Roadside target fusion method and device, electronic equipment and computer program product

Through the target trajectory prediction technology combined with observation data, the problem of poor target perception accuracy of the roadside fusion system in the under-rod area is solved, achieving more stable target fusion and lower latency data delivery.

CN118965277AInactive Publication Date: 2024-11-15ZHIDAO NETWORK TECH (BEIJING) CO LTD
View PDF 10 Cites 0 Cited by

Patent Information

Application Number
CN202411426175.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-14
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing roadside fusion system has poor target perception accuracy in the under-pole area, resulting in frequent heading abnormalities, lateral jumps and tracking discontinuities in the target, affecting the delivery of downstream data.

Method used

Using target trajectory prediction technology, we obtain target observation data at the current moment and target prediction data at the previous moment output by a single rod on the roadside, determine the effective state of the data, and perform target fusion based on the effective state, use prediction algorithms to predict, and improve the pole target fusion accuracy and stability in the lower area.

Benefits of technology

Through the combination of target prediction data and observation data, abnormal situations are filtered, target fusion accuracy and stability in the area under the roadside rod are improved, and the fusion link delay is reduced, thereby improving the real-timeness of the data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965277A_ABST
    Figure CN118965277A_ABST
Patent Text Reader

Abstract

The invention discloses a roadside target fusion method and device, electronic equipment and a computer program product, and the method comprises the steps: obtaining target observation data at a current moment and target prediction data at a previous moment, which are outputted by a roadside single pole; determining the effective state of the target observation data at the current moment and the effective state of the target prediction data at the previous moment; and according to the effective state of the target observation data at the current moment and the effective state of the target prediction data at the previous moment, performing target fusion of the roadside single pole to obtain a target fusion result of the roadside single pole at the current moment. The target track prediction technology is utilized, the target prediction data is provided for the roadside fusion system, the abnormal condition of the perception result of the camera under the roadside rod can be filtered out, the precision of the area target fusion under the roadside rod is improved, and the target prediction data and the target observation data are combined for use, so that the fusion precision is improved. And the stability of the target fusion of the area under the roadside rod is further improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of roadside target fusion, and in particular to a roadside target fusion method, device, electronic equipment, and computer program product. Background Art

[0002] A typical roadside fusion system will align and fuse the target perception results of each sensor based on sensors such as lidar, cameras, and millimeter-wave radar installed on the pole to achieve data output to the downstream.

[0003] The current fusion solution mainly integrates and aligns the perception targets of the LiDAR and the camera. However, since the perception range of the LiDAR generally only covers 35 to 150 meters from the road pole, the area under the pole, including the forward and reverse directions, is only covered by the fisheye camera within a range of 35*2=70 meters. At the same time, the edge area of ​​the fisheye camera is severely distorted, and there are problems such as dynamic occlusion, resulting in poor perception accuracy of the fisheye camera under the pole and many target anomalies, which in turn leads to poor overall fusion effect under the pole. For example, the target frequently has abnormal heading, lateral jumps, and discontinuous tracking, which affects the delivery of downstream data. Summary of the invention

[0004] The embodiments of the present application provide a roadside target fusion method, device, electronic device, and computer program product to improve the accuracy and stability of roadside target fusion.

[0005] The present application embodiment adopts the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a roadside target fusion method, wherein the roadside target fusion method includes:

[0007] Obtain the current target observation data and the previous target prediction data output by the roadside single pole;

[0008] Determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment;

[0009] According to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, target fusion of the roadside single pole is performed to obtain the target fusion result of the roadside single pole at the current moment.

[0010] Optionally, the target observation data at the current moment includes the target observation position at the current moment, and the obtaining of the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole includes:

[0011] Determine the corresponding roadside associated sensor according to the target observation position at the current moment;

[0012] Determine whether the target enters a preset roadside area according to the target observation position at the current moment and the corresponding roadside associated sensor;

[0013] When the target enters the preset roadside area, the target observation data at the current moment and the target prediction data at the previous moment are obtained.

[0014] Optionally, determining the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment includes:

[0015] Obtain target track data;

[0016] Determining whether a preset steady-state condition is met at the current moment according to the target track data;

[0017] When the preset steady-state condition is met, the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment are determined.

[0018] Optionally, determining the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment includes:

[0019] Determine the change amount of the target observation data at the current moment according to the target observation data at the current moment;

[0020] Determine the effective state of the target observation data at the current moment according to the change amount of the target observation data at the current moment;

[0021] When the valid state of the target observation data at the current moment is abnormal, determining the change amount of the target prediction data at the previous moment according to the target prediction data at the previous moment;

[0022] The validity state of the target prediction data at the previous moment is determined according to the change amount of the target prediction data at the previous moment.

[0023] Optionally, the target observation data includes a target observation heading, the target prediction data includes a target prediction heading, the target fusion result includes a fused target heading, and the target fusion of the roadside single pole is performed according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and the target fusion result of the roadside single pole at the current moment is obtained, which includes:

[0024] If the validity state of the target observation heading at the current moment is normal, taking the target observation heading at the current moment as the fused target heading at the current moment;

[0025] If the effective state of the target observed heading at the current moment is abnormal and the effective state of the target predicted heading at the previous moment is normal, the target predicted heading at the previous moment is used as the fused target heading at the current moment;

[0026] If the valid state of the target observed heading at the current moment and the valid state of the target predicted heading at the previous moment are both abnormal, the fused target heading at the previous moment is obtained as the fused target heading at the current moment.

