Data verification method and device, electronic equipment and autonomous vehicle

Abstract

The present disclosure provides a data verification method and device, electronic equipment and autonomous vehicle, relating to the technical field of autonomous driving, and especially relating to the application fields of obstacle identification, obstacle tracking, sensor data association, sensor data fusion and the like. The specific implementation scheme is as follows: acquiring current frame data collected by a first target sensor in a plurality of sensors; wherein the plurality of sensors are installed on a target object; taking each sensor in the plurality of sensors except the first target sensor as a second target sensor to acquire a plurality of reference data for representing the similarity between the current frame data and historical frame data collected by the second target sensor; wherein the historical frame data is derived from global obstacle tracking information of the target object; fusing the plurality of reference data to obtain a verification result for the current frame data; wherein the verification result is used to represent whether the current frame data can be associated with the historical frame data.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of autonomous driving, in particular to the application fields of obstacle identification, obstacle tracking, sensor data association, sensor data fusion, etc., and specifically relates to a data verification method and device, an electronic device, and an autonomous driving vehicle. BACKGROUND

[0002] In the technical field of autonomous driving, sensor data association is a crucial step, which aims to improve the ability of autonomous driving systems to identify and track obstacles, so as to ensure that the autonomous driving system can make relatively accurate driving decisions. SUMMARY

[0003] The present disclosure provides a data verification method and device, an electronic device, and an autonomous driving vehicle.

[0004] According to a first aspect of the present disclosure, a data verification method is provided, comprising:

[0005] obtaining current frame data collected by a first target sensor in a plurality of sensors; wherein the plurality of sensors are installed on a target object;

[0006] respectively taking each sensor in the plurality of sensors except the first target sensor as a second target sensor to obtain a plurality of reference data for representing the similarity between the current frame data and historical frame data collected by the second target sensor; wherein the historical frame data is derived from global obstacle tracking information of the target object;

[0007] fusing the plurality of reference data to obtain a verification result for the current frame data; wherein the verification result is used to represent whether the current frame data can be associated with the historical frame data.

[0008] According to a second aspect of the present disclosure, a data verification device is provided, comprising:

[0009] a first data acquisition unit configured to obtain current frame data collected by a first target sensor in a plurality of sensors; wherein the plurality of sensors are installed on a target object;

[0010] a second data acquisition unit configured to respectively take each sensor in the plurality of sensors except the first target sensor as a second target sensor to obtain a plurality of reference data for representing the similarity between the current frame data and historical frame data collected by the second target sensor; wherein the historical frame data is derived from global obstacle tracking information of the target object;

[0011] a data fusion unit configured to fuse the plurality of reference data to obtain a verification result for the current frame data; wherein the verification result is used to represent whether the current frame data can be associated with the historical frame data.

[0012] According to a third aspect of the present disclosure, an electronic device is provided, comprising:

[0013] at least one processor;

[0014] a memory connected with the at least one processor in communication;

[0015] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method provided in the first aspect of the present disclosure.

[0016] According to a fourth aspect of the present disclosure, an autonomous vehicle is provided, comprising the electronic device provided in the third aspect of the present disclosure.

[0017] According to a fifth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to make the computer perform the method provided in the first aspect of the present disclosure.

[0018] According to a sixth aspect of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the method provided in the first aspect of the present disclosure.

[0019] The present disclosure can improve the accuracy of sensor data association.

[0020] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0021] The accompanying drawings are used to better understand the present scheme and do not limit the present disclosure. Among them:

[0022] Figure 1 A flowchart of a data verification method provided by an embodiment of the present disclosure;

[0023] Figure 2 A schematic diagram of the probability distribution of the width difference between the first bounding box and the second bounding box provided by an embodiment of the present disclosure;

[0024] Figure 3 A schematic diagram of the probability distribution of the height difference between the first bounding box and the second bounding box provided by an embodiment of the present disclosure;

[0025] Figure 4 A schematic diagram of the probability distribution of the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension provided by an embodiment of the present disclosure;

[0026] Figure 5 A probability partial case diagram of a first overflow amount of a plurality of feature points in a first direction relative to a third bounding box is provided for an embodiment of the present disclosure;

[0027] Figure 6 Another probability partial case diagram of a first overflow amount of a plurality of feature points in a first direction relative to a third bounding box is provided for an embodiment of the present disclosure;

[0028] Figure 7 A probability partial case diagram of a second overflow amount of a plurality of feature points in a second direction relative to a third bounding box is provided for an embodiment of the present disclosure;

[0029] Figure 8 Another probability partial case diagram of a second overflow amount of a plurality of feature points in a second direction relative to a third bounding box is provided for an embodiment of the present disclosure;

[0030] Figure 9 A probability distribution case diagram of reference data used to represent similarity between current frame data and historical frame data in a position evaluation dimension is provided for an embodiment of the present disclosure;

[0031] Figure 10 A probability distribution case diagram of a first initial distance of a fifth obstacle and a sixth obstacle in a two-dimensional plane is provided for an embodiment of the present disclosure;

[0032] Figure 11 A probability distribution case diagram of a second initial distance of a fifth obstacle and a sixth obstacle in a three-dimensional space is provided for an embodiment of the present disclosure;

[0033] Figure 12 A probability distribution case diagram of a third initial distance of a seventh obstacle and an eighth obstacle in a three-dimensional space is provided for an embodiment of the present disclosure;

[0034] Figure 13 Another probability distribution case diagram of a third initial distance of a seventh obstacle and an eighth obstacle in a three-dimensional space is provided for an embodiment of the present disclosure;

[0035] Figure 14 A probability distribution case diagram of reference data used to represent similarity between current frame data and historical frame data in a second three-dimensional distance evaluation dimension is provided for an embodiment of the present disclosure;

[0036] Figure 15 A probability distribution case diagram of an initial speed difference between a first driving speed and a second driving speed is provided for an embodiment of the present disclosure;

[0037] Figure 16A probability distribution diagram of reference data for representing similarity between current frame data and historical frame data in a speed evaluation dimension is provided for an embodiment of the present disclosure.

[0038] Figure 17 A probability distribution diagram of a fourth initial distance is provided for an embodiment of the present disclosure.

[0039] Figure 18 A probability distribution diagram of reference data for representing similarity between current frame data and historical frame data in a first projection position evaluation dimension is provided for an embodiment of the present disclosure.

[0040] Figure 19 A probability distribution diagram of a fifth initial distance is provided for an embodiment of the present disclosure.

[0041] Figure 20 A probability distribution diagram of reference data for representing similarity between current frame data and historical frame data in a second projection position evaluation dimension is provided for an embodiment of the present disclosure.

[0042] Figure 21 A scene diagram of a data verification method is provided for an embodiment of the present disclosure.

[0043] Figure 22 A schematic structural block diagram of a data verification apparatus is provided for an embodiment of the present disclosure.

[0044] Figure 23 A schematic structural block diagram of an electronic device is provided for an embodiment of the present disclosure. DETAILED DESCRIPTION

[0045] Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, which include various details of the embodiments of the present disclosure to help in understanding, which should be considered in the context of the present disclosure only. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the present disclosure. Also, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0046] As described in the background, in the field of autonomous driving technology, sensor data association is a crucial step, which aims to improve the ability of autonomous driving systems to identify and track obstacles, so as to ensure that the autonomous driving system can make relatively accurate driving decisions.

[0047] However, at present, the sensor data association is usually achieved by manually designing (Hand-Craft) features, similar to the following expression of "data collected by sensor A-data collected by sensor B".

[0048] wherein, the sensor B can be used to characterize each sensor installed on the autonomous vehicle except the sensor A. For example, the sensor A is Lidar, then the sensor B can be used to characterize Camera and Radar; for another example, the sensor A is Camera, then the sensor B can be used to characterize Lidar and Radar; for yet another example, the sensor A is Radar, then the sensor B can be used to characterize Lidar and Camera.

[0049] In addition, the data collected by the sensor A can be specifically the current frame data collected by the sensor A, that is, the latest sensor data collected by the sensor A, which is used to characterize the latest obstacle state perceived by the sensor A; the data collected by the sensor B can be specifically the historical frame data collected by the sensor B, that is, the historical sensor data collected by the sensor B, which is used to characterize the historical obstacle state perceived by the sensor B. Wherein, the data collected by the sensor B is derived from the global obstacle tracking information (Global_Track) of the autonomous vehicle. Since all the sensor data stored in the global obstacle tracking information of the autonomous vehicle has been associated, it can be determined that the data collected by the sensor B is also associated sensor data.

[0050] Further, when the sensor A is Lidar, it can be determined whether the data collected by the sensor A (that is, the data collected by the Lidar) can be associated with the data collected by the sensor B (including the data collected by the Camera and the data collected by the Radar) based on the following manners:

[0051] (1) Lidar collected data—Camera collected data

[0052] In the case that the shape similarity between the obstacle characterized by the Lidar collected data and the obstacle characterized by the Camera collected data is greater than a first preset similarity threshold, the first evaluation result is determined to be true; in the case that the shape similarity between the obstacle characterized by the Lidar collected data and the obstacle characterized by the Camera collected data is less than or equal to the first preset similarity threshold, the first evaluation result is determined to be false.

[0053] In the case that the Lidar collected data (three-dimensional point cloud data) completely falls within the bounding box of the obstacle characterized by the Camera collected data, the second evaluation result is determined to be true; in the case that the Lidar collected data does not completely fall within the bounding box of the obstacle characterized by the Camera collected data, the second evaluation result is determined to be false.

[0054] (2) Lidar collected data—Radar collected data

[0055] In a case where the planar distance between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is less than a first preset distance threshold, the third evaluation result is determined to be true; in a case where the planar distance between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is greater than or equal to the first preset distance threshold, the third evaluation result is determined to be false.

[0056] In a case where the speed difference between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is less than a first preset speed difference threshold, the fourth evaluation result is determined to be true; in a case where the speed difference between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is greater than or equal to the first preset speed difference threshold, the fourth evaluation result is determined to be false.

[0057] In a case where the first evaluation result, the second evaluation result, the third evaluation result and the fourth evaluation result are all true, it is determined that the data collected by sensor A can be associated with the data collected by sensor B; in a case where at least one of the first evaluation result, the second evaluation result, the third evaluation result and the fourth evaluation result is false, it is determined that the data collected by sensor A cannot be associated with the data collected by sensor B.

[0058] In a case where sensor A is a camera, whether the data collected by sensor A (i.e., the data collected by the camera) can be associated with the data collected by sensor B (including the data collected by the laser radar and the data collected by the radar) can be determined based on the following manners:

[0059] (1) Camera collected data - laser radar collected data

[0060] In a case where the shape similarity between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the camera is greater than a first preset similarity threshold, the first evaluation result is determined to be true; in a case where the shape similarity between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the camera is less than or equal to the first preset similarity threshold, the first evaluation result is determined to be false.

[0061] In a case where the data collected by the laser radar is completely within the bounding box of the obstacle represented by the data collected by the camera, the second evaluation result is determined to be true; in a case where the data collected by the laser radar is not completely within the bounding box of the obstacle represented by the data collected by the camera, the second evaluation result is determined to be false.

[0062] (2) Camera collected data - radar collected data

[0063] In a case where the three-dimensional distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is less than a second preset distance threshold, the fifth evaluation result is determined to be true; in a case where the three-dimensional distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is greater than or equal to the second preset distance threshold, the fifth evaluation result is determined to be false.

[0064] In a case where the speed difference between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is less than a second preset speed difference threshold, the sixth evaluation result is determined to be true; in a case where the speed difference between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is greater than or equal to the second preset speed difference threshold, the sixth evaluation result is determined to be false.

[0065] In a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the first direction is less than a third preset distance threshold, the seventh evaluation result is determined to be true; in a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the first direction is greater than or equal to the third distance threshold, the seventh evaluation result is determined to be false.

[0066] In a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the second direction is less than a fourth distance threshold, the eighth evaluation result is determined to be true; in a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the second direction is greater than or equal to the third distance threshold, the eighth evaluation result is determined to be false.

[0067] In a case where the first evaluation result, the second evaluation result, the fifth evaluation result, the sixth evaluation result, the seventh evaluation result and the eighth evaluation result are all true, it is determined that the data collected by the sensor A can be associated with the data collected by the sensor B; in a case where at least one of the first evaluation result, the second evaluation result, the fifth evaluation result, the sixth evaluation result, the seventh evaluation result and the eighth evaluation result is false, it is determined that the data collected by the sensor A cannot be associated with the data collected by the sensor B.

[0068] In a case where the sensor A is a radar, whether the data collected by the sensor A (i.e., the data collected by the radar) can be associated with the data collected by the sensor B (including the data collected by the laser radar and the camera) can be determined based on the following manners:

[0069] (1) The data collected by the radar - the data collected by the laser radar

[0070] In a case where the planar distance between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is less than a first preset distance threshold, the third evaluation result is determined to be true; in a case where the planar distance between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is greater than or equal to the first preset distance threshold, the third evaluation result is determined to be false.

[0071] In a case where the speed difference between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is less than a first preset speed difference threshold, the fourth evaluation result is determined to be true; in a case where the speed difference between the obstacle represented by the data collected by the laser radar and the obstacle represented by the data collected by the radar is greater than or equal to the first preset speed difference threshold, the fourth evaluation result is determined to be false.

[0072] (2) Radar-collected data—camera-collected data

[0073] In a case where the three-dimensional distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is less than a second preset distance threshold, the fifth evaluation result is determined to be true; in a case where the three-dimensional distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is greater than or equal to the second preset distance threshold, the fifth evaluation result is determined to be false.

[0074] In a case where the speed difference between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is less than a second preset speed difference threshold, the sixth evaluation result is determined to be true; in a case where the speed difference between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar is greater than or equal to the second preset speed difference threshold, the sixth evaluation result is determined to be false.

[0075] In a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the first direction is less than a third preset distance threshold, the seventh evaluation result is determined to be true; in a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the first direction is greater than or equal to the third distance threshold, the seventh evaluation result is determined to be false.

[0076] In a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the second direction is less than a fourth distance threshold, the eighth evaluation result is determined to be true; in a case where the distance between the obstacle represented by the data collected by the camera and the obstacle represented by the data collected by the radar in the second direction is greater than or equal to the third distance threshold, the eighth evaluation result is determined to be false.

[0077] In a case where the third evaluation result, the fourth evaluation result, the fifth evaluation result, the sixth evaluation result, the seventh evaluation result, and the eighth evaluation result are all true, it is determined that the data collected by the sensor A can be associated with the data collected by the sensor B; in a case where at least one of the third evaluation result, the fourth evaluation result, the fifth evaluation result, the sixth evaluation result, the seventh evaluation result, and the eighth evaluation result is false, it is determined that the data collected by the sensor A cannot be associated with the data collected by the sensor B.

[0078] However, the inventors have found that the above manner actually adopts a simple condition superposition rule to determine whether the data collected by the sensor A can be associated with the data collected by the sensor B. Specifically, in a case where the sensor A is a laser radar, only in a case where the first evaluation result, the second evaluation result, the third evaluation result, and the fourth evaluation result are all true, it is determined that the data collected by the sensor A can be associated with the data collected by the sensor B; in a case where the sensor A is a camera, only in a case where the first evaluation result, the second evaluation result, the fifth evaluation result, the sixth evaluation result, the seventh evaluation result, and the eighth evaluation result are all true, it is determined that the data collected by the sensor A can be associated with the data collected by the sensor B; in a case where the sensor A is a radar, only in a case where the third evaluation result, the fourth evaluation result, the fifth evaluation result, the sixth evaluation result, the seventh evaluation result, and the eighth evaluation result are all true, it is determined that the data collected by the sensor A can be associated with the data collected by the sensor B. In this process, as long as one evaluation result is wrong, a misassociation or a missed association will occur, and thus the accuracy of sensor data association cannot be ensured.

