For example, the aforementioned positioning device could be a lidar, a camera, or the like.

[0113] Method (6): If a positioning device is installed in the driving environment where the target object is located, the first position of the target object can be obtained by the positioning device.

[0114] Method (7): If a positioning sensor is installed in the target object, the first position of the target object can be obtained by the positioning sensor mentioned above.

[0115] Method (8): The first position of the target object at the target acquisition time can also be obtained through step S102A1 in subsequent embodiments, which will not be described in detail here.

[0116] Step S102B: Obtain the second position of the target object at the second acquisition time of the second point cloud data.

[0117] In the second point cloud data, the time difference between the second acquisition time corresponding to the second preset laser point and the first acquisition time is less than the preset time difference.

[0118] The preset time difference can be 25 milliseconds, 50 milliseconds, or other durations.

[0119] The aforementioned second preset laser point can be the laser point with the earliest acquisition time in the second point cloud data, the laser point with the latest acquisition time in the second point cloud data, or any other laser point in the second point cloud data.

[0120] The aforementioned second point cloud data can be point cloud data acquired before or after the first acquisition time. Furthermore, if multiple point cloud data acquired by the LiDAR have a time difference less than a preset time difference between their acquisition time and the first acquisition time, multiple second point cloud data can be determined. For each second point cloud data, its acquisition time can be either before or after the first acquisition time. For each second point cloud data, the second position of the target object at the acquisition time of that second point cloud data can be obtained.

[0121] Specifically, this step is similar to step S102A above, except that the time corresponding to the obtained position of the target object is different, while the method of obtaining the position of the target object is the same, so it will not be described again here.

[0122] Step S102C: Based on the first position and the second position, calculate the target velocity of the target object at the target acquisition time.

[0123] Since the first position corresponds to the target acquisition time and the second position corresponds to the second acquisition time, based on the first and second positions, the displacement of the target object between the target acquisition time and the second acquisition time can be determined. This displacement is the displacement of the target object due to its own movement between the two acquisition times.

[0124] In one scenario, the preset time difference is typically small, for example, 30 milliseconds. This means the time difference between the first and second acquisition moments is small. Furthermore, the acquisition time for a single frame of point cloud data by a LiDAR is usually short, typically 50 milliseconds. Therefore, the time difference between the target acquisition moment and the first acquisition moment is also small, and consequently, the time difference between the target acquisition moment and the second acquisition moment is also small. Based on this, it can be assumed that the target velocity remains constant between the target acquisition moment and the second acquisition moment. Therefore, the velocity obtained by dividing the displacement of the target object by the time difference between the target acquisition moment and the second acquisition moment is the target velocity of the target object at the first acquisition moment.

[0125] In another scenario, if there are multiple second point cloud data points, then the number of the aforementioned second acquisition moments is multiple. Furthermore, since the second acquisition moment may precede or follow the first acquisition moment, for multiple second point cloud data points, it is possible that all second acquisition moments precede the first acquisition moment, or all second acquisition moments follow the first acquisition moment, or some second acquisition moments precede the first acquisition moment and some second acquisition moments follow the first acquisition moment.

[0126] In this scenario, for each second acquisition moment, a second position of the target object can be obtained at that second acquisition moment; therefore, there may be multiple second positions. Based on the first position and the multiple second positions, the target velocity of the target object at the target acquisition moment can be calculated.

[0127] Since the time difference between the target acquisition time and the second acquisition time is small, the motion of the target object between the target acquisition time and the second acquisition time can be approximated as uniform motion or uniformly accelerated motion. Based on the first position and multiple second positions, multiple velocities of the target object between the target acquisition time and multiple different second acquisition times can be calculated. If the calculated multiple velocities are the same, the motion of the target object can be considered as uniform motion, and in this case, the calculated velocity is determined as the target velocity of the target object at the target acquisition time. If the calculated multiple velocities are different, the acceleration of the target object can be further calculated based on the calculated multiple velocities, and then the target velocity of the target object at the target acquisition time can be calculated based on the acceleration of the target object and the aforementioned multiple different velocities.

