Object matching method, device and electronic equipment

Abstract

The present invention provides an object matching method, device and electronic device. The method comprises: obtaining first result data of M traffic participant objects output by roadside equipment, and obtaining second result data reported by a target vehicle, where M is a positive integer; based on the first result data and the second result data, determining N first similarities corresponding to N traffic participant objects, where the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, where N is a positive integer less than or equal to M; based on the N first similarities, determining a target traffic participant object that matches the target vehicle from the N traffic participant objects. The embodiment of the present invention can improve the accuracy of vehicle information perception and vehicle tracking.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of vehicle-road collaboration technology, and in particular, to an object matching method, device and electronic device. Background Art

[0002] In vehicle perception technology, roadside equipment uses its detection methods to perceive the real-time status information of surrounding traffic participants, which can track vehicles and provide this real-time status information to vehicles to help them detect the environment around them.

[0003] At present, vehicle perception calculations are mainly concentrated in roadside equipment and are relatively independent of the vehicle, which will affect the accuracy of vehicle information perception and vehicle tracking. Summary of the invention

[0004] The embodiments of the present invention provide an object matching method, device and electronic device to solve the problem in the prior art that vehicle perception calculation is mainly concentrated in roadside equipment, which is relatively independent of the vehicle, thus affecting the accuracy of vehicle information perception and vehicle tracking.

[0005] In a first aspect, an embodiment of the present invention provides an object matching method, the method comprising:

[0006] Obtaining first result data of M traffic participants output by the roadside equipment, and obtaining second result data reported by the target vehicle, where M is a positive integer;

[0007] Based on the first result data and the second result data, determine N first similarities corresponding to N traffic participant objects, the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, and N is a positive integer less than or equal to M;

[0008] Based on the N first similarities, a target traffic participant object matching the target vehicle is determined from the N traffic participant objects.

[0009] In a second aspect, an embodiment of the present invention provides an object matching device, the device comprising:

[0010] An acquisition module, used to acquire first result data of M traffic participants output by the roadside equipment, and to acquire second result data reported by the target vehicle, where M is a positive integer;

[0011] A first determination module is used to determine N first similarities corresponding to N traffic participant objects based on the first result data and the second result data, wherein the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, and N is a positive integer less than or equal to M;

[0012] The second determination module is used to determine a target traffic participant object matching the target vehicle from the N traffic participant objects based on the N first similarities.

[0013] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the above-mentioned object matching method when executed by the processor.

[0014] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned object matching method are implemented.

[0015] In an embodiment of the present invention, the first result data of M traffic participants output by the roadside equipment and the second result data reported by the target vehicle are obtained; based on the first result data and the second result data, N first similarities corresponding to the N traffic participants are determined, and the N traffic participants are screened from the M traffic participants; based on the N first similarities, a target traffic participant matching the target vehicle is determined from the N traffic participants. In this way, the similarity between the target vehicle and the N traffic participants is calculated based on the first result data of the M traffic participants output by the roadside equipment and the second result data reported by the target vehicle, and the target traffic participant matching the target vehicle is screened for the target vehicle by the similarity. The target traffic participant can be used as the basis for information perception and tracking of the target vehicle, thereby improving the accuracy of vehicle information perception and vehicle tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0017] Figure 1 is a flow chart of an object matching method provided by an embodiment of the present invention;

[0018] Figure 2 It is a schematic diagram of the system structure applied to edge cloud to achieve object matching;

[0019] Figure 3 is a schematic diagram of the structure of an object matching device provided by an embodiment of the present invention;

[0020] Figure 4 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] At present, vehicle perception calculations are mainly concentrated in roadside equipment and are relatively independent of the vehicle. At this time, there is a gap between the vehicle's perception information and the roadside equipment's perception information. The vehicle may not be able to distinguish between the traffic participants perceived by the roadside equipment and itself, which will affect the accuracy of vehicle information perception and vehicle tracking.

[0022] At the same time, in some application scenarios, it is necessary to identify and track traffic participants. When there are multiple sets of roadside equipment covering the vehicle, how to quickly and effectively track the vehicle continuously is also a problem that needs to be solved urgently.

[0023] Based on this, an embodiment of the present invention provides a new object matching solution to solve the above problem.

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The object matching method provided by the embodiment of the present invention is first described below.

[0026] It should be noted that the object matching method provided in the embodiment of the present invention relates to the field of vehicle-road collaborative technology, which can be widely used in scenarios such as assisted driving and automatic driving. The method can be executed by the object matching device of the embodiment of the present invention. The object matching device can be configured in any electronic device to execute the object matching method. The electronic device can be a server or a terminal, which is not specifically limited here. In the following embodiments, the electronic device will be explained by taking the server as an example. The server can be distributed on the edge cloud side, which is a cloud data center distributed on the edge side of the network.

[0027] See also Figure 1 , the figure shows a schematic flow chart of the object matching method provided by an embodiment of the present invention.

[0028] like Figure 1 As shown, the method may include the following steps:

[0029] Step 101: obtain first result data of M traffic participants output by roadside equipment, and obtain second result data reported by the target vehicle.

[0030] Wherein, M is a positive integer.

[0031] In this step, the roadside equipment may refer to equipment that is installed on the roadside and has certain perception capabilities and can perceive information related to traffic participants on the road. It may be intelligent perception equipment such as high-definition cameras, lidars, microwave radars, etc.

[0032] The first result data may be data output by the roadside device and uploaded to the edge cloud side, and the first result data may include location information of each traffic participant collected by the roadside device, such as longitude and latitude, shape information, and heading angle information, etc. The traffic participant may be a motor vehicle, a non-motor vehicle, or a pedestrian, etc.

[0033] Typically, the data collected by roadside equipment is usually image data or point cloud data. Uploading these forms of data to the edge cloud will cause pressure on data transmission. The roadside equipment can identify the collected data and obtain the first result data of each traffic participant.