[0079] To solve the above problems, the present embodiment provides a data verification method which can be applied to an electronic device. The electronic device can be an automatic driving controller. Hereinafter, the data verification method provided by the present embodiment will be described with reference to the flowchart shown in FIG. 1. Figure 1 It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described in the flowchart can also be executed in other orders.

[0080] In step S101, current frame data collected by a first target sensor in a plurality of sensors is obtained.

[0081] The plurality of sensors are installed on a target object, for example, an automatic driving vehicle, an aircraft, a ship, a robot, etc. Here, the plurality of sensors can only include a laser radar and a camera, can only include a laser radar and a radar, can only include a radar and a camera, or can simultaneously include a laser radar, a camera, and a radar.

[0082] In the embodiments of the present disclosure, the first target sensor can be any one of the plurality of sensors. For example, when the plurality of sensors simultaneously include a laser radar, a camera and a radar, the first target sensor can be any one of the laser radar, the camera and the radar, and the current frame data can be sensor data newly collected by the first target sensor, which is used to represent the obstacle state newly perceived by the first target sensor. In addition, it can be understood that in the embodiments of the present disclosure, when the first sensor is a laser radar, the current frame data collected by the first sensor can be three-dimensional point cloud data; when the first sensor is a camera, the current frame data collected by the first sensor can be image data; and when the first sensor is a radar, the current frame data collected by the first sensor can include distance information, speed information, angle information, etc.

[0083] In step S102, each sensor except the first target sensor in the plurality of sensors is taken as a second target sensor to obtain a plurality of reference data used to represent the similarity between the current frame data and historical frame data collected by the second target sensor.

[0084] In the embodiments of the present disclosure, the historical frame data can be sensor data historically collected by the second target sensor, which is used to represent the obstacle state historically perceived by the second target sensor.

[0085] For example, the current frame data is sensor data collected by the first target sensor at time T, which is used to represent the obstacle state perceived by the first target sensor at time T; and the historical frame data can be sensor data collected by the second target sensor at time T-1, which is used to represent the obstacle state perceived by the second target sensor at time T-1.

[0086] In addition, it should be noted that in the embodiments of the present disclosure, the historical frame data is derived from the global obstacle tracking information of the target object. Since all the sensor data stored in the global obstacle tracking information of the target object has been associated, it can be determined that the historical frame data is also associated sensor data. For example, the historical frame data is sensor data collected by the second target sensor at time T-1, and it can be understood that the historical frame data is associated sensor data means that the historical frame data has been associated with sensor data collected by each third target sensor except the second target sensor in the plurality of sensors at time T-2. The sensor data collected by each third target sensor except the second target sensor in the plurality of sensors at time T-2 is also derived from the global obstacle tracking information of the target object.

[0087] It should be further noted that in the embodiments of the present disclosure, the plurality of reference data can represent the similarity between the current frame data and the historical frame data collected by the second target sensor from different evaluation dimensions. Based on this, it can be understood that in the embodiments of the present disclosure, for each reference data in the plurality of reference data, it can be a scalar calculated based on the current frame data and the historical frame data under a certain evaluation dimension, to represent the similarity between the current frame data and the historical frame data under the evaluation dimension.

[0088] In step S103, the plurality of reference data is fused to obtain a verification result for the current frame data.

[0089] The verification result is used to represent whether the current frame data can be associated with the historical frame data.

[0090] In an example, a first verification result for the current frame data can be obtained in a case where the verification result is greater than the overall similarity threshold, to represent that the current frame data can be associated with the historical frame data; and a second verification result for the current frame data can be obtained in a case where the verification result is less than or equal to the overall similarity threshold, to represent that the current frame data cannot be associated with the historical frame data. The overall similarity threshold can be set according to actual application requirements, and the embodiments of the present disclosure do not limit this.

[0091] By using the data verification method provided in the embodiments of the present disclosure, the current frame data collected by the first target sensor in the plurality of sensors can be obtained; each sensor in the plurality of sensors except the first target sensor is taken as the second target sensor to obtain a plurality of reference data representing the similarity between the current frame data and the historical frame data collected by the second target sensor; and the plurality of reference data is fused to obtain a verification result for the current frame data. The key is to obtain the plurality of reference data representing the similarity between the current frame data and the historical frame data, and fuse the plurality of reference data to obtain the verification result for the current frame data, to represent whether the current frame data can be associated with the historical frame data. Based on this, by using the data verification method provided in the embodiments of the present disclosure, even if a small amount of reference data is wrong, it does not necessarily lead to misassociation or missed association. Compared with the way of using a simple condition superposition rule to judge whether the current frame data can be associated with the historical frame data in the prior art, the fault tolerance is higher, and therefore the accuracy of sensor data association can be improved.

[0092] In some optional embodiments, the "obtaining a plurality of reference data representing the similarity between the current frame data and the historical frame data collected by the second target sensor" in step S102 can include:

[0093] Based on the first target sensor and the second target sensor, a plurality of evaluation dimensions are determined;

[0094] respectively, as a target evaluation dimension to obtain reference data under the target evaluation dimension for characterizing the similarity between the current frame data and the historical frame data collected by the second target sensor.

[0095] In an example, the plurality of evaluation dimensions includes a shape evaluation dimension and a position evaluation dimension in a case where the first target sensor and the second target sensor are any combination of a lidar and a camera (e.g., the first target sensor is a lidar and the second target sensor is a camera; or the first target sensor is a camera and the second target sensor is a lidar); the plurality of evaluation dimensions includes a planar distance evaluation dimension and a first three-dimensional distance evaluation dimension in a case where the first target sensor and the second target sensor are any combination of a lidar and a radar (e.g., the first target sensor is a lidar and the second target sensor is a radar; or the first target sensor is a radar and the second target sensor is a lidar); and the plurality of evaluation dimensions includes a second three-dimensional distance evaluation dimension, a velocity evaluation dimension, a first projected position evaluation dimension, and a second projected position evaluation dimension in a case where the first target sensor and the second target sensor are any combination of a radar and a camera (e.g., the first target sensor is a radar and the second target sensor is a camera; or the first target sensor is a camera and the second target sensor is a radar).

[0096] The reference data under the shape evaluation dimension for characterizing the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor can be specifically used to characterize the shape similarity between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data.

[0097] The reference data under the position evaluation dimension for characterizing the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor can be specifically used to characterize the position similarity between the third obstacle represented by the first perception data and the fourth obstacle represented by the second perception data. Here, the first perception data is the data collected by the lidar in the current frame data and the historical frame data; and the second perception data is the data collected by the camera in the current frame data and the historical frame data.

[0098] The reference data used to represent the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor in the plane distance evaluation dimension can be specifically used to represent the plane distance similarity between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data. Here, the greater the plane distance between the fifth obstacle and the sixth obstacle, the smaller the plane distance similarity between the fifth obstacle and the sixth obstacle is represented; the smaller the plane distance between the fifth obstacle and the sixth obstacle, the greater the plane distance similarity between the fifth obstacle and the sixth obstacle is represented.

[0099] The reference data used to represent the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor in the first three-dimensional distance evaluation dimension can be specifically used to represent the three-dimensional distance similarity between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data. Here, the greater the three-dimensional distance between the fifth obstacle and the sixth obstacle, the smaller the three-dimensional distance similarity between the fifth obstacle and the sixth obstacle is represented; the smaller the three-dimensional distance between the fifth obstacle and the sixth obstacle, the greater the three-dimensional distance similarity between the fifth obstacle and the sixth obstacle is represented.

[0100] The reference data used to represent the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor in the second three-dimensional distance evaluation dimension can be specifically used to represent the three-dimensional distance similarity between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data. Here, the greater the three-dimensional distance between the seventh obstacle and the eighth obstacle, the smaller the three-dimensional distance similarity between the seventh obstacle and the eighth obstacle is represented; the smaller the three-dimensional distance between the seventh obstacle and the eighth obstacle, the greater the three-dimensional distance similarity between the seventh obstacle and the eighth obstacle is represented.

[0101] The reference data used to represent the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor in the speed evaluation dimension can be specifically used to represent the speed similarity between the ninth obstacle represented by the third perception data and the tenth obstacle represented by the fourth perception data. Here, the third perception data is the data collected by the radar in the current frame data and the historical frame data; the fourth perception data is the data collected by the camera in the current frame data and the historical frame data.

[0102] The reference data used to represent the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor in the first projection position evaluation dimension can be specifically used to represent one-way distance similarity of the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data in the first direction (for example, the X-axis direction of the world coordinate system). Here, the greater the one-way distance of the seventh obstacle and the eighth obstacle in the first direction, the smaller the one-way distance similarity of the seventh obstacle and the eighth obstacle in the first direction is represented; the smaller the one-way distance of the seventh obstacle and the eighth obstacle in the first direction, the greater the one-way distance similarity of the seventh obstacle and the eighth obstacle in the first direction is represented.

[0103] The reference data used to represent the similarity between the current frame data collected by the first target sensor and the historical frame data collected by the second target sensor in the second projection position evaluation dimension can be specifically used to represent one-way distance similarity of the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data in the second direction (for example, the Y-axis direction of the world coordinate system). Here, the greater the one-way distance of the seventh obstacle and the eighth obstacle in the second direction, the smaller the one-way distance similarity of the seventh obstacle and the eighth obstacle in the second direction is represented; the smaller the one-way distance of the seventh obstacle and the eighth obstacle in the second direction, the greater the one-way distance similarity of the seventh obstacle and the eighth obstacle in the second direction is represented.

[0104] Further, in the embodiments of the present disclosure, each of the plurality of evaluation dimensions has a corresponding evaluation strategy for indicating how to obtain the reference data used to represent the similarity between the current frame data and the historical frame data collected by the second target sensor in the evaluation dimension. Based on this, in an example, after determining the plurality of evaluation dimensions based on the first target sensor and the second target sensor, each of the plurality of evaluation dimensions can be taken as a target evaluation dimension, and the reference data used to represent the similarity between the current frame data and the historical frame data collected by the second target sensor in the target evaluation dimension can be obtained by the following way:

[0105] determining an evaluation strategy corresponding to the target evaluation dimension;

[0106] mapping the current frame data and the historical frame data collected by the second target sensor according to the evaluation strategy corresponding to the target evaluation dimension, to obtain the reference data used to represent the similarity between the current frame data and the historical frame data in the target evaluation dimension.

[0107] Wherein, the mapping of the current frame data and the historical frame data collected by the second target sensor can be a comparison of the current frame data and the historical frame data. For example, the comparison can be performed after spatial mapping of the current frame data and the historical frame data.

[0108] By the above manner, in the embodiments of the present disclosure, the plurality of evaluation dimensions can be determined based on the first target sensor and the second target sensor; each of the plurality of evaluation dimensions is taken as a target evaluation dimension to obtain reference data for representing similarity between the current frame data and the historical frame data collected by the second target sensor in the target evaluation dimension. That is, in the embodiments of the present disclosure, the plurality of reference data for representing similarity between the current frame data and the historical frame data are obtained from different evaluation dimensions, so that the current frame data and the historical frame data can be analyzed more finely, and these evaluation dimensions have relatively reliable similarity representation capability, so that more accurate reference data for representing similarity between the current frame data and the historical frame data can be ensured, thereby further improving the accuracy of sensor data association.

[0109] In addition, each of the plurality of evaluation dimensions has a corresponding evaluation strategy for indicating how to obtain the reference data for representing similarity between the current frame data and the historical frame data in the evaluation dimension. Therefore, after the plurality of evaluation dimensions are determined based on the first target sensor and the second target sensor, each of the plurality of evaluation dimensions can be taken as a target evaluation dimension, the evaluation strategy corresponding to the target evaluation dimension can be determined, and the current frame data and the historical frame data can be mapped according to the evaluation strategy corresponding to the target evaluation dimension to obtain the reference data for representing similarity between the current frame data and the historical frame data in the target evaluation dimension. In this way, the efficiency of obtaining the reference data for representing similarity between the current frame data and the historical frame data in the target evaluation dimension can be improved, thereby improving the execution efficiency of the data verification method.

[0110] In an optional implementation, the first target sensor and the second target sensor are any combination of a laser radar and a camera, and the target evaluation dimension is a shape evaluation dimension. The operation of "mapping the current frame data and the historical frame data collected by the second target sensor to obtain reference data for representing similarity between the current frame data and the historical frame data in the target evaluation dimension" can include:

[0111] mapping the current frame data and the historical frame data collected by the second target sensor to obtain shape difference between a first obstacle represented by the current frame data and a second obstacle represented by the historical frame data;

[0112] obtaining reference data for representing similarity between the current frame data and the historical frame data in the shape evaluation dimension based on the shape difference.

[0113] The mapping of the current frame data and the historical frame data collected by the second target sensor can be performed after spatial mapping of the current frame data and the historical frame data. For example, the current frame data and the historical frame data can be mapped to the same world coordinate system and then compared.

[0114] Further, in an example, the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data can include:

[0115] mapping the current frame data and the historical frame data collected by the second target sensor to obtain a width difference of the first bounding box and the second bounding box and a height difference of the first bounding box and the second bounding box, wherein the first bounding box is a bounding box of the first obstacle represented by the current frame data, and the second bounding box is a bounding box of the second obstacle represented by the historical frame data.

[0116] standardizing the width difference and the height difference in a conversion manner from a Gaussian distribution to a Chi-square distribution to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data.

[0117] The width difference of the first bounding box and the second bounding box obtained by mapping the current frame data and the historical frame data collected by the second target sensor can be represented as:

[0118] Box_diff(w)=(w1-w2) / ((wl+wc) / 2)

[0119] Here, Box_diff(w) is used to represent the width difference of the first bounding box and the second bounding box, w1 is used to represent the width of the first bounding box, and w2 is used to represent the width of the second bounding box.

[0120] In addition, it should be noted that in the present disclosure, the width difference of the first bounding box and the second bounding box is subject to a Gaussian distribution in consideration of the theoretical jitter range of the width difference of the first bounding box and the second bounding box. Specifically, in an actual application example, the probability distribution of the width difference of the first bounding box and the second bounding box is subject to a Gaussian distribution, which can be as shown in Figure 2 , where the abscissa represents the width difference (unit: m), and the ordinate represents the probability density. Figure 2 The theoretical jitter range of the width difference of the first bounding box and the second bounding box can be determined according to experimental values, and the present disclosure does not limit this.

[0121] Wherein, the height difference between the first bounding box and the second bounding box obtained by mapping the current frame data and the historical frame data collected by the second target sensor can be represented as:

[0122] Box_diff(h)=(h1-h2) / ((hl+hc) / 2)

[0123] Here, Box_diff(h) is used to represent the height difference between the first bounding box and the second bounding box; h1 is used to represent the height of the first bounding box; h2 is used to represent the height of the second bounding box.

[0124] In addition, it needs to be explained that in the disclosure embodiment, the height difference between the first bounding box and the second bounding box is subject to Gaussian distribution in the case of considering the theoretical jitter range of the height difference between the first bounding box and the second bounding box. Specifically, in an actual application example, the probability distribution of the height difference between the first bounding box and the second bounding box is subject to Gaussian distribution, which can be as shown in Figure 3 Figure 3 Wherein, the abscissa represents the height difference (unit: m), and the ordinate represents the probability density. The theoretical jitter range of the height difference between the first bounding box and the second bounding box can be determined according to the experimental value, and the disclosure embodiment does not limit this.