[0027] Optionally, the target observation data includes a target observation lateral position, the target prediction data includes a target prediction lateral position, the target fusion result includes a fused target lateral position, and the target fusion of the roadside single pole is performed according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and the target fusion result of the roadside single pole at the current moment is obtained, which includes:

[0028] If the validity state of the target observed lateral position at the current moment is normal, taking the target observed lateral position at the current moment as the fused target lateral position at the current moment;

[0029] If the effective state of the target observed lateral position at the current moment is abnormal and the effective state of the target predicted lateral position at the previous moment is normal, then the fused target lateral position at the current moment is calculated according to the observation direction at the current moment and the predicted modulus at the previous moment;

[0030] If the valid state of the target observed lateral position at the current moment and the valid state of the target predicted lateral position at the previous moment are both abnormal, the fused target lateral position at the current moment is calculated based on the observation direction at the current moment and the default modulus.

[0031] Optionally, after performing target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment to obtain the target fusion result of the roadside single pole at the current moment, the roadside target fusion method further includes:

[0032] According to the target fusion result at the current moment, a target trajectory prediction algorithm is used to perform prediction to obtain a first prediction result at the current moment, where the first prediction result refers to a result obtained by predicting the time difference between adjacent moments;

[0033] When the target observation data at the next moment is obtained, target fusion is performed according to the first prediction result at the current moment and the target observation data at the next moment to obtain the target fusion result at the next moment.

[0034] Optionally, after performing target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment to obtain the target fusion result of the roadside single pole at the current moment, the roadside target fusion method further includes:

[0035] According to the target fusion result at the current moment, a target trajectory prediction algorithm is used to perform prediction to obtain a second prediction result at the current moment, where the second prediction result refers to a result obtained based on the fusion delay time prediction;

[0036] The second prediction result at the current moment is sent to a downstream module.

[0037] In a second aspect, an embodiment of the present application further provides a roadside target fusion device, wherein the roadside target fusion device comprises:

[0038] An acquisition unit, used to acquire the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole;

[0039] A determination unit, used to determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment;

[0040] The fusion unit is used to perform target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, so as to obtain the target fusion result of the roadside single pole at the current moment.

[0041] In a third aspect, an embodiment of the present application further provides an electronic device, including:

[0042] Processor; and

[0043] A memory arranged to store computer executable instructions, which, when executed, cause the processor to perform any of the roadside object fusion methods described above.

[0044] In a fourth aspect, an embodiment of the present application further provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements any of the aforementioned roadside target fusion methods.

[0045] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: the roadside target fusion method in the embodiments of the present application first obtains the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole; then determines the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment; finally, according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, the target fusion of the roadside single pole is performed to obtain the target fusion result of the roadside single pole at the current moment. The roadside target fusion method in the embodiments of the present application utilizes the target trajectory prediction technology to provide target prediction data for the roadside fusion system, which can filter out abnormal situations of the camera perception results under the roadside pole, improve the accuracy of target fusion in the area under the roadside pole, and the target prediction data is used in combination with the target observation data to further improve the stability of target fusion in the area under the roadside pole. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0047] Figure 1 A schematic diagram of a flow chart of a roadside target fusion method in an embodiment of the present application;

[0048] Figure 2 This is a schematic diagram of an overall process of roadside target fusion in an embodiment of the present application;

[0049] Figure 3 This is a schematic diagram of a target fusion link in an embodiment of the present application;

[0050] Figure 4 This is a schematic diagram of the structure of a roadside target fusion device in an embodiment of the present application;

[0051] Figure 5 This is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0053] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0054] The present application embodiment provides a roadside target fusion method, such as Figure 1 As shown, a schematic flow chart of a roadside target fusion method in an embodiment of the present application is provided, and the roadside target fusion method at least includes the following steps S110 to S130:

[0055] Step S110, obtaining the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole.

[0056] The roadside target fusion method of the embodiment of the present application can be executed by a roadside edge computing (MEC) module deployed on any roadside pole. When fusing the targets perceived by the roadside pole, it is necessary to first obtain the target observation data at the current moment output by the roadside pole. The target observation data at the current moment refers to the current target information perceived by the perception module on the roadside pole, such as a lidar, camera, etc., which may include, for example, the perceived target position, type, heading and other information.

[0057] In addition, it is also necessary to obtain the target prediction data of the previous moment. The target prediction data of the previous moment refers to the result obtained by predicting the target information of the current moment at the previous moment, including the predicted position, heading and other information of the target at the current moment. It can be based on the target fusion result of the previous moment, combined with a certain prediction algorithm or rule prediction, as long as the accuracy can meet the downstream delivery requirements. For example, the trajectory prediction algorithm currently used in the vehicle side can be used, and the average prediction accuracy is within 0.2 meters to meet the downstream delivery needs. Of course, the specific prediction method can be flexibly selected by technicians in this field according to actual needs, and no specific limitation is made here.