[0034] Afterwards, the first result data can be uploaded to the edge cloud side. Accordingly, the edge cloud can receive the first result data reported by the roadside device. Compared with reporting the collected data such as point cloud or image to the edge cloud side, reporting the result data to the edge cloud side can reduce the pressure of data transmission.

[0035] In an optional implementation, the first result data may be transmitted to the edge cloud side in the form of structured data, and the structured data may be a list or a data object, etc. At the same time, in the structured data, the information of the traffic participant object may be arranged in a certain order to perform structural alignment with the second result data, so that the data matching of the first result data and the second result data can be utilized to determine the target traffic participant object that matches the target vehicle. For example, in the structured data, the latitude and longitude, shape information, and heading angle information may be arranged in order, and when performing data matching later, the corresponding information may be obtained in this order for data matching.

[0036] The target vehicle may refer to any vehicle, and the target vehicle may be referred to as a main vehicle, and the target vehicle has the ability to obtain the second result data.

[0037] For example, the second result data may include vehicle basic data, which may include the latitude and longitude, shape information, and heading angle information of the target vehicle. The target vehicle may be equipped with on-board equipment or other equipment, which can obtain the vehicle basic data and report it. Correspondingly, the edge cloud can receive the vehicle basic data reported by the target vehicle.

[0038] For another example, the second result data may include both vehicle basic data and vehicle perception data. The vehicle perception data may be perception result data of the target vehicle for surrounding traffic participants. The perception result data may include relative position information between the target vehicle and nearby traffic participants, such as relative distance and relative heading angle information, etc. The target vehicle may be equipped with a perception device, and the perception result data may be obtained through the perception device.

[0039] The target vehicle may also transmit the second result data to the edge cloud side in the form of structured data, and the structured data may be a list or a data object, etc. At the same time, in the structured data, the information of the target vehicle may be arranged in a certain order to perform structural alignment with the first result data, so that the data matching of the first result data and the second result data is utilized to determine the target traffic participant matching the target vehicle.

[0040] In addition, in order to perform data matching more accurately, it is necessary to unify data reference standards, such as unified data positioning reference standards. In an optional implementation, the first result data and the second result data can both be uploaded in a unified geocentric coordinate system -84 (WGS-84).

[0041] Step 102, based on the first result data and the second result data, determine N first similarities corresponding to N traffic participant objects, the N traffic participant objects are screened from the M traffic participant objects, and a first similarity is the similarity between a traffic participant object and the target vehicle.

[0042] Wherein, N is a positive integer less than or equal to M.

[0043] In this step, N traffic participant objects can be screened from M traffic participant objects. Specifically, traffic participant objects around the target vehicle can be screened from the M traffic participant objects, and traffic participant objects far away from the target vehicle can be filtered out to reduce the amount of matching calculations.

[0044] A screening electronic fence can be established with the latitude and longitude position of the target vehicle in the vehicle basic data as the center and a radius of a preset width, such as a radius twice the width of the road where the target vehicle is located, so as to screen out the traffic participants located within the electronic fence from the M traffic participants, that is, to screen out the traffic participants whose latitude and longitude positions in the first result data are within the electronic fence.

[0045] After the N traffic participant objects are screened out, data matching may be performed based on the result data corresponding to the N traffic participant objects in the first result data and the second result data to determine N first similarities corresponding to the N traffic participant objects.

[0046] In an optional implementation, if the screening result includes only one traffic participant object, and the traffic participant object is a vehicle, it can be considered that the traffic participant object matches the target vehicle, that is, the traffic participant object is determined to be the main vehicle, and a matching relationship can be established between the two, which can reduce the amount of matching calculations. If the screening result includes multiple traffic participants, data matching is performed based on the result data corresponding to the N traffic participant objects in the first result data and the second result data to determine the N first similarities corresponding to the N traffic participant objects, and based on the N first similarities, the target traffic participant object that has a matching relationship with the target vehicle among the N traffic participant objects is determined.

[0047] Specifically, for each of the N traffic participant objects, based on the result data and the second result data of the traffic participant object, the traffic participant object can be matched with the target vehicle in terms of type, position, shape, heading angle, etc. to determine the first similarity between the traffic participant object and the target vehicle.

[0048] Step 103: Based on the N first similarities, determine a target traffic participant object that matches the target vehicle from the N traffic participant objects.

[0049] In this step, the N first similarities may be sorted in descending order, and the traffic participant object corresponding to the first similarity ranked first, that is, the largest first similarity, is determined as the target traffic participant object matched by the target vehicle.

[0050] Taking into account the heavy traffic volume in some intersection scenarios, under the influence of transmission delay, equipment collection accuracy, etc., when the edge cloud matches the result data received at the same time, it is very likely that there will be multiple traffic participants with close similarity to the target vehicle. Alternatively, under the influence of processing accuracy, there may be actual traffic participants that are calculated to have a matching relationship with the target vehicle, but their similarity with the target vehicle is not the greatest.

[0051] In this case, the first Q traffic participant objects corresponding to the first similarities may be obtained, that is, a plurality of traffic participant objects similar to the target vehicle may be screened out from the N traffic participant objects.

[0052] Afterwards, data matching may be performed again on the first result data and the second result data for the Q traffic participant objects, so as to determine a target traffic participant object that matches the target vehicle from among the Q traffic participant objects.

[0053] The matching relationship between the target vehicle and the target traffic participant can be cached in the edge cloud. The edge cloud can use the matching relationship to locate and track the vehicle, or it can be sent to the target vehicle to assist the vehicle in distinguishing between the traffic participant perceived by the roadside equipment and itself, so as to better realize information perception.