[0125] Based on the above, it can be understood that in the disclosure embodiment, after obtaining the width difference between the first bounding box and the second bounding box and the height difference between the first bounding box and the second bounding box, the width difference and the height difference can be standardized according to the conversion from Gaussian distribution to chi-square distribution, to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data. The process can be represented as:

[0126] Box_diff(shape)=

[0127] [Box_diff(w) 2 / dev(Box_diff(w)) 2 ]+[Box_diff(h) 2 / dev(Box_diff(h)) 2 ]

[0128] ​wherein Box_diff(shape) is used to represent shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data; Box_diff(w) is used to represent width difference; dev(Box_diff(w)) is used to represent standard deviation of the width difference; Box_diff(h) is used to represent height difference; and dev(Box_diff(h)) is used to represent standard deviation of the height difference.

[0129] Here, the standard deviation of the width difference between the first bounding box and the second bounding box can be obtained by calculation, for example, after obtaining the probability distribution of the width difference (e.g., the probability distribution shown in FIG. 6), the standard deviation of the width difference can be calculated based on the probability distribution of the width difference, which will not be repeated here; similarly, the standard deviation of the height difference can be obtained by calculation, for example, after obtaining the probability distribution of the height difference (e.g., the probability distribution shown in FIG. 7), the standard deviation of the height difference can be calculated based on the probability distribution of the height difference, which will also not be repeated here. Figure 2 Figure 3

[0130] Further, in an example, after obtaining the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data, the reference data in the shape evaluation dimension for representing the similarity between the current frame data and the historical frame data can be obtained based on the shape difference in the following manner:

[0131] obtaining a first chi-square statistic for the shape difference;

[0132] obtaining the reference data in the shape evaluation dimension for representing the similarity between the current frame data and the historical frame data based on the first chi-square statistic.

[0133] In a specific example, the above process can be represented as:

[0134] Rdl = 1 - ChiSquaredCdf2TableFun(Box_diff(shape)

[0135] ​​wherein, Rd1 is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension; ChiSquaredCdf2TableFun(Box_diff(shape)) is used to represent the first chi-square statistic for the shape difference based on the shape difference (i.e., Box_diff(shape)) and is obtained from the chi-square distribution table with 2 degrees of freedom. In an example, the first chi-square statistic for the shape difference based on the shape difference and obtained from the chi-square distribution table with 2 degrees of freedom can be obtained by interpolation method to improve the efficiency of the first chi-square statistic.

[0136] In addition, in an actual application example, in the case that the probability distribution of the width difference between the first bounding box and the second bounding box is as shown in Figure 2 , and the probability distribution of the height difference between the first bounding box and the second bounding box is as shown in Figure 3 , the probability distribution of the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension can be as shown in Figure 4 , wherein the abscissa is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension, and the ordinate is used to represent the probability density. Figure 4

[0137] In the above manner, in the embodiments of the present disclosure, the current frame data and the historical frame data collected by the second target sensor can be mapped to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data, and the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension can be obtained based on the shape difference. This process does not involve complex data processing procedures, and therefore, the efficiency of obtaining the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension can be improved, thereby improving the execution efficiency of the data verification method. Moreover, the above process involves standardization processing of related data (for example, the width difference and the height difference), and therefore, the accuracy of the reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension can be improved, thereby further improving the accuracy of the sensor data association.

[0138] In an optional implementation, the first target sensor and the second target sensor are any combination of a laser radar and a camera, and the target evaluation dimension is a position evaluation dimension. The mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the reference data used to represent the similarity between the current frame data and the historical frame data in the target evaluation dimension can include:

[0139] ​The current frame data and the data collected by the laser radar in the historical frame data collected by the second target sensor are determined as first perception data, and the data collected by the camera in the current frame data and the historical frame data are determined as second perception data;

[0140] The first perception data and the second perception data are mapped to obtain a first plane distance of a third obstacle represented by the first perception data and a fourth obstacle represented by the second perception data in a two-dimensional plane.

[0141] Based on the first plane distance, reference data for representing the similarity between the current frame data and the historical frame data in the position evaluation dimension is obtained.

[0142] The mapping of the first perception data and the second perception data can be performed after spatial mapping of the first perception data and the second perception data. For example, the first perception data and the second perception data can be mapped to the same world coordinate system and then compared.

[0143] Further, in an example, "mapping the first perception data and the second perception data to obtain a first plane distance of a third obstacle represented by the first perception data and a fourth obstacle represented by the second perception data in a two-dimensional plane" can include:

[0144] The first perception data and the second perception data are mapped to determine a plurality of feature points in the first perception data outside the third bounding box; wherein the third bounding box is a bounding box of the fourth obstacle represented by the second perception data;

[0145] The first overflow amount of the plurality of feature points relative to the third bounding box in the first direction and the second overflow amount of the plurality of feature points relative to the third bounding box in the second direction are obtained;

[0146] The first overflow amount and the second overflow amount are standardized according to a conversion mode from Gaussian distribution to Chi-square distribution to obtain a first plane distance of a third obstacle represented by the first perception data and a fourth obstacle in a two-dimensional plane.

[0147] In a specific example, after determining the plurality of feature points in the first perception data outside the third bounding box, a first shortest distance of each feature point in the plurality of feature points relative to the third bounding box in the first direction (for example, the X-axis direction of the world coordinate system) can be obtained to obtain a plurality of first shortest distances corresponding to the plurality of feature points, and the mean of the plurality of first shortest distances is taken as the first overflow amount of the plurality of feature points relative to the third bounding box in the first direction.

[0148] Furthermore, it should be noted that in this embodiment of the disclosure, considering the theoretical jitter range of the first overflow amount of multiple feature points relative to the third bounding box in the first direction, a new probability distribution following a Gaussian distribution can be obtained by symmetrically processing the probability distribution of the first overflow amount of multiple feature points relative to the third bounding box in the first direction. Specifically, in a practical application example, the probability distribution of the first overflow amount of multiple feature points relative to the third bounding box in the first direction can be as follows: Figure 5 As shown ( Figure 5 In the diagram, the horizontal axis represents the first overflow (in meters), and the vertical axis represents the probability density. After symmetry processing, we can obtain the following: Figure 6 The new probability distribution shown follows a Gaussian distribution. Figure 6 In the diagram, the horizontal axis represents the first overflow amount (in meters), and the vertical axis represents the probability density. The theoretical jitter range of the first overflow amount of multiple feature points in the first direction relative to the third bounding box can be determined based on experimental values, and this embodiment does not impose any limitations on this.

[0149] Similarly, after determining multiple feature points located outside the third bounding box in the first perception data, the second shortest distance of each feature point relative to the third bounding box in the second direction (e.g., the Y-axis direction of the world coordinate system) can be obtained to obtain multiple second shortest distances corresponding one-to-one with the multiple feature points, and the average of the multiple second shortest distances is used as the second overflow amount of the multiple feature points relative to the third bounding box in the second direction.

[0150] Furthermore, it should be noted that in this embodiment of the disclosure, considering the theoretical jitter range of the second overflow amount of multiple feature points in the second direction relative to the third bounding box, symmetrical processing of the probability distribution of the second overflow amount of multiple feature points in the second direction relative to the third bounding box can yield a new probability distribution that follows a Gaussian distribution. Specifically, in a practical application example, the probability distribution of the second overflow amount of multiple feature points in the second direction relative to the third bounding box can be as follows: Figure 7 As shown ( Figure 7 In the diagram, the horizontal axis represents the second spillover amount (in meters), and the vertical axis represents the probability density. After symmetry processing, we can obtain the following: Figure 8 The new probability distribution shown follows a Gaussian distribution. Figure 8 In the diagram, the horizontal axis represents the first overflow amount (in meters), and the vertical axis represents the probability density. The theoretical jitter range of the second overflow amount of multiple feature points in the second direction relative to the third bounding box can be determined based on experimental values, and this embodiment does not impose any limitations on this.

[0151] Based on the above, it can be understood that in this embodiment of the disclosure, after obtaining the first overflow amount of multiple feature points in the first direction relative to the third bounding box, and the second overflow amount of multiple feature points in the second direction relative to the third bounding box, the first and second overflow amounts can be standardized according to the transformation method from Gaussian distribution to chi-square distribution to obtain the first planar distance in the two-dimensional plane represented by the first perception data between the third obstacle and the fourth obstacle. This process can be characterized as follows:

[0152] Box_diff(Location) =

[0153] [Box_diff(x) 2 / dev(Box_diff(x)) 2 ]+[Box_diff(y) 2 / dev(Box_diff(y)) 2 ]

[0154] Wherein, Box_diff(Location) is used to characterize the first plane distance in the two-dimensional plane between the third obstacle and the fourth obstacle as represented by the first perception data; Box_diff(x) is used to characterize the first overflow amount; dev(Box_diff(x)) is used to characterize the standard deviation of the first overflow amount; Box_diff(y) is used to characterize the second overflow amount; dev(Box_diff(y)) is used to characterize the standard deviation of the second overflow amount.

[0155] Here, the standard deviation of the first overflow can be obtained through calculation, for example, in the probability part of obtaining the first overflow (e.g., Figure 6 After obtaining the probability partial case shown, the standard deviation of the first overflow can be calculated based on the probability partial case of the first overflow, which will not be elaborated here; similarly, after obtaining the probability partial case of the second overflow (e.g., Figure 8 After the probability partial case shown, the standard deviation of the second overflow can be calculated based on the probability partial case of the second overflow, which will not be elaborated here.

[0156] Furthermore, in one example, after obtaining the first planar distance in the two-dimensional plane between the third and fourth obstacles represented by the first perception data, reference data for representing the similarity between the current frame data and historical frame data in the position evaluation dimension can be obtained based on the first planar distance in the following way:

[0157] Obtain the second chi-square statistic for the distance from the first plane;

[0158] Based on the second chi-square statistic, reference data for representing similarity between the current frame data and the historical frame data in the position evaluation dimension is obtained.

[0159] In a specific example, the above process can be represented as:

[0160] Rd2 = 1 - ChiSquaredCdf2TableFun(Box_diff(Location)

[0161] Wherein, Rd2 is used to represent reference data for representing similarity between the current frame data and the historical frame data in the position evaluation dimension; ChiSquaredCdf2TableFun(Box_diff(Location) is used to represent the second chi-square statistic for the first plane distance (i.e., Box_diff(Location)) obtained from the chi-square distribution table with 2 degrees of freedom. In an example, the second chi-square statistic for the first plane distance obtained from the chi-square distribution table with 2 degrees of freedom can be based on the first plane distance by interpolation method to improve the lookup efficiency of the second chi-square statistic.

[0162] In addition, in an actual application example, the first overflow amount of the plurality of feature points relative to the third bounding box in the first direction is as shown in Figure 6 The probability distribution of the reference data for representing similarity between the current frame data and the historical frame data in the position evaluation dimension can be as shown in Figure 8 Figure 9 Figure 9 Wherein, the horizontal coordinate is used to represent the reference data for representing similarity between the current frame data and the historical frame data in the position evaluation dimension, and the vertical coordinate is used to represent the probability density.

[0163] ​​By means of the above, in the embodiments of the present disclosure, the data collected by the laser radar in the current frame data and the historical frame data collected by the second target sensor can be determined as the first perception data, and the data collected by the camera in the current frame data and the historical frame data can be determined as the second perception data, and the first perception data and the second perception data are mapped to obtain the first plane distance of the third obstacle represented by the first perception data and the fourth obstacle represented by the second perception data in the two-dimensional plane, and then the first plane distance is used to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the position evaluation dimension. The process does not involve a complex data processing flow, so the efficiency of obtaining the reference data for representing the similarity between the current frame data and the historical frame data in the position evaluation dimension can be improved, thereby improving the execution efficiency of the data verification method. Moreover, the above process involves standardization processing of related data (for example, the first overflow amount and the second overflow amount), so the accuracy of the reference data for representing the similarity between the current frame data and the historical frame data in the position evaluation dimension can be improved, thereby further improving the accuracy of the sensor data association.

[0164] In an optional embodiment, the first target sensor and the second target sensor are any combination of a laser radar and a radar, and the target evaluation dimension is a plane distance evaluation dimension. The mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the target evaluation dimension can include:

[0165] Mapping the current frame data and the historical frame data collected by the second target sensor to obtain a first initial distance of a fifth obstacle represented by the current frame data and a sixth obstacle represented by the historical frame data in a two-dimensional plane;

[0166] Standardizing the first initial distance to obtain a second plane distance;

[0167] Based on the second plane distance, the reference data for representing the similarity between the current frame data and the historical frame data in the plane distance evaluation dimension is obtained.

[0168] In an example, the mapping of the current frame data and the historical frame data collected by the second target sensor can be performed after spatial mapping of the current frame data and the historical frame data. For example, the current frame data and the historical frame data can be mapped to the same world coordinate system for comparison.

[0169] Further, in an example, the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain a first initial distance of a fifth obstacle represented by the current frame data and a sixth obstacle represented by the historical frame data in a two-dimensional plane can include:

[0170] Determine the first target vertex on the fourth bounding box that is closest to the reference position; wherein, the fourth bounding box is the bounding box of the fifth obstacle represented by the current frame data;

[0171] Determine the second target vertex that is closest to the reference position on the fifth bounding box; wherein, the fifth bounding box is the bounding box of the sixth obstacle represented by the historical frame data collected by the second target sensor;

[0172] Map the first target vertex and the second target vertex to obtain the interval distance between the first target vertex and the second target vertex, which is used as the first initial distance between the fifth obstacle and the sixth obstacle in the two-dimensional plane.

[0173] The reference location can be the ground plane.

[0174] After determining the first target vertex closest to the reference position on the fourth bounding box and the second target vertex closest to the reference position on the fifth bounding box, the Euclidean distance between the first target vertex and the second target vertex can be calculated as the interval distance between the first target vertex and the second target vertex obtained by mapping the first target vertex and the second target vertex. That is, it serves as the first initial distance between the fifth obstacle and the sixth obstacle in the two-dimensional plane.

[0175] Furthermore, it should be noted that, in this embodiment of the present disclosure, considering the theoretical jitter range of the first initial distance between the fifth and sixth obstacles in the two-dimensional plane, in a practical application example, the probability distribution of the first initial distance between the fifth and sixth obstacles in the two-dimensional plane can be as follows: Figure 10 As shown ( Figure 10 In the diagram, the horizontal axis represents the first initial distance (in meters), and the vertical axis represents the probability density. The theoretical fluctuation range of the first initial distance between the fifth and sixth obstacles in the two-dimensional plane can be determined based on experimental values, and this embodiment does not impose any limitations on this.

[0176] After obtaining the first initial distance in the two-dimensional plane between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data, the first initial distance can be standardized to obtain the second planar distance. This process can be characterized as follows:

[0177] SPd(Postion21)=Pd(Postion21) / Lnv21

[0178] wherein, SPd(Postion21) is used to represent the second plane distance; Pd(Postion21) is used to represent the first initial distance; and Lnv21 is used to represent the first normalized parameter. Here, the first normalized parameter can be a value greater than the first initial distance, which can be set according to actual application requirements, and the embodiments of the present disclosure do not make any limitation in this regard.

[0179] After the first initial distance is normalized to obtain the second plane distance, the reference data for representing the similarity between the current frame data and the historical frame data in the plane distance evaluation dimension can be obtained based on the second plane distance. This process can be represented as:

[0180] Rd3 = 1 - SPd(Postion21)

[0181] wherein, Rd3 is used to represent the reference data for representing the similarity between the current frame data and the historical frame data in the plane distance evaluation dimension; and SPd(Postion21) is used to represent the second plane distance.