[0128] Speed ​​or acceleration can be calculated using existing technologies, which will not be elaborated here.

[0129] Step S102D: Based on the driving speed of the driving object and the target speed at the target acquisition time, determine the relative speed of the target object relative to the driving object at the target acquisition time.

[0130] Specifically, the target speed can be subtracted from the speed of the moving object, and the result of the subtraction is the relative speed of the target object relative to the moving object.

[0131] In one embodiment of the present invention, the relative speed of the target object relative to the moving object at the target acquisition time can also be obtained based on multiple first point cloud data. For each first point cloud data, the target laser point corresponding to the target object in the first point cloud data can be determined. Then, based on the position information of the target laser point in the multiple first point cloud data, the target object is tracked to determine the motion trajectory of the target object. The target speed of the target object is then determined based on the motion trajectory of the target object, and thus the relative speed of the target object relative to the moving object at the first acquisition time is determined.

[0132] The aforementioned methods for achieving target tracking can be Kalman filtering, particle filtering, or other existing techniques for target tracking.

[0133] As can be seen from the above, in the solution provided by the embodiments of the present invention, by obtaining the first position of the target object at the target acquisition time and the second position at the second acquisition time, since the positional difference between the first position and the second position is the displacement of the target object between the target acquisition time and the second acquisition time, the target speed of the target object at the target acquisition time can be accurately calculated based on the first position and the second position. Furthermore, based on the driving speed of the driving object at the target acquisition time and the calculated target speed, the relative speed of the target object relative to the driving object at the target acquisition time can be accurately determined, thereby accurately adjusting the positional information of the target laser point in the target point cloud data.

[0134] In one embodiment of the present invention, step S102A can be implemented by the following step S102A1.

[0135] Step S102A1: Determine the first position of the target object based on the position information of the target laser point and the first position information of the moving object at the target acquisition time.

[0136] The first position information of the driving object can be used to characterize the absolute position of the driving object. Based on the first position information of the driving object, the position of the driving object in the driving environment in which the driving object is located can be determined.

[0137] As can be seen from step S103 above, the position information of the target laser point can be used to characterize the relative position of the target object with respect to the moving object. Based on the position information of the target laser point, the relative position of the target object with respect to the moving object can be determined.

[0138] In this case, by determining the relative position of the target object with respect to the moving object, and combining this with the absolute position of the moving object itself, the initial position of the target object can be determined.

[0139] Specifically, the geometric shape of the target object can be determined based on the position information of each target laser point, the geometric center of the geometric shape can be calculated, and the relative position of the target object relative to the moving object can be obtained based on the position information represented by the geometric center.

[0140] In one embodiment of the present invention, the first position information of the driving object can be obtained by measuring a positioning sensor installed in the driving object, or by measuring a positioning device installed in the driving environment where the driving object is located.

[0141] In one embodiment of the present invention, the position information of the target laser point can be represented in the form of three-dimensional coordinates in a spatial coordinate system. The origin of the three-dimensional coordinate system is located at the position of the driving object. The coordinate on each coordinate axis can be represented as the distance between the point on the target object corresponding to the target laser point and the driving object in the direction of the coordinate axis. Based on the three-dimensional coordinates of each target laser point, the average coordinates, which are the three-dimensional coordinates of the geometric center, can be calculated. The average coordinates can also be regarded as the three-dimensional coordinates of the target object in the spatial coordinate system. The position information reflected by the average coordinates is the relative position of the target object relative to the driving object.

[0142] For example, if the three-dimensional coordinates of target laser point a are determined to be (15, 48, 42), the three-dimensional coordinates of target laser point b are (36, 78, 3), and the three-dimensional coordinates of target laser point c are (6, 90, 33), then based on the three-dimensional coordinates of these three target laser points, the coordinates of the above geometric center on the three coordinate axes can be calculated as (15+36+6) / 3=19, (48+78+90) / 3=72, and (42+3+33) / 3=26. Therefore, the three-dimensional coordinates of the above geometric center are (19, 72, 26).