[0054] In this embodiment, the first result data of M traffic participants output by the roadside equipment and the second result data reported by the target vehicle are obtained; based on the first result data and the second result data, N first similarities corresponding to the N traffic participants are determined, and the N traffic participants are screened from the M traffic participants; based on the N first similarities, a target traffic participant matching the target vehicle is determined from the N traffic participants. In this way, the similarity between the target vehicle and the N traffic participants is calculated based on the first result data of the M traffic participants output by the roadside equipment and the second result data reported by the target vehicle, and the target traffic participant matching the target vehicle is screened for the target vehicle by the similarity. The target traffic participant can be used as the basis for information perception and tracking of the target vehicle, thereby improving the accuracy of vehicle information perception and vehicle tracking.

[0055] Optionally, the first result data includes first perception result data of each of the N traffic participants in K dimensions; the second result data includes basic vehicle data of the target vehicle in the K dimensions, where K is a positive integer; when N is greater than 1, step 102 specifically includes:

[0056] For each of the N traffic participant objects, based on the first perception result data and the vehicle basic data, determine K second similarities corresponding to the traffic participant object, where one second similarity is a similarity between the traffic participant object and the target vehicle in one dimension;

[0057] The K second similarities corresponding to each traffic participant object are weighted based on the preset weight information to obtain the N first similarities corresponding to the N traffic participant objects.

[0058] In this embodiment, if the screening results include multiple traffic participant objects, that is, N is greater than 1, the target vehicle can be matched with the traffic participant objects in the screening results, and the first result data can include the first perception result data of the traffic participant in K dimensions, and the first perception result data can include several descriptive parameters such as geometric shape information, type information, position information, and heading angle information of the traffic participant object, and the second result data can include the vehicle basic data of the target vehicle in K dimensions, and the vehicle basic data can include several descriptive parameters such as geometric shape information, type information, position information, and heading angle information of the target vehicle.

[0059] For each traffic participant and for each dimension of the traffic participant, the similarity between the traffic participant and the target vehicle in the dimension can be calculated based on the data of the first perception result data of the traffic participant in the dimension and the data of the target vehicle in the vehicle basic data in the dimension, to obtain a second similarity. And for different dimensions, the same or different similarity calculation methods can be used to calculate K second similarities between the traffic participant and the target vehicle in K dimensions. The following uses different similarity calculation methods to calculate K second similarities as an example for explanation.

[0060] Assuming that the traffic participant is P1 and the target vehicle is P2, the similarity under the type dimension, namely the type similarity, can be calculated. N traffic participant objects can be screened out from the M traffic participant objects. These N traffic participant objects can all be motor vehicles, and the target vehicle is a motor vehicle. In the process of matching P1 and P2, the type similarity indicates that the traffic participant object is similar to the target vehicle.

[0061] The similarity in the shape dimension, namely the shape similarity, can be calculated using the following formula (1).

[0062]

[0063] Among them, in the above formula (1), SIM sha (P1, P2) is the shape similarity between P1 and P2, width1, length1 and height1 are the length, width and height of P1 respectively, and width2, length2 and height2 are the length, width and height of P2 respectively. That is, the above formula (1) compares the length, width and height of P1 with the length, width and height of P2 to calculate the shape similarity between P1 and P2.

[0064] The similarity in the location dimension, namely, location similarity, can be calculated using the following formula (2).

[0065]

[0066] Among them, in the above formula (2), SIMdis (P1, P2) is the position similarity between P1 and P2, D 1,2 The Euclidean distance between P1 and P2 can be calculated based on the longitude and latitude of P1 and the longitude and latitude of P2 using the distance calculation formula. That is, by calculating the Euclidean distance between P1 and P2, the position similarity between P1 and P2 is obtained according to the Euclidean distance calculation.

[0067] The similarity in the heading angle dimension, namely the heading angle similarity, can be calculated using the following formula (3).

[0068]

[0069] Among them, in the above formula (3), SIM phi (P1, P2) is the heading angle similarity between P1 and P2, phi 1,2 is the heading angle deviation between P1 and P2. The heading angle deviation between P1 and P2 can be calculated, and the heading angle similarity between P1 and P2 can be calculated according to the heading angle deviation between the two.

[0070] Afterwards, for each traffic participant object, the following formula (4) can be used to perform weighted processing on the K second similarities corresponding to the traffic participant object based on preset weight information to obtain the first similarity corresponding to the traffic participant object.

[0071]

[0072] In the above formula (4), SIM(P1, P2) is the comprehensive similarity between P1 and P2 in multiple dimensions, SIM i (P1, P2) is the similarity between P1 and P2 in the i-th dimension, w i is the weight of the similarity under the i-th dimension. The preset weight information can be set according to actual conditions and is not specifically limited here.

[0073] In this embodiment, for each of the N traffic participants, based on the first perception result data and the vehicle basic data, K second similarities corresponding to the traffic participant are determined, and one second similarity is the similarity between the traffic participant and the target vehicle in one dimension; the K second similarities corresponding to each traffic participant are weighted based on the preset weight information to obtain N first similarities corresponding to the N traffic participants. In this way, a more accurate similarity calculation result between the traffic participant and the target vehicle can be obtained, making it more accurate when determining the effective cooperative vehicle of the main vehicle.

[0074] Optionally, step 103 specifically includes:

[0075] Filter out Q first target similarities from the N first similarities, each first target similarity is greater than other similarities among the N first similarities except the Q first target similarities, where Q is a positive integer;

[0076] Based on the Q first target similarities, a target traffic participant object matching the target vehicle is determined from the Q traffic participant objects corresponding to the Q first target similarities.

[0077] In this implementation, the N first similarities may be sorted in descending order, and the traffic participant object corresponding to the first similarity ranked first, that is, the largest first similarity, may be determined as the target traffic participant object matched by the target vehicle.