[0182] In the embodiments of the present disclosure, the current frame data and the historical frame data collected by the second target sensor can be mapped to obtain the first initial distance of the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data in the two-dimensional plane, the first initial distance is normalized to obtain the second plane distance, and the reference data for representing the similarity between the current frame data and the historical frame data in the plane distance evaluation dimension is obtained based on the second plane distance. This process does not involve complex data processing procedures, so it can improve the efficiency of obtaining the reference data for representing the similarity between the current frame data and the historical frame data in the plane distance evaluation dimension, thereby improving the execution efficiency of the data verification method. Moreover, the above process involves the normalization of related data (for example, the first initial distance), so it can also improve the accuracy of the reference data for representing the similarity between the current frame data and the historical frame data in the plane distance evaluation dimension, thereby further improving the accuracy of the sensor data association.

[0183] In an optional embodiment, the first target sensor and the second target sensor are any combination of lidar and radar, and the target evaluation dimension is the first three-dimensional distance evaluation dimension. The process of mapping the current frame data and the historical frame data collected by the second target sensor to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the target evaluation dimension can include:

[0184] The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the second initial distance of the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data in the three-dimensional space.

[0185] The second initial distance is normalized to obtain a first three-dimensional distance;

[0186] Based on the first three-dimensional distance, reference data for representing similarity between the current frame data and the historical frame data in a first three-dimensional distance evaluation dimension is obtained.

[0187] The mapping of the current frame data and the historical frame data collected by the second target sensor can be performed after spatial mapping of the current frame data and the historical frame data. For example, the current frame data and the historical frame data can be mapped to the same world coordinate system and then compared.

[0188] Further, in an example, "mapping the current frame data and the historical frame data collected by the second target sensor to obtain a second initial distance between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data in a three-dimensional space" can include:

[0189] determining a first target center point of the fifth obstacle represented by the current frame data;

[0190] determining a second target center point of the sixth obstacle represented by the historical frame data collected by the second target sensor;

[0191] mapping the first target center point and the second target center point to obtain an interval distance between the first target center point and the second target center point as the second initial distance between the fifth obstacle and the sixth obstacle in the three-dimensional space.

[0192] In a specific example, at least one of a Voxel Grid Filter (VGF) algorithm, a clustering algorithm (for example, a K-Means clustering algorithm, a Density-Based Spatial Clustering of Applications with Noise (DBSCAN) clustering algorithm), a Random Sample Consensus (RANSAC) algorithm, a Principal Component Analysis (PCA) algorithm, etc. can be used to determine the center point of the obstacle represented by the data collected by the laser radar in the current frame data and the historical frame data. In another specific example, at least one of a target detection algorithm, a Kalman filtering algorithm, a clustering algorithm, a fusion algorithm, etc. can be used to determine the center point of the obstacle represented by the data collected by the radar in the current frame data and the historical frame data.

[0193] After determining the first target center point of the fifth obstacle represented by the current frame data and determining the second target center point of the sixth obstacle represented by the historical frame data collected by the second target sensor, a Euclidean distance between the first target center point and the second target center point can be calculated as a mapping of the first target center point and the second target center point to obtain an interval distance between the first target center point and the second target center point, that is, a second initial distance of the fifth obstacle and the sixth obstacle in the three-dimensional space.

[0194] In addition, it needs to be explained that in the embodiments of the present disclosure, in the case of considering the theoretical jitter range of the second initial distance of the fifth obstacle and the sixth obstacle in the three-dimensional space, in an actual application example, the probability distribution of the second initial distance of the fifth obstacle and the sixth obstacle in the three-dimensional space can be as shown in Figure 11 Figure 11 , where the abscissa is used to represent the second initial distance (unit: m), and the ordinate is used to represent the probability density. Wherein, the theoretical jitter range of the second initial distance of the fifth obstacle and the sixth obstacle in the three-dimensional space can be determined according to experimental values, and the embodiments of the present disclosure are not limited thereto.

[0195] After obtaining the second initial distance of the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data in the three-dimensional space, the second initial distance can be standardized to obtain a first three-dimensional distance. This process can be represented as:

[0196] SPd(Postion22) = Pd(Postion22) / Lnv22

[0197] Wherein, SPd(Postion22) is used to represent the first three-dimensional distance; Pd(Postion22) is used to represent the second initial distance; Lnv22 is used to represent the second standardization parameter. Here, the second standardization parameter can be a value greater than the second initial distance, which can be set according to actual application requirements, and the embodiments of the present disclosure are not limited thereto.

[0198] After the second initial distance is standardized to obtain the first three-dimensional distance, the reference data for representing the similarity between the current frame data and the historical frame data in the first three-dimensional distance evaluation dimension can be obtained based on the first three-dimensional distance. This process can be represented as:

[0199] Rd4 = 1 - SPd(Postion22)

[0200] Wherein, Rd4 is used to represent the reference data for representing the similarity between the current frame data and the historical frame data in the first three-dimensional distance evaluation dimension; SPd(Postion22) is used to represent the first three-dimensional distance.​

[0201] By the above manner, in the embodiment of the disclosure, the current frame data and the historical frame data collected by the second target sensor can be mapped to obtain a second initial distance of a fifth obstacle represented by the current frame data and a sixth obstacle represented by the historical frame data in a three-dimensional space, the second initial distance is standardized to obtain a first three-dimensional distance, and the first three-dimensional distance is used to obtain reference data for representing similarity between the current frame data and the historical frame data in a first three-dimensional distance evaluation dimension. The process does not involve a complex data processing flow, so that the efficiency of obtaining the reference data for representing the similarity between the current frame data and the historical frame data in the first three-dimensional distance evaluation dimension can be improved, thereby improving the execution efficiency of the data verification method. Moreover, the above process involves standardization of related data (for example, the second initial distance), so that the accuracy of the reference data for representing the similarity between the current frame data and the historical frame data in the first three-dimensional distance evaluation dimension can be improved, thereby further improving the accuracy of sensor data association.

[0202] In an optional embodiment, the first target sensor and the second target sensor are any combination of a radar and a camera, and the target evaluation dimension is a second three-dimensional distance evaluation dimension. The process of "mapping the current frame data and the historical frame data collected by the second target sensor to obtain reference data for representing similarity between the current frame data and the historical frame data in the target evaluation dimension" can include:

[0203] Mapping the current frame data and the historical frame data collected by the second target sensor to obtain a third initial distance of a seventh obstacle represented by the current frame data and an eighth obstacle represented by the historical frame data in a three-dimensional space;

[0204] Standardizing the third initial distance to obtain a second three-dimensional distance;

[0205] Based on the second three-dimensional distance, reference data for representing similarity between the current frame data and the historical frame data in a second three-dimensional distance evaluation dimension is obtained.

[0206] In the above process, the mapping of the current frame data and the historical frame data collected by the second target sensor can be performed after spatial mapping of the current frame data and the historical frame data. For example, the current frame data and the historical frame data can be mapped to the same world coordinate system for comparison.

[0207] Further, in an example, "mapping the current frame data and the historical frame data collected by the second target sensor to obtain a third initial distance of a seventh obstacle represented by the current frame data and an eighth obstacle represented by the historical frame data in a three-dimensional space" can include:

[0208] Determine the center point of the third target of the seventh obstacle represented by the current frame data;

[0209] Determine the center point of the fourth target of the eighth obstacle as represented by the historical frame data collected by the second target sensor;

[0210] Map the center points of the third and fourth targets to obtain the distance between them, which serves as the third initial distance between the seventh and eighth obstacles in three-dimensional space.

[0211] In one specific example, at least one of the following algorithms can be used: target detection algorithm, Kalman filter algorithm, clustering algorithm, fusion algorithm, etc., to determine the center point of the obstacle represented by the data collected by the radar in the current frame data and historical frame data; in another specific example, at least one of the following algorithms can be used: contour detection and shape analysis, target detection algorithm, semantic segmentation algorithm, etc., to determine the center point of the obstacle represented by the data collected by the camera in the current frame data and historical frame data.

[0212] After determining the third target center point of the seventh obstacle represented by the current frame data and the fourth target center point of the eighth obstacle represented by the historical frame data collected by the second target sensor, the Euclidean distance between the third target center point and the fourth target center point can be calculated as the interval distance between the third target center point and the fourth target center point obtained by mapping the third target center point and the fourth target center point. That is, it serves as the third initial distance between the seventh obstacle and the eighth obstacle in three-dimensional space.

[0213] Furthermore, it should be noted that, in this embodiment of the disclosure, considering the theoretical jitter range of the third initial distance between the seventh and eighth obstacles in three-dimensional space, in a practical application example, the probability distribution of the third initial distance between the seventh and eighth obstacles in three-dimensional space can be as follows: Figure 12 As shown ( Figure 12 In the diagram, the horizontal axis represents the third initial distance (in meters), and the vertical axis represents the probability density. Figure 12 Perform long-tail removal operations on the probability distribution shown (e.g., in...) Figure 12 In the process of removing the portion of the third initial distance greater than or equal to 6.5m and then redistributing the probabilities, we can obtain... Figure 13 This is used to characterize the probability distribution of the third initial distance between the seventh and eighth obstacles in three-dimensional space. Figure 13In the figure, the abscissa is used to represent the third initial distance (in m), and the ordinate is used to represent the probability density. The theoretical fluctuation range of the third initial distance of the seventh obstacle and the eighth obstacle in the three-dimensional space can be determined according to experimental values, and the embodiments of the present disclosure do not limit this.

[0214] After obtaining the third initial distance of the seventh obstacle and the eighth obstacle in the three-dimensional space represented by the current frame data and the historical frame data, the third initial distance can be standardized to obtain a second three-dimensional distance. This process can be represented as:

[0215] SPd(Postion31) = Pd(Postion31) / Lnv31

[0216] Wherein, SPd(Postion31) is used to represent the second three-dimensional distance; Pd(Postion31) is used to represent the third initial distance; Lnv31 is used to represent the third standardization parameter. Here, the third standardization parameter can be a value greater than the second three-dimensional distance, which can be set according to actual application requirements, and the embodiments of the present disclosure do not limit this.

[0217] After the third initial distance is standardized to obtain the second three-dimensional distance, the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension needs to be obtained based on the second three-dimensional distance. In an example, "obtaining reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension based on the second three-dimensional distance" can include:

[0218] In the case where the second three-dimensional distance is less than the first distance threshold, the first similarity representation value is determined as the first intermediate data;

[0219] Or, in the case where the second three-dimensional distance is greater than or equal to the first distance threshold, the second three-dimensional distance is processed by the Welsh loss function to obtain the first intermediate data;

[0220] Based on the first intermediate data, the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension is obtained.

[0221] Wherein, the first distance threshold and the first similarity representation value can be set according to actual application requirements, for example, the first distance threshold can be set to 0.05m, and the first similarity representation value can be set to a probability value close to 1, for example, the first similarity representation value can be set to 1-(1e-6), and the embodiments of the present disclosure do not limit this.

[0222] In the case that the second three-dimensional distance is greater than or equal to the first distance threshold, the second three-dimensional distance can be processed by a Welsh loss function to obtain first intermediate data. This process can be represented as:

[0223] ID1 = Exp[(SPd(Postion31)-Dth31) / Dsc31]

[0224] where ID1 is used to represent the first intermediate data obtained by processing the second three-dimensional distance by the Welsh loss function; SPd(Postion31) is used to represent the second three-dimensional distance; Dth31 is used to represent the first distance threshold; and Dsc31 is used to represent the first step parameter. Here, the first step parameter can be set according to actual application requirements, for example, the first step parameter can be set to 0.05 m, and the embodiments of the present disclosure do not limit this.

[0225] In the above example, in the case that the second three-dimensional distance is less than the first distance threshold, the first similarity representation value (i.e., a probability value close to 1) can be taken as the first intermediate data to represent that the current frame data and the historical frame data have high similarity in the second three-dimensional distance evaluation dimension; in the case that the second three-dimensional distance is greater than or equal to the first distance threshold, the second three-dimensional distance is processed by the Welsh loss function to obtain the first intermediate data, and specifically, the second three-dimensional distance can be sharply attenuated in an exponential decay manner to obtain a smaller first intermediate data to represent that the current frame data and the historical frame data have low similarity in the second three-dimensional distance evaluation dimension.

[0226] After obtaining the first intermediate data, the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension can be obtained based on the first intermediate data in the following manner:

[0227] In the case that the first intermediate data is less than or equal to the first similarity threshold, the first intermediate data is determined as the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension;

[0228] Or, in the case that the first intermediate data is greater than the first similarity threshold, the first intermediate data is compressed by a scaling function to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension.

[0229] where the first similarity threshold can be set according to actual application requirements, for example, the first distance threshold can be set to 0.5, and the embodiments of the present disclosure do not limit this.

[0230] Furthermore, in a specific example, "when the first intermediate data is less than or equal to the first similarity threshold, the first intermediate data is determined as reference data for characterizing the similarity between the current frame data and historical frame data in the second three-dimensional distance evaluation dimension" can be characterized as:

[0231] Rd5 = ID1

[0232] Rd5 is used to characterize reference data in the second three-dimensional distance evaluation dimension, which is used to characterize the similarity between the current frame data and the historical frame data; ID1 is used to characterize the first intermediate data.

[0233] "When the first intermediate data is greater than the first similarity threshold, the first intermediate data is compressed using a scaling function to obtain reference data for characterizing the similarity between the current frame data and historical frame data in the second three-dimensional distance evaluation dimension." This can be characterized as follows:

[0234] Rd5=[(ID1-Pth1) / (1-Pth1)](Pmax-Pth1)+Pth1

[0235] Wherein, Rd5 is used to characterize reference data used to characterize the similarity between the current frame data and historical frame data in the second three-dimensional distance evaluation dimension; ID1 is used to characterize the first intermediate data; Pth1 is used to characterize the first scaling threshold; and Pmax1 is used to characterize the first maximum scaling probability value. Here, the first scaling threshold and the first maximum scaling probability value can be set according to actual application requirements. For example, the first scaling threshold can be set to 0.5, and the first maximum scaling probability value can be set to 0.7. This embodiment of the present disclosure does not limit this.

[0236] Furthermore, in a practical example, the probability distribution of the third initial distance in three-dimensional space between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data is as follows: Figure 13 In the case shown, the probability distribution of the reference data used to characterize the similarity between the current frame data and historical frame data in the second three-dimensional distance evaluation dimension can be as follows: Figure 14 As shown ( Figure 14 In the diagram, the horizontal axis represents reference data used in the position evaluation dimension to represent the similarity between the current frame data and historical frame data in the second three-dimensional distance evaluation dimension, and the vertical axis represents the probability density.

[0237] By the above method, in the embodiment of the present disclosure, the current frame data and the historical frame data collected by the second target sensor can be mapped to obtain a third initial distance of the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data in the three-dimensional space, and the third initial distance is standardized to obtain a second three-dimensional distance. Based on the second three-dimensional distance, reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension is obtained. This process does not involve complex data processing procedures, so it can improve the efficiency of obtaining reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension, thereby improving the execution efficiency of the data verification method. Moreover, the above process involves standardization of related data (for example, the third initial distance), so it can also improve the accuracy of the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension, thereby further improving the accuracy of sensor data association.

[0238] In addition, by the above method, in the embodiment of the present disclosure, when the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension is obtained based on the second three-dimensional distance, the related data (for example, the second three-dimensional distance and the first intermediate data) is optimized by using the Welsh loss function and the scaling function in turn. Finally, the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension is obtained, so the robustness and rationality of the reference data for representing the similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension can be improved.

