Method for recognizing a moving object and related apparatus

Abstract

The application discloses a method for identifying a moving target and a related device. The method comprises the following steps: merging first point cloud data and second point cloud data to obtain third point cloud data; the second point cloud data is collected at a next time of the first point cloud data; clustering the third point cloud data to obtain clustered third point cloud data; splitting the clustered third point cloud data to obtain fourth point cloud data and fifth point cloud data; and determining a moving target in the point cloud data based on the fourth point cloud data and the fifth point cloud data. In the foregoing manner, the clustering effect and the splitting accuracy can be improved, and the accuracy of determining the moving target is improved, so as to facilitate subsequent processing based on the moving target.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of point cloud processing, in particular to a motion target recognition method and related device. BACKGROUND

[0002] Point cloud data is usually obtained by a radar sensor, and a motion target can be determined by corresponding point cloud data. For example, in the field of unmanned driving, the main task of a perception system is to perceive the position, speed, category, behavior prediction and other information of a target by collecting point cloud data using a radar sensor, among which the most important is the detection and tracking of a motion target. If the perception of the motion target is wrong, such as a static target A being recognized as dynamic, the subsequent trajectory planning module of the unmanned driving may plan a trajectory to the position of A, thereby causing a traffic accident. Similarly, if a dynamic target is recognized as static, it may cause an emergency brake, thereby affecting the driving experience. SUMMARY

[0003] To solve the above problems, the present application provides a motion target recognition method and related device, which can improve the clustering effect and the accuracy of splitting, and further improve the accuracy of determining the motion target, so as to facilitate subsequent processing based on the motion target.

[0004] To solve the above technical problems, one technical solution adopted by the present application is to provide a motion target recognition method, which comprises: merging first point cloud data and second point cloud data to obtain third point cloud data; the second point cloud data is collected at the next time of the first point cloud data; clustering the third point cloud data to obtain clustered third point cloud data; splitting the clustered third point cloud data to obtain fourth point cloud data and fifth point cloud data; and determining a motion target in the point cloud data based on the fourth point cloud data and the fifth point cloud data.

[0005] Among them, the third point cloud data is clustered to obtain the clustered third point cloud data, which comprises: clustering the third point cloud data to obtain at least one clustered point cloud; each point cloud cluster corresponds to a target object; and the clustered third point cloud data is split to obtain the fourth point cloud data and the fifth point cloud data, which comprises: splitting each point cloud cluster to obtain a first point cloud cluster corresponding to the first point cloud data and a second point cloud cluster corresponding to the second point cloud data.

[0006] Among them, each point cloud cluster is split to obtain the first point cloud cluster corresponding to the first point cloud data and the second point cloud cluster corresponding to the second point cloud data, which comprises: obtaining the time sequence number of the data points in each point cloud cluster; and obtaining the first point cloud cluster and the second point cloud cluster based on the point cloud with the same time sequence number.

[0007] The determining of the moving target in the point cloud data based on the fourth point cloud data and the fifth point cloud data comprises: determining whether the corresponding target object is a moving target based on the first point cloud cluster and the second point cloud cluster.

[0008] The determining of the moving target based on the first point cloud cluster and the second point cloud cluster comprises: determining the coincidence of the first point cloud cluster and the second point cloud cluster; and determining whether the corresponding target object is a moving target based on the coincidence.

[0009] Before the clustering of the third point cloud data to obtain the clustered third point cloud data, the method further comprises: performing a filtering operation on the third point cloud data to remove ground plane points in the third point cloud data; and the clustering of the third point cloud data to obtain the clustered third point cloud data comprises: clustering the third point cloud data after the removal to obtain the clustered third point cloud data.

[0010] The merging of the first point cloud data and the second point cloud data to obtain the third point cloud data comprises: projecting the first point cloud data into the second point cloud data to obtain the third point cloud data.

[0011] The projecting of the first point cloud data into the second point cloud data to obtain the third point cloud data comprises: obtaining first position information of an unmanned vehicle corresponding to the first point cloud data and second position information of the unmanned vehicle corresponding to the second point cloud data; performing coordinate system conversion on the first point cloud data by using the second position information and the first position information; and projecting the first point cloud data after the coordinate system conversion into the second point cloud data to obtain the third point cloud data.

