Target identification method and device, vehicle and storage medium

Abstract

The invention provides a target recognition method and device, a vehicle and a storage medium, and the method comprises the steps: obtaining first point cloud data collected at a target moment, and carrying out the recognition of a first vehicle target in at least one vehicle target in the first point cloud data according to the similarity between the first vehicle target and a second vehicle target, the split count of the first vehicle target at the previous moment is updated, the split count used for indicating the possibility that the first vehicle target is a split target at the target moment is obtained, and according to the ratio of the split count of the first vehicle target at the target moment to the total amount of the processed point cloud data corresponding to the first vehicle target, the split target is obtained; and determining whether the first vehicle target is the split target, and identifying whether the first vehicle target is the split target based on the ratio of the split count corresponding to the first vehicle target determined by the similarity to the total amount of the point cloud data corresponding to the processed first vehicle target at the target moment, thereby reducing the occupation of the memory space, and improving the reliability of the vehicle target. And the split target identification efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a target recognition method, device, vehicle and storage medium. Background Art

[0002] Millimeter-wave radar is the main sensor in autonomous driving, but its resolution performance is poor and it often cannot accurately give the size of the target. In particular, for some long vehicle targets, the target splitting problem is prone to occur, that is, a vehicle target is split into multiple targets, including split targets and main targets. Among them, the main target is the target closer to the radar, and the split target is the target farther away from the radar.

[0003] In the related art, when performing target recognition, the radar needs to recognize split targets based on data from multiple historical moments, which increases memory consumption and reduces the efficiency of split target recognition. Summary of the invention

[0004] The present application aims to solve one of the technical problems in the related art at least to some extent.

[0005] To this end, the present application proposes a target recognition method, device, vehicle and storage medium, which identifies whether the first vehicle target is a split target based on the ratio between the split count corresponding to the first vehicle target determined by similarity and the total amount of point cloud data corresponding to the first vehicle target processed at the target moment. It does not require more memory space, reduces memory consumption, and improves the efficiency of split target recognition.

[0006] In one aspect, an embodiment of the present application provides a target recognition method, including:

[0007] Acquire first point cloud data collected by the radar at the target time; wherein the first point cloud data includes at least one vehicle target;

[0008] For a first vehicle target among the at least one vehicle target, determining a similarity between the first vehicle target and a second vehicle target among the at least one vehicle target;

[0009] According to the similarity between the first vehicle target and the second vehicle target, updating the split count corresponding to the first vehicle target at the previous moment of the target moment, and obtaining the split count corresponding to the first vehicle target at the target moment; wherein the split count is used to indicate the possibility that the first vehicle target is a split target;

[0010] Whether the first vehicle target is a split target is determined according to a ratio between a split count corresponding to the first vehicle target at the target time and a total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

[0011] Another aspect of the present application provides a target recognition device, including:

[0012] An acquisition module, used to acquire first point cloud data collected by the radar at the target time; wherein the first point cloud data includes at least one vehicle target;

[0013] A first determining module, configured to determine, for a first vehicle target among the at least one vehicle target, a similarity between the first vehicle target and a second vehicle target among the at least one vehicle target;

[0014] An updating module, configured to update a split count corresponding to the first vehicle target at a moment before the target moment according to the similarity between the first vehicle target and the second vehicle target, to obtain a split count corresponding to the first vehicle target at the target moment; wherein the split count is used to indicate the possibility that the first vehicle target is a split target;

[0015] The second determination module is used to determine whether the first vehicle target is a split target according to the ratio between the split count corresponding to the first vehicle target at the target time and the total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

[0016] Another aspect of the present application provides a vehicle, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method described in the first aspect or the method described in the second aspect is implemented.

[0017] Another aspect of the present application provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method described in the first aspect or the method described in the second aspect is implemented.

[0018] Another aspect of the present application provides a computer program product on which a computer program is stored. When the program is executed by a processor, the method described in the first aspect or the second aspect is implemented.

[0019] The target recognition method, device, vehicle and storage medium proposed in the present application obtain at least one vehicle target included in the first point cloud data collected by the radar at the target moment, determine the similarity between the first vehicle target and the second vehicle target for the first vehicle target among the at least one vehicle target, update the split count corresponding to the first vehicle target at the previous moment of the target moment according to the similarity, obtain the split count for indicating the possibility that the first vehicle target is a split target at the target moment, determine whether the first vehicle target is a split target according to the ratio between the split count corresponding to the first vehicle target at the target moment and the total amount of point cloud data corresponding to the first vehicle target that has been processed, and identify whether the first vehicle target is a split target based on the ratio between the split count corresponding to the first vehicle target determined based on the similarity and the total amount of point cloud data corresponding to the first vehicle target that has been processed at the target moment, without occupying more memory space, thereby improving the efficiency of split target recognition.

[0020] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0022] Figure 1 A flowchart of a target recognition method provided in an embodiment of the present application;

[0023] Figure 2 A schematic diagram of a vehicle splitting scenario provided in an embodiment of the present application;

[0024] Figure 3 A flowchart of another target recognition method provided in an embodiment of the present application;

[0025] Figure 4 A flowchart of another target recognition method provided in an embodiment of the present application;

[0026] Figure 5 A schematic diagram of a vehicle turning right provided in an embodiment of the present application;

[0027] Figure 6 A schematic diagram of the structure of a target recognition device provided in an embodiment of the present application;

[0028] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0030] The target recognition method, device, vehicle and storage medium of the embodiments of the present application are described below with reference to the accompanying drawings.

[0031] Figure 1 A flowchart of a target recognition method provided in an embodiment of the present application.

[0032] The target recognition method of the embodiment of the present application is executed by a target recognition device, which can be set in an on-board device of a vehicle and is not limited in the present embodiment.

[0033] like Figure 1 As shown, the method may include the following steps:

[0034] Step 101, obtaining first point cloud data collected by the radar at the target time.

[0035] In the embodiment of the present application, the target time is any time during the vehicle's driving process. According to the point cloud data collected by the radar in the vehicle (self-vehicle), the vehicle information in front of or around the self-vehicle can be identified. Since the radar collects point cloud data according to the set sampling period, the point cloud data collected at each moment is different. For identification, the point cloud data collected at the target moment is called the first point cloud data.