[0078] Taking into account the heavy traffic volume in some intersection scenarios, under the influence of transmission delay, equipment collection accuracy, etc., when the edge cloud matches the result data received at the same time, it is very likely that there will be multiple traffic participants with close similarity to the target vehicle. Alternatively, under the influence of processing accuracy, there may be actual traffic participants that are calculated to have a matching relationship with the target vehicle, but their similarity with the target vehicle is not the greatest.

[0079] In this case, the first Q traffic participant objects corresponding to the first similarities may be obtained, that is, a plurality of traffic participant objects similar to the target vehicle may be screened out from the N traffic participant objects.

[0080] Afterwards, data matching may be performed again on the first result data and the second result data for the Q traffic participant objects, so as to determine a target traffic participant object that matches the target vehicle from among the Q traffic participant objects.

[0081] In this way, based on the N first similarities, traffic participant objects that are relatively similar to the target vehicle can be screened out from the N traffic participant objects, and finally the target traffic participant objects that match the target vehicle can be screened out from the traffic participant objects that are relatively similar to the target vehicle. Therefore, through multiple screenings, traffic participant objects with sufficient correlation with the main vehicle can be obtained, the matching accuracy can be improved, and comprehensive and effective tracking of the main vehicle can be achieved.

[0082] Optionally, the second result data includes second perception result data of the target vehicle for traffic participants within a preset range; when Q is greater than 1, determining a target traffic participant matching the target vehicle from Q traffic participants corresponding to the Q first target similarities based on the Q first target similarities includes:

[0083] Determine Q first relative reference objects corresponding to the Q first target similarities based on the first result data, where one first relative reference object is a relative reference object of a traffic participant object corresponding to one first target similarity;

[0084] Determine a second relative reference object of the target vehicle based on the second perception result data, wherein the relative relationship between the first relative reference object and the traffic participant object corresponding to the first relative reference object is the same as the target relative relationship, and the target relative relationship is the relative relationship between the second relative reference object and the target vehicle;

[0085] Based on the first result data and the second perception result data, determine Q third similarities corresponding to the Q first relative reference objects, where one third similarity is a similarity between a first relative reference object and the second relative reference object;

[0086] For each first target similarity, the first target similarity is merged with a third similarity corresponding to the first target similarity to obtain a second target similarity;

[0087] Based on the Q second target similarities, a target traffic participant object matching the target vehicle is determined from the Q traffic participant objects.

[0088] In this embodiment, when Q is greater than 1, the vehicle perception result data and the roadside perception result data can be combined to further match the Q traffic participants with the target vehicle to determine the traffic participants that are sufficiently associated with the main vehicle.

[0089] Specifically, for each of the Q traffic participants, the first relative reference object of the traffic participant can be obtained based on the roadside sensing result data, and the first relative reference object can be a traffic participant around the traffic participant. Based on the roadside sensing result data, a traffic participant that has a certain relative relationship with the traffic participant can be obtained, such as the relative relationship can be that the traffic participant is in the same lane as the traffic participant and is located in the forward position of the driving direction of the traffic participant, and the forward position is adjacent to the position of the traffic participant, that is, the first relative reference object is the forward vehicle in the same lane as the traffic participant, which can be the first vehicle in front or the second vehicle in front, which is not specifically limited here.

[0090] The second relative reference object of the target vehicle can be obtained based on the vehicle perception result data, and the second relative reference object can be a traffic participant object around the target vehicle. Based on the vehicle perception result data, a traffic participant object that has a certain relative relationship with the target vehicle can be obtained, such as a traffic participant object that is in the same lane as the target vehicle and is located in a forward position in the driving direction of the target vehicle, and the forward position is adjacent to the position of the target vehicle, that is, the second relative reference object is a forward vehicle in the same lane as the target vehicle, which can be the first vehicle in front or the second vehicle in front, and no specific limitation is made here.

[0091] The relative relationship between the first relative reference object and the traffic participant object corresponding to the first relative reference object is the same as the target relative relationship, and the target relative relationship is the relative relationship between the second relative reference object and the target vehicle.

[0092] In this way, if the target vehicle matches the traffic participant object, the first relative reference object of the traffic participant object and the second relative reference object of the target vehicle also match. The similarity between the target vehicle and the traffic participant object and the similarity between the first relative reference object and the second relative reference object are superimposed to achieve the effect of increasing the similarity.

[0093] Therefore, based on the first result data and the second perception result data, Q third similarities corresponding to the Q first relative reference objects can be determined, and one third similarity is the similarity between a first relative reference object and the second relative reference object. Specifically, for each of the Q first relative reference objects, the similarity between the first relative reference object and the second relative reference object can be determined based on the first result data and the second perception result data. The similarity can include relative distance similarity and / or relative angle similarity, and the third similarity corresponding to the first relative reference object is determined based on the similarity.

[0094] Assuming that the first relative reference object is denoted by P1' and the second relative reference object is denoted by P2', the relative distance similarity between P1' and P2' can be calculated using the following formula (5).

[0095]

[0096] Among them, in the above formula (5), SIM D (P'1, P'2) is the relative distance similarity between P1' and P2', D'1 is the Euclidean distance between P1 and P1', and D'2 is the Euclidean distance between P2 and P2'.

[0097] The relative angle similarity between P1' and P2' can be calculated using the following formula (6).

[0098] SIM A (P'1,P'2)=cos(|180°-|A'2-A'1||) (6)

[0099] Among them, in the above formula (6), SIM A (P'1, P'2) is the relative angle similarity between P1' and P2', A'1 is the relative angle between P1 and P1', and A'2 is the relative angle between P2 and P2'.

[0100] The relative similarity between P1' and P2', ie, the third similarity corresponding to the first relative reference object, can be calculated by combining the relative distance similarity between P1' and P2' and the relative angle similarity between P1' and P2' using the following formula (7).