[0143] In another embodiment of the present invention, when the above-mentioned position information is represented in the form of coordinates, for each coordinate axis, the weighted average, maximum, minimum or median of the coordinates of each target laser point on the coordinate axis can also be determined as the coordinates of the target object on the coordinate axis, thereby obtaining the coordinates representing the relative position between the target object and the driving object.

[0144] In one embodiment of the present invention, step S102B can be implemented by the following step S102B1.

[0145] Step S102B1: Determine the first laser point corresponding to the target object in the second point cloud data, and determine the second position of the target object based on the position information of the first laser point and the second position information of the moving object at the second acquisition time.

[0146] The method for determining the first laser point in this step is similar to the method for determining the target laser point in step S101 above. The first laser point can be determined by feature extraction, point cloud clustering, or other methods, which will not be elaborated here.

[0147] The method for determining the second position of the target object in this step is similar to the method for determining the first position of the target object in step S102A1 above. First, the relative position of the target object with respect to the driving object is determined, and then the absolute position of the driving object itself is combined to determine the second position of the target object. This will not be elaborated further here.

[0148] In one embodiment of the present invention, the first position and the second position can be obtained by simultaneously applying steps S102A1 and S102B1 respectively; or the first position can be obtained by step S102A1 alone, and the second position can be obtained by other means other than step S102B1; or the second position can be obtained by step S102B1 alone, and the first position can be obtained by other means other than step S102A1.

[0149] As can be seen from the above, in the solution provided by the embodiments of the present invention, since the position information of the target laser point represents the relative position of the target object relative to the moving object at the target acquisition time, and the first position information of the moving object is used to represent the absolute position of the moving object, the first position of the target object can be accurately determined based on the position information of the target laser point and the first position information of the moving object at the target acquisition time. Similarly, the position information of the first laser point represents the relative position of the target object relative to the moving object at the second acquisition time, and the second position of the target object can be accurately determined based on the position information of the first laser point and the second position information of the moving object at the second acquisition time. Based on the accurate first and second positions, the accurate relative velocity of the target object relative to the moving object can be obtained, thereby enabling accurate adjustment of the position information of the target laser point and improving the accuracy of the point cloud data.

[0150] In one embodiment of the present invention, see Figure 4 It provides a flowchart of the fourth point cloud data processing method, which is consistent with the aforementioned Figure 1 Compared to the illustrated embodiment, in this embodiment, the above step S102 can be achieved through the following step S102E.

[0151] Step S102E: When it is determined that the target object is stationary, the speed that is the same in magnitude but opposite in direction to the speed of the moving object at the time of target acquisition is determined as the relative speed of the target object relative to the moving object.

[0152] The phrase "the target object is in a stationary state" means that the target object remains stationary and its own velocity is 0.

[0153] You can determine whether a target object is stationary in the following two ways.

[0154] Method 1: Since static point cloud data is point cloud data used to describe objects in a static state, when the first point cloud data mentioned above is static point cloud data, the target object can be considered to be in a static state.

[0155] Method 2: If the position difference between the first position and the second position is less than the preset position threshold, the target object can also be considered to be in a stationary state.

[0156] Specifically, the relative speed of the target object relative to the moving object is the target speed of the target object minus the moving object's speed. When the target object is stationary, its target speed is 0. Subtracting the moving object's speed from the target speed results in a speed that is the same magnitude but opposite in direction to the moving object's speed. Therefore, the speed that is the same magnitude but opposite in direction to the moving object's speed at the time of target acquisition is determined as the relative speed of the target object relative to the moving object.

[0157] For example, if the speed of the moving object at the time of target acquisition is 60 km / h and the direction of this speed is due east, then when the target object is stationary, the relative speed of the target object relative to the moving object is also 60 km / h and the direction of this relative speed is due west.

[0158] As can be seen from the above, in the solution provided by the embodiments of the present invention, when the target object is stationary, the target speed of the target object is 0, and the relative speed of the target object relative to the moving object is the speed with the same magnitude but opposite direction to the moving object's speed. Therefore, the speed with the same magnitude but opposite direction to the moving object's speed at the time of target acquisition can be directly determined as the relative speed of the target object relative to the moving object. Applying the point cloud data processing solution provided by the embodiments of the present invention can improve the efficiency of point cloud data processing.