[0058] Step S120, determining the validity status of the target observation data at the current moment and the validity status of the target prediction data at the previous moment.

[0059] In the scenario of single-pole target fusion on the roadside, since a certain area under the pole belongs to the blind spot of the lidar, it is impossible to obtain complete and reliable lidar observation data, and only fisheye camera observation data can be obtained. Simply relying on fisheye camera observation data for single-pole target fusion will result in poor perception accuracy due to severe distortion of the edge area perceived by the fisheye camera and dynamic occlusion, which will cause the fusion system to be extremely unstable in the area under the pole, such as frequent abnormal heading, lateral jumps, discontinuous tracking, etc.

[0060] Based on this, in order to improve and enhance the accuracy and stability of target fusion in the area under the pole in the above-mentioned roadside single-pole target fusion scenario, the embodiment of the present application further introduces target prediction information, and determines the optimal target fusion strategy currently adopted by comprehensively considering the valid status of the target observation data at the current moment and the valid status of the target prediction data at the previous moment, thereby improving the accuracy and stability of target fusion in the roadside single-pole fusion scenario.

[0061] The judgment of the valid state of data mainly measures whether the accuracy of the data meets the requirements of fusion accuracy. The specific judgment method can be flexibly set by technicians in this field based on the needs of the actual scenario, and no specific limitation is made here.

[0062] Step S130, performing target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, to obtain the target fusion result of the roadside single pole at the current moment.

[0063] Based on the valid status of the target observation data at the current moment and the valid status of the target prediction data at the previous moment determined in the above steps, the target observation data at the current moment and the target prediction data at the previous moment can be more fully utilized for target fusion. The target observation data and the target prediction data complement each other and no longer rely solely on the target observation data, thereby improving the accuracy of target fusion.

[0064] The roadside target fusion method of the embodiment of the present application utilizes target trajectory prediction technology to provide target prediction data for the roadside fusion system, which can filter out abnormal situations in the perception results of the camera under the roadside pole, improve the accuracy of target fusion in the area under the roadside pole, and the target prediction data is used in combination with the target observation data to further improve the stability of target fusion in the area under the roadside pole.

[0065] In some embodiments of the present application, the target observation data at the current moment includes the target observation position at the current moment, and obtaining the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole includes: determining the corresponding roadside associated sensor according to the target observation position at the current moment; determining whether the target enters a preset roadside area according to the target observation position at the current moment and the corresponding roadside associated sensor; and when the target enters the preset roadside area, obtaining the target observation data at the current moment and the target prediction data at the previous moment.

[0066] As mentioned above, the target fusion strategy implemented by the embodiment of the present application through the additional target prediction data is mainly to compensate for the problem of low target fusion accuracy caused by only inaccurate fisheye camera observation data in the area under the pole. Therefore, before triggering the above-mentioned target fusion logic, the embodiment of the present application can first determine whether the target has entered a specific area under the pole, that is, the blind spot of the laser radar. The actual range of the area under the pole can be determined based on the actual deployment of the roadside sensors.

[0067] When making a specific judgment, on the one hand, the distance from the target to the nearest road pole ahead can be calculated based on the observed target position, and the distance from the target to the road pole ahead can be compared with the range of the area under the pole set in advance to determine whether it has entered the area under the pole. On the other hand, since the perception ranges of roadside sensors deployed on different road sections are also different, the actual blind area ranges corresponding to the roadside sensors may also be different. Therefore, the unique identification information of the sensor associated with the current position of the target, such as the slot position of the sensor, can also be combined to determine the blind area range of the current sensor, so as to compare the distance from the above target to the nearest road pole ahead with the corresponding sensor blind area range to determine whether it has entered the blind area.

[0068] In order to further improve the accuracy of the judgment, the output of the sensor observation data can be further continuously monitored based on the above judgment. For example, if the laser radar observation data is not obtained for a period of time, it means that the laser radar has entered the observation blind spot.

[0069] In some embodiments of the present application, determining the valid status of the target observation data at the current moment and the valid status of the target prediction data at the previous moment includes: acquiring target track data; determining whether a preset steady-state condition is met at the current moment based on the target track data; and when the preset steady-state condition is met, determining the valid status of the target observation data at the current moment and the valid status of the target prediction data at the previous moment.

[0070] The target track data refers to the historical track of the target's navigation, which can be understood as the track point data passed by the target. When obtaining the target track data, the unique identification code UUID of the predicted target can be used to complete the pairing with the target track data and find the one-to-one correspondence between the target prediction data and the track data.

[0071] Considering that the target track has a new construction process, when the track is first constructed, the tracking distance is relatively short and the data volume is insufficient, which often does not meet the requirements for subsequent target fusion based on target observation data and target prediction data. In addition, in the roadside single-pole target fusion scenario of the embodiment of the present application, the targets focused on are mainly various types of vehicle targets, because the movement of these targets is relatively stable and the trajectory is easier to predict.