[0101]

[0102] In the above formula (7), SIM(P'1, P'2) is the relative similarity between P1' and P2', w i ' is the weight of the relative distance similarity or similar angle similarity, SIM i (P'1, P'2) is the relative distance similarity or relative angle similarity.

[0103] For each of the Q traffic participant objects, the first target similarity corresponding to the traffic participant object and the third similarity corresponding to the traffic participant object may be fused to obtain a second target similarity.

[0104] The first target similarity and the third similarity can be fused using the following formula (8).

[0105] SIM1=SIM(P1,P2)+W*SIM(P'1,P'2) (8)

[0106] In the above formula (8), SIM1 is the second target similarity, and W is the weight of the relative similarity, which can be set according to actual conditions and is not specifically limited here.

[0107] The Q second target similarities may be sorted in descending order, and the traffic participant object corresponding to the second target similarity that is ranked the highest, that is, the largest second target similarity, is determined as the target traffic participant object matched by the target vehicle.

[0108] In order to ensure the accuracy of the matching, multiple rounds of matching can be performed by combining the result data uploaded at multiple times. Finally, based on the second target similarity determined by multiple rounds of matching, the target traffic participant matching the target vehicle is determined. That is, after the initial recognition of SIM1, the data matching is repeated in the result data uploaded at the next two times to obtain SIM2 and SIM3. After that, the similarities obtained by matching these three times are calculated using the formula The final similarity SIM is obtained by summarizing, and the traffic participant object corresponding to the maximum similarity in the final similarity SIM is determined as the target traffic participant object matching the target vehicle.

[0109] In this embodiment, by obtaining the first relative reference object of the traffic participant object and the second relative reference object of the target vehicle, the relative similarity between the two relative reference objects is determined, and the similarity between the traffic participant object and the target vehicle and the relative similarity between the two relative reference objects are superimposed to obtain the final similarity between the traffic participant object and the target vehicle. Based on the final similarity, the target traffic participant object that matches the target vehicle is screened, so that the traffic participant object with sufficient correlation with the main vehicle can be matched through the two relative reference objects, thereby further improving the accuracy of the matching and achieving comprehensive and effective tracking of the main vehicle.

[0110] Optionally, the target relative relationship is:

[0111] is in the same lane as the target vehicle and is located at a first forward position in the travel direction of the target vehicle, the first forward position being adjacent to the position of the target vehicle; or

[0112] At a second forward position relative to a preset direction of the target vehicle, the second forward position is adjacent to a position where the target vehicle is located.

[0113] In this implementation manner, there may be two types of target relative relationships, and the first relative reference object and the second relative reference object may be acquired in two ways.

[0114] The first method is to introduce high-precision maps, use them to locate the lane and position of traffic participants in the roadside perception data, and further calculate the preceding vehicle P1' in the same lane of traffic participant P1. At the same time, locate the lane and position of the target vehicle P2 and the traffic participants in the vehicle perception data in the high-precision map, and calculate the preceding vehicle P2' in the same lane of P2.

[0115] The second is that, for each of the Q traffic participants, the first forward vehicle P1' rotated clockwise in the due north direction of the traffic participant P1 identified by the roadside equipment is calculated, that is, the first forward vehicle P1' relative to the preset direction of the traffic participant P1, such as the due north direction; the first forward vehicle P2' rotated clockwise in the due north direction of the traffic participant identified by the main vehicle is P2', that is, the first forward vehicle P2' relative to the preset direction of the target vehicle P2, such as the due north direction.

[0116] In this embodiment, by setting the target relative relationship to be in the same lane as the target vehicle and at a first forward position in the driving direction of the target vehicle, the first forward position is adjacent to the position of the target vehicle; or, at a second forward position relative to a preset direction of the target vehicle, the second forward position is adjacent to the position of the target vehicle, so that traffic participants that are around the target vehicle and related to the target vehicle can be obtained as relative reference objects of the target vehicle, thereby improving the accuracy of obtaining the relative reference objects and further improving the accuracy of matching.

[0117] Optionally, determining a target traffic participant object matching the target vehicle from the Q traffic participant objects based on the Q second target similarities includes:

[0118] For each of the Q traffic participants, the second target similarities corresponding to the traffic participant at multiple times are merged to obtain a third target similarity corresponding to the traffic participant;

[0119] The traffic participant object corresponding to the largest third target similarity among the Q third target similarities is determined as the target traffic participant object matched by the target vehicle.

[0120] In this embodiment, multiple rounds of matching can be performed in combination with the result data uploaded at multiple times, and finally the target traffic participant matching the target vehicle can be determined based on the second target similarity determined by multiple rounds of matching. That is, after the initial identification of SIM1, the data matching is repeated in the result data uploaded at the next two times to obtain SIM2 and SIM3. After that, the similarities obtained by matching these three times are calculated using the formula The final similarity SIM, i.e., the third target similarity, is obtained by summarizing, and the traffic participant object corresponding to the maximum similarity in the final similarity SIM is determined as the target traffic participant object matching the target vehicle, thereby further improving the matching accuracy.

[0121] Optionally, after determining a target traffic participant object matching the target vehicle from the N traffic participant objects based on the N first similarities, the method further includes:

[0122] The first identification of the target traffic participant object is associated with the second identification of the target vehicle to obtain a target association relationship.

[0123] In this implementation, the first identifier of the target traffic participant object observed by the roadside device can be set to deviceId+ptcId, where deviceId is the ID of the roadside device, and ptcId is the ID of the target traffic participant object identified and tracked by the roadside device. The first identifier can be associated with the second identifier reported by the target vehicle itself to obtain a target association relationship.

[0124] The target association relationship is cached in the edge cloud system. It serves as the basis for the subsequent edge cloud to provide services to the main vehicle, and can realize continuous tracking of the vehicle in the edge cloud based on the ID reported by the vehicle itself. Subsequent tracking can rely on the perception ability of the roadside equipment. The roadside equipment can always use the cached target association relationship to realize continuous tracking of the main vehicle without losing the traffic participants.