[0036] The first point cloud data includes at least one vehicle target. The at least one vehicle target may belong to the same vehicle, or each vehicle target may be a different vehicle. As an example, Figure 2 A schematic diagram of a vehicle splitting scenario provided in an embodiment of the present application, such as Figure 2 As shown in the figure, there is a large truck in front of the ego vehicle. The vehicle is relatively long. The data collected by the radar of the ego vehicle can identify the large truck in front as two vehicle targets, among which target B close to the ego vehicle is the main target, and target A far away from the ego vehicle is the split target. In fact, both the split target and the main target indicate the large truck. Therefore, it is necessary to identify the split target and delete the split target so that the number of targets identified in the scenarios of visual recognition and radar recognition is consistent, thereby improving the accuracy of target recognition or obstacle recognition.

[0037] Step 102 : for a first vehicle target among at least one vehicle target, determine the similarity between the first vehicle target and a second vehicle target among at least one vehicle target.

[0038] The first vehicle target is one of the at least one vehicle target. The first vehicle target can be a complete vehicle or one of multiple vehicle targets that a large vehicle is split into. The first vehicle target is a vehicle target that is currently being identified as a split target, that is, a potential split target to be identified. Therefore, other vehicle targets other than the first vehicle target are potential main targets, called second vehicle targets, and the second vehicle targets can be one or more. The first vehicle target and the second vehicle target are different vehicle targets in the at least one vehicle target.

[0039] In the embodiment of the present application, since the split target and the main target are obtained by splitting the same vehicle target, they belong to the same vehicle target, and the split target and the main target belonging to the same vehicle must have a high similarity in position and speed. Therefore, for a potential split target first vehicle target, if there is a second vehicle target whose similarity with the first vehicle target is higher than a threshold, then the possibility that the first vehicle target is a split target is high, otherwise, the possibility is low. Among them, the similarity between the first vehicle target and the second vehicle target includes position similarity and speed similarity. For each first vehicle target, in the vehicle coordinate system, the position similarity between the first vehicle target and the second vehicle target is determined, the speed information of the first vehicle target and the speed information of the second vehicle target are obtained, and the speed similarity between the first vehicle target and the second vehicle target is determined based on the speed information of the first vehicle target and the speed information of the second vehicle target. Among them, the vehicle coordinate system is the coordinate system of the vehicle on which the radar is installed.

[0040] Step 103: based on the similarity between the first vehicle target and the second vehicle target, the split count corresponding to the first vehicle target at the previous moment of the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment.

[0041] The split count is used to indicate the possibility that the first vehicle target is a split target. The higher the value of the split count is, the higher the possibility that the first vehicle target is a split target is, and vice versa.

[0042] In the embodiment of the present application, the split count of the first vehicle target has different values ​​at different times. During the time period in which the radar can continuously collect the first vehicle target, the value of the split count of the first vehicle target during the time period changes. If there is a third vehicle target in the second vehicle target whose similarity with the first vehicle target is greater than a threshold, it means that the third vehicle target may be the main target of the first vehicle target, and the possibility that the first vehicle target is a split target increases, and the split count corresponding to the first vehicle target at the previous moment of the target time is increased; if there is no third vehicle target in the second vehicle target whose similarity with the first vehicle target is greater than a threshold, it means that there is no first vehicle target. If the main target of the first vehicle target is the main target of the first vehicle target, the possibility that the first vehicle target is a split target is reduced, and the split count corresponding to the first vehicle target at the previous moment of the target moment is reduced. At each moment, based on the first point cloud data of the first vehicle target collected at the moment, the similarity between the first vehicle target and the second vehicle target is determined, and based on the similarity, it is determined whether the split count of the first vehicle target determined at the historical moment is increased or decreased, thereby improving the accuracy of the split count corresponding to the first vehicle target. The split count is used to indicate the possibility that the first vehicle target is a split target, and there is no need to pre-store the historical point cloud data including the first vehicle target, thereby reducing memory usage and improving processing efficiency.

[0043] Step 104 , determining whether the first vehicle target is a split target according to a ratio between a split count corresponding to the first vehicle target at the target time and a total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

[0044] In an embodiment of the present application, the total amount of point cloud data corresponding to the first vehicle target that has been processed at the target moment indicates the length of the life cycle corresponding to the first vehicle target. The greater the total amount of point cloud data corresponding to the first vehicle target, the longer the length of the life cycle corresponding to the first vehicle target. The ratio between the split count corresponding to the first vehicle target at the target moment and the total amount of point cloud data corresponding to the first vehicle target that has been processed at the target moment is compared with a set threshold. If the ratio is greater than the set threshold, the recognition result is considered to be that the first vehicle target is a split target; if the ratio is less than or equal to the set threshold, the recognition result is considered to be that the first vehicle target is not a split target, and the recognition of the split target is completed while reducing memory consumption.

[0045] It should be understood that each point cloud data in the total amount of point cloud data includes the first vehicle target.

[0046] The size of the set threshold may be related to the total amount of point cloud data and may be dynamically changed. When the total amount of point cloud data is small, that is, less than or equal to the quantity threshold, the set threshold may be larger, for example, greater than a certain set value; when the total amount of point cloud data is greater than the quantity threshold, the set threshold may be smaller, that is, less than a certain set value. Because the small amount of point cloud data means that target tracking is not yet stable, the first vehicle target may be allowed not to be identified as a split target to improve the reliability of identification.

[0047] It should be noted that steps 102 to 104 may be repeated multiple times to identify each first vehicle target in the first point cloud data, and delete each first vehicle target identified as a split target, so as to improve the accuracy of target identification based on data collected by the radar.