[0159] In one embodiment of the present invention, see Figure 5It provides a flowchart of the fifth point cloud data processing method, which is consistent with the aforementioned Figure 1 Compared to the illustrated embodiment, in this embodiment, before determining the target laser point of the corresponding target object in the first point cloud data to be processed in step S101, the above method further includes the following step S104.

[0160] Step S104: Remove the laser points corresponding to the preset objects from the first point cloud data.

[0161] The aforementioned preset objects can be buildings, trees, hills, or ground in the driving environment where the driving object is located.

[0162] Specifically, the laser points corresponding to the preset objects can be determined first in the first point cloud data, and then the determined laser points can be removed from the first point cloud data.

[0163] In one embodiment of the present invention, feature extraction can be performed on the first point cloud data based on the features of a preset object, and laser points in the first point cloud data that match the object features of the preset object can be determined based on the extracted features. Then, the determined laser points are removed from the first point cloud data.

[0164] In one embodiment of the present invention, when the preset object is the ground in the driving environment where the driving object is located, the laser point corresponding to the ground in the first point cloud data can be referred to as a ground point.

[0165] Specifically, since the ground is often relatively flat in real-world environments, the vertical distance between the lidar and the ground in the direction perpendicular to the ground is usually constant. This vertical distance can be used to determine the ground points in the first point cloud data, thereby removing the determined ground points from the first point cloud data.

[0166] For example, when the location information of the laser point is represented in the form of three-dimensional coordinates in a three-dimensional spatial coordinate system, the direction of one coordinate axis Z of the three-dimensional spatial coordinate system can be a direction perpendicular to the ground and upward. In this case, the height of the laser radar installed on the moving object above the ground can be measured in advance, and then the three-dimensional coordinates of the laser point with the same Z-axis coordinate as the above height can be found. The laser point is determined as the ground point, and the determined ground point is removed from the first point cloud data.

[0167] In addition, ground point filtering methods in existing technologies can be used to remove ground points from the first point cloud data, which will not be described in detail here.

[0168] As can be seen from the above, removing the laser points corresponding to the preset objects from the first point cloud data can reduce the number of laser points contained in the first point cloud data, thereby reducing the workload when processing the first point cloud data subsequently. Therefore, applying the solution provided by the embodiments of the present invention can improve the efficiency of point cloud data processing.

[0169] In one embodiment of the present invention, see Figure 6 It provides a flowchart of the sixth point cloud data processing method, which is consistent with the aforementioned Figure 1 Compared to the illustrated embodiment, in this embodiment, the above step S101 can be achieved through the following step S101A.

[0170] Step S101A: Perform point cloud clustering processing on the first point cloud data to be processed, and determine the laser points belonging to the same cluster as the target laser points of the corresponding target object.

[0171] The purpose of point cloud clustering is to identify laser points corresponding to the same object in point cloud data as belonging to the same cluster, and to ensure that laser points corresponding to different objects belong to different clusters. There are various ways to implement point cloud clustering, such as Euclidean clustering, density clustering, and hypervolume clustering in existing technologies, which will not be detailed here.

[0172] By performing point cloud clustering on the target point cloud data, laser points belonging to the same cluster can be identified. These laser points can then be designated as target laser points for the corresponding target objects. Operations such as calculating relative velocity and adjusting position information can then be performed based on the identified target laser points without further identifying the target objects. Therefore, as long as laser points belonging to the same cluster are obtained, they can be identified as target laser points for the corresponding target objects.

[0173] As can be seen from the above, in the solution provided by the embodiments of the present invention, point cloud clustering processing of the target point cloud data can identify the laser points belonging to the same cluster as the target laser points of the corresponding target object. Since the laser points reflected by the target object and received by the lidar are usually quite close together when collecting point cloud data of the target object, meaning the distance between each target laser point is small, and point cloud clustering processing can determine laser points belonging to the same cluster based on the distance between the laser points, identifying the laser points belonging to the same cluster obtained from point cloud clustering processing of the target point cloud data as the target laser points of the corresponding target object can accurately determine the target laser points of the corresponding target object.