[0239] In an optional implementation, the first target sensor and the second target sensor are any combination of radar and camera, and the target evaluation dimension is the speed evaluation dimension. The step of "mapping the current frame data and the historical frame data collected by the second target sensor to obtain reference data for representing the similarity between the current frame data and the historical frame data in the target evaluation dimension" can include:

[0240] determining the data collected by the radar in the current frame data and the historical frame data collected by the second target sensor as third perception data, and determining the data collected by the camera in the current frame data and the historical frame data as fourth perception data;

[0241] obtaining a first driving speed of a ninth obstacle represented by the third perception data;

[0242] obtaining a second driving speed of a tenth obstacle represented by the fourth perception data;

[0243] The first driving speed and the second driving speed are mapped to obtain reference data in a speed evaluation dimension for representing similarity between the current frame data and the historical frame data.

[0244] The mapping of the first driving speed and the second driving speed can be determining a correlation between the first driving speed and the second driving speed, so as to obtain, based on the correlation between the first driving speed and the second driving speed, the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data.

[0245] Further, in an example, the mapping of the first driving speed and the second driving speed to obtain the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data can include:

[0246] (1) The first driving speed and the second driving speed are mapped to determine, in a case where the first driving speed and the second driving speed meet a first preset relationship, a second similarity representation value as the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data.

[0247] The first preset relationship can be that the first driving speed is less than a first speed threshold, and the second driving speed is greater than a second speed threshold, and the first speed threshold is less than the second speed threshold. Here, the first speed threshold and the second speed threshold can be set according to actual application requirements, for example, the first speed threshold can be set to 0.1 m / s, and the second speed threshold can be set to 1 m / s, and the embodiments of the present disclosure are also not limited to this. In addition, in the embodiments of the present disclosure, the second similarity representation value can be set according to actual application requirements, for example, the first speed threshold can be set to 0.00001, and the embodiments of the present disclosure are also not limited to this.

[0248] Further, in a specific example, the determination of the second similarity representation value as the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data in a case where the first driving speed and the second driving speed meet the first preset relationship can be represented as:

[0249] Rd6=SCV2

[0250] Wherein, Rd6 is used to represent the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data; SCV2 is used to represent the second similarity representation value.

[0251] (2) mapping the first driving speed and the second driving speed to obtain reference data in a speed evaluation dimension for representing similarity between the current frame data and the historical frame data based on an initial speed difference between the first driving speed and the second driving speed, in a case where it is determined that the first driving speed and the second driving speed meet the second preset relationship.

[0252] The second preset relationship can be that the second driving speed is greater than a second speed threshold, and at least one of the first driving speed and the second driving speed is greater than a third speed threshold, and the third speed threshold is greater than the second speed threshold. Here, the third speed threshold can be set according to actual application requirements, for example, the third speed threshold can be set to 2 m / s, and the embodiments of the present disclosure are not limited to this.

[0253] Further, in a specific example, "obtaining reference data in a speed evaluation dimension for representing similarity between the current frame data and the historical frame data based on an initial speed difference between the first driving speed and the second driving speed" can include:

[0254] standardizing the initial speed difference between the first driving speed and the second driving speed to obtain a target speed difference between the first driving speed and the second driving speed;

[0255] obtaining reference data in a speed evaluation dimension for representing similarity between the current frame data and the historical frame data based on the target speed difference.

[0256] In a more specific example, "standardizing the initial speed difference between the first driving speed and the second driving speed to obtain a target speed difference between the first driving speed and the second driving speed" can be represented as:

[0257] Target_diff(Speed) = M_diff(Speed) / Max(Speed)

[0258] Target_diff(Speed) is used to represent the target speed difference between the first driving speed and the second driving speed; M_diff(Speed) is used to represent the initial speed difference between the first driving speed and the second driving speed; and Max(Speed) is used to represent the driving speed with a larger value between the first driving speed and the second driving speed.

[0259] In addition, it should be noted that in the embodiments of the present disclosure, in the case of considering the theoretical jitter range of the initial speed difference between the first driving speed and the second driving speed, in an actual application example, the probability distribution of the initial speed difference between the first driving speed and the second driving speed can be as shown in Figure 15 Figure 15 ​In the figure, the abscissa is used to represent the initial speed difference (in m / s), and the ordinate is used to represent the probability density. The theoretical jitter range of the initial speed difference between the first driving speed and the second driving speed can be determined according to experimental values, and the embodiments of the present disclosure do not limit this.

[0260] It should be further pointed out that, in the embodiments of the present disclosure, based on the probability distribution of the initial speed difference between the first driving speed and the second driving speed, the probability distribution of the target speed difference between the first driving speed and the second driving speed can be obtained, which is not repeated here.

[0261] In a more specific example, "obtaining reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension based on the target speed difference" can include:

[0262] Obtaining a third chi-square statistic for the target speed difference;

[0263] Based on the third chi-square statistic, obtaining second intermediate data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension;

[0264] In the case where the second intermediate data is less than or equal to the second similarity threshold, the second intermediate data is determined as the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension;

[0265] Or, in the case where the second intermediate data is greater than the second similarity threshold, the second intermediate data is compressed by a scaling function to obtain the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension.

[0266] Wherein, "obtaining a third chi-square statistic for the target speed difference, and obtaining second intermediate data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension based on the third chi-square statistic" can be represented as:

[0267] ID2 = ChiSquaredCdf1TableFun[Target_diff(Speed)

[0268] 1-ChiSquaredCdf1TableFun[Target_diff(Speed) 2 / dev(Target_diff(Speed)) 2 ]

[0269] ID2, Target_diff(Speed), dev(Target_diff(Speed)), ChiSquaredCdf1TableFun[Target_diff(Speed) 2 dev(Target_diff(Speed)) 2 ] and Rd6 = [(ID2-Pth2) / (1-Pth2)](Pmax2-Pth2)+Pth2 2 dev(Target_diff(Speed)) 2 ), a third Chi-square statistic for the final speed is found from a Chi-square distribution table with 1 degree of freedom. For example, the third Chi-square statistic for the final speed is found from a Chi-square distribution table with 1 degree of freedom based on the final speed difference by interpolation to improve the efficiency of finding the third Chi-square statistic.

[0270] The standard deviation of the target speed difference can be obtained by calculation, for example, after obtaining the probability distribution of the target speed difference, the standard deviation of the target speed difference can be calculated based on the probability distribution of the target speed difference, which is not described here.

[0271] Thereafter, the second intermediate data can be determined as reference data for representing the similarity between the current frame data and the historical frame data in the speed evaluation dimension when the second intermediate data is less than or equal to the second similarity threshold. This process can be represented as:

[0272] Rd6 = ID2

[0273] Wherein, Rd6 is used to represent the reference data for representing the similarity between the current frame data and the historical frame data in the speed evaluation dimension; ID2 is used to represent the second intermediate data.

[0274] The second intermediate data can also be compressed by a scaling function to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the speed evaluation dimension when the second intermediate data is greater than the second similarity threshold. This process can be represented as:

[0275] Rd6 = [(ID2-Pth2) / (1-Pth2)](Pmax2-Pth2)+Pth2

[0276] wherein, Rd6 is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension; ID2 is used to represent the second intermediate data; Pth2 is used to represent the second scaling threshold; and Pmax2 is used to represent the second maximum scaling probability value. Here, the second scaling threshold and the second maximum scaling probability value can be set according to actual application requirements, for example, the second scaling threshold can be set to 0.5, and the second maximum scaling probability value can be set to 0.7, and the embodiments of the present disclosure do not make any restrictions in this regard.

[0277] (3) mapping the first driving speed and the second driving speed to determine the third similarity representation value as the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension in a case where the first driving speed and the second driving speed do not meet the first preset relationship and the second preset relationship.

[0278] wherein, the third similarity representation value is greater than the second similarity representation value. Here, the third similarity representation value can be set according to actual application requirements, for example, the third similarity representation value can be set to 0.5, and the embodiments of the present disclosure do not make any restrictions in this regard.

[0279] In a specific example, “determining the third similarity representation value as the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension in a case where the first driving speed and the second driving speed do not meet the first preset relationship and the second preset relationship” can be represented as:

[0280] Rd7 = SCV3

[0281] wherein, Rd7 is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension; and SCV3 is used to represent the third similarity representation value.

[0282] In addition, in an actual example, in a case where the probability distribution of the initial speed difference between the first driving speed and the second driving speed is as shown in Figure 15 , the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension can be as shown in Figure 16 . Figure 16 wherein, the abscissa is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the speed evaluation dimension, and the ordinate is used to represent the probability density.

[0283] By the above manner, in the embodiment of the present disclosure, the data collected by the radar in the current frame data and the historical frame data collected by the second target sensor can be determined as the third perception data, and the data collected by the camera in the current frame data and the historical frame data can be determined as the fourth perception data, the first driving speed of the ninth obstacle represented by the third perception data is obtained, and the second driving speed of the tenth obstacle represented by the fourth perception data is obtained, then the first driving speed and the second driving speed are mapped to obtain the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data. The process does not involve a complex data processing flow, so that the efficiency of obtaining the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data can be improved, thereby improving the execution efficiency of the data verification method. Moreover, when the first driving speed and the second driving speed are mapped to obtain the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data, the initial speed difference between the first driving speed and the second driving speed is also standardized, so that the accuracy of the reference data in the speed evaluation dimension for representing the similarity between the current frame data and the historical frame data can be improved, thereby further improving the accuracy of the sensor data association.

[0284] In an optional implementation, the first target sensor and the second target sensor are any combination of a radar and a camera, and the target evaluation dimension is the first projection position evaluation dimension. The "mapping the current frame data and the historical frame data collected by the second target sensor to obtain the reference data in the target evaluation dimension for representing the similarity between the current frame data and the historical frame data" can include:

[0285] Mapping the current frame data and the historical frame data collected by the second target sensor to obtain the fourth initial distance in the first direction (for example, the X-axis direction of the world coordinate system) between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor as the first one-way distance.

[0286] Based on the first one-way distance, the reference data in the first projection position evaluation dimension for representing the similarity between the current frame data and the historical frame data is obtained.

[0287] Wherein, the mapping of the current frame data and the historical frame data collected by the second target sensor can be performed after the spatial mapping of the current frame data and the historical frame data. For example, the current frame data and the historical frame data can be mapped to the same world coordinate system for comparison.

[0288] Furthermore, in one example, "mapping the current frame data and the historical frame data collected by the second target sensor to obtain the fourth initial distance in the first direction between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor" may include:

[0289] Determine the center point of the third target of the seventh obstacle represented by the current frame data;

[0290] Determine the center point of the fourth target of the eighth obstacle as represented by the historical frame data collected by the second target sensor;

[0291] Mapping the center points of the third and fourth targets yields the fourth initial distance between the seventh and eighth obstacles in the first direction.

[0292] In one specific example, at least one of the following algorithms can be used: target detection algorithm, Kalman filter algorithm, clustering algorithm, fusion algorithm, etc., to determine the center point of the obstacle represented by the data collected by the radar in the current frame data and historical frame data; in another specific example, at least one of the following algorithms can be used: contour detection and shape analysis, target detection algorithm, semantic segmentation algorithm, etc., to determine the center point of the obstacle represented by the data collected by the camera in the current frame data and historical frame data.

[0293] After determining the third target center point of the seventh obstacle represented by the current frame data and the fourth target center point of the eighth obstacle represented by the historical frame data collected by the second target sensor, the third target center point and the fourth target center point can be mapped to obtain the fourth initial distance between the seventh obstacle and the eighth obstacle in the first direction, which is used as the first unidirectional distance.

[0294] Furthermore, it should be noted that, considering the theoretical jitter range of the fourth initial distance, in a practical application example, the probability distribution of the fourth initial distance follows a Gaussian distribution, as shown in the following example. Figure 17 As shown ( Figure 17 In the diagram, the horizontal axis represents the fourth initial distance (in meters), and the vertical axis represents the probability density. The theoretical fluctuation range of the fourth initial distance can be determined based on experimental values, and this embodiment does not impose any limitations on it.

[0295] In one example, "obtaining reference data to characterize the similarity between the current frame data and historical frame data under the first projection position evaluation dimension based on the first unidirectional distance" may include:

[0296] If the first one-way distance is less than or equal to the second distance threshold, the fourth similarity representation value is determined as the third intermediate data.

[0297] Or, in the case that the first one-way distance is greater than the second distance threshold, the first one-way distance is processed by a Welsh loss function to obtain third intermediate data.

[0298] Based on the third intermediate data, reference data for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension is obtained.

[0299] The second distance threshold and the fourth similarity representation value can be set according to actual application requirements, for example, the second distance threshold can be set to 0.05 m, and the fourth similarity representation value is set to a probability value close to 1, for example, the fourth similarity representation value can be set to 1-(1e-6), and the embodiments of the present disclosure are not limited to this.

[0300] In the case that the first one-way distance is greater than the second distance threshold, the first one-way distance can be processed by a Welsh loss function to obtain third intermediate data. This process can be represented as:

[0301] ID3=Exp[(SPd(Postion33)-Dth33) / Dsc33]

[0302] Wherein, ID3 is used to represent the third intermediate data obtained by processing the first one-way distance by the Welsh loss function; SPd(Postion33) is used to represent the first one-way distance; Dth33 is used to represent the second distance threshold; Dsc33 is used to represent the second step parameter. Here, the second step parameter can be set according to actual application requirements, for example, the second step parameter can be set to 0.05 m, and the embodiments of the present disclosure are not limited to this.

[0303] In the above example, in the case that the first one-way distance is less than the second distance threshold, the fourth similarity representation value (i.e., a probability value close to 1) can be taken as the third intermediate data to represent that the current frame data and the historical frame data have high similarity in the first projection position evaluation dimension; in the case that the first one-way distance is greater than or equal to the second distance threshold, the first one-way distance is processed by a Welsh loss function to obtain third intermediate data, specifically, the first one-way distance can be sharply attenuated in an exponential decay manner to obtain a smaller third intermediate data to represent that the current frame data and the historical frame data have low similarity in the first projection position evaluation dimension.

[0304] After obtaining the third intermediate data, the reference data for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension can be obtained based on the third intermediate data in the following way:

[0305] In a case that the third intermediate data is less than or equal to a third similarity threshold, the third intermediate data is determined as reference data used for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension;

[0306] Alternatively, in a case that the third intermediate data is greater than the third similarity threshold, the third intermediate data is compressed by a scaling function to obtain the reference data used for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension.

[0307] The third similarity threshold can be set according to actual application requirements, for example, the third distance threshold can be set to 0.5, which is not limited by the embodiments of the present disclosure.

[0308] Further, in a specific example, "in a case that the third intermediate data is less than or equal to a third similarity threshold, the third intermediate data is determined as reference data used for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension" can be represented as:

[0309] Rd7=ID3

[0310] Wherein, Rd7 is used to represent the reference data used for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension; ID3 is used to represent the third intermediate data.

[0311] "In a case that the third intermediate data is greater than the third similarity threshold, the third intermediate data is compressed by a scaling function to obtain the reference data used for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension" can be represented as:

[0312] Rd7= [(ID3-Pth3) / (1-Pth3)](Pmax3-Pth3)+Pth3

[0313] Wherein, Rd7 is used to represent the reference data used for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension; ID3 is used to represent the third intermediate data; Pth3 is used to represent the third scaling threshold; Pmax3 is used to represent the third maximum scaling probability value. Here, the third scaling threshold and the third maximum scaling probability value can be set according to actual application requirements, for example, the third scaling threshold can be set to 0.5, and the third maximum scaling probability value can be set to 0.7, which is not limited by the embodiments of the present disclosure.