[0048] In the target recognition method of the embodiment of the present application, first point cloud data collected by the radar at the target moment is obtained, wherein the first point cloud data includes at least one vehicle target, and for the first vehicle target among the at least one vehicle target, the similarity between the first vehicle target and the second vehicle target is determined, and based on the similarity between the first vehicle target and the second vehicle target, the split count corresponding to the first vehicle target at the previous moment of the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment, wherein the split count is used to indicate the possibility that the first vehicle target is a split target, and based on the ratio between the split count corresponding to the first vehicle target at the target moment and the total amount of point cloud data corresponding to the first vehicle target processed at the target moment, it is determined whether the first vehicle target is a split target, and based on the ratio between the split count corresponding to the first vehicle target determined based on the similarity and the total amount of point cloud data corresponding to the first vehicle target processed at the target moment, it is identified whether the first vehicle target is a split target, without occupying more memory space, thereby reducing memory consumption and improving the efficiency of split target recognition.

[0049] Based on the above embodiments, Figure 3 A flowchart of another target recognition method provided in an embodiment of the present application is shown as follows: Figure 3 As shown, the method comprises the following steps:

[0050] Step 301, obtaining the first point cloud data collected by the radar at the target time.

[0051] Step 302 : For a first vehicle target among at least one vehicle target, determine the similarity between the first vehicle target and a second vehicle target among at least one vehicle target.

[0052] Among them, step 301 and step 302 can refer to the explanation in the above-mentioned embodiment, and the principles are the same, so they will not be repeated here.

[0053] Step 303: Determine an offset of a split count corresponding to the first vehicle target in the first point cloud data according to the similarity between the first vehicle target and the second vehicle target.

[0054] In the embodiment of the present application, the second vehicle target may be one or more. If the second vehicle target is one, the first vehicle target and the second vehicle target are compared for similarity. If the first vehicle target and the second vehicle target have similarity, the second vehicle target is determined to be the main target of the first vehicle target, and the first vehicle target is considered to be a possible split target. The offset of the split count corresponding to the determined first vehicle target is a positive value, which is used to increase the split count corresponding to the first vehicle target. On the contrary, the offset is a negative value, which is used to reduce the split count corresponding to the first vehicle target. If the second vehicle target is multiple, the first vehicle target and each second vehicle target are compared for similarity. If the first vehicle target and any second vehicle target have similarity, it is determined that there is a second vehicle target as the main target of the first vehicle target, and the first vehicle target is considered to be a possible split target. The offset of the split count corresponding to the determined first vehicle target is a positive value, which is used to increase the split count corresponding to the first vehicle target. On the contrary, the offset is a negative value, which is used to reduce the split count corresponding to the first vehicle target.

[0055] According to the similarity between the first vehicle object and the second vehicle object, it is determined whether there is a main object similar to the first vehicle object from among the second vehicle objects.

[0056] In the embodiment of the present application, similarity includes position similarity and speed similarity. If there is a second vehicle target among the second vehicle targets that has a similar position and speed to the first vehicle target, the second vehicle target is called the main target of the first vehicle target.

[0057] In one scenario, in response to the presence of a main target, a first offset of a split count corresponding to a first vehicle target in the first point cloud data is determined; wherein the first offset is greater than zero to increase the possibility that the first vehicle target is a split target.

[0058] In another scenario, in response to the absence of a main target, a second offset of a split count corresponding to a first vehicle target in the first point cloud data is determined; wherein the second offset is less than zero to reduce the possibility that the first vehicle target is a split target.

[0059] The first offset and the second offset may be set according to requirements, or determined based on the degree of similarity, which will be specifically described in subsequent embodiments and will not be repeated here.

[0060] Step 304: Update the split count corresponding to the first vehicle target at the moment before the target moment according to the offset to obtain the split count corresponding to the first vehicle target at the target moment.

[0061] The second point cloud data collected at the previous moment includes the first vehicle target.

[0062] In an embodiment of the present application, the split count corresponding to the first vehicle target at the moment before the target moment is added to the offset of the split count at the target moment to obtain the split count corresponding to the first vehicle target at the target moment, so as to update the split count corresponding to the first vehicle target at each moment of the first vehicle target, so as to continuously update the split count corresponding to the first vehicle target during the detection cycle of the first vehicle target, thereby improving the accuracy of the split count corresponding to the first vehicle target.

[0063] In one scenario, if the target moment is the first moment, that is, the current first point cloud data is the first frame data collected by the radar, the split count of the first vehicle target is set to 0, and the split count of the first vehicle target is updated in the point cloud data including the first vehicle target collected at each subsequent moment, so as to realize the continuous accumulation and update of the split count throughout the life cycle of the first vehicle target, and improve the accuracy of identifying whether the first vehicle target is a split target through the multiple frames of point cloud data included in the entire life cycle.

[0064] In another scenario, if the target moment is not the first moment, that is, the first point cloud data is not the point cloud data of the first frame collected by the radar, then the split count S1 corresponding to the first vehicle target at the moment before the target moment is obtained, and the offset K of the target moment and the split count S1 are added to realize the update of the split count S1 corresponding to the moment before the target moment, and the split count S2 corresponding to the first vehicle target at the target moment is obtained = S1 + K. Among them, K can be a set value greater than zero, which can be set according to needs, or determined based on the degree of similarity, which will be explained in the specific subsequent embodiments and will not be repeated here.

[0065] Step 305 , determining whether the first vehicle target is a split target according to a ratio between a split count corresponding to the first vehicle target at the target time and a total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

[0066] Among them, step 305 can refer to the explanation in the above-mentioned embodiment, and the principle is the same, which will not be repeated here.

[0067] In the target recognition method of the embodiment of the present application, based on the similarity between the first vehicle target and the second vehicle target, the offset of the split count corresponding to the first vehicle target in the first point cloud data is determined, and based on the offset, the split count corresponding to the first vehicle target at the previous moment of the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment, and based on the ratio between the split count corresponding to the first vehicle target at the target moment and the total amount of point cloud data corresponding to the first vehicle target processed at the target moment, it is determined whether the first vehicle target is a split target. There is no need to store and process historical point cloud data, which reduces the memory consumption of multi-frame point cloud data, and the processing results based on the historical point cloud data are converted into the split count corresponding to the first vehicle target. By updating the split count in each frame, the memory consumption is reduced on the basis of satisfying the processing based on multi-frame point cloud data, and the processing efficiency is improved.