[0174] Corresponding to the above-described point cloud data processing method, this embodiment of the invention also provides a point cloud data processing device.

[0175] See Figure 7A schematic diagram of the structure of a first point cloud data processing device is provided. The device includes:

[0176] The target determination module 701 is used to determine the target laser point of the corresponding target object in the first point cloud data to be processed, wherein the first point cloud data is the point cloud data collected by the lidar installed on the moving object;

[0177] The speed acquisition module 702 is used to acquire the relative speed of the target object relative to the moving object at the target acquisition time corresponding to the target laser point;

[0178] The position adjustment module 703 is used to adjust the position information of the target laser point in the opposite direction to the direction of the relative velocity based on the acquisition time difference and the relative velocity, wherein the acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data, and the first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data.

[0179] As can be seen from the above, in the solution provided by the embodiments of the present invention, both the target acquisition time and the first acquisition time are the acquisition times of a portion of the laser points in the first point cloud data acquired by the lidar. Since the time required for the lidar to acquire one frame of the first point cloud data is often short, the acquisition time difference between the target acquisition time and the first acquisition time is often small. The aforementioned relative speed is the speed of the target object relative to the moving object at the target acquisition time. Since the aforementioned acquisition time difference of the first point cloud data is usually small, the speed of the target object relative to the moving object between the target acquisition time and the first acquisition time can be approximated as the aforementioned relative speed. Therefore, based on the acquisition time difference and the relative speed, the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time can be obtained. Adjusting the position information of the target laser points in the direction opposite to the direction of the relative speed according to the aforementioned relative displacement can offset the offset of the target laser points in the first point cloud data caused by the relative displacement between the target object and the moving object, so that the laser points in the adjusted first point cloud data are closer to the laser points that the lidar can acquire at the first acquisition time. Therefore, the solution provided by the embodiments of the present invention can improve the accuracy of point cloud data.

[0180] In one embodiment of the present invention, the position adjustment module 703 is specifically used for:

[0181] The relative velocity is decomposed into longitudinal velocity and lateral velocity, wherein the direction of the longitudinal velocity is the same as the direction of travel of the moving object, and the direction of the lateral velocity is perpendicular to the direction of travel of the moving object;

[0182] Based on the acquisition time difference and the longitudinal velocity, a longitudinal compensation amount is obtained, and the position information of the target laser point is adjusted based on the longitudinal compensation amount in the direction opposite to the direction of the longitudinal velocity.

[0183] Based on the acquisition time difference and the lateral velocity, a lateral compensation amount is obtained, and the position information of the target laser point is adjusted based on the lateral compensation amount in the direction opposite to the direction of the lateral velocity.

[0184] As can be seen from the above, in the solution provided by the embodiments of the present invention, the relative velocity is decomposed into longitudinal velocity and lateral velocity, and the position information of the target laser point is adjusted twice in the directions opposite to the longitudinal velocity and the directions opposite to the lateral velocity, respectively. This offsets the offset of the target laser point in the first point cloud data caused by the relative displacement between the target object and the moving object in the two directions, thereby improving the accuracy of the point cloud data.

[0185] In one embodiment of the present invention, see Figure 8 A schematic diagram of the structure of a second type of point cloud data processing device is provided, which is similar to the aforementioned Figure 7 Compared to the illustrated embodiment, in this embodiment, the speed acquisition module 702 includes:

[0186] The first position acquisition submodule 702A is used to acquire the first position of the target object at the target acquisition time.

[0187] The second position acquisition submodule 702B is used to obtain the second position of the target object at the second acquisition time of the second point cloud data, wherein the time difference between the second acquisition time corresponding to the second preset laser point in the second point cloud data and the first acquisition time is less than the preset time difference.

[0188] The target velocity calculation submodule 702C is used to calculate the target velocity of the target object at the target acquisition time based on the first position and the second position;

[0189] The relative speed determination submodule 702D is used to determine the relative speed of the target object relative to the driving object at the target acquisition time based on the driving speed of the driving object at the target acquisition time and the target speed.