[0314] In addition, in an actual example, in the case of probability distribution of the fourth initial distance as Figure 17In the case shown, the probability distribution of the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension can be as shown in Figure 18 Figure 18 In the case shown, the horizontal coordinate is used to represent the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension, and the vertical coordinate is used to represent the probability density.

[0315] In the above manner, in the embodiments of the present disclosure, the current frame data and the historical frame data collected by the second target sensor can be mapped to obtain a fourth initial distance between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor in the first direction as a first one-way distance, and the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension is obtained based on the first one-way distance. This process does not involve complex data processing procedures, and therefore can improve the efficiency of obtaining the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension, thereby improving the execution efficiency of the data verification method.

[0316] In addition, in the above manner, in the embodiments of the present disclosure, when the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension is obtained based on the first one-way distance, the related data (for example, the first one-way distance and the third intermediate data) is optimized by using the Welsh loss function and the scaling function in sequence, and finally the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension is obtained. Therefore, the robustness and rationality of the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension can be improved.

[0317] In an optional implementation, the first target sensor and the second target sensor are any combination of a radar and a camera, and the target evaluation dimension is the second projection position evaluation dimension. The operation of "mapping the current frame data and the historical frame data collected by the second target sensor to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the target evaluation dimension" can include:

[0318] Mapping the current frame data and the historical frame data collected by the second target sensor to obtain a fifth initial distance between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor in a second direction (for example, the Y-axis direction of the world coordinate system);

[0319] Standardizing the fifth initial distance to obtain a second one-way distance;

[0320] ​Based on the second unidirectional distance, reference data is obtained in the second projection position evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

[0321] Mapping the current frame data to the historical frame data acquired by the second target sensor can be done by comparing the current frame data and the historical frame data after spatial mapping. For example, the current frame data and the historical frame data can be mapped to the same world coordinate system before comparison.

[0322] Furthermore, in one example, "mapping the current frame data and the historical frame data collected by the second target sensor to obtain the fifth initial distance in the second direction between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor" may include:

[0323] Determine the center point of the third target of the seventh obstacle represented by the current frame data;

[0324] Determine the center point of the fourth target of the eighth obstacle as represented by the historical frame data collected by the second target sensor;

[0325] Mapping the center points of the third and fourth targets yields the fifth initial distance between the seventh and eighth obstacles in the second direction.

[0326] In one specific example, at least one of the following algorithms can be used: target detection algorithm, Kalman filter algorithm, clustering algorithm, fusion algorithm, etc., to determine the center point of the obstacle represented by the data collected by the radar in the current frame data and historical frame data; in another specific example, at least one of the following algorithms can be used: contour detection and shape analysis, target detection algorithm, semantic segmentation algorithm, etc., to determine the center point of the obstacle represented by the data collected by the camera in the current frame data and historical frame data.

[0327] After determining the third target center point of the seventh obstacle represented by the current frame data and the fourth target center point of the eighth obstacle represented by the historical frame data collected by the second target sensor, the third target center point and the fourth target center point can be mapped to obtain the fifth initial distance between the seventh obstacle and the eighth obstacle in the second direction.

[0328] Furthermore, it should be noted that, considering the theoretical jitter range of the fifth initial distance, in a practical application example, the probability distribution of the fifth initial distance can be as follows: Figure 19 As shown ( Figure 19 In the diagram, the horizontal axis represents the fifth initial distance (in meters), and the vertical axis represents the probability density. The theoretical fluctuation range of the fifth initial distance can be determined based on experimental values, and this embodiment does not impose any limitations on it.

[0329] In an example, "standardizing the fifth initial distance to obtain a second standardized one-way distance" can include:

[0330] standardizing the fifth initial distance according to a conversion manner from Gaussian distribution to Chi-square distribution to obtain the second one-way distance.

[0331] In a specific example, the above process can be characterized as:

[0332] D_diff(Postion) = [D_diff(yd) 2 dev(D_diff(yd)) 2 ]

[0333] wherein D_diff(Postion) is used to represent the second one-way distance; D_diff(yd) is used to represent the fifth initial distance; and dev(D_diff(yd)) is used to represent the standard deviation of the fifth initial distance.

[0334] Here, the standard deviation of the fifth initial distance can be obtained by calculation, for example, after obtaining the probability distribution of the fifth initial distance, the standard deviation of the fifth initial distance can be calculated based on the probability distribution of the fifth initial distance, which is not described herein.

[0335] Further, in an example, after obtaining the second one-way distance, the reference data for representing the similarity between the current frame data and the historical frame data under the second projection position evaluation dimension can be obtained based on the second one-way distance in the following manner:

[0336] obtaining a fourth Chi-square statistic for the second one-way distance;

[0337] obtaining the reference data for representing the similarity between the current frame data and the historical frame data under the second projection position evaluation dimension based on the fourth Chi-square statistic.

[0338] In a specific example, the above process can be characterized as:

[0339] Rd8 = 1 - ChiSquaredCdf1TableFun(D_diff(Postion))

[0340] wherein, Rd8 is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the second projection position evaluation dimension; ChiSquaredCdf1TableFun(D_diff(Postion)) is used to represent the fourth chi-square statistic for the second one-way distance found from the chi-square distribution table with 1 degree of freedom based on the second one-way distance (i.e., D_diff(Postion)). In an example, the fourth chi-square statistic for the second one-way distance found from the chi-square distribution table with 1 degree of freedom based on the second one-way distance can be obtained by interpolation method to improve the efficiency of finding the fourth chi-square statistic.

[0341] In addition, in an actual example, in the case that the probability distribution of the fifth initial distance is as shown in FIG. 8B, Figure 19 the probability distribution of the reference data used to represent the similarity between the current frame data and the historical frame data in the second projection position evaluation dimension can be as shown in FIG. 8C, Figure 20 wherein, the abscissa is used to represent the reference data used to represent the similarity between the current frame data and the historical frame data in the second projection position evaluation dimension, and the ordinate is used to represent the probability density. Figure 20

[0342] In the above manner, in the embodiments of the present disclosure, the current frame data and the historical frame data collected by the second target sensor can be mapped to obtain the fifth initial distance of the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor in the second direction, and the fifth initial distance is standardized to obtain the second one-way distance, and then the reference data used to represent the similarity between the current frame data and the historical frame data in the second projection position evaluation dimension is obtained based on the second one-way distance. This process does not involve complex data processing procedures, and therefore, the efficiency of obtaining the reference data used to represent the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension can be improved, thereby improving the execution efficiency of the data verification method.

[0343] In some optional embodiments, step S103, i.e., "fusing the plurality of reference data to obtain the verification result for the current frame data" can include:

[0344] converting each of the plurality of reference data to obtain a plurality of log probabilities corresponding to the plurality of reference data one by one;

[0345] calculating the sum of the plurality of log probabilities as an overall log probability;

[0346] converting the overall log probability to a probability type parameter as the verification result for the current frame data.

[0347] ​In an example, "transforming each of the plurality of reference data into a log-odds to obtain a plurality of log-odds" can be represented as:

[0348] Log odds i = log(R di / (1-R di ) where Log odds i is used to represent the log-odds obtained by transforming the i th reference data in the plurality of reference data; R di is used to represent the i th reference data in the plurality of reference data.

[0349] In an example, after obtaining the plurality of log-odds corresponding to the plurality of reference data, the sum of the plurality of log-odds can be calculated as an overall log-odds, denoted as Log odds, and the overall log-odds can be transformed into a probability-type parameter as a verification result for the current frame data. Wherein, "transforming the log-odds value into a probability-type parameter as a verification result for the current frame data" can be represented as:

[0350] RD = e -Log odds Log odds -Log odds +1)

[0351] wherein, RD is used to represent the verification result for the current frame data; Log odds is used to represent the overall log-odds.

[0352] In the above manner, in the embodiments of the present disclosure, each of the plurality of reference data can be transformed to obtain a plurality of log-odds corresponding to the plurality of reference data, and the sum of the plurality of log-odds can be calculated as an overall log-odds, and then the log-odds value can be transformed into a probability-type parameter as a verification result for the current frame data. In this way, not only can the verification result for the current frame data be limited to a probability value in the interval [0, 1] to meet the actual application requirements, but also the complexity of the calculation can be reduced, thereby improving the execution efficiency of the data verification method.

[0353] In the following, step S103 will be described in conjunction with a specific example.

[0354] In this example, the plurality of sensors simultaneously includes a laser radar, a camera, and a radar, the first target sensor is the laser radar, and each sensor in the plurality of sensors except the first target sensor is taken as the second target sensor, i.e., the camera and the radar are taken as the second target sensor, respectively.

[0355] Suppose, when the camera is taken as the second target sensor, the following can be obtained:

[0356] (1) The reference data used to represent the similarity between the current frame data and the historical frame data in the shape evaluation dimension, denoted as Rd1;

[0357] (2) reference data for representing similarity between the current frame data and the historical frame data in the position evaluation dimension, denoted as Rd2.

[0358] Suppose that the radar is taken as the second target sensor, the following can be obtained:

[0359] (1) reference data for representing similarity between the current frame data and the historical frame data in the planar distance evaluation dimension, denoted as Rd3;

[0360] (2) reference data for representing similarity between the current frame data and the historical frame data in the first three-dimensional distance evaluation dimension, denoted as Rd4.

[0361] Thereafter, the reference data Rd1 is converted to obtain the log odds Log odds1; the reference data Rd2 is converted to obtain the log odds Log odds2; the reference data Rd3 is converted to obtain the log odds Log odds3; and the reference data Rd4 is converted to obtain the log odds Log odds4.

[0362] Then, the sum of the plurality of log odds is calculated as the overall log odds: Log odds=Log odds1+Log odds2+Log odds3+Log odds4.

[0363]

[0364] Finally, the overall log odds is converted to a probability type parameter as the verification result for the current frame data:

[0365] RD=e -Log odds / (e -Log odds +1)

[0366] After obtaining the verification result for the current frame data, a first verification result for the current frame data can be obtained in a case where the verification result is greater than the overall similarity threshold, to represent that the current frame data can be associated with the historical frame data; and a second verification result for the current frame data can be obtained in a case where the verification result is less than or equal to the overall similarity threshold, to represent that the current frame data cannot be associated with the historical frame data.

[0367] Please refer to Figure 21 for a scene schematic diagram of a data verification method provided by an embodiment of the present disclosure.

[0368] As described above, the data verification method provided by the embodiment of the present disclosure is applied to an electronic device. The electronic device can be a server, a workstation, a large computer, a general computer or other similar computing devices.​

[0369] The electronic device is used for:

[0370] obtaining current frame data collected by a first target sensor in a plurality of sensors; wherein the plurality of sensors are installed on a target object;

[0371] respectively taking each sensor except the first target sensor in the plurality of sensors as a second target sensor to obtain a plurality of reference data for representing similarity between the current frame data and historical frame data collected by the second target sensor; wherein the historical frame data is derived from global obstacle tracking information of the target object;

[0372] fusing the plurality of reference data to obtain a verification result for the current frame data; wherein the verification result is used to represent whether the current frame data can be associated with the historical frame data.

[0373] It should be noted that in the embodiments of the present disclosure, Figure 4 the scene schematic diagram shown is only illustrative and not restrictive, and those skilled in the art can make various obvious changes and / or replacements based on Figure 4 the examples, and the technical solutions obtained still belong to the disclosure range of the embodiments of the present disclosure.

[0374] In order to better implement the data verification method, the embodiments of the present disclosure also provide a data verification device 2200 which can be integrated in an electronic device. The electronic device can be a server, a workstation, a large computer, a general computer or other similar computing device. Hereinafter, the data verification device 2200 provided by the embodiments of the present disclosure will be described in conjunction with the schematic block diagram shown in the figure. Figure 22

[0375] The data verification device 2200 comprises:

[0376] a first data acquisition unit 2201 configured to obtain current frame data collected by a first target sensor in a plurality of sensors; wherein the plurality of sensors are installed on a target object;

[0377] a second data acquisition unit 2202 configured to respectively take each sensor except the first target sensor in the plurality of sensors as a second target sensor to obtain a plurality of reference data for representing similarity between the current frame data and historical frame data collected by the second target sensor; wherein the historical frame data is derived from global obstacle tracking information of the target object;

[0378] a data fusion unit 2203 configured to fuse the plurality of reference data to obtain a verification result for the current frame data; wherein the verification result is used to represent whether the current frame data can be associated with the historical frame data.

[0379] ​In some optional embodiments, the second data obtaining unit 2202 is configured to:

[0380] determine a plurality of evaluation dimensions based on the first target sensor and the second target sensor;

[0381] respectively take each of the plurality of evaluation dimensions as a target evaluation dimension to obtain reference data under the target evaluation dimension for characterizing similarity between the current frame data and historical frame data collected by the second target sensor.

[0382] In some optional embodiments, the second data obtaining unit 2202 is configured to perform one of:

[0383] In the case where the first target sensor and the second target sensor are any combination of a lidar and a camera, the plurality of evaluation dimensions include a shape evaluation dimension and a position evaluation dimension;

[0384] In the case where the first target sensor and the second target sensor are any combination of a lidar and a radar, the plurality of evaluation dimensions include a plane distance evaluation dimension and a first three-dimensional distance evaluation dimension;

[0385] In the case where the first target sensor and the second target sensor are any combination of a radar and a camera, the plurality of evaluation dimensions include a second three-dimensional distance evaluation dimension, a velocity evaluation dimension, a first projection position evaluation dimension, and a second projection position evaluation dimension.

[0386] In some optional embodiments, the second data obtaining unit 2202 is configured to:

[0387] determine an evaluation strategy corresponding to the target evaluation dimension;

[0388] perform mapping on the current frame data and the historical frame data collected by the second target sensor according to the evaluation strategy corresponding to the target evaluation dimension, to obtain reference data under the target evaluation dimension for characterizing similarity between the current frame data and the historical frame data.

[0389] In some optional embodiments, the first target sensor and the second target sensor are any combination of a lidar and a camera; and the target evaluation dimension is a shape evaluation dimension. Based on this, the second data obtaining unit 2202 is configured to:

[0390] perform mapping on the current frame data and the historical frame data collected by the second target sensor, to obtain shape difference between a first obstacle represented by the current frame data and a second obstacle represented by the historical frame data;

[0391] obtain reference data under the shape evaluation dimension for characterizing similarity between the current frame data and the historical frame data based on the shape difference.

[0392] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0393] map the current frame data and the historical frame data collected by the second target sensor to obtain a width difference between a first bounding box and a second bounding box, and a height difference between the first bounding box and the second bounding box; the first bounding box is a bounding box of a first obstacle represented by the current frame data; the second bounding box is a bounding box of a second obstacle represented by the historical frame data;

[0394] standardize the width difference and the height difference according to a conversion manner from a Gaussian distribution to a Chi-square distribution, to obtain a shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data.

[0395] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0396] obtain a first Chi-square statistic for the shape difference;

[0397] obtain, based on the first Chi-square statistic, reference data in a shape evaluation dimension for representing similarity between the current frame data and the historical frame data.

[0398] In some optional embodiments, the first target sensor and the second target sensor are any combination of a laser radar and a camera; and the target evaluation dimension is a position evaluation dimension. Based on this, the second data acquisition unit 2202 is configured to:

[0399] determine, as first perception data, data collected by the laser radar in the current frame data and the historical frame data collected by the second target sensor, and determine, as second perception data, data collected by the camera in the current frame data and the historical frame data;

[0400] map the first perception data and the second perception data to obtain a first plane distance between a third obstacle represented by the first perception data and a fourth obstacle represented by the second perception data in a two-dimensional plane;

[0401] obtain, based on the first plane distance, reference data in the position evaluation dimension for representing similarity between the current frame data and the historical frame data.