[0125] In some complex scenarios, such as some complex intersections, when there are multiple roadside devices, each time a vehicle enters the perception range of a roadside device, it can establish an association between the perceived traffic participants and the main vehicle. Since the main vehicle ID always remains unchanged, continuous tracking across roadside devices at complex intersections can be achieved, thereby realizing trajectory restoration.

[0126] In addition, the edge cloud can also send the target association relationship to the target vehicle so that the target vehicle can distinguish its own perception result data from the perception result data of other traffic participants based on the perception result data received from the roadside equipment, thereby improving the vehicle's information perception ability.

[0127] Optionally, before determining N first similarities corresponding to N traffic participant objects based on the first result data and the second result data, the method further includes:

[0128] Performing alignment processing on the first result data and the second result data to obtain aligned first result data and second result data, wherein the alignment processing includes at least one of data time domain alignment, data space domain alignment, and data structure alignment;

[0129] The determining, based on the first result data and the second result data, N first similarities corresponding to the N traffic participant objects includes:

[0130] Based on the aligned first result data and the second result data, N first similarities corresponding to the N traffic participant objects are determined.

[0131] In this implementation, before data matching, the first result data and the second result data may be aligned so that the multi-source data (including vehicle basic data and perception result data of roadside equipment) are incorporated into a unified reference framework.

[0132] Specifically, through filtering and smoothing and estimating the motion parameters of the entity, the multi-source data that are not synchronized in the time domain and belong to different coordinate systems in the spatial domain can be aligned in time and space, so as to incorporate the multi-source data into a unified reference frame and improve the accuracy of data matching. At the same time, the first result data and the second result data can be structurally aligned to make use of the matching of the two data.

[0133] See also Figure 2 , Figure 2 This is a schematic diagram of the system structure used for edge cloud object matching. Figure 2 As shown, the system may include roadside equipment, a target vehicle and an edge cloud. The roadside equipment may upload its perception result data to the edge cloud, and the target vehicle may upload vehicle basic data and its perception result data to the edge cloud. The edge cloud may perform multiple data matches based on the perception result data of the roadside equipment, the vehicle basic data and the perception result data of the target vehicle to determine the target traffic participant object that matches the target vehicle, and establish an association relationship between the target vehicle and the target traffic participant object. The association relationship may be cached to the edge cloud and used for vehicle tracking, or the association relationship may be sent to the target vehicle so that the target vehicle can distinguish its own perception result data from the perception result data of other traffic participants.

[0134] The object matching device provided by the embodiment of the present invention is described below.

[0135] See also Figure 3 , the figure shows a schematic diagram of the structure of the object matching device provided by an embodiment of the present invention.

[0136] like Figure 3 As shown, the object matching device 300 includes:

[0137] The acquisition module 301 is used to acquire the first result data of M traffic participants output by the roadside equipment, and acquire the second result data reported by the target vehicle, where M is a positive integer;

[0138] A first determination module 302 is used to determine N first similarities corresponding to N traffic participant objects based on the first result data and the second result data, wherein the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, and N is a positive integer less than or equal to M;

[0139] The second determination module 303 is used to determine a target traffic participant object matching the target vehicle from the N traffic participant objects based on the N first similarities.

[0140] Optionally, the first result data includes first perception result data of each of the N traffic participants in K dimensions; the second result data includes basic vehicle data of the target vehicle in the K dimensions, where K is a positive integer; when N is greater than 1, the first determination module 302 includes:

[0141] A first determination submodule is used to determine, for each of the N traffic participant objects, K second similarities corresponding to the traffic participant object based on the first perception result data and the vehicle basic data, where one second similarity is a similarity between the traffic participant object and the target vehicle in one dimension;

[0142] The weighted processing submodule is used to perform weighted processing on the K second similarities corresponding to each traffic participant object based on preset weight information to obtain the N first similarities corresponding to the N traffic participant objects.

[0143] Optionally, the second determining module 303 includes:

[0144] A screening submodule, configured to screen out Q first target similarities from the N first similarities, each first target similarity being greater than other similarities among the N first similarities except the Q first target similarities, where Q is a positive integer;

[0145] The second determination submodule is used to determine, based on the Q first target similarities, a target traffic participant object that matches the target vehicle from among the Q traffic participant objects corresponding to the Q first target similarities.

[0146] Optionally, the second result data includes second perception result data of the target vehicle for traffic participants within a preset range; when Q is greater than 1, the second determination submodule includes:

[0147] A first determination unit is used to determine Q first relative reference objects corresponding to the Q first target similarities based on the first result data, where a first relative reference object is a relative reference object of a traffic participant object corresponding to a first target similarity;

[0148] A second determination unit is used to determine a second relative reference object of the target vehicle based on the second perception result data, wherein a relative relationship between the first relative reference object and a traffic participant object corresponding to the first relative reference object is the same as a target relative relationship, and the target relative relationship is a relative relationship between the second relative reference object and the target vehicle;

[0149] A third determining unit, configured to determine, based on the first result data and the second perception result data, Q third similarities corresponding to the Q first relative reference objects, where one third similarity is a similarity between a first relative reference object and the second relative reference object;

[0150] a fusion unit, configured to fuse, for each first target similarity, the first target similarity with a third similarity corresponding to the first target similarity to obtain a second target similarity;

[0151] The fourth determination unit is used to determine a target traffic participant object that matches the target vehicle from the Q traffic participant objects based on the Q second target similarities.

[0152] Optionally, the target relative relationship is:

[0153] is in the same lane as the target vehicle and is located at a first forward position in the travel direction of the target vehicle, the first forward position being adjacent to the position of the target vehicle; or

[0154] At a second forward position relative to a preset direction of the target vehicle, the second forward position is adjacent to a position where the target vehicle is located.