[0012] To solve the above technical problems, another technical solution adopted by the present application is to provide a vehicle-mounted control system, which comprises: a radar sensor configured to collect first point cloud data and second point cloud data; and a processor connected to the radar sensor and configured to implement the method provided in the above technical solution.

[0013] To solve the above technical problems, another technical solution adopted by the present application is to provide an unmanned vehicle, which comprises the vehicle-mounted control system provided in the above technical solution.

[0014] To solve the above technical problems, another technical solution adopted by the present application is to provide a computer-readable storage medium configured to store a computer program, wherein the computer program, when executed by a processor, is configured to implement the method provided in the above technical solution.

[0015] The beneficial effects of the embodiments of the present application are: different from the prior art, the motion target recognition method provided by the present application, the method comprises: merging first point cloud data and second point cloud data to obtain third point cloud data; the second point cloud data is collected at the next moment of the first point cloud data; clustering the third point cloud data to obtain the clustered third point cloud data; split the clustered third point cloud data to obtain fourth point cloud data and fifth point cloud data; determine the motion target in the point cloud data based on the fourth point cloud data and the fifth point cloud data. Through the above-mentioned mode, on the one hand, the adjacent time point cloud data is merged, and the merged data is clustered, the simultaneous clustering of the two time point cloud data is completed, because the merged point cloud data is more, the clustering effect can be improved, on the other hand, the clustered third point cloud data is split, the accuracy of the split can be improved, and the accuracy of determining the motion target is improved, so as to facilitate the subsequent processing based on the motion target. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Among them:

[0017] Figure 1 is a flowchart of the first embodiment of the motion target recognition method provided by the present application;

[0018] Figure 2 is a flowchart of the second embodiment of the motion target recognition method provided by the present application;

[0019] Figure 3 is a flowchart of an embodiment of step 23 provided by the present application;

[0020] Figure 4 is a flowchart of an embodiment of step 24 provided by the present application;

[0021] Figure 5 is a flowchart of the third embodiment of the motion target recognition method provided by the present application;

[0022] Figure 6 is a flowchart of an embodiment of step 51 provided by the present application;

[0023] Figure 7 is a structural schematic diagram of an embodiment of the vehicle-mounted control system provided by the present application;

[0024] Figure 8 is a structural schematic diagram of an embodiment of the unmanned vehicle provided by the present application;

[0025] Figure 9 FIG. 1 is a structural schematic diagram of an embodiment of the computer readable storage medium provided in the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application. It can be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings, but not all the structures. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0027] Reference herein to “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearance of the phrase in various places in the specification is not necessarily all referring to the same embodiment, nor is it necessarily referring to a particular alternative embodiment. It will be explicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0028] Reference is made to Figure 1 , Figure 1 FIG. 1 is a flow schematic diagram of a first embodiment of the method for identifying a moving target provided in the present application. The method comprises:

[0029] Step 11: merging the first point cloud data and the second point cloud data to obtain third point cloud data; the second point cloud data is collected at a next time instant of the first point cloud data.

[0030] In some embodiments, the point cloud data is collected by using a radar sensor. The radar sensor can be a laser radar. For example, a mechanical laser radar or a solid-state laser radar. Therefore, the radar sensor can collect corresponding point cloud data at different time instants. For example, the second point cloud data is collected at t-1 time instant, and the first point cloud data is collected at t time instant.

[0031] When the collection frequency of the radar sensor is high, the interval between each time instant is small, and then the point cloud data collected at adjacent time instants can be used to determine whether the state of the target object corresponding to the point cloud data is a moving state. The target object belonging to the moving state is determined as a moving target.

[0032] In step 11, the way of merging the first point cloud data and the second point cloud data can be projecting the first point cloud data onto the second point cloud data to obtain the third point cloud data. Or projecting the second point cloud data onto the first point cloud data to obtain the third point cloud data.

[0033] In some embodiments, if the radar sensor is in a fixed position and does not produce displacement in the time interval, the manner of merging the first point cloud data and the second point cloud data can be to project the first point cloud data onto the second point cloud data, or to project the second point cloud data onto the first point cloud data, to obtain third point cloud data.