[0190] As can be seen from the above, in the solution provided by the embodiments of the present invention, by obtaining the first position of the target object at the target acquisition time and the second position at the second acquisition time, since the positional difference between the first position and the second position is the displacement of the target object between the target acquisition time and the second acquisition time, the target speed of the target object at the target acquisition time can be accurately calculated based on the first position and the second position. Furthermore, based on the driving speed of the driving object at the target acquisition time and the calculated target speed, the relative speed of the target object relative to the driving object at the target acquisition time can be accurately determined, thereby accurately adjusting the positional information of the target laser point in the target point cloud data.

[0191] In one embodiment of the present invention, the first position obtaining submodule 702A is specifically used for:

[0192] Based on the position information of the target laser point and the first position information of the moving object at the target acquisition time, the first position of the target object is determined.

[0193] In one embodiment of the present invention, the second position obtaining submodule 702B is specifically used for:

[0194] The first laser point corresponding to the target object in the second point cloud data is determined, and the second position of the target object is determined based on the position information of the first laser point and the second position information of the moving object at the second acquisition time.

[0195] In one implementation,

[0196] The first position acquisition submodule 702A is specifically used to: determine the first position of the target object based on the position information of the target laser point and the first position information of the moving object at the target acquisition time;

[0197] The second position acquisition submodule 702B is specifically used to: determine the first laser point corresponding to the target object in the second point cloud data, and determine the second position of the target object based on the position information of the first laser point and the second position information of the driving object at the second acquisition time.

[0198] As can be seen from the above, in the solution provided by the embodiments of the present invention, since the position information of the target laser point represents the relative position of the target object relative to the moving object at the target acquisition time, and the first position information of the moving object is used to represent the absolute position of the moving object, the first position of the target object can be accurately determined based on the position information of the target laser point and the first position information of the moving object at the target acquisition time. Similarly, the position information of the first laser point represents the relative position of the target object relative to the moving object at the second acquisition time, and the second position of the target object can be accurately determined based on the position information of the first laser point and the second position information of the moving object at the second acquisition time. Based on the accurate first and second positions, the accurate relative velocity of the target object relative to the moving object can be obtained, thereby enabling accurate adjustment of the position information of the target laser point and improving the accuracy of the point cloud data.

[0199] In one embodiment of the present invention, the speed acquisition module 702 is specifically used for:

[0200] If the target object is determined to be stationary, the speed of the target object relative to the moving object is determined to be the speed that is the same in magnitude but opposite in direction as the speed of the moving object at the time of target acquisition.

[0201] As can be seen from the above, in the solution provided by the embodiments of the present invention, when the target object is stationary, the target speed of the target object is 0, and the relative speed of the target object relative to the moving object is the speed with the same magnitude but opposite direction to the moving object's speed. Therefore, the speed with the same magnitude but opposite direction to the moving object's speed at the time of target acquisition can be directly determined as the relative speed of the target object relative to the moving object. Applying the point cloud data processing solution provided by the embodiments of the present invention can improve the efficiency of point cloud data processing.

[0202] In one embodiment of the present invention, see Figure 9 A structural schematic diagram of a third type of point cloud data processing device is provided, which is similar to the aforementioned Figure 7 Compared to the illustrated embodiment, in this embodiment, the device further includes:

[0203] The laser point removal module 704 is used to remove the laser points corresponding to the preset object in the first point cloud data before determining the target laser points of the corresponding target object in the first point cloud data to be processed.

[0204] As can be seen from the above, removing the laser points corresponding to the preset objects from the first point cloud data can reduce the number of laser points contained in the first point cloud data, thereby reducing the workload when processing the first point cloud data subsequently. Therefore, applying the solution provided by the embodiments of the present invention can improve the efficiency of point cloud data processing.

[0205] In one embodiment of the present invention, the target determination module 701 is specifically used for:

[0206] The first point cloud data to be processed is subjected to point cloud clustering processing, and the laser points belonging to the same cluster are identified as the target laser points of the corresponding target objects.