[0155] Optionally, the fourth determining unit is specifically configured to:

[0156] For each of the Q traffic participants, the second target similarities corresponding to the traffic participant at multiple times are merged to obtain a third target similarity corresponding to the traffic participant;

[0157] The traffic participant object corresponding to the largest third target similarity among the Q third target similarities is determined as the target traffic participant object matched by the target vehicle.

[0158] Optionally, the device further comprises:

[0159] The association module is used to associate the first identification of the target traffic participant object with the second identification of the target vehicle to obtain a target association relationship.

[0160] Optionally, the device further comprises:

[0161] an alignment processing module, used for performing alignment processing on the first result data and the second result data to obtain aligned first result data and second result data, wherein the alignment processing includes at least one of data time domain alignment, data space domain alignment, and data structure alignment;

[0162] The first determination module 302 is specifically configured to determine N first similarities corresponding to N traffic participant objects based on the aligned first result data and the second result data.

[0163] The object matching device 300 can implement each process implemented in the above-mentioned object matching method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described here.

[0164] The electronic device provided by the embodiment of the present invention is described below.

[0165] See also Figure 4 , the figure shows a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 4 As shown, the electronic device 400 includes: a processor 401 , a memory 402 , a user interface 403 and a bus interface 404 .

[0166] The processor 401 is used to read the program in the memory 402 and execute the following process:

[0167] Obtaining first result data of M traffic participants output by the roadside equipment, and obtaining second result data reported by the target vehicle, where M is a positive integer;

[0168] Based on the first result data and the second result data, determine N first similarities corresponding to N traffic participant objects, the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, and N is a positive integer less than or equal to M;

[0169] Based on the N first similarities, a target traffic participant object matching the target vehicle is determined from the N traffic participant objects.

[0170] exist Figure 4In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by processor 401 and various circuits of memory represented by memory 402 are linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface 404 provides an interface. For different user devices, the user interface 403 may also be an interface that can be connected to external or internal devices, and the connected devices include but are not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0171] The processor 401 is responsible for managing the bus architecture and general processing, and the memory 402 can store data used by the processor 401 when performing operations.

[0172] Optionally, the first result data includes first perception result data of each of the N traffic participants in K dimensions; the second result data includes basic vehicle data of the target vehicle in the K dimensions, where K is a positive integer; when N is greater than 1, the processor 401 is further configured to:

[0173] For each of the N traffic participant objects, based on the first perception result data and the vehicle basic data, determine K second similarities corresponding to the traffic participant object, where one second similarity is a similarity between the traffic participant object and the target vehicle in one dimension;

[0174] The K second similarities corresponding to each traffic participant object are weighted based on the preset weight information to obtain the N first similarities corresponding to the N traffic participant objects.

[0175] Optionally, the processor 401 is further configured to:

[0176] Filter out Q first target similarities from the N first similarities, each first target similarity is greater than other similarities among the N first similarities except the Q first target similarities, where Q is a positive integer;

[0177] Based on the Q first target similarities, a target traffic participant object matching the target vehicle is determined from the Q traffic participant objects corresponding to the Q first target similarities.

[0178] Optionally, the second result data includes second perception result data of the target vehicle for traffic participants within a preset range; when Q is greater than 1, the processor 401 is further configured to:

[0179] Determine Q first relative reference objects corresponding to the Q first target similarities based on the first result data, where one first relative reference object is a relative reference object of a traffic participant object corresponding to one first target similarity;

[0180] Determine a second relative reference object of the target vehicle based on the second perception result data, wherein the relative relationship between the first relative reference object and the traffic participant object corresponding to the first relative reference object is the same as the target relative relationship, and the target relative relationship is the relative relationship between the second relative reference object and the target vehicle;

[0181] Based on the first result data and the second perception result data, determine Q third similarities corresponding to the Q first relative reference objects, where one third similarity is a similarity between a first relative reference object and the second relative reference object;

[0182] For each first target similarity, the first target similarity is merged with a third similarity corresponding to the first target similarity to obtain a second target similarity;

[0183] Based on the Q second target similarities, a target traffic participant object matching the target vehicle is determined from the Q traffic participant objects.

[0184] Optionally, the target relative relationship is:

[0185] is in the same lane as the target vehicle and is located at a first forward position in the travel direction of the target vehicle, the first forward position being adjacent to the position of the target vehicle; or

[0186] At a second forward position relative to a preset direction of the target vehicle, the second forward position is adjacent to a position where the target vehicle is located.

[0187] Optionally, the processor 401 is further configured to:

[0188] For each of the Q traffic participants, the second target similarities corresponding to the traffic participant at multiple times are merged to obtain a third target similarity corresponding to the traffic participant;

[0189] The traffic participant object corresponding to the largest third target similarity among the Q third target similarities is determined as the target traffic participant object matched by the target vehicle.

[0190] Optionally, the processor 401 is further configured to:

[0191] The first identification of the target traffic participant object is associated with the second identification of the target vehicle to obtain a target association relationship.

[0192] Optionally, the processor 401 is further configured to:

[0193] Performing alignment processing on the first result data and the second result data to obtain aligned first result data and second result data, wherein the alignment processing includes at least one of data time domain alignment, data space domain alignment, and data structure alignment;

[0194] Based on the aligned first result data and the second result data, N first similarities corresponding to the N traffic participant objects are determined.