[0068] Based on the above embodiments, the present application provides another target recognition method. Figure 4 A flowchart of another target recognition method provided in an embodiment of the present application is shown as follows: Figure 4 As shown, the method comprises the following steps:

[0069] Step 401, obtaining the first point cloud data collected by the radar at the target time.

[0070] Among them, step 401 can refer to the explanation in the above-mentioned embodiment, and the principle is the same, which will not be repeated here.

[0071] Step 402, obtaining the turning angle information of the vehicle on which the radar is installed.

[0072] In an embodiment of the present application, the turning angle information of the front wheels of the vehicle can be measured based on the speed measuring device in the vehicle, wherein the turning angle information is the steering and steering angle during the operation of the vehicle, which can indicate the driving state of the vehicle, that is, the steering state, such as a left turn state, a straight state, or a right turn state.

[0073] Step 403: Determine the driving status of a first vehicle target and a second vehicle target traveling in the same direction as the vehicle according to the turning angle information.

[0074] As an example, Figure 5 A schematic diagram of a vehicle turning right provided in an embodiment of the present application, such as Figure 5 As shown in the figure, the turning angle information of the front wheels of the vehicle, that is, the self-vehicle, indicates that the self-vehicle is traveling on a road where it needs to turn right, so the large vehicle traveling in the same direction obtained by the radar scanning of the self-vehicle also needs to turn right. Therefore, based on the turning angle information of the self-vehicle, the driving status of the first vehicle target and the second vehicle target traveling in the same direction as the self-vehicle can be determined. Figure 5As shown, the driving state of the first vehicle target and the second vehicle target is a right turn state, that is, the vehicle turns right. There is one first vehicle target and one second vehicle target. If vehicle target A is the first vehicle target, then vehicle target B is the second vehicle target; if vehicle target B is the first vehicle target, then vehicle target A is the second vehicle target.

[0075] Step 404 : for a first vehicle target among the at least one vehicle target, determine the position similarity between the first vehicle target and a second vehicle target in a driving state.

[0076] The position similarity includes the similarity of the lateral position and the similarity of the longitudinal position. In the embodiment of the present application, the position similarity is described in the vehicle coordinate system of the vehicle.

[0077] In an embodiment of the present application, in a driving state, the position of the first vehicle target in the lateral direction and the longitudinal direction, as well as the position of the second vehicle target in the lateral direction and the longitudinal direction in the vehicle coordinate system are acquired.

[0078] For the position similarity in the longitudinal direction, the position similarity between the first vehicle target and the second vehicle target in the longitudinal direction is determined based on the position of the second vehicle target in the longitudinal direction and the position of the first vehicle target in the longitudinal direction. Specifically, the difference obtained by subtracting the position of the first vehicle target in the longitudinal direction from the position of each second vehicle target in the longitudinal direction is compared with the distance threshold. If any difference is less than the distance threshold, it is considered that there is a second vehicle target that is similar to the first vehicle target in the longitudinal direction; if each difference is greater than or equal to the distance threshold, it is considered that there is no second vehicle target that is similar to the first vehicle target in the longitudinal direction.

[0079] For the position similarity in the lateral direction, the position interval corresponding to the driving state is determined based on the position of the first vehicle target in the lateral direction and the set position offset, and the position similarity of the first vehicle target and the second vehicle target in the lateral direction is determined based on the position interval corresponding to the driving state and the position of the second vehicle target in the lateral direction. When the vehicle is driving, especially when driving on a curve, even different parts of the same vehicle may have large position deviations. When driving on a curve, the lateral position difference between the front and rear of the vehicle will be further widened, and it is difficult to ensure the effect of simple position similarity judgment. Therefore, in the embodiment of the present application, for different driving states, a position interval corresponding to the driving state is determined based on the position offset. The position interval is the position interval of the main target corresponding to the first vehicle target when the first vehicle target is used as a possible split target under different driving states. That is to say, in this driving state, if there is no second vehicle target in the corresponding position interval, there is no similarity between the position of each second vehicle target in the lateral direction and the position of the first vehicle target in the lateral direction, that is, there is no second vehicle target that is the main target of the first vehicle target; if there is a second vehicle target in the corresponding position interval, there is similarity between the position of the second vehicle target in the lateral direction and the position of the first vehicle target in the lateral direction, that is, the second vehicle target is the main target of the first vehicle target, which improves the accuracy of recognition in different scenarios. The position intervals of different scenarios are explained below:

[0080] In the first scenario, the driving state is a straight-ahead state, and the first position interval corresponding to the straight-ahead state is determined according to the position of the first vehicle target in the lateral direction and the set first position offset; wherein the lower limit value of the first position interval is less than the value of the position of the first vehicle target in the lateral direction, and the upper limit value of the first position interval is greater than the value of the position of the first vehicle target in the lateral direction.

[0081] Then, the position of the second vehicle target in the lateral direction is compared with the first position interval. In response to the position of the second vehicle target in the lateral direction being in the first position interval, it is determined that the second vehicle target and the first vehicle target are similar in position in the lateral direction; in response to the position of the second vehicle target in the lateral direction not being in the first position interval, it is determined that the second vehicle target and the first vehicle target are not similar in position in the lateral direction. In the case of multiple second vehicle targets, if there is a second vehicle target among the multiple second vehicle targets whose position in the lateral direction is within an equal set range on the left and right sides of the first vehicle target, it is considered that the second vehicle and the first vehicle target are similar in lateral position in the straight-ahead state, and the second vehicle target is the main target of the first vehicle target.

[0082] As an example, the driving state is a straight state. In the vehicle coordinate system of the vehicle, the position of the vehicle in the lateral direction is X0, and the set position offset is X1. Then the first position interval determined in the straight state is [X0-X1, X0+X1].

[0083] For example, the position of the first vehicle target in the lateral direction (corresponding to the horizontal axis in the vehicle coordinate system) is a coordinate point of 10 meters, and the first position interval is [8, 12].