[0034] In some embodiments, if the radar sensor produces displacement in the time interval, the coordinate information before and after displacement needs to be determined, and the first point cloud data and the second point cloud data are merged based on the coordinate information before and after displacement. For example, the radar sensor is arranged on an unmanned vehicle, and the vehicle collects point cloud data by using the radar sensor during movement.

[0035] Step 12: clustering the third point cloud data to obtain clustered third point cloud data.

[0036] When collecting point cloud data by using a radar sensor, all target objects within the collection range of the radar sensor will have corresponding data points. Since some data points correspond to the same target object, the point cloud data can be clustered to divide the point cloud data into data points corresponding to each target object.

[0037] Among them, the manner of clustering the third point cloud data can adopt any method in K-Means clustering, mean shift clustering, density-based clustering method, Gaussian mixture model (GMM) expectation maximization (EM) clustering, agglomerative hierarchical clustering, graph community detection, connected component clustering method, DBSCAN algorithm. The data points in the third point cloud data are clustered into different point cloud clusters. The purpose is to cluster the data points belonging to the same target object together.

[0038] Step 13: splitting the clustered third point cloud data to obtain fourth point cloud data and fifth point cloud data.

[0039] Among them, the clustered third point cloud data clusters the data points originally belonging to the first point cloud data and the data points originally belonging to the second point cloud data together according to the same characteristics. Therefore, these clustered data points need to be split to distinguish the fourth point cloud data corresponding to the first point cloud data and the fifth point cloud data corresponding to the second point cloud data.

[0040] Since the clustered third point cloud data is split, the fourth point cloud data and the fifth point cloud data after splitting also have point cloud clusters of the corresponding time after clustering.

[0041] Step 14: determining a moving target in the point cloud data based on the fourth point cloud data and the fifth point cloud data.

[0042] Because the splitting into the fourth point cloud data and the fifth point cloud data. That is, each clustered target object is split into two point cloud clusters. One point cloud cluster corresponds to the fourth point cloud data, and one point cloud cluster corresponds to the fifth point cloud data. The two point cloud clusters are at respective positions in the world coordinate system, and whether the target object is moving can be determined by the positional relationship of the point cloud clusters. For example, if the target object is stationary, the two point cloud clusters have a very high degree of coincidence. Conversely, if the target object is moving, there is some interval between the two point cloud clusters.

[0043] Therefore, whether the clustered target object is a moving target can be distinguished in this way.

[0044] In an application scenario, the embodiment is applied to an unmanned vehicle, and the unmanned vehicle is provided with a corresponding radar sensor. After the control system of the unmanned vehicle determines the moving target in the above manner, the driving route and driving speed of the vehicle can be planned according to the moving target.

[0045] In the embodiment, the first point cloud data and the second point cloud data are merged to obtain third point cloud data. The second point cloud data is collected at a next time instant of the first point cloud data. The third point cloud data is clustered to obtain clustered third point cloud data. The clustered third point cloud data is split to obtain fourth point cloud data and fifth point cloud data. The moving target in the point cloud data is determined based on the fourth point cloud data and the fifth point cloud data. On the one hand, the adjacent time instant point cloud data is merged, and the merged data is clustered to complete the simultaneous clustering of the two time instant point cloud data. Because the merged point cloud data is more, the clustering effect can be improved. On the other hand, the clustered third point cloud data is split to improve the accuracy of the splitting, and then the accuracy of determining the moving target is improved, so as to facilitate subsequent processing based on the moving target.

[0046] Referring to Figure 2 , Figure 2 is a flowchart of a second embodiment of the method for identifying a moving target provided by the present application. The method comprises the following steps:

[0047] Step 21: merging the first point cloud data and the second point cloud data to obtain third point cloud data. The second point cloud data is collected at a next time instant of the first point cloud data.

[0048] Step 22: clustering the third point cloud data to obtain at least one clustered point cloud cluster. Each point cloud cluster corresponds to a target object.

[0049] Step 21 and step 22 have the same or similar technical solutions as the above embodiments, and will not be described here.