[0207] As can be seen from the above, in the solution provided by the embodiments of the present invention, point cloud clustering processing of the target point cloud data can identify the laser points belonging to the same cluster as the target laser points of the corresponding target object. Since the laser points reflected by the target object and received by the lidar are usually quite close together when collecting point cloud data of the target object, meaning the distance between each target laser point is small, and point cloud clustering processing can determine laser points belonging to the same cluster based on the distance between the laser points, identifying the laser points belonging to the same cluster obtained from point cloud clustering processing of the target point cloud data as the target laser points of the corresponding target object can accurately determine the target laser points of the corresponding target object.

[0208] This invention also provides an electronic device, such as... Figure 10 As shown, it includes a processor 1001, a communication interface 1002, a memory 1003, and a communication bus 1004, wherein the processor 1001, the communication interface 1002, and the memory 1003 communicate with each other through the communication bus 1004.

[0209] Memory 1003 is used to store computer programs;

[0210] When processor 1001 executes a program stored in memory 1003, it performs the following steps:

[0211] In the first point cloud data to be processed, the target laser point corresponding to the target object is determined, wherein the first point cloud data is the point cloud data collected by the lidar installed on the moving object;

[0212] The relative velocity of the target object to the moving object is obtained at the target acquisition time corresponding to the target laser point;

[0213] Based on the acquisition time difference and the relative velocity, the position information of the target laser point is adjusted in the direction opposite to the direction of the relative velocity. The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data.

[0214] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0215] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0216] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0217] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0218] In another embodiment of the present invention, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the point cloud data processing methods described above.

[0219] In another embodiment of the present invention, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the point cloud data processing methods described above.

[0220] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0221] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0222] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, electronic devices, computer-readable storage media, and computer program products are basically similar to the method embodiments, and therefore the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0223] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. A point cloud data processing method, characterized in that, The method includes: In the first point cloud data to be processed, the target laser points corresponding to the target objects are determined. The first point cloud data is the point cloud data collected by the lidar installed on the driving object. The number of target laser points corresponding to one target object is multiple. The relative velocity of the target object relative to the moving object is obtained at the target acquisition time corresponding to the target laser point. The target acquisition time is different for different laser points in the target laser points. The relative velocity is: the relative velocity of the target object relative to the moving object at the target acquisition time of any laser point in the target laser points. Multiplying the acquisition time difference by the relative velocity yields the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time. When the position information is represented in three-dimensional coordinates in a spatial coordinate system with the moving object's location as the origin, subtracting the coordinates of the target laser point and the relative displacement along the same coordinate axis in their three-dimensional coordinates allows for adjustment of the target laser point's position information in the direction opposite to the relative velocity. The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data, where the first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data. A small acquisition time difference between the target acquisition time and the first acquisition time allows for the adjustment of the target laser point's position information in the direction opposite to the relative velocity based on the relative displacement. This offset of the target laser point in the first point cloud data caused by the relative displacement between the target object and the moving object can be counteracted, making the laser point in the adjusted first point cloud data closer to the laser point that the lidar could acquire at the first acquisition time.

2. The method according to claim 1, characterized in that, The process of obtaining the relative velocity of the target object with respect to the moving object at the target acquisition time corresponding to the target laser point includes: Obtain the first position of the target object at the target acquisition time; The second position of the target object is obtained at the second acquisition time of the second point cloud data, wherein the time difference between the second acquisition time corresponding to the second preset laser point in the second point cloud data and the first acquisition time is less than the preset time difference. Based on the first position and the second position, calculate the target velocity of the target object at the target acquisition time; Based on the driving speed of the driving object at the target acquisition time and the target speed, the relative speed of the target object with respect to the driving object at the target acquisition time is determined.

3. The method according to claim 2, characterized in that, Obtaining the first position of the target object at the target acquisition time includes: Based on the position information of the target laser point and the first position information of the moving object at the target acquisition time, the first position of the target object is determined; and / or Obtaining the second position of the target object at the second acquisition time of the second point cloud data includes: The first laser point corresponding to the target object in the second point cloud data is determined, and the second position of the target object is determined based on the position information of the first laser point and the second position information of the moving object at the second acquisition time.