[0084] In the second scenario, the driving state is a left turn state. According to the position of the first vehicle target in the lateral direction and the set second position offset, the first reference position of the first vehicle target in the lateral direction is determined, that is, the position of the first vehicle target in the lateral direction is offset to the right by the second position offset, that is, the position of the first vehicle target in the lateral direction is added to the second position offset, and the second position offset is a value greater than zero. Then, according to the first reference position and the set first position offset, the second position interval corresponding to the left turn state is determined; wherein the lower limit value of the second position interval is less than the value of the first reference position, and the upper limit value of the second position interval is greater than the value of the first reference position, that is, the second position interval is a symmetrical position centered on the first reference position. Then, the position of the second vehicle target in the lateral direction is compared with the second position interval. In response to the position of the second vehicle target in the lateral direction being in the second position interval, it is determined that the second vehicle target and the first vehicle target are similar in position in the lateral direction; in response to the position of the second vehicle target in the lateral direction not being in the second position interval, it is determined that the second vehicle target and the first vehicle target are not similar in position in the lateral direction. In the case where there is at least one second vehicle target, among the at least one second vehicle target, if there is a second vehicle target whose position in the lateral direction is within the second position interval relatively to the right side of the first vehicle target, it is considered that the second vehicle and the first vehicle target are similar in lateral position in the left turn state, and the second vehicle target is the main target of the first vehicle target.

[0085] As an example, the driving state is a left turn state. In the vehicle coordinate system of the vehicle, the position of the vehicle in the lateral direction is X0, and the set second position offset is X2. The determined first reference position is X0+X2, and then, based on the first reference position and the first position offset, the position interval in the left turn state is determined to be [X0+X2-X1, X0+X2+X1].

[0086] For example, if the position of the first vehicle target in the lateral direction (corresponding to the horizontal axis in the vehicle coordinate system) is a coordinate point of 10 meters, and the second position offset is 1, then the first reference position is determined to be a coordinate point of 11 meters, and the first position offset is 2, then the second position interval is determined to be [9, 13]. In the third scenario, the second reference position of the first vehicle target in the lateral direction is determined according to the position of the first vehicle target in the lateral direction and the set second position offset; the third position interval corresponding to the right turn state is determined according to the second reference position and the set first position offset; wherein the lower limit value of the third position interval is less than the value of the second reference position, and the upper limit value of the third position interval is greater than the value of the second reference position. Then, the position of the second vehicle target in the lateral direction is compared with the third position interval, and in response to the position of the second vehicle target in the lateral direction being in the third position interval, it is determined that the second vehicle target and the first vehicle target are similar in position in the lateral direction; in response to the position of the second vehicle target in the lateral direction not being in the third position interval, it is determined that the second vehicle target and the first vehicle target are not similar in position in the lateral direction. In the case where there is at least one second vehicle target, among the at least one second vehicle target, if there is a second vehicle target whose position in the lateral direction is within the third position interval relatively to the left of the first vehicle target, it is considered that in the right turn state, the second vehicle and the first vehicle target are similar in lateral position, and the second vehicle target is the main target of the first vehicle target.

[0087] As an example, the driving state is a right turn state. In the vehicle coordinate system of the vehicle, the position of the vehicle in the lateral direction is X0, and the set second position offset is X2. The determined second reference position is X0-X2, and then, based on the second reference position and the first position offset, the position interval in the right turn state is determined to be [X0-X2-X1, X0-X2+X1].

[0088] For example, the position of the first vehicle target in the lateral direction (corresponding to the horizontal axis in the vehicle coordinate system) is a coordinate point of 10 meters, and the second position offset is 1, then the first reference position is determined to be a coordinate point of 9 meters, and the first position offset is 2, then the second position interval is determined to be [7, 11].

[0089] The value of the first position offset and the value of the second position offset may be determined based on prior experience or may be set based on the accuracy requirement of the recognition scene, which is not limited in this embodiment.

[0090] Optionally, when determining the position similarity between the first vehicle target and the second vehicle target, the same point on the first vehicle target and the second vehicle target may be selected. For example, when calculating the position similarity in the longitudinal direction, the farthest point or the closest point in the longitudinal length may be selected.

[0091] Step 405, obtaining speed information of the first vehicle target and speed information of the second vehicle target.

[0092] As an implementation method, the speed information of the first vehicle target and the speed information of the second vehicle target are measured by a speed measuring device installed on the vehicle, and the speed information is the speed information of the vehicle's driving direction.

[0093] As another implementation method, the speed information measured by the speed measuring device on the first vehicle target can be sent to the vehicle of the own vehicle via wireless communication, and the speed information measured by the speed measuring device on the second vehicle target can be sent to the vehicle of the own vehicle via wireless communication.

[0094] Step 406 : Determine the speed similarity between the first vehicle target and the second vehicle target based on the speed information of the first vehicle target and the speed information of the second vehicle target.

[0095] In the embodiment of the present application, when there is at least one second vehicle target, for each second vehicle target, the speed information of the first vehicle target and the speed information of the second vehicle target are compared. If the difference between the speed information of the first vehicle target and the speed information of the second vehicle target is less than or equal to the speed threshold, it is determined that the speeds of the first vehicle target and the second vehicle target are similar; if the difference between the speed information of the first vehicle target and the speed information of the second vehicle target is greater than the speed threshold, it is determined that the speeds of the first vehicle target and the second vehicle target are not similar. Among them, the speed threshold can be 0.5m / s. Since the split target and the main target belong to the same vehicle, as long as the difference between the speeds is greater than the speed threshold, it is considered that the speeds of the first vehicle target and the second vehicle target are not similar, and the second vehicle target cannot be the main target of the first vehicle target. Step 407, based on the position similarity and speed similarity between the first vehicle target and the second vehicle target, the split count corresponding to the first vehicle target at the previous moment of the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment.

[0096] The split count is used to indicate the possibility that the first vehicle target is a split target.