[0050] In some embodiments, the radar sensor is arranged on an unmanned vehicle, and the radar sensor collects corresponding data points to form point cloud data during the driving of the vehicle. For example, the corresponding target objects in the point cloud data can include trees, lanes, other vehicles, and buildings. Therefore, during clustering, the data points are aggregated according to the characteristics of the target objects to form corresponding point cloud clusters. That is, each target object can be represented by a point cloud cluster.

[0051] Step 23: splitting each point cloud cluster to obtain a first point cloud cluster corresponding to the first point cloud data and a second point cloud cluster corresponding to the second point cloud data.

[0052] Therefore, the point cloud cluster after clustering includes data points at two time points, and the point cloud cluster needs to be split to form point cloud clusters at different time points to facilitate the determination of the moving target.

[0053] In some embodiments, referring to Figure 3 , step 23 can be the following process:

[0054] Step 231: obtaining the time sequence number of the data points in each point cloud cluster.

[0055] During the collection of point cloud data, each data point in the point cloud data has a corresponding time sequence number. For example, the second point cloud data is collected at time t-1, and the time sequence number of the data points in the second point cloud data can be set to t-1. The first point cloud data is collected at time t, and the time sequence number of the data points in the first point cloud data can be set to t.

[0056] Thus, the time sequence number of the data points in each point cloud cluster after clustering can be obtained.

[0057] Step 232: obtaining the first point cloud cluster and the second point cloud cluster based on the point cloud with the same time sequence number.

[0058] The data points with the same time sequence number are determined as a class, and these data points can still represent the target object at the time sequence number after clustering.

[0059] For example, each point cloud cluster after clustering is classified according to the time sequence number t-1 and the time sequence number t to obtain the first point cloud cluster corresponding to the first point cloud data and the second point cloud cluster corresponding to the second point cloud data. For example, there are point cloud cluster A, point cloud cluster B, point cloud cluster C, and point cloud cluster D after clustering. After obtaining the time sequence number of the data points in point cloud cluster A, point cloud cluster B, point cloud cluster C, and point cloud cluster D, point cloud cluster A is split into point cloud cluster A1 and point cloud cluster A2, point cloud cluster B is split into point cloud cluster B1 and point cloud cluster B2, point cloud cluster C is split into point cloud cluster C1 and point cloud cluster C2, and point cloud cluster D is split into point cloud cluster D1 and point cloud cluster D2.

[0060] The point cloud cluster A1, the point cloud cluster B1, the point cloud cluster C1 and the point cloud cluster D1 correspond to the first point cloud data. The point cloud cluster A2, the point cloud cluster B2, the point cloud cluster C2 and the point cloud cluster D2 correspond to the second point cloud data.

[0061] Step 24: determining whether the target object corresponding to the first point cloud cluster and the second point cloud cluster is a moving target based on the first point cloud cluster and the second point cloud cluster.

[0062] In some embodiments, an ICP (Iterative Closest Point) algorithm can be used to determine whether the target object corresponding to the first point cloud cluster and the second point cloud cluster is a moving target.

[0063] In some embodiments, a Chamfer Distance algorithm can be used to determine whether the target object corresponding to the first point cloud cluster and the second point cloud cluster is a moving target. In some embodiments, referring to Figure 4 , step 24 can be the following flow:

[0064] Step 241: determining the coincidence of the first point cloud cluster and the second point cloud cluster.

[0065] For example, the projection direction corresponding to the first point cloud cluster and the second point cloud cluster is determined, the first point cloud cluster and the second point cloud cluster are projected onto the projection direction to obtain corresponding vectors. The coincidence between the vectors can be determined. For example, the intersection-over-union of the vectors is determined.

[0066] Step 242: determining whether the target object corresponding to the first point cloud cluster and the second point cloud cluster is a moving target based on the coincidence.

[0067] A coincidence threshold is set. When the coincidence is greater than the threshold, it is determined that the target object corresponding to the first point cloud cluster and the second point cloud cluster is not a moving target. When the coincidence is less than or equal to the threshold, it is determined that the target object corresponding to the first point cloud cluster and the second point cloud cluster is a moving target.