4. The method according to claim 1, characterized in that, The process of obtaining the relative velocity of the target object with respect to the moving object at the target acquisition time corresponding to the target laser point includes: If the target object is determined to be stationary, the speed of the target object relative to the moving object is determined to be the speed that is the same in magnitude but opposite in direction as the speed of the moving object at the time of target acquisition.

5. The method according to claim 1, characterized in that, Before determining the target laser point corresponding to the target object in the first point cloud data to be processed, the method further includes: Remove the laser points corresponding to the preset objects from the first point cloud data.

6. The method according to claim 1, characterized in that, The step of determining the target laser point corresponding to the target object in the first point cloud data to be processed includes: The first point cloud data to be processed is subjected to point cloud clustering processing, and the laser points belonging to the same cluster are identified as the target laser points of the corresponding target objects.

7. A point cloud data processing device, characterized in that, The device includes: The target determination module is used to determine the target laser points of the corresponding target object in the first point cloud data to be processed, wherein the first point cloud data is the point cloud data collected by the lidar installed on the moving object, and the number of target laser points corresponding to one target object is multiple; The speed acquisition module is used to obtain the relative speed of the target object relative to the moving object at the target acquisition time corresponding to the target laser point. The target acquisition time is different for different laser points in the target laser points. The relative speed is: the relative speed of the target object relative to the moving object at the target acquisition time of any laser point in the target laser points. The position adjustment module is used to multiply the acquisition time difference by the relative velocity to obtain the relative displacement of the target object relative to the moving object between the target acquisition time and the first acquisition time. When the position information is represented in three-dimensional coordinates in a spatial coordinate system with the moving object's location as the origin, the module subtracts the coordinates of the target laser point and the relative displacement from the coordinates of the same axis in the three-dimensional coordinates. This adjusts the position information of the target laser point in the direction opposite to the relative velocity. The acquisition time difference is the time difference between the target acquisition time and the first acquisition time of the first point cloud data. The first acquisition time is the acquisition time corresponding to the first preset laser point in the first point cloud data. The acquisition time difference between the target acquisition time and the first acquisition time is small. Adjusting the position information of the target laser point in the direction opposite to the relative velocity based on the relative displacement can offset the offset of the target laser point in the first point cloud data caused by the relative displacement between the target object and the moving object, making the laser point in the adjusted first point cloud data closer to the laser point that the lidar could acquire at the first acquisition time.

8. The apparatus according to claim 7, characterized in that, The speed acquisition module includes: The first position acquisition submodule is used to obtain the first position of the target object at the target acquisition time. The second position acquisition submodule is used to obtain the second position of the target object at the second acquisition time of the second point cloud data, wherein the time difference between the second acquisition time corresponding to the second preset laser point in the second point cloud data and the first acquisition time is less than the preset time difference. The target velocity calculation submodule is used to calculate the target velocity of the target object at the target acquisition time based on the first position and the second position; The relative speed determination submodule is used to determine the relative speed of the target object relative to the driving object at the target acquisition time based on the driving speed of the driving object at the target acquisition time and the target speed.

9. The apparatus according to claim 8, characterized in that, The first position acquisition submodule is specifically used for: Based on the position information of the target laser point and the first position information of the moving object at the target acquisition time, the first position of the target object is determined; and / or The second position obtains the sub-module, specifically used for: The first laser point corresponding to the target object in the second point cloud data is determined, and the second position of the target object is determined based on the position information of the first laser point and the second position information of the moving object at the second acquisition time.

10. The apparatus according to claim 7, characterized in that, The speed acquisition module is specifically used for: If the target object is determined to be stationary, the speed of the target object relative to the moving object is determined to be the speed that is the same in magnitude but opposite in direction as the speed of the moving object at the time of target acquisition.

11. The apparatus according to claim 7, characterized in that, The device further includes: The laser point removal module is used to remove laser points corresponding to a preset object in the first point cloud data before determining the target laser points of the corresponding target object in the first point cloud data to be processed.

12. The apparatus according to claim 7, characterized in that, The target determination module is specifically used for: The first point cloud data to be processed is subjected to point cloud clustering processing, and the laser points belonging to the same cluster are identified as the target laser points of the corresponding target objects.

13. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-6.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.

Images