[0097] In one implementation of the embodiment of the present application, according to the position similarity and speed similarity between the first vehicle target and at least one second vehicle target, it is determined from at least one second vehicle target whether there is a main target with a similar position and speed to the first vehicle target. In one scenario, in response to the existence of the main target, according to the position similarity between the first vehicle target and the main target, and the speed similarity between the first vehicle target and the main target, a first offset of the split count corresponding to the first vehicle target in the first point cloud data is determined, and the first offset is greater than zero, wherein the greater the similarity, the greater the value of the first offset. As an implementation, if the position similarity between the first vehicle target and the main target is greater than a set first position similarity threshold, and the speed similarity between the first vehicle target and the main target is greater than a set first speed similarity threshold, then the value of the first offset is determined to be 2; if the position similarity between the first vehicle target and the main target is greater than a set second position similarity threshold but less than a set first position similarity threshold, and the speed similarity between the first vehicle target and the main target is greater than a set second speed similarity threshold but less than a set first speed similarity threshold, then the value of the first offset is determined to be 1; wherein the second position similarity threshold is less than the first position similarity threshold, and the second speed similarity threshold is less than the first speed similarity threshold. The degree of similarity may be divided based on demand, which is not limited in this embodiment.

[0098] In another scenario, in response to the absence of a main target, that is, the positions and speeds of the first target vehicle and each second vehicle target are not similar, a second offset of the split count corresponding to the first vehicle target in the first point cloud data is determined, and the second offset is less than zero, for example, the second offset is -2. If the positions of the first target vehicle and at least one second vehicle target are similar, but the speeds are not similar, or the speeds of the first target vehicle and at least one second vehicle target are similar, but the positions are not similar, the second offset of the split count corresponding to the first vehicle target in the first point cloud data is determined to be, for example, -1.

[0099] Step 408 , determining whether the first vehicle target is a split target recognition result according to the ratio between the split count corresponding to the first vehicle target at the target time and the total amount of point cloud data corresponding to the first vehicle target processed at the target time.

[0100] Among them, step 408 can refer to the explanation in the above-mentioned embodiment, and the principle is the same, which will not be repeated here.

[0101] In the target recognition method of the embodiment of the present application, a multi-frame similarity statistics method is adopted, but the similarity statistics of the whole life cycle of the first vehicle target are performed instead of a historical moment, wherein the whole life cycle refers to the time period from the first appearance of the first vehicle target in the point cloud data frame of the radar of the own vehicle to its disappearance in a frame of point cloud data, that is, based on the split count of the first vehicle target to be identified determined by multiple frames, it is determined whether the first vehicle target is a split target, and no additional historical information is saved to reduce memory consumption. When performing similarity statistics, adaptive but loose position similarity statistics are performed, and strict speed similarity statistics are performed at the same time to adapt to situations where large vehicles / curves may cause large differences in the positions of split targets and main targets, thereby improving the accuracy of split target recognition.

[0102] In order to implement the above embodiment, the embodiment of the present application also proposes a target recognition device.

[0103] Figure 6 A schematic diagram of the structure of a target recognition device provided in an embodiment of the present application.

[0104] like Figure 6 As shown, the device may include:

[0105] An acquisition module 61 is used to acquire first point cloud data collected by the radar at a target time; wherein the first point cloud data includes at least one vehicle target;

[0106] A first determination module 62, configured to determine, for a first vehicle target among the at least one vehicle target, a similarity between the first vehicle target and a second vehicle target among the at least one vehicle target;

[0107] An updating module 63 is used to update the split count corresponding to the first vehicle target at a moment before the target moment according to the similarity between the first vehicle target and the second vehicle target, so as to obtain the split count corresponding to the first vehicle target at the target moment; wherein the split count is used to indicate the possibility that the first vehicle target is a split target;

[0108] The second determination module 64 is used to determine whether the first vehicle target is a split target according to the ratio between the split count corresponding to the first vehicle target at the target time and the total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

[0109] Further, in an implementation of the embodiment of the present application, the updating module 63 is specifically configured to:

[0110] Determining an offset of a split count corresponding to the first vehicle target in the first point cloud data according to a similarity between the first vehicle target and the second vehicle target;

[0111] According to the offset, the split count corresponding to the first vehicle target at the previous moment of the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment; wherein the second point cloud data collected at the previous moment includes the first vehicle target.

[0112] In one implementation of the embodiment of the present application, the updating module 63 is specifically configured to:

[0113] According to the similarity between the first vehicle target and the second vehicle target, determining whether there is a main target similar to the first vehicle target from among the second vehicle targets;

[0114] In response to the presence of the main target, determining a first offset of a split count corresponding to the first vehicle target in the first point cloud data; the first offset being greater than zero;

[0115] In response to the absence of the main target, a second offset of the split count corresponding to the first vehicle target in the first point cloud data is determined; the second offset is less than zero.

[0116] In an implementation of the embodiment of the present application, the first determining module 62 is specifically configured to:

[0117] Acquiring turning angle information of a vehicle on which the radar is installed;

[0118] Determining the driving states of the first vehicle target and the second vehicle target traveling in the same direction as the vehicle according to the turning angle information;

[0119] For each first vehicle target of at least one vehicle target, determining a position similarity between the first vehicle target and the second vehicle target in the driving state;

[0120] Acquiring speed information of the first vehicle target and speed information of the second vehicle target;

[0121] Based on the speed information of the first vehicle target and the speed information of the second vehicle target, speed similarity between the first vehicle target and the second vehicle target is determined.

[0122] In an implementation of the embodiment of the present application, the first determining module 62 is further configured to:

[0123] In the driving state, acquiring a position of the first vehicle target in a lateral direction and a position of the first vehicle target in a longitudinal direction, and a position of the second vehicle target in a lateral direction and a position of the second vehicle target in a vehicle coordinate system;

[0124] Determining the position similarity between the first vehicle target and the second vehicle target in the longitudinal direction according to the position of the second vehicle target in the longitudinal direction and the position of the first vehicle target in the longitudinal direction;

[0125] determining a position interval corresponding to the driving state according to the position of the first vehicle target in the lateral direction and a set position offset;

[0126] The position similarity between the first vehicle target and the second vehicle target in the lateral direction is determined according to the position interval corresponding to the driving state and the position of the second vehicle target in the lateral direction.