[0072] Based on this, the embodiment of the present application can first perform the above-mentioned steady-state condition judgment, including the judgment of tracking distance and target type, etc. before performing the effective state judgment. If the tracking distance is greater than a certain tracking distance threshold and the target type also belongs to the pre-set vehicle target type, it is considered that the steady-state condition requirements are currently met, and the subsequent effective state judgment and target fusion processing logic of the target observation data and target prediction data can be performed. Otherwise, the steady-state condition is not met and the observation data is still used for fusion.

[0073] In some embodiments of the present application, determining the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment includes: determining the change in the target observation data at the current moment based on the target observation data at the current moment; determining the valid state of the target observation data at the current moment based on the change in the target observation data at the current moment; when the valid state of the target observation data at the current moment is abnormal, determining the change in the target prediction data at the previous moment based on the target prediction data at the previous moment; determining the valid state of the target prediction data at the previous moment based on the change in the target prediction data at the previous moment.

[0074] The change in the target observation data at the current moment refers to the change in the target observation data between the current moment and the previous moment, for example, it may include the change in the angle of the target observation heading and the change in the target observation lateral position. Under normal circumstances, the change in heading and position between two adjacent frames will not be large. Therefore, if the change in the angle of the target observation heading is less than the corresponding threshold requirement, the target observation heading at the current moment is considered to be normal data, otherwise it is abnormal data. If the change in the target observation lateral position is less than the corresponding threshold requirement, the target observation lateral position at the current moment is considered to be normal data, otherwise it is abnormal data.

[0075] If the target observation data at the current moment is normal data, it means that the target observation data at the current moment meets the accuracy requirements of subsequent target fusion and can be used for subsequent fusion processing. In this case, the valid state of the target prediction data at the previous moment can no longer be judged. However, if the target observation data at the current moment is abnormal data, it means that the target observation data at the current moment is inaccurate and is not suitable as the basis for subsequent target fusion. At this time, the valid state of the target prediction data at the previous moment can be further judged, in the hope that the target fusion can be performed with the help of the target prediction data at the previous moment, thereby ensuring the accuracy and stability of target fusion.

[0076] Similar to the judgment logic of the valid state of the target observation data at the current moment, the change amount of the target prediction data at the previous moment refers to the change amount between the target prediction data at the previous moment and the moment before that. If the angle change of the target prediction heading is less than the corresponding threshold requirement, the target prediction heading at the previous moment is considered to be normal data, otherwise it is abnormal data. If the change of the target prediction lateral position is less than the corresponding threshold requirement, the target prediction lateral position at the previous moment is considered to be normal data, otherwise it is abnormal data.

[0077] In some embodiments of the present application, the target observation data includes a target observation heading, the target prediction data includes a target prediction heading, and the target fusion result includes a fused target heading. The target fusion of the roadside single pole is performed according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and the target fusion result of the roadside single pole at the current moment is obtained, including: if the valid state of the target observation heading at the current moment is normal, then the target observation heading at the current moment is used as the fused target heading at the current moment; if the valid state of the target observation heading at the current moment is abnormal and the valid state of the target prediction heading at the previous moment is normal, then the target prediction heading at the previous moment is used as the fused target heading at the current moment; if the valid state of the target observation heading at the current moment and the valid state of the target prediction heading at the previous moment are both abnormal, then the fused target heading at the previous moment is obtained as the fused target heading at the current moment.

[0078] When the target fusion of the roadside single pole is performed based on the effective state of the target observation heading at the current moment and the effective state of the target prediction heading at the previous moment, different target fusion strategies can be adopted according to the specific effective conditions of the data, mainly including the following three situations:

[0079] 1) The effective state of the target observation heading at the current moment is normal: Since the target observation heading is directly sensed by the sensor, when the target observation heading is normal data, the target observation heading data is relatively more credible and reliable. Therefore, in this case, the target observation heading at the current moment can be directly used as the fused target heading at the current moment;

[0080] 2) The effective state of the target observation heading at the current moment is abnormal, and the effective state of the target prediction heading at the previous moment is normal: If the target observation heading at the current moment is invalid, and the target prediction heading at this moment meets the accuracy requirement, the target prediction heading at the previous moment can be directly used as the fused target heading at the current moment;

[0081] 3) The valid state of the target observed heading at the current moment and the valid state of the target predicted heading at the previous moment are both abnormal: At this time, the target observed heading and the target predicted heading data are both invalid. In order to avoid jumps in the fused target heading, the fused target heading data at the previous moment can be obtained as the fused target heading at the current moment, thereby ensuring the stability of the target heading fusion.

[0082] In some embodiments of the present application, the target observation data includes the target observation lateral position, the target prediction data includes the target predicted lateral position, and the target fusion result includes the fused target lateral position. The target fusion of the roadside single pole is performed according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and the target fusion result of the roadside single pole at the current moment is obtained, including: if the valid state of the target observation lateral position at the current moment is normal, then the target observation lateral position at the current moment is used as the fused target lateral position at the current moment; if the valid state of the target observation lateral position at the current moment is abnormal and the valid state of the target predicted lateral position at the previous moment is normal, then the fused target lateral position at the current moment is calculated according to the observation direction at the current moment and the predicted modulus at the previous moment; if the valid state of the target observation lateral position at the current moment and the valid state of the target predicted lateral position at the previous moment are both abnormal, then the fused target lateral position at the current moment is calculated according to the observation direction at the current moment and the default modulus.