[0068] Taking the intersection-over-union as an example, the greater the intersection-over-union, the higher the coincidence degree, which means that the probability of the target object not moving is greater. The smaller the intersection-over-union, the lower the coincidence degree, which means that the probability of the target object moving is greater. Therefore, an intersection-over-union threshold can be set. When the intersection-over-union obtained by determining the coincidence is greater than the intersection-over-union threshold, it is determined that the target object is a fixed object and has not moved. When the intersection-over-union obtained by determining the coincidence is less than or equal to the intersection-over-union threshold, it is determined that the target object is a moving object.

[0069] In the embodiment, the first point cloud cluster and the second point cloud cluster are split from the clustered point cloud data, and the determination of the motion target is performed on the first point cloud cluster and the second point cloud cluster. The accuracy of the split is ensured, and the accuracy of the determination of the motion target is improved, so as to facilitate subsequent processing based on the motion target.

[0070] Referring to Figure 5 , Figure 5 is a flowchart of a third embodiment of the motion target recognition method provided in the application. The method comprises the following steps:

[0071] Step 51: merging the first point cloud data and the second point cloud data to obtain third point cloud data; the second point cloud data is collected at the next time of the first point cloud data.

[0072] In some embodiments, the first point cloud data can be projected into the second point cloud data to obtain the third point cloud data.

[0073] In an application scenario, referring to Figure 6 , step 51 can adopt the following flow:

[0074] Step 511: obtaining first position information of an unmanned vehicle corresponding to the first point cloud data, and second position information of the unmanned vehicle corresponding to the second point cloud data.

[0075] Because the unmanned vehicle itself is also in motion during the driving process, the coordinate conversion based on the pose of the unmanned vehicle is needed to eliminate the error of itself.

[0076] Step 512: performing coordinate system conversion on the first point cloud data by using the second position information and the first position information.

[0077] At this time, the second position information can be taken as a reference to perform coordinate conversion on the first point cloud data by using the first position information.

[0078] First, the conversion relationship between the first position information and the second position information is determined, such as determining a conversion matrix. The coordinate system conversion is performed on the first point cloud data by using the conversion matrix to obtain the coordinates of the second position information corresponding to the first point cloud data.

[0079] Step 513: projecting the first point cloud data after the coordinate system conversion into the second point cloud data to obtain the third point cloud data.

[0080] Because the second point cloud data is collected at the second position information, the first point cloud data after the coordinate system conversion can be projected into the second point cloud data, and the coordinate systems are the same.

[0081] Step 52: performing a filtering operation on the third point cloud data to remove ground plane points in the third point cloud data.

[0082] Optionally, any ground plane segmentation algorithm, such as a plane estimation method, can be used to remove data points belonging to the ground plane in the point cloud data, so that only non-ground plane points are left.

[0083] Step 53: clustering the third point cloud data after removal to obtain clustered third point cloud data.

[0084] Step 54: splitting the clustered third point cloud data to obtain fourth point cloud data and fifth point cloud data.

[0085] Step 55: determining a moving target in the point cloud data based on the fourth point cloud data and the fifth point cloud data.

[0086] Steps 53-55 have the same or similar technical solutions as any of the above embodiments, and will not be described here.

[0087] In this embodiment, merging adjacent time point cloud data and clustering the merged data are used to complete simultaneous clustering of two time point cloud data. Since the merged point cloud data is more, the clustering effect can be improved. On the other hand, splitting the clustered third point cloud data can improve the accuracy of splitting, and further improve the accuracy of determining the moving target, so as to facilitate subsequent processing based on the moving target.

[0088] Referring to Figure 7 , Figure 7 is a structural schematic diagram of an embodiment of a vehicle-mounted control system provided by the present application. The vehicle-mounted control system 70 includes a radar sensor 71 and a processor 72.

[0089] The radar sensor 71 is configured to collect first point cloud data and second point cloud data. The processor 72 is connected with the radar sensor 71 and configured to perform the following steps:

[0090] merging the first point cloud data and the second point cloud data to obtain third point cloud data; the second point cloud data is collected at a next time of the first point cloud data; clustering the third point cloud data to obtain clustered third point cloud data; splitting the clustered third point cloud data to obtain fourth point cloud data and fifth point cloud data; and determining a moving target in the point cloud data based on the fourth point cloud data and the fifth point cloud data.