[0127] In an implementation of the embodiment of the present application, the driving state is a straight state, and the first determining module 62 is further configured to:

[0128] According to the position of the first vehicle target in the lateral direction and the set first position offset, the first position interval corresponding to the straight-ahead state is determined; wherein the lower limit value of the first position interval is less than the value of the position of the first vehicle target in the lateral direction, and the upper limit value of the first position interval is greater than the value of the position of the first vehicle target in the lateral direction.

[0129] In an implementation of the embodiment of the present application, the driving state is a left turn state, and the first determining module 62 is further configured to:

[0130] Determining a first reference position of the first vehicle target in the lateral direction according to the position of the first vehicle target in the lateral direction and a set second position offset;

[0131] According to the first reference position and the set first position offset, the second position interval corresponding to the left turn state is determined; wherein the lower limit value of the second position interval is less than the value of the first reference position, and the upper limit value of the second position interval is greater than the value of the first reference position.

[0132] In an implementation of the embodiment of the present application, the driving state is a right turn state, and the first determining module 62 is further configured to:

[0133] determining a second reference position of the first vehicle target in the lateral direction according to the position of the first vehicle target in the lateral direction and a set second position offset;

[0134] According to the second reference position and the set first position offset, a third position interval corresponding to the right turn state is determined; wherein a lower limit value of the third position interval is less than a value of the second reference position, and an upper limit value of the third position interval is greater than a value of the second reference position.

[0135] In an implementation of the embodiment of the present application, the first determining module 62 is further configured to:

[0136] comparing the position of the second vehicle target in the lateral direction with the first position interval;

[0137] In response to the position of the second vehicle target in the lateral direction being within the first position interval, determining that the second vehicle target and the first vehicle target are similar in position in the lateral direction;

[0138] In response to the position of the second vehicle target in the lateral direction not being within the first position interval, it is determined that the second vehicle target and the first vehicle target are not similar in position in the lateral direction.

[0139] It should be noted that the above explanation of the method embodiment is also applicable to the device of this embodiment, and will not be repeated here.

[0140] In the target recognition device of the embodiment of the present application, first point cloud data collected at a target moment is obtained, wherein the first point cloud data includes at least one vehicle target, and for each first vehicle target of the at least one vehicle target, the similarity between the first vehicle target and the second vehicle target is determined, and according to the similarity between the first vehicle target and the second vehicle target, the split count corresponding to the first vehicle target at a moment before the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment, wherein the split count is used to indicate the possibility that the first vehicle target is a split target, and according to the ratio between the split count corresponding to the first vehicle target at the target moment and the total amount of point cloud data corresponding to the first vehicle target processed at the target moment, an identification result of whether the first vehicle target is a split target is determined, and based on the ratio between the split count corresponding to the first vehicle target determined by similarity and the total amount of point cloud data corresponding to the first vehicle target processed at the target moment, whether the first vehicle target is a split target is identified, without occupying more memory space, reducing memory consumption, and improving the efficiency of split target recognition.

[0141] In order to implement the above embodiments, the present application also proposes a vehicle, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method described in the above method embodiments is implemented.

[0142] In order to implement the above embodiments, the present application also proposes a non-temporary computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method described in the above method embodiments is implemented.

[0143] In order to implement the above embodiments, the present application also proposes a computer program product on which a computer program is stored. When the computer program is executed by a processor, the method described in the above method embodiments is implemented.

[0144] Figure 7 A schematic diagram of a vehicle structure provided in an embodiment of the present application. For example, vehicle 700 may be a hybrid vehicle, a non-hybrid vehicle, an electric vehicle, a fuel cell vehicle, or other types of vehicles. Vehicle 700 may be an autonomous vehicle, a semi-autonomous vehicle, or a non-autonomous vehicle.

[0145] Reference Figure 7 , the vehicle 700 may include various subsystems, for example, an infotainment system 710, a perception system 720, a decision control system 730, a drive system 740, and a computing platform 750. The vehicle 700 may also include more or fewer subsystems, and each subsystem may include multiple components. In addition, each subsystem and each component of the vehicle 700 may be interconnected by wire or wireless means.

[0146] In some embodiments, the infotainment system 710 may include a communication system, an entertainment system, and a navigation system, etc.

[0147] The perception system 720 may include several sensors for sensing information about the environment around the vehicle 700. For example, the perception system 720 may include a global positioning system (the global positioning system may be a GPS system, or a Beidou system or other positioning systems), an inertial measurement unit (IMU), a laser radar, a millimeter wave radar, an ultrasonic radar, and a camera.

[0148] The decision control system 730 may include a computing system, a vehicle controller, a steering system, a throttle, and a braking system.

[0149] The drive system 740 may include components that provide power movement for the vehicle 700. In one embodiment, the drive system 740 may include an engine, an energy source, a transmission system, and wheels. The engine may be one or a combination of an internal combustion engine, an electric motor, and an air compression engine. The engine is capable of converting energy provided by the energy source into mechanical energy.

[0150] Some or all functions of the vehicle 700 are controlled by a computing platform 750. The computing platform 750 may include at least one processor 751 and a memory 752, and the processor 751 may execute instructions 753 stored in the memory 752.

[0151] The processor 751 may be any conventional processor, such as a commercially available CPU. The processor may also include a graphics processor (Graphic Process Unit, GPU), a field programmable gate array (Field Programmable Gate Array, FPGA), a system on chip (System on Chip, SOC), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC) or a combination thereof.

[0152] Memory 752 can be implemented by any type of volatile or nonvolatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0153] In addition to instructions 753 , memory 752 may also store data, such as road maps, route information, vehicle location, direction, speed, etc. The data stored in memory 752 may be used by computing platform 750 .

[0154] In the embodiment of the present disclosure, the processor 751 may execute instruction 753 to complete all or part of the steps of the above method embodiment.