[0083] When the target fusion of the roadside single pole is performed based on the effective state of the target observed lateral position at the current moment and the effective state of the target predicted heading at the previous moment, different target fusion strategies can be adopted according to the specific effective conditions of the data, mainly including the following three situations:

[0084] 1) The effective state of the target observed lateral position at the current moment is normal: Since the target observed lateral position is sensed by the sensor, when the target observed lateral position is normal data, the target observed lateral position data is relatively more credible and reliable. Therefore, in this case, the target observed lateral position at the current moment can be directly used as the fused target lateral position at the current moment;

[0085] 2) The effective state of the target observed lateral position at the current moment is abnormal, and the effective state of the target predicted lateral position at the previous moment is normal: If the target observed lateral position at the current moment is invalid, but the target predicted lateral position at the previous moment is available, and the observation direction at the current moment is valid, the fused target lateral position at the current moment can be calculated based on the observation direction at the current moment and the predicted modulus at the previous moment;

[0086] 3) The valid state of the target observed lateral position at the current moment and the valid state of the target predicted lateral position at the previous moment are both abnormal: At this time, the target observed lateral position and the target predicted lateral position are both invalid. In order to avoid jumps in the fused target lateral position, when the observation direction at the current moment is valid, the fused target lateral position at the current moment can be calculated based on the observation direction at the current moment and the fixed modulus set in advance, that is, the lateral position is smooth along the observation direction, thereby ensuring the stability of the target lateral position fusion.

[0087] It should be noted that the main considerations for target fusion processing in the aforementioned embodiments are the target's heading and lateral position information. This is because in the scenario of single-pole fusion on the roadside, target abnormalities in the area under the pole are mainly due to large deviations in the target's heading and lateral position information. This is a design based on actual scenario conditions.

[0088] In some embodiments of the present application, after performing target fusion of a roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment to obtain the target fusion result of the roadside single pole at the current moment, the roadside target fusion method further includes: based on the target fusion result at the current moment, using a target trajectory prediction algorithm to perform prediction to obtain a first prediction result at the current moment, wherein the first prediction result refers to a result obtained by predicting based on the time difference between adjacent moments; when the target observation data at the next moment is obtained, performing target fusion based on the first prediction result at the current moment and the target observation data at the next moment to obtain the target fusion result at the next moment.

[0089] After obtaining the target fusion result at the current moment, the target fusion result at the current moment can also be used as the basis for predicting the target information at the next moment, because the target fusion result at the current moment is the fused result that meets the accuracy requirements. Therefore, based on the target fusion result at the current moment, the target trajectory prediction algorithm can be used for prediction to obtain the first prediction result at the current moment. The prediction logic here is mainly based on the time difference between adjacent moments. The first prediction result at the current moment obtained in this way refers to the information of the target at the next moment predicted at the current moment.

[0090] When the target fusion proceeds to the next moment, the target observation data of the next moment can be obtained, and then combined with the first prediction result of the current moment for target fusion. The specific logic is the same as the logic of fusing the target observation data of the current moment and the target prediction data of the previous moment in the aforementioned embodiments, and will not be repeated here.

[0091] In some embodiments of the present application, after performing target fusion of a roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment to obtain the target fusion result of the roadside single pole at the current moment, the roadside target fusion method further includes: based on the target fusion result at the current moment, using a target trajectory prediction algorithm to perform prediction to obtain a second prediction result at the current moment, the second prediction result refers to a result obtained based on the fusion delay time prediction; and sending the second prediction result at the current moment to a downstream module.

[0092] Taking into account the large delay of the existing roadside perception fusion link system, the delay here refers to the total time required from the exposure of sensors such as cameras to a series of processing such as the driving module, perception module, single-pole fusion module, etc.

[0093] Based on this, the embodiment of the present application further uses the target prediction data to perform delay compensation, that is, based on the target fusion result at the current moment, the target trajectory prediction algorithm is also used for prediction to obtain the second prediction result at the current moment. Different from the prediction logic in the above embodiment, the predicted time length in the embodiment of the present application is mainly set based on factors such as the fusion delay time and the prediction accuracy of the prediction algorithm.

[0094] For example, the delay time of the roadside perception fusion link system is about 250~300ms. If the prediction accuracy requirements are met, the prediction time length can be set to 200ms. Then the second prediction result at the current moment refers to the target information predicted 200ms later at the current moment, and then the target information predicted 200ms later at the current moment is directly sent to the downstream module such as the multi-pole fusion module. In this way, the data received by the downstream module is equivalent to the data that compensates the target data delay time by 200ms. This delay processing method can greatly reduce the delay time with little or no loss of accuracy, and improve the real-time nature of the data.