[0091] It can be understood that the processor 72 is also configured to implement the technical solutions of any of the above embodiments, which will not be described here.

[0092] The vehicle-mounted control system 70 can determine a driving route and a driving speed at the next moment based on the determined moving target.

[0093] In other embodiments, the processor 72 is also connected with a memory (not shown in the figure). The memory is used to store a computer program, which, when executed by the processor 72, is used to implement the technical solutions of any of the above embodiments, which will not be repeated here.

[0094] Referring to Figure 8 , Figure 8 is a structural schematic diagram of an embodiment of the unmanned vehicle provided in the present application. The unmanned vehicle 80 comprises a vehicle-mounted control system 70.

[0095] Referring to Figure 9 , Figure 9 is a structural schematic diagram of an embodiment of the computer readable storage medium provided in the present application. The computer readable storage medium 90 is used to store a computer program 91, which, when executed by a processor, is used to implement the following method:

[0096] The first point cloud data and the second point cloud data are merged to obtain third point cloud data; the second point cloud data is collected at the next moment of the first point cloud data; the third point cloud data is clustered to obtain clustered third point cloud data; and the clustered third point cloud data is split to obtain fourth point cloud data and fifth point cloud data; and a moving target in the point cloud data is determined based on the fourth point cloud data and the fifth point cloud data.

[0097] It can be understood that the computer program 91, when executed by the processor, is also used to implement the technical solutions of any of the above embodiments, which will not be repeated here.

[0098] In the several embodiments provided in the present application, it should be understood that the disclosed method and device can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the circuit or unit is only a logical function division. There can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0099] The units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e. they can be located in one place, or can be distributed on a plurality of network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the present embodiment scheme.

[0100] In addition, each function unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software function unit.

[0101] The above is only the embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation according to the content of the specification and drawings of the present application, or direct or indirect application in other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A method of identifying a moving object, characterized by, The method comprises: merging the first point cloud data and the second point cloud data to obtain third point cloud data; the second point cloud data is collected at a next time of the first point cloud data; clustering the third point cloud data to obtain at least one clustered point cloud cluster; each point cloud cluster corresponds to a target object; splitting each point cloud cluster to obtain a first point cloud cluster corresponding to the first point cloud data and a second point cloud cluster corresponding to the second point cloud data; determining the coincidence of the first point cloud cluster and the second point cloud cluster; determining whether the corresponding target object is a moving target based on the coincidence.

2. The method of claim 1, wherein, The splitting of each point cloud cluster to obtain the first point cloud cluster corresponding to the first point cloud data and the second point cloud cluster corresponding to the second point cloud data comprises: obtaining the time sequence number of the data points in each point cloud cluster; obtaining the first point cloud cluster and the second point cloud cluster based on the point cloud with the same time sequence number.

3. The method of claim 1, wherein, Before the clustering of the third point cloud data to obtain at least one clustered point cloud cluster, it comprises: performing a filtering operation on the third point cloud data to remove the ground plane point cloud in the third point cloud data; The clustering of the third point cloud data to obtain at least one clustered point cloud cluster comprises: clustering the third point cloud data after removal to obtain at least one clustered point cloud cluster.

4. The method of claim 1, wherein, The merging of the first point cloud data and the second point cloud data to obtain the third point cloud data comprises: projecting the first point cloud data into the second point cloud data to obtain the third point cloud data.

5. The method of claim 4, wherein, The projecting of the first point cloud data into the second point cloud data to obtain the third point cloud data comprises: obtaining the first position information of the unmanned vehicle corresponding to the first point cloud data and the second position information of the unmanned vehicle corresponding to the second point cloud data; performing coordinate system conversion on the first point cloud data using the second position information and the first position information; projecting the first point cloud data after coordinate system conversion into the second point cloud data to obtain the third point cloud data.

6. An in-vehicle control system characterized by comprising: The vehicle-mounted control system comprises: a radar sensor for collecting first point cloud data and second point cloud data; a processor connected with the radar sensor for implementing the method of any one of claims 1-5.

7. An unmanned vehicle, characterized in that The unmanned vehicle comprises the vehicle-mounted control system of claim 6.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium is used to store a computer program, which is used to implement the method of any one of claims 1-5 when executed by a processor.

Images