[0402] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0403] map the first perception data and the second perception data to determine a plurality of feature points in the first perception data that are located outside a third bounding box; the third bounding box is a bounding box of the fourth obstacle represented by the second perception data;

[0404] obtain a first overflow amount of the plurality of feature points in the first direction relative to the third bounding box, and a second overflow amount of the plurality of feature points in the second direction relative to the third bounding box;

[0405] normalize the first overflow amount and the second overflow amount according to a conversion mode from a Gaussian distribution to a Chi-square distribution, to obtain a first plane distance of the third obstacle and the fourth obstacle in the two-dimensional plane represented by the first perception data.

[0406] In some optional embodiments, the second data obtaining unit 2202 is configured to:

[0407] obtain a second Chi-square statistic for the first plane distance;

[0408] obtain, based on the second Chi-square statistic, reference data in the position evaluation dimension for representing the similarity between the current frame data and the historical frame data.

[0409] In some optional embodiments, the first target sensor and the second target sensor are any combination of a laser radar and a radar; and the target evaluation dimension is a plane distance evaluation dimension. Based on this, the second data obtaining unit 2202 is configured to:

[0410] map the current frame data and the historical frame data collected by the second target sensor to obtain a first initial distance of a fifth obstacle represented by the current frame data and a sixth obstacle represented by the historical frame data in a two-dimensional plane;

[0411] normalize the first initial distance to obtain a second plane distance;

[0412] obtain, based on the second plane distance, reference data in the plane distance evaluation dimension for representing the similarity between the current frame data and the historical frame data.

[0413] In some optional embodiments, the second data obtaining unit 2202 is configured to:

[0414] determine a first target vertex closest to the reference position on a fourth bounding box; wherein the fourth bounding box is a bounding box of the fifth obstacle represented by the current frame data;

[0415] determine a second target vertex closest to the reference position on a fifth bounding box; wherein the fifth bounding box is a bounding box of the sixth obstacle represented by the historical frame data collected by the second target sensor;

[0416] map the first target vertex and the second target vertex to obtain an interval distance between the first target vertex and the second target vertex as the first initial distance of the fifth obstacle and the sixth obstacle in the two-dimensional plane.

[0417] In some optional embodiments, the first target sensor and the second target sensor are any combination of a laser radar and a radar; and the target evaluation dimension is a first three-dimensional distance evaluation dimension. Based on this, the second data acquisition unit 2202 is configured to:

[0418] mapping the current frame data and the historical frame data collected by the second target sensor to obtain a second initial distance between a fifth obstacle represented by the current frame data and a sixth obstacle represented by the historical frame data in a three-dimensional space;

[0419] performing standardization processing on the second initial distance to obtain a first three-dimensional distance;

[0420] based on the first three-dimensional distance, obtaining reference data for representing similarity between the current frame data and the historical frame data in the first three-dimensional distance evaluation dimension.

[0421] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0422] determining a first target center point of the fifth obstacle represented by the current frame data;

[0423] determining a second target center point of the sixth obstacle represented by the historical frame data collected by the second target sensor;

[0424] mapping the first target center point and the second target center point to obtain an interval distance between the first target center point and the second target center point as the second initial distance between the fifth obstacle and the sixth obstacle in the three-dimensional space.

[0425] In some optional embodiments, the first target sensor and the second target sensor are any combination of a radar and a camera; and the target evaluation dimension is a second three-dimensional distance evaluation dimension. Based on this, the second data acquisition unit 2202 is configured to:

[0426] mapping the current frame data and the historical frame data collected by the second target sensor to obtain a third initial distance between a seventh obstacle represented by the current frame data and an eighth obstacle represented by the historical frame data in a three-dimensional space;

[0427] performing standardization processing on the third initial distance to obtain a second three-dimensional distance;

[0428] based on the second three-dimensional distance, obtaining reference data for representing similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension.

[0429] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0430] determining a third target center point of the seventh obstacle represented by the current frame data;

[0431] determining a fourth target center point of an eighth obstacle represented by the historical frame data collected by the second target sensor;

[0432] mapping the third target center point and the fourth target center point to obtain an interval distance between the third target center point and the fourth target center point as a third initial distance of the seventh obstacle and the eighth obstacle in the three-dimensional space.

[0433] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0434] determine the first intermediate data as the first similarity representation value in a case where the second three-dimensional distance is less than the first distance threshold value;

[0435] or, in a case where the second three-dimensional distance is greater than or equal to the first distance threshold value, process the second three-dimensional distance by a Welsh loss function to obtain the first intermediate data;

[0436] obtain reference data for representing similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension based on the first intermediate data.

[0437] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0438] determine the first intermediate data as the reference data for representing similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension in a case where the first intermediate data is less than or equal to the first similarity threshold value;

[0439] or, in a case where the first intermediate data is greater than the first similarity threshold value, compress the first intermediate data by a scaling function to obtain the reference data for representing similarity between the current frame data and the historical frame data in the second three-dimensional distance evaluation dimension.

[0440] In some optional embodiments, the first target sensor and the second target sensor are any combination of a radar and a camera; and the target evaluation dimension is a speed evaluation dimension. Based on this, the second data acquisition unit 2202 is configured to:

[0441] determine data collected by the radar in the current frame data and the historical frame data collected by the second target sensor as third perception data, and determine data collected by the camera in the current frame data and the historical frame data as fourth perception data;

[0442] obtain a first driving speed of a ninth obstacle represented by the third perception data;

[0443] obtain a second driving speed of a tenth obstacle represented by the fourth perception data;

[0444] The first driving speed and the second driving speed are mapped to obtain reference data in a speed evaluation dimension for representing similarity between the current frame data and the historical frame data.

[0445] In some optional embodiments, the second data acquisition unit 2202 is configured to perform one of the following:

[0446] The first driving speed and the second driving speed are mapped to determine, in a case where the first driving speed and the second driving speed meet a first preset relationship, the second similarity representation value as reference data in the speed evaluation dimension for representing similarity between the current frame data and the historical frame data; wherein the first preset relationship is that the first driving speed is less than a first speed threshold and the second driving speed is greater than a second speed threshold; the first speed threshold is less than the second speed threshold.

[0447] The first driving speed and the second driving speed are mapped to determine, in a case where the first driving speed and the second driving speed meet a second preset relationship, reference data in the speed evaluation dimension for representing similarity between the current frame data and the historical frame data based on an initial speed difference between the first driving speed and the second driving speed; wherein the second preset relationship is that the second driving speed is greater than the first speed threshold and at least one of the first driving speed and the second driving speed is greater than a third speed threshold; the third speed threshold is greater than the second speed threshold.

[0448] The first driving speed and the second driving speed are mapped to determine, in a case where the first driving speed and the second driving speed do not meet the first preset relationship and the second preset relationship, a third similarity representation value as reference data in the speed evaluation dimension for representing similarity between the current frame data and the historical frame data; wherein the third similarity representation value is greater than the second similarity representation value.

[0449] In some optional embodiments, the second data acquisition unit 2202 is configured to perform one of the following:

[0450] The initial speed difference between the first driving speed and the second driving speed is standardized to obtain a target speed difference between the first driving speed and the second driving speed.

[0451] The target speed difference is used to obtain reference data in the speed evaluation dimension for representing similarity between the current frame data and the historical frame data.

[0452] In some optional embodiments, the second data acquisition unit 2202 is configured to perform one of the following:

[0453] A third chi-square statistic is obtained for the target speed difference.

[0454] Based on the third chi-square statistics, second intermediate data for representing similarity between the current frame data and the historical frame data in the speed evaluation dimension is obtained.

[0455] In a case where the second intermediate data is less than or equal to a second similarity threshold, the second intermediate data is determined as reference data for representing similarity between the current frame data and the historical frame data in the speed evaluation dimension.

[0456] Or, in a case where the second intermediate data is greater than the second similarity threshold, the second intermediate data is compressed by a scaling function to obtain reference data for representing similarity between the current frame data and the historical frame data in the speed evaluation dimension.

[0457] In some optional embodiments, the first target sensor and the second target sensor are any combination of radars and cameras; and the target evaluation dimension is the first projection position evaluation dimension. Based on this, the second data acquisition unit 2202 is configured to:

[0458] map the current frame data and the historical frame data collected by the second target sensor to obtain a fourth initial distance between the seventh obstacle represented by the current frame data and an eighth obstacle represented by the historical frame data collected by the second target sensor in the first direction;

[0459] standardize the fourth initial distance to obtain a first one-way distance;

[0460] Based on the first one-way distance, reference data for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension is obtained.

[0461] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0462] determine a third target center point of the seventh obstacle represented by the current frame data;

[0463] determine a fourth target center point of the eighth obstacle represented by the historical frame data collected by the second target sensor;

[0464] map the third target center point and the fourth target center point to obtain the fourth initial distance between the seventh obstacle and the eighth obstacle in the first direction.

[0465] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0466] In a case where the first one-way distance is less than or equal to a second distance threshold, the fourth similarity representation value is determined as the third intermediate data;

[0467] Or, in a case where the first one-way distance is greater than the second distance threshold, the first one-way distance is processed by a Welsh loss function to obtain third intermediate data;

[0468] Based on the third intermediate data, reference data for representing similarity between the current frame data and the historical frame data in the first projection position evaluation dimension is obtained.

[0469] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0470] In a case where the third intermediate data is less than or equal to the third similarity threshold, the third intermediate data is determined as the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension;

[0471] Or, in a case where the third intermediate data is greater than the third similarity threshold, the third intermediate data is compressed by a scaling function to obtain the reference data for representing the similarity between the current frame data and the historical frame data in the first projection position evaluation dimension.

[0472] In some optional embodiments, the first target sensor and the second target sensor are any combination of a radar and a camera; and the target evaluation dimension is the second projection position evaluation dimension. Based on this, the second data acquisition unit 2202 is configured to:

[0473] Map the current frame data and the historical frame data collected by the second target sensor to obtain a fifth initial distance between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor in the second direction;

[0474] Standardize the fifth initial distance to obtain a second one-way distance;

[0475] Based on the second one-way distance, reference data for representing similarity between the current frame data and the historical frame data in the second projection position evaluation dimension is obtained.

[0476] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0477] Determine a third target center point of the seventh obstacle represented by the current frame data;

[0478] Determine a fourth target center point of the eighth obstacle represented by the historical frame data collected by the second target sensor;

[0479] Map the third target center point and the fourth target center point to obtain the fifth initial distance between the seventh obstacle and the eighth obstacle in the second direction.

[0480] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0481] The fifth initial distance is normalized according to a conversion mode from a Gaussian distribution to a Chi-square distribution, to obtain a second one-way distance.

[0482] In some optional embodiments, the second data acquisition unit 2202 is configured to:

[0483] A fourth Chi-square statistic for the second one-way distance is obtained.

[0484] Based on the fourth Chi-square statistic, reference data for representing similarity between the current frame data and the historical frame data in the second projection position evaluation dimension are obtained.

[0485] In some optional embodiments, the data fusion unit is configured to:

[0486] Each of the plurality of reference data is converted to obtain a plurality of log probabilities corresponding to the plurality of reference data one by one.

[0487] The sum of the plurality of log probabilities is calculated as an overall log probability.

[0488] The overall log probability is converted to a probability type parameter as a verification result for the current frame data.

[0489] In the embodiments of the present disclosure, the specific functions and examples of each unit in the data verification apparatus 2200 can be referred to the related descriptions of the corresponding steps in the foregoing data verification method embodiments, which will not be repeated here.

[0490] In the technical solutions of the present disclosure, the acquisition, storage and application of user personal information involved comply with relevant laws and regulations and do not violate public order and good customs.

[0491] According to the embodiments of the present disclosure, the present disclosure further provides an electronic device, an autonomous vehicle, a readable storage medium and a computer program product.

[0492] Figure 23 A schematic structural block diagram of an example electronic device 2300 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit implementations of the present disclosure described and / or claimed in this document.

[0493] like Figure 23 As shown, the electronic device 2300 includes a computing unit 2301, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 2302 or a computer program loaded from a storage unit 2308 into a random access memory (RAM) 2303. The RAM 2303 may also store various programs and data required for the operation of the electronic device 2300. The computing unit 2301, ROM 2302, and RAM 2303 are interconnected via a bus 2304. An input / output (I / O) interface 2305 is also connected to the bus 2304.

[0494] Multiple components in electronic device 2300 are connected to I / O interface 2305, including: input unit 2306, such as keyboard, mouse, etc.; output unit 2307, such as various types of renderers, speakers, etc.; storage unit 2308, such as disk, optical disk, etc.; and communication unit 2309, such as network card, modem, wireless transceiver, etc. Communication unit 2309 allows electronic device 2300 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0495] The computing unit 2301 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 2301 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Process (DSP), and any appropriate processors, controllers, microcontrollers, and the like. The computing unit 2301 performs various methods and processes described above, e.g., the data verification method. For example, in some embodiments, the data verification method can be implemented as a computer software program tangibly embodied in a machine-readable medium, e.g., the storage unit 2308. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 2300 via the ROM 2302 and / or the communication unit 2309. When the computer program is loaded onto the RAM 2303 and executed by the computing unit 2301, one or more steps of the data verification method described above can be performed. Alternatively, in other embodiments, the computing unit 2301 can be configured as the data verification method by any other appropriate means, e.g., by means of firmware.

[0496] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a Field Programmable Gate Array (FPGA), an Application Specific Integrated Circuit (ASIC), an Application Specific Standard Product (ASSP), a System on Chip (SOC), a Complex Programmable Logic Device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0497] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces the functions / operations specified in the flowcharts and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0498] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable storage medium can include, without limitation, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk, RAM, ROM, erasable programmable read-only memory (EPROM) or flash memory, fiber optics, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0499] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a rendering device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor for conveying information to a user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0500] The systems and techniques described here can be implemented in a computing system that includes a back end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front end component, e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here, or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0501] The computer system can include a user terminal and a server. The user terminal and the server are generally remote from each other and typically interact through a communication network. The relationship of user terminal and server is one of user and server, e.g., a client-server relationship. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0502] The embodiments of the present disclosure further provide an automatic driving vehicle, including the electronic device 2300.

[0503] The embodiments of the present disclosure further provide a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the data verification method.

[0504] The embodiments of the present disclosure further provide a computer program product, including a computer program, which, when executed by a processor, implements the data verification method.

[0505] It should be understood that the various forms of flow shown above can be used to reorder, add, or delete steps. For example, the steps described in the present disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, which are not limited herein. In addition, in the present disclosure, relationship terms such as "first", "second", "third", etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. In addition, in the present disclosure, "a plurality of" can be understood as at least two.

[0506] The foregoing detailed description has set forth various embodiments of the devices and / or methods via the use of specific terminology. However, embodiments thereof can be practiced with the exact description not being presented in detail. The term "device" as used herein should be interpreted to include devices and / or methods. The embodiments disclosed herein are illustrative of the principles of the present disclosure. Numerous modifications and adaptations will be readily apparent to those skilled in the art without departing from the spirit and scope of the present disclosure.