[0095] In some embodiments of the present application, the roadside target fusion method also includes: when the target observation data at the current moment cannot be obtained, determining whether the preset steady-state condition is met at the current moment based on the target track data; when the preset steady-state condition is met, determining the validity status of the target prediction data at the previous moment; when the validity status of the target prediction data at the previous moment is valid, determining the target fusion result of the roadside single pole at the current moment based on the target prediction data at the previous moment.

[0096] In some special scenarios, if no observation data can be obtained in the area under the pole, such as the boundary area of ​​the edge of the camera field of view, the target is blocked or the large target is truncated, etc., causing no sensor to observe the target, certain strategies can be adopted to continue to supplement the target trajectory, thereby improving the continuity of target tracking.

[0097] Specifically, it is possible to first determine whether the steady-state condition is currently met. The steady-state condition here can be determined based on the current tracking distance, target type, target speed, etc. The determination of the tracking distance and target type can refer to the description in the aforementioned embodiments. The determination of the target speed is mainly because if the target movement speed is slow, it will lead to a large subsequent prediction deviation. Therefore, the target speed also needs to be greater than a certain threshold requirement to meet the steady-state condition.

[0098] When the preset steady-state conditions are met, the validity status of the target prediction data at the previous moment is further determined. If valid, the target fusion result of the roadside single pole at the current moment can be determined based on the target prediction data at the previous moment. Here, the calculation of the target lateral position and longitudinal position is mainly carried out. The lateral position can directly adopt the target predicted lateral position at the previous moment, and the longitudinal position can be multiplied by a custom coefficient K based on the target predicted longitudinal position at the previous moment, that is, the target predicted longitudinal position at the previous moment is appropriately considered instead of being fully adopted, because the prediction of the lateral position is more accurate than that of the longitudinal position.

[0099] In order to facilitate the understanding of the above embodiments, Figure 2 As shown, a schematic diagram of the overall process of roadside target fusion in an embodiment of the present application is provided. First, the target track data is obtained, and it is determined whether the steady-state condition is met at the current moment according to the target heading data. If not, the target fusion is performed according to the original logic. If it is met, the target observation data at the current moment and the target prediction data at the previous moment are further obtained, so that different target fusion strategies are adopted according to the effective status of the target observation data at the current moment and the target prediction data at the previous moment, so as to obtain the target fusion result at the current moment.

[0100] After obtaining the target fusion result at the current moment, the target prediction at the next moment can be performed based on the target fusion result at the current moment, and the first prediction result at the current moment can be obtained as the basis for the target fusion at the next moment. In addition, the target prediction of the fusion delay time can be performed based on the target fusion result at the current moment, and the second prediction result at the current moment can be obtained. The second prediction result at the current moment is sent to the downstream module, thereby reducing the delay time of the fusion link.

[0101] like Figure 3As shown, a schematic diagram of a target fusion link in an embodiment of the present application is provided. First, at time k-1, the input of the target fusion at time k-1 is the target observation data at time k-1 and the target prediction data at time k-2, and the target fusion result at time k-1 is obtained. Then, on the one hand, the target fusion result at time k-1 is input into the target prediction algorithm for target prediction at the next moment, and the first prediction result at time k-1 is obtained as one of the inputs of the target fusion at time k. On the other hand, the target fusion result at time k-1 is input into the target prediction algorithm for fusion delay time prediction, and the second prediction result at time k-1 is obtained. The second prediction result at time k-1 is directly sent to the downstream module.

[0102] Similarly, when it reaches time k, the input of the target fusion at time k is the target observation data at time k and the target prediction data at time k-1, and the target fusion result at time k is obtained. Then, on the one hand, the target fusion result at time k is input into the target prediction algorithm for target prediction at the next time, and the first prediction result at time k is obtained as one of the inputs of the target fusion at time k+1. On the other hand, the target fusion result at time k is input into the target prediction algorithm for fusion delay time prediction, and the second prediction result at time k is obtained. The second prediction result at time k is directly sent to the downstream module.

[0103] By analogy, a data loop of prediction -> fusion -> prediction is formed, thus optimizing the fusion effect and link delay.

[0104] In summary, the roadside target fusion method of the present application has achieved at least the following technical effects:

[0105] 1) The target trajectory prediction technology is used to provide target prediction data for the roadside fusion system, which can filter out abnormalities in the perception results of the camera under the roadside pole, improve the accuracy of target fusion in the area under the roadside pole, and combine the target prediction data with the target observation data to further improve the stability of target fusion in the area under the roadside pole;

[0106] 2) Taking into account the delay time of the fusion link and the accuracy of the prediction algorithm, the target trajectory prediction technology is used to predict the target data forward for a period of time, which can greatly reduce the delay time with little or no loss of accuracy and improve the real-time performance of the data.

[0107] The present application embodiment also provides a roadside target fusion device 400, such as Figure 4 As shown, a schematic diagram of the structure of a roadside target fusion device in an embodiment of the present application is provided. The roadside target fusion device 400 includes: an acquisition unit 410, a determination unit 420 and a fusion unit 430, wherein:

[0108] An acquisition unit 410 is used to acquire the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole;

[0109] A determination unit 420, configured to determine a valid state of the target observation data at the current moment and a valid state of the target prediction data at the previous moment;

[0110] The fusion unit 430 is used to perform target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, so as to obtain the target fusion result of the roadside single pole at the current moment.