[0155] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0156] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0157] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0158] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0159] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0160] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0161] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0162] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A target recognition method, characterized in that: include: Acquire first point cloud data collected by the radar at the target time; wherein the first point cloud data includes at least one vehicle target; For a first vehicle target among the at least one vehicle target, determining a similarity between the first vehicle target and a second vehicle target among the at least one vehicle target; According to the similarity between the first vehicle target and the second vehicle target, updating the split count corresponding to the first vehicle target at the previous moment of the target moment, and obtaining the split count corresponding to the first vehicle target at the target moment; wherein the split count is used to indicate the possibility that the first vehicle target is a split target; Whether the first vehicle target is a split target is determined according to a ratio between a split count corresponding to the first vehicle target at the target time and a total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

2. The method according to claim 1, characterized in that The updating, according to the similarity between the first vehicle target and the second vehicle target, of the split count corresponding to the first vehicle target at the previous moment of the target moment to obtain the split count corresponding to the first vehicle target at the target moment includes: Determining an offset of a split count corresponding to the first vehicle target in the first point cloud data according to a similarity between the first vehicle target and the second vehicle target; According to the offset, the split count corresponding to the first vehicle target at the previous moment of the target moment is updated to obtain the split count corresponding to the first vehicle target at the target moment; wherein the second point cloud data collected at the previous moment includes the first vehicle target.

3. The method according to claim 2, characterized in that The determining, according to the similarity between the first vehicle target and the second vehicle target, an offset of a split count corresponding to the first vehicle target in the first point cloud data includes: According to the similarity between the first vehicle target and the second vehicle target, determining whether there is a main target similar to the first vehicle target from among the second vehicle targets; In response to the presence of the main target, determining a first offset of a split count corresponding to the first vehicle target in the first point cloud data; the first offset being greater than zero; In response to the absence of the main target, a second offset of the split count corresponding to the first vehicle target in the first point cloud data is determined; the second offset is less than zero.

4. The method according to any one of claims 1 to 3, characterized in that: The determining, for each first vehicle target among the at least one vehicle target, a similarity between the first vehicle target and a second vehicle target among the at least one vehicle target comprises: Acquiring turning angle information of a vehicle on which the radar is installed; Determining the driving states of the first vehicle target and the second vehicle target traveling in the same direction as the vehicle according to the turning angle information; For each first vehicle target of at least one vehicle target, determining a position similarity between the first vehicle target and the second vehicle target in the driving state; Acquiring speed information of the first vehicle target and speed information of the second vehicle target; Based on the speed information of the first vehicle target and the speed information of the second vehicle target, speed similarity between the first vehicle target and the second vehicle target is determined.

5. The method according to claim 4, characterized in that Determining the position similarity between the first vehicle target and the second vehicle target in the driving state includes: In the driving state, acquiring a position of the first vehicle target in a lateral direction and a position of the first vehicle target in a longitudinal direction, and a position of the second vehicle target in a lateral direction and a position of the second vehicle target in a vehicle coordinate system; Determining the position similarity between the first vehicle target and the second vehicle target in the longitudinal direction according to the position of the second vehicle target in the longitudinal direction and the position of the first vehicle target in the longitudinal direction; determining a position interval corresponding to the driving state according to the position of the first vehicle target in the lateral direction and a set position offset; The position similarity between the first vehicle target and the second vehicle target in the lateral direction is determined according to the position interval corresponding to the driving state and the position of the second vehicle target in the lateral direction.

6. The method according to claim 5, characterized in that The driving state is a straight-ahead state, and determining a position interval corresponding to the driving state according to a position of the first vehicle target in a lateral direction and a set position offset includes: According to the position of the first vehicle target in the lateral direction and the set first position offset, the first position interval corresponding to the straight-ahead state is determined; wherein the lower limit value of the first position interval is less than the value of the position of the first vehicle target in the lateral direction, and the upper limit value of the first position interval is greater than the value of the position of the first vehicle target in the lateral direction.

7. The method according to claim 5, characterized in that The driving state is a left turn state, and determining a position interval corresponding to the driving state according to the position of the first vehicle target in the lateral direction and a set position offset includes: Determining a first reference position of the first vehicle target in the lateral direction according to the position of the first vehicle target in the lateral direction and a set second position offset; According to the first reference position and the set first position offset, the second position interval corresponding to the left turn state is determined; wherein the lower limit value of the second position interval is less than the value of the first reference position, and the upper limit value of the second position interval is greater than the value of the first reference position.

8. The method according to claim 5, characterized in that The driving state is a right turn state, and determining a position interval corresponding to the driving state according to a position of the first vehicle target in a lateral direction and a set position offset includes: Determining a second reference position of the first vehicle target in the lateral direction according to the position of the first vehicle target in the lateral direction and a set second position offset; According to the second reference position and the set first position offset, a third position interval corresponding to the right turn state is determined; wherein a lower limit value of the third position interval is less than a value of the second reference position, and an upper limit value of the third position interval is greater than a value of the second reference position.

9. The method according to claim 6, characterized in that The determining, based on the position interval corresponding to the driving state and the position of the second vehicle target in the lateral direction, the position similarity between the first vehicle target and the second vehicle target in the lateral direction comprises: comparing the position of the second vehicle target in the lateral direction with the first position interval; In response to the position of the second vehicle target in the lateral direction being within the first position interval, determining that the second vehicle target and the first vehicle target are similar in position in the lateral direction; In response to the position of the second vehicle target in the lateral direction not being within the first position interval, it is determined that the second vehicle target and the first vehicle target are not similar in position in the lateral direction.

10. A target recognition device, characterized in that: include: An acquisition module, used to acquire first point cloud data collected by the radar at the target time; wherein the first point cloud data includes at least one vehicle target; A first determination module, configured to determine, for a first vehicle target among the at least one vehicle target, a similarity between the first vehicle target and a second vehicle target among the at least one vehicle target; An updating module, configured to update a split count corresponding to the first vehicle target at a moment before the target moment according to the similarity between the first vehicle target and the second vehicle target, to obtain a split count corresponding to the first vehicle target at the target moment; wherein the split count is used to indicate the possibility that the first vehicle target is a split target; The second determination module is used to determine whether the first vehicle target is a split target according to the ratio between the split count corresponding to the first vehicle target at the target time and the total amount of point cloud data corresponding to the first vehicle target that has been processed at the target time.

11. A vehicle, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

12. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images