Claims

1. A data validation method, comprising: Acquire the current frame data collected by the first target sensor among multiple sensors; wherein, the multiple sensors are installed on the target object; Each of the plurality of sensors, except for the first target sensor, is designated as a second target sensor to obtain a plurality of reference data for characterizing the similarity between the current frame data and the historical frame data collected by the second target sensor; wherein, the historical frame data originates from the global obstacle tracking information of the target object, and the global obstacle tracking information stores sensor data that has already been associated; The multiple reference data are fused to obtain a verification result for the current frame data; wherein the verification result is used to characterize whether the current frame data can be associated with the historical frame data.

2. The method according to claim 1, wherein, The acquisition of multiple reference data used to characterize the similarity between the current frame data and historical frame data acquired by the second target sensor includes: Based on the first target sensor and the second target sensor, multiple evaluation dimensions are determined; Each of the multiple evaluation dimensions is taken as a target evaluation dimension to obtain reference data under the target evaluation dimension to characterize the similarity between the current frame data and the historical frame data collected by the second target sensor.

3. The method according to claim 2, wherein, Based on the first target sensor and the second target sensor, multiple evaluation dimensions are determined, including one of the following: When the first target sensor and the second target sensor are either a combination of LiDAR and a camera, the plurality of evaluation dimensions are determined to include shape evaluation dimension and position evaluation dimension; When the first target sensor and the second target sensor are either a combination of lidar and radar, the plurality of evaluation dimensions are determined to include a planar distance evaluation dimension and a first three-dimensional distance evaluation dimension; When the first target sensor and the second target sensor are either a radar or a camera, the plurality of evaluation dimensions are determined to include a second three-dimensional distance evaluation dimension, a velocity evaluation dimension, a first projection position evaluation dimension, and a second projection position evaluation dimension.

4. The method according to claim 2 or 3, wherein, The reference data used to characterize the similarity between the current frame data and the historical frame data collected by the second target sensor under the target evaluation dimension includes: Determine the evaluation strategy corresponding to the target evaluation dimensions; According to the evaluation strategy corresponding to the target evaluation dimension, the current frame data and the historical frame data collected by the second target sensor are mapped to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data.

5. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of LiDAR and a camera; the target evaluation dimension is a shape evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension for characterizing the similarity between the current frame data and the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data; Based on the shape differences, reference data is obtained under the shape evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

6. The method according to claim 5, wherein, The mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the width difference between the first bounding box and the second bounding box, and the height difference between the first bounding box and the second bounding box; wherein, the first bounding box is the bounding box of the first obstacle represented by the current frame data; and the second bounding box is the bounding box of the second obstacle represented by the historical frame data. The width and height differences are standardized according to the transformation method from Gaussian distribution to chi-square distribution to obtain the shape difference between the first obstacle represented by the current frame data and the second obstacle represented by the historical frame data.

7. The method according to claim 6, wherein, The reference data obtained based on the shape difference, used to characterize the similarity between the current frame data and the historical frame data in the shape evaluation dimension, includes: Obtain the first chi-square statistic for the shape difference; Based on the first chi-square statistic, reference data is obtained under the shape evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

8. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of LiDAR and a camera; the target evaluation dimension is a position evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The data collected by the lidar in the current frame data and the historical frame data collected by the second target sensor are determined as the first perception data, and the data collected by the camera in the current frame data and the historical frame data are determined as the second perception data; The first perception data and the second perception data are mapped to obtain the first plane distance in a two-dimensional plane between the third obstacle represented by the first perception data and the fourth obstacle represented by the second perception data. Based on the first planar distance, reference data is obtained under the position evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

9. The method according to claim 8, wherein, The mapping of the first sensing data and the second sensing data to obtain the first planar distance in a two-dimensional plane between the third obstacle represented by the first sensing data and the fourth obstacle represented by the second sensing data includes: The first perception data and the second perception data are mapped to determine multiple feature points in the first perception data that are located outside the third bounding box; wherein, the third bounding box is the bounding box of the fourth obstacle represented by the second perception data; Obtain a first overflow amount of the plurality of feature points relative to the third bounding box in a first direction, and a second overflow amount of the plurality of feature points relative to the third bounding box in a second direction; Following the transformation from Gaussian distribution to chi-square distribution, the first overflow and the second overflow are standardized to obtain the first planar distance in the two-dimensional plane between the third obstacle and the fourth obstacle represented by the first perception data.

10. The method according to claim 9, wherein, The reference data obtained based on the first planar distance, used to characterize the similarity between the current frame data and the historical frame data under the position evaluation dimension, includes: Obtain the second chi-square statistic for the first planar distance; Based on the second chi-square statistic, reference data is obtained under the location evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

11. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of lidar and radar; the target evaluation dimension is a planar distance evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the first initial distance in the two-dimensional plane between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data. The first initial distance is standardized to obtain the second planar distance; Based on the second planar distance, reference data is obtained under the planar distance evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

12. The method according to claim 11, wherein, The step of mapping the current frame data and the historical frame data collected by the second target sensor to obtain the first initial distance in the two-dimensional plane between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data includes: Determine the first target vertex on the fourth bounding box that is closest to the reference position; wherein, the fourth bounding box is the bounding box of the fifth obstacle represented by the current frame data; Determine the second target vertex on the fifth bounding box that is closest to the reference position; wherein, the fifth bounding box is the bounding box of the sixth obstacle represented by the historical frame data collected by the second target sensor; Map the first target vertex and the second target vertex to obtain the interval distance between the first target vertex and the second target vertex, which is used as the first initial distance between the fifth obstacle and the sixth obstacle in the two-dimensional plane.

13. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of lidar and radar; the target evaluation dimension is a first three-dimensional distance evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the second initial distance in three-dimensional space between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data. The second initial distance is standardized to obtain the first three-dimensional distance; Based on the first three-dimensional distance, reference data is obtained under the first three-dimensional distance evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

14. The method according to claim 13, wherein, The step of mapping the current frame data and the historical frame data collected by the second target sensor to obtain the second initial distance in three-dimensional space between the fifth obstacle represented by the current frame data and the sixth obstacle represented by the historical frame data includes: Determine the first target center point of the fifth obstacle represented by the current frame data; Determine the second target center point of the sixth obstacle as represented by the historical frame data collected by the second target sensor; Mapping the first target center point and the second target center point yields the interval distance between them, which serves as the second initial distance between the fifth obstacle and the sixth obstacle in three-dimensional space.

15. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of radar and a camera; the target evaluation dimension is a second three-dimensional distance evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the third initial distance in three-dimensional space between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data. The third initial distance is standardized to obtain the second three-dimensional distance; Based on the second three-dimensional distance, reference data is obtained under the second three-dimensional distance evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

16. The method according to claim 15, wherein, The mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the third initial distance in three-dimensional space between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data includes: Determine the third target center point of the seventh obstacle represented by the current frame data; Determine the center point of the fourth target of the eighth obstacle as represented by the historical frame data collected by the second target sensor; The center point of the third target and the center point of the fourth target are mapped to obtain the interval distance between the center point of the third target and the center point of the fourth target, which is used as the third initial distance between the seventh obstacle and the eighth obstacle in three-dimensional space.

17. The method according to claim 15, wherein, The reference data obtained based on the second three-dimensional distance, used to characterize the similarity between the current frame data and the historical frame data under the second three-dimensional distance evaluation dimension, includes: If the second three-dimensional distance is less than the first distance threshold, the first similarity representation value is determined as the first intermediate data; Alternatively, if the second three-dimensional distance is greater than or equal to the first distance threshold, the second three-dimensional distance is processed using the Welsh loss function to obtain the first intermediate data; Based on the first intermediate data, reference data is obtained in the second three-dimensional distance evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

18. The method according to claim 17, wherein, The reference data obtained based on the first intermediate data, in the second three-dimensional distance evaluation dimension, for characterizing the similarity between the current frame data and the historical frame data, includes: If the first intermediate data is less than or equal to the first similarity threshold, the first intermediate data is determined as reference data for characterizing the similarity between the current frame data and the historical frame data under the second three-dimensional distance evaluation dimension; Alternatively, if the first intermediate data is greater than the first similarity threshold, the first intermediate data can be compressed using a scaling function to obtain reference data in the second three-dimensional distance evaluation dimension used to characterize the similarity between the current frame data and the historical frame data.

19. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of radar and a camera; the target evaluation dimension is a velocity evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The data collected by radar in the current frame data and the historical frame data collected by the second target sensor are determined as the third perception data, and the data collected by the camera in the current frame data and the historical frame data are determined as the fourth perception data; Obtain the first speed of the ninth obstacle as represented by the third perception data; Obtain the second speed of the tenth obstacle as represented by the fourth perception data; The first driving speed and the second driving speed are mapped to obtain reference data under the speed evaluation dimension used to characterize the similarity between the current frame data and the historical frame data.

20. The method according to claim 19, wherein, The mapping between the first driving speed and the second driving speed to obtain reference data under the speed evaluation dimension for characterizing the similarity between the current frame data and the historical frame data includes one of the following: The first driving speed and the second driving speed are mapped so that, if the first driving speed and the second driving speed meet a first preset relationship, the second similarity characterization value is determined as reference data under the speed evaluation dimension to characterize the similarity between the current frame data and the historical frame data; wherein, the first preset relationship is that the first driving speed is less than a first speed threshold and the second driving speed is greater than a second speed threshold; the first speed threshold is less than the second speed threshold; The first driving speed and the second driving speed are mapped to obtain reference data for characterizing the similarity between the current frame data and the historical frame data under the speed evaluation dimension, based on the initial speed difference between the first driving speed and the second driving speed, provided that the first driving speed and the second driving speed meet a second preset relationship; wherein, the second preset relationship is that the second driving speed is greater than a first speed threshold, and at least one of the first driving speed and the second driving speed is greater than a third speed threshold; the third speed threshold is greater than the second speed threshold; The first driving speed and the second driving speed are mapped so that, if the first driving speed and the second driving speed do not conform to the first preset relationship and the second preset relationship, the third similarity characterization value is determined as the reference data used to characterize the similarity between the current frame data and the historical frame data under the speed evaluation dimension; wherein, the third similarity characterization value is greater than the second similarity characterization value.

21. The method according to claim 20, wherein, The reference data used to characterize the similarity between the current frame data and the historical frame data under the speed evaluation dimension, based on the initial speed difference between the first driving speed and the second driving speed, includes: The initial speed difference between the first driving speed and the second driving speed is standardized to obtain the target speed difference between the first driving speed and the second driving speed; Based on the target speed difference, reference data is obtained under the speed evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

22. The method according to claim 21, wherein, The reference data obtained based on the target speed difference, used to characterize the similarity between the current frame data and the historical frame data under the speed evaluation dimension, includes: Obtain the third chi-square statistic for the difference in the target velocity; Based on the third chi-square statistic, second intermediate data is obtained under the velocity evaluation dimension to characterize the similarity between the current frame data and the historical frame data; If the second intermediate data is less than or equal to the second similarity threshold, the second intermediate data is determined as reference data for characterizing the similarity between the current frame data and the historical frame data under the speed evaluation dimension; Alternatively, if the second intermediate data is greater than the second similarity threshold, the second intermediate data can be compressed using a scaling function to obtain reference data under the speed evaluation dimension used to characterize the similarity between the current frame data and the historical frame data.

23. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of radar and a camera; the target evaluation dimension is a first projection position evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the fourth initial distance in the first direction between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor, which is used as the first unidirectional distance. Based on the first unidirectional distance, reference data is obtained under the first projection position evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

24. The method according to claim 23, wherein, The mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the fourth initial distance in the first direction between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor includes: Determine the third target center point of the seventh obstacle represented by the current frame data; Determine the center point of the fourth target of the eighth obstacle as represented by the historical frame data collected by the second target sensor; Mapping the center points of the third and fourth targets yields the fourth initial distance between the seventh and eighth obstacles in the first direction.

25. The method according to claim 23, wherein, The reference data used to characterize the similarity between the current frame data and the historical frame data under the first projection position evaluation dimension, based on the first unidirectional distance, includes: If the first one-way distance is less than or equal to the second distance threshold, the fourth similarity representation value is determined as the third intermediate data; Alternatively, if the first one-way distance is greater than the second distance threshold, the first one-way distance can be processed using the Wales loss function to obtain the third intermediate data. Based on the third intermediate data, reference data is obtained under the first projection position evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

26. The method of claim 25, wherein, The reference data obtained based on the third intermediate data, used to characterize the similarity between the current frame data and the historical frame data under the first projection position evaluation dimension, includes: If the third intermediate data is less than or equal to the third similarity threshold, the third intermediate data is determined as reference data for characterizing the similarity between the current frame data and the historical frame data under the first projection position evaluation dimension; Alternatively, if the third intermediate data is greater than the third similarity threshold, the third intermediate data is compressed using a scaling function to obtain reference data in the first projection position evaluation dimension used to characterize the similarity between the current frame data and the historical frame data.

27. The method according to claim 4, wherein, The first target sensor and the second target sensor are either a combination of radar and a camera; the target evaluation dimension is a second projection position evaluation dimension; the mapping of the current frame data and the historical frame data collected by the second target sensor to obtain reference data under the target evaluation dimension used to characterize the similarity between the current frame data and the historical frame data includes: The current frame data and the historical frame data collected by the second target sensor are mapped to obtain the fifth initial distance in the second direction between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor. The fifth initial distance is standardized to obtain the second one-way distance; Based on the second unidirectional distance, reference data is obtained under the second projection position evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

28. The method according to claim 27, wherein, The mapping of the current frame data and the historical frame data collected by the second target sensor to obtain the fifth initial distance in the second direction between the seventh obstacle represented by the current frame data and the eighth obstacle represented by the historical frame data collected by the second target sensor includes: Determine the third target center point of the seventh obstacle represented by the current frame data; Determine the center point of the fourth target of the eighth obstacle as represented by the historical frame data collected by the second target sensor; Mapping the center points of the third and fourth targets yields the fifth initial distance between the seventh and eighth obstacles in the second direction.

29. The method according to claim 27, wherein, The standardization process for the fifth initial distance to obtain the second standardized one-way distance includes: The fifth initial distance is standardized according to the transformation method from Gaussian distribution to chi-square distribution to obtain the second one-way distance.

30. The method according to claim 29, wherein, The reference data used to characterize the similarity between the current frame data and the historical frame data under the second projection position evaluation dimension, based on the second unidirectional distance, includes: Obtain the fourth chi-square statistic for the second one-way distance; Based on the fourth chi-square statistic, reference data is obtained under the second projection position evaluation dimension to characterize the similarity between the current frame data and the historical frame data.

31. The method according to claim 1, wherein, The process of fusing the multiple reference data to obtain a verification result for the current frame data includes: Each of the multiple reference data is transformed to obtain multiple log probabilities that correspond one-to-one with the multiple reference data. Calculate the sum of the plurality of log probabilities as the overall log probability; The overall logarithmic probability is converted into a probabilistic parameter to serve as the verification result for the current frame data.

32. A data verification device, comprising: The first data acquisition unit is used to acquire the current frame data collected by the first target sensor among multiple sensors; wherein, the multiple sensors are installed on the target object; The second data acquisition unit is used to take each of the plurality of sensors other than the first target sensor as a second target sensor to acquire a plurality of reference data for characterizing the similarity between the current frame data and the historical frame data collected by the second target sensor; wherein the historical frame data comes from the global obstacle tracking information of the target object, and the global obstacle tracking information stores sensor data that has been associated. A data fusion unit is used to fuse the multiple reference data to obtain a verification result for the current frame data; wherein the verification result is used to characterize whether the current frame data can be associated with the historical frame data.

33. An electronic device, comprising: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 31.

34. An autonomous vehicle comprising the electronic device of claim 33.

35. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1 to 31.

36. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 31.

Images