[0111] In some embodiments of the present application, the target track data at the current moment includes the target position at the current moment, and the acquisition unit 410 is specifically used to: determine the corresponding roadside associated sensor according to the target observation position at the current moment; determine whether the target enters a preset roadside area according to the target observation position at the current moment and the corresponding roadside associated sensor; when the target enters the preset roadside area, obtain the target observation data at the current moment and the target prediction data at the previous moment.

[0112] In some embodiments of the present application, the determination unit 420 is specifically used to: obtain target track data; determine whether the preset steady-state condition is met at the current moment based on the target track data; and when the preset steady-state condition is met, determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment.

[0113] In some embodiments of the present application, the determination unit 420 is specifically used to: determine the change amount of the target observation data at the current moment based on the target observation data at the current moment; determine the valid state of the target observation data at the current moment based on the change amount of the target observation data at the current moment; when the valid state of the target observation data at the current moment is abnormal, determine the change amount of the target prediction data at the previous moment based on the target prediction data at the previous moment; determine the valid state of the target prediction data at the previous moment based on the change amount of the target prediction data at the previous moment.

[0114] In some embodiments of the present application, the target observation data includes the target observation heading, the target prediction data includes the target prediction heading, and the target fusion result includes the fused target heading, and the fusion unit 430 is specifically used to: if the valid state of the target observation heading at the current moment is normal, then use the target observation heading at the current moment as the fused target heading at the current moment; if the valid state of the target observation heading at the current moment is abnormal and the valid state of the target prediction heading at the previous moment is normal, then use the target prediction heading at the previous moment as the fused target heading at the current moment; if the valid state of the target observation heading at the current moment and the valid state of the target prediction heading at the previous moment are both abnormal, then obtain the fused target heading at the previous moment as the fused target heading at the current moment.

[0115] In some embodiments of the present application, the target observation data includes the target observation lateral position, the target prediction data includes the target predicted lateral position, and the target fusion result includes the fused target lateral position, and the fusion unit 430 is specifically used to: if the effective state of the target observation lateral position at the current moment is normal, then use the target observation lateral position at the current moment as the fused target lateral position at the current moment; if the effective state of the target observation lateral position at the current moment is abnormal and the effective state of the target predicted lateral position at the previous moment is normal, then calculate the fused target lateral position at the current moment according to the observation direction at the current moment and the predicted modulus at the previous moment; if the effective state of the target observation lateral position at the current moment and the effective state of the target predicted lateral position at the previous moment are both abnormal, then calculate the fused target lateral position at the current moment according to the observation direction at the current moment and the default modulus.

[0116] In some embodiments of the present application, the roadside target fusion device also includes: a first prediction unit, which is used to perform target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and after obtaining the target fusion result of the roadside single pole at the current moment, perform prediction based on the target fusion result at the current moment using a target trajectory prediction algorithm to obtain a first prediction result at the current moment, and the first prediction result refers to a result obtained by predicting based on the time difference between adjacent moments; the fusion unit 430 is also used to perform target fusion based on the first prediction result at the current moment and the target observation data at the next moment when the target observation data at the next moment is obtained, to obtain the target fusion result at the next moment.

[0117] In some embodiments of the present application, the roadside target fusion device 400 also includes: a second prediction unit, which is used to perform target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and after obtaining the target fusion result of the roadside single pole at the current moment, perform prediction based on the target fusion result at the current moment using a target trajectory prediction algorithm to obtain a second prediction result at the current moment, wherein the second prediction result refers to a result obtained based on the fusion delay time prediction; a sending unit, which is used to send the second prediction result at the current moment to a downstream module.

[0118] It can be understood that the above-mentioned roadside target fusion device can implement the various steps of the roadside target fusion method provided in the aforementioned embodiment, and the relevant explanations about the roadside target fusion method are applicable to the roadside target fusion device and will not be repeated here.

[0119] Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 5 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0120] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0121] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0122] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a roadside target fusion device at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations:

[0123] Obtain the current target observation data and the previous target prediction data output by the roadside single pole;

[0124] Determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment;

[0125] According to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, target fusion of the roadside single pole is performed to obtain the target fusion result of the roadside single pole at the current moment.

[0126] The above application Figure 1 The method performed by the roadside target fusion device disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0127] The present application also provides a computer program product, which stores one or more programs, wherein the one or more programs include instructions, which, when executed by an electronic device including multiple application programs, enable the electronic device to execute Figure 1 The method performed by the roadside target fusion device in the illustrated embodiment is specifically used to perform:

[0128] Obtain the current target observation data and the previous target prediction data output by the roadside single pole;

[0129] Determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment;

[0130] According to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, target fusion of the roadside single pole is performed to obtain the target fusion result of the roadside single pole at the current moment.

[0131] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0132] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0133] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0134] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0135] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0136] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0137] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0138] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0139] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0140] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A roadside object fusion method, in, The roadside target fusion method comprises: Obtain the current target observation data and the previous target prediction data output by the roadside single pole; Determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment; According to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, target fusion of the roadside single pole is performed to obtain the target fusion result of the roadside single pole at the current moment.