[0195] Preferably, an embodiment of the present invention further provides an electronic device, comprising a processor 401, a memory 402, and a computer program stored in the memory 402 and executable on the processor 401. When the computer program is executed by the processor 401, each process of the above-mentioned object matching method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0196] The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each process of the above-mentioned object matching method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0197] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0198] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0199] In the embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0200] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0201] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0202] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0203] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An object matching method, characterized in that: The method comprises: Acquire first result data of M traffic participants output by the roadside equipment, and acquire second result data reported by the target vehicle, where M is a positive integer; the second result data includes second perception result data of the target vehicle for the traffic participants within a preset range; Based on the first result data and the second result data, determine N first similarities corresponding to N traffic participant objects, the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, and N is a positive integer less than M; Filter out Q first target similarities from the N first similarities, each first target similarity is greater than other similarities among the N first similarities except the Q first target similarities, where Q is a positive integer; Based on the Q first target similarities, determining a target traffic participant object matching the target vehicle from the Q traffic participant objects corresponding to the Q first target similarities; In the case where Q is greater than 1, determining a target traffic participant object matching the target vehicle from Q traffic participant objects corresponding to the Q first target similarities based on the Q first target similarities includes: Determine Q first relative reference objects corresponding to the Q first target similarities based on the first result data, where one first relative reference object is a relative reference object of a traffic participant object corresponding to one first target similarity; Determine a second relative reference object of the target vehicle based on the second perception result data, wherein the relative relationship between the first relative reference object and the traffic participant object corresponding to the first relative reference object is the same as the target relative relationship, and the target relative relationship is the relative relationship between the second relative reference object and the target vehicle; Based on the first result data and the second perception result data, determine Q third similarities corresponding to the Q first relative reference objects, where one third similarity is a similarity between a first relative reference object and the second relative reference object; For each first target similarity, the first target similarity is merged with a third similarity corresponding to the first target similarity to obtain a second target similarity; Based on the Q second target similarities, a target traffic participant object matching the target vehicle is determined from the Q traffic participant objects.

2. The method according to claim 1, characterized in that: The first result data includes first perception result data of each of the N traffic participants in K dimensions; the second result data includes basic vehicle data of the target vehicle in the K dimensions, where K is a positive integer; when N is greater than 1, determining N first similarities corresponding to the N traffic participants based on the first result data and the second result data includes: For each of the N traffic participant objects, based on the first perception result data and the vehicle basic data, determine K second similarities corresponding to the traffic participant object, where one second similarity is a similarity between the traffic participant object and the target vehicle in one dimension; The K second similarities corresponding to each traffic participant object are weighted based on the preset weight information to obtain the N first similarities corresponding to the N traffic participant objects.

3. The method according to claim 1, characterized in that The relative relationship of the targets is: Being in the same lane as the target vehicle and being located at a first forward position in the travel direction of the target vehicle, the first forward position being adjacent to the position of the target vehicle; or, At a second forward position relative to a preset direction of the target vehicle, the second forward position is adjacent to a position where the target vehicle is located.

4. The method according to claim 1, characterized in that: The step of determining a target traffic participant object matching the target vehicle from the Q traffic participant objects based on the Q second target similarities includes: For each of the Q traffic participants, the second target similarities corresponding to the traffic participant at multiple times are merged to obtain a third target similarity corresponding to the traffic participant; The traffic participant object corresponding to the largest third target similarity among the Q third target similarities is determined as the target traffic participant object matched by the target vehicle.

5. The method according to claim 1, characterized in that After determining a target traffic participant object matching the target vehicle from the N traffic participant objects based on the N first similarities, the method further includes: The first identification of the target traffic participant object is associated with the second identification of the target vehicle to obtain a target association relationship.

6. The method according to claim 1, characterized in that Before determining N first similarities corresponding to N traffic participant objects based on the first result data and the second result data, the method further includes: Performing alignment processing on the first result data and the second result data to obtain aligned first result data and second result data, wherein the alignment processing includes at least one of data time domain alignment, data space domain alignment, and data structure alignment; The determining, based on the first result data and the second result data, N first similarities corresponding to the N traffic participant objects includes: Based on the aligned first result data and the second result data, N first similarities corresponding to the N traffic participant objects are determined.

7. An object matching device, characterized in that: The device comprises: An acquisition module is used to acquire first result data of M traffic participants output by the roadside equipment, and to acquire second result data reported by the target vehicle, where M is a positive integer; the second result data includes second perception result data of the target vehicle for the traffic participants within a preset range; A first determination module is used to determine N first similarities corresponding to N traffic participant objects based on the first result data and the second result data, wherein the N traffic participant objects are screened from the M traffic participant objects, a first similarity is a similarity between a traffic participant object and the target vehicle, and N is a positive integer less than M; The second determination module includes: A screening submodule, configured to screen out Q first target similarities from the N first similarities, each first target similarity being greater than other similarities among the N first similarities except the Q first target similarities, where Q is a positive integer; A second determination submodule is used to determine, based on the Q first target similarities, a target traffic participant object that matches the target vehicle from the Q traffic participant objects corresponding to the Q first target similarities; When Q is greater than 1, the second determination submodule includes: A first determination unit is used to determine Q first relative reference objects corresponding to the Q first target similarities based on the first result data, where one first relative reference object is a relative reference object of a traffic participant object corresponding to one first target similarity; A second determination unit is used to determine a second relative reference object of the target vehicle based on the second perception result data, wherein a relative relationship between the first relative reference object and a traffic participant object corresponding to the first relative reference object is the same as a target relative relationship, and the target relative relationship is a relative relationship between the second relative reference object and the target vehicle; A third determining unit, configured to determine, based on the first result data and the second perception result data, Q third similarities corresponding to the Q first relative reference objects, where one third similarity is a similarity between a first relative reference object and the second relative reference object; a fusion unit, configured to fuse, for each first target similarity, the first target similarity with a third similarity corresponding to the first target similarity to obtain a second target similarity; The fourth determination unit is used to determine a target traffic participant object that matches the target vehicle from the Q traffic participant objects based on the Q second target similarities.

8. An electronic device, characterized in that: The electronic device comprises: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the object matching method according to any one of claims 1 to 6 when executed by the processor.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the object matching method according to any one of claims 1 to 6 are implemented.

Images