2. The roadside target fusion method according to claim 1, in, The target observation data at the current moment includes the target observation position at the current moment, and the target observation data at the current moment output by the roadside single pole and the target prediction data at the previous moment are obtained, including: Determine the corresponding roadside associated sensor according to the target observation position at the current moment; Determine whether the target enters a preset roadside area according to the target observation position at the current moment and the corresponding roadside associated sensor; When the target enters the preset roadside area, the target observation data at the current moment and the target prediction data at the previous moment are obtained.

3. The roadside target fusion method according to claim 1, in, Determining the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment includes: Obtain target track data; Determining whether a preset steady-state condition is met at the current moment according to the target track data; When the preset steady-state condition is met, the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment are determined.

4. The roadside target fusion method according to claim 1 or 3, in, Determining the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment includes: Determine the change amount of the target observation data at the current moment according to the target observation data at the current moment; Determine the effective state of the target observation data at the current moment according to the change amount of the target observation data at the current moment; When the valid state of the target observation data at the current moment is abnormal, determining the change amount of the target prediction data at the previous moment according to the target prediction data at the previous moment; The validity state of the target prediction data at the previous moment is determined according to the change amount of the target prediction data at the previous moment.

5. The roadside target fusion method as claimed in claim 1, in, The target observation data includes the target observation heading, the target prediction data includes the target prediction heading, the target fusion result includes the fused target heading, and the target fusion of the roadside single pole is performed according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, and the target fusion result of the roadside single pole at the current moment is obtained, which includes: If the validity state of the target observation heading at the current moment is normal, taking the target observation heading at the current moment as the fused target heading at the current moment; If the effective state of the target observed heading at the current moment is abnormal and the effective state of the target predicted heading at the previous moment is normal, the target predicted heading at the previous moment is used as the fused target heading at the current moment; If the valid state of the target observed heading at the current moment and the valid state of the target predicted heading at the previous moment are both abnormal, the fused target heading at the previous moment is obtained as the fused target heading at the current moment.

6. The roadside target fusion method according to claim 1, in, The target observation data includes a target observation lateral position, the target prediction data includes a target prediction lateral position, the target fusion result includes a fused target lateral position, and the target fusion result of the roadside single pole at the current moment is obtained by performing target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment. If the validity state of the target observed lateral position at the current moment is normal, taking the target observed lateral position at the current moment as the fused target lateral position at the current moment; If the effective state of the target observed lateral position at the current moment is abnormal and the effective state of the target predicted lateral position at the previous moment is normal, then the fused target lateral position at the current moment is calculated according to the observation direction at the current moment and the predicted modulus at the previous moment; If the valid state of the target observed lateral position at the current moment and the valid state of the target predicted lateral position at the previous moment are both abnormal, the fused target lateral position at the current moment is calculated based on the observation direction at the current moment and the default modulus.

7. The roadside target fusion method according to claim 1, in, After performing target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment to obtain the target fusion result of the roadside single pole at the current moment, the roadside target fusion method further includes: According to the target fusion result at the current moment, a target trajectory prediction algorithm is used to perform prediction to obtain a first prediction result at the current moment, where the first prediction result refers to a result obtained by predicting the time difference between adjacent moments; When the target observation data at the next moment is obtained, target fusion is performed according to the first prediction result at the current moment and the target observation data at the next moment to obtain the target fusion result at the next moment.

8. The roadside target fusion method as claimed in claim 1, in, After performing target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment to obtain the target fusion result of the roadside single pole at the current moment, the roadside target fusion method further includes: According to the target fusion result at the current moment, a target trajectory prediction algorithm is used to perform prediction to obtain a second prediction result at the current moment, where the second prediction result refers to a result obtained based on the fusion delay time prediction; The second prediction result at the current moment is sent to a downstream module.

9. A roadside target fusion device, in, The roadside target fusion device comprises: An acquisition unit, used to acquire the target observation data at the current moment and the target prediction data at the previous moment output by the roadside single pole; A determination unit, used to determine the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment; The fusion unit is used to perform target fusion of the roadside single pole according to the valid state of the target observation data at the current moment and the valid state of the target prediction data at the previous moment, so as to obtain the target fusion result of the roadside single pole at the current moment.

10. An electronic device, include: processor; as well as A memory arranged to store computer executable instructions, wherein when the executable instructions are executed, the processor executes the roadside target fusion method according to any one of claims 1 to 8.

11. A computer program product, comprising a computer program or instructions, wherein when the computer program or instructions are executed by a processor, the roadside target fusion method according to any one of claims 1 to 8 is implemented.

Citation Information

Patent Citations

  • Data fusion method, device and system based on multiple roadside sensors, and medium

    CN114065876A

  • Fusion positioning method and device of vehicle, electronic equipment and storage medium

    CN114777796A

  • Fusion positioning method and device of automatic driving vehicle and electronic equipment

    CN114894214A

  • Multi-sensor target fusion method and device, electronic equipment and storage medium

    CN115494494A

  • Target fusion method and device, electronic equipment and computer readable storage medium

    CN115719375A