Method and device for screening occlusion-free target, electronic equipment and storage medium

Abstract

This application relates to the field of target detection technology, and provides a method, apparatus, electronic device, and storage medium for screening unobstructed targets. The method includes: acquiring three-dimensional sensing data collected by a sensing device; performing target detection on the three-dimensional sensing data to determine the three-dimensional detection boxes corresponding to each target object contained in the three-dimensional sensing data; determining the occlusion angle and importance of each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object; and screening unobstructed targets from the target objects based on the occlusion angle and importance of each target object. Therefore, by distinguishing between occluded and unobstructed targets in the three-dimensional target detection results, the accuracy of target detection is improved, and the reliability of subsequent system decisions or data analysis relying on the target detection results is enhanced.

Description

Technical Field

[0001] This application belongs to the field of target detection technology, and in particular relates to a method, apparatus, electronic device, storage medium and computer program product for screening unobstructed targets. Background Technology

[0002] In fields such as assisted driving, autonomous driving, and road monitoring, it is typically necessary to collect data using various sensing devices such as radars, sensors, and cameras to provide data support for the system's functions. Taking autonomous vehicles as an example, vehicles can use the sensing data collected by various sensing devices installed in the vehicle to detect targets, and use the target detection results to monitor the road environment or make further decisions or data analyses. For 3D sensing data (such as laser point clouds, depth images, and 3D reconstructed images) collected by sensing devices such as LiDAR and depth cameras, 3D target detection algorithms can also be applied to achieve target detection on the 3D sensing data.

[0003] In related technologies, in 3D target detection, some targets are often occluded, which reduces the accuracy of target detection and affects the reliability of subsequent system decisions or data analysis. Summary of the Invention

[0004] This application provides a method, apparatus, electronic device, storage medium, and computer program product for screening unobstructed targets, which can effectively identify unobstructed and occluded targets in the target detection results, improve the accuracy of target detection, and enhance the reliability of subsequent system decisions or data analysis.

[0005] In a first aspect, embodiments of this application provide a method for screening unobstructed targets, comprising: acquiring three-dimensional sensing data collected by a sensing device; performing target detection on the three-dimensional sensing data to determine the three-dimensional detection boxes corresponding to each target object contained in the three-dimensional sensing data; determining the occlusion angle and importance of each target object according to the vertex coordinates of the three-dimensional detection box corresponding to each target object; and screening unobstructed targets from each target object according to the occlusion angle and importance of each target object.

[0006] In one possible implementation of the first aspect, the above-mentioned determination of the occlusion angle and importance of each target object based on the vertex coordinates of the 3D detection box corresponding to each target object includes:

[0007] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the top view vertex coordinates of the 3D detection box corresponding to each target object.

[0008] Based on the top view vertex coordinates and origin coordinates of the 3D detection box corresponding to each target object, the occlusion angle corresponding to each target object is determined. The origin coordinates refer to the coordinates of the sensing device in the 3D sensing data.

[0009] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the center distance corresponding to each target object;

[0010] The importance of each target object is determined based on its center distance.

[0011] Optionally, in another possible implementation of the first aspect, the number of the aforementioned targets is N, where N is an integer greater than 1; correspondingly, the determination of the occlusion angle corresponding to each target based on the top view vertex coordinates and origin coordinates of the 3D detection box corresponding to each target includes:

[0012] Based on the coordinates of each top view vertex and the origin corresponding to the i-th target object, determine the ray between each top view vertex and the origin, where i is an integer greater than or equal to 1 and less than or equal to N;

[0013] The maximum angle between each ray is determined as the blocking angle corresponding to the i-th target.

[0014] Optionally, in another possible implementation of the first aspect, the determination of the center distance corresponding to each target object based on the vertex coordinates of the 3D detection box corresponding to each target object includes:

[0015] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the coordinates of the center point corresponding to each target object;

[0016] The distance between the center point coordinates of each target object and the origin coordinates is determined as the center distance for each target object.

[0017] Optionally, in another possible implementation of the first aspect, the determination of the importance of each target object based on the center distance of each target object includes:

[0018] Obtain the maximum sensing distance corresponding to the sensing device;

[0019] The importance of each target is determined based on its center distance and maximum sensing distance.

[0020] Optionally, in another possible implementation of the first aspect, the number of the aforementioned target objects is N, where N is an integer greater than 1; correspondingly, the above-mentioned selection of unobstructed targets from among the target objects based on the occlusion angle and importance of each target object includes:

[0021] Based on the importance of each target, sort the targets in descending order;

[0022] Based on the occlusion angles corresponding to the i-th target and the j-th target, determine the angle intersection-union ratio between the i-th target and the j-th target. Here, the importance of the i-th target is greater than the importance of the j-th target, and the importance of both the i-th and j-th target is not 0. i is an integer greater than or equal to 1 and less than N, and j is an integer greater than i and less than or equal to N.

[0023] If the angle cross-union ratio between the i-th target and the j-th target is greater than the cross-union ratio threshold, the importance of the j-th target is set to 0.

[0024] Each target object with a non-zero importance value is identified as an unobstructed target.

[0025] Optionally, in another possible implementation of the first aspect, determining the angle intersection-over-union ratio between the i-th target and the j-th target based on the occlusion angles corresponding to the i-th target and the j-th target includes:

[0026] Transform the occlusion angles corresponding to the i-th target and the j-th target into polar coordinates to represent them, so as to determine the starting angle and ending angle of the first occlusion angle corresponding to the i-th target, and the starting angle and ending angle of the second occlusion angle corresponding to the j-th target.

[0027] Based on the starting angle of the first occlusion angle, the ending angle of the first occlusion angle, the starting angle of the second occlusion angle, and the ending angle of the second occlusion angle, determine the overlap angle and the angle union between the i-th target and the j-th target;

[0028] The angle intersection-union ratio between the i-th and j-th target objects is determined based on the ratio between the overlap angle and the angle union between the i-th and j-th target objects.

[0029] Secondly, embodiments of this application provide a device for screening unobstructed targets, comprising: an acquisition module for acquiring three-dimensional sensing data collected by a sensing device; a first determination module for performing target detection on the three-dimensional sensing data to determine the three-dimensional detection boxes corresponding to each target object contained in the three-dimensional sensing data; a second determination module for determining the occlusion angle and importance of each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object; and a screening module for screening unobstructed targets from each target object based on the occlusion angle and importance of each target object.

[0030] In one possible implementation of the second aspect, the second determining module includes:

[0031] The first determining unit is used to determine the top view vertex coordinates of the three-dimensional detection box corresponding to each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object.

[0032] The second determining unit is used to determine the occlusion angle corresponding to each target object based on the top view vertex coordinates and origin coordinates of the three-dimensional detection box corresponding to each target object. The origin coordinates refer to the coordinates of the sensing device in the three-dimensional sensing data.

[0033] The third determining unit is used to determine the center distance of each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object.

[0034] The fourth determining unit is used to determine the importance of each target object based on the center distance of each target object.

[0035] Optionally, in another possible implementation of the second aspect, the number of the aforementioned target objects is N, where N is an integer greater than 1; correspondingly, the aforementioned second determining unit is specifically used for:

[0036] Based on the coordinates of each top view vertex and the origin corresponding to the i-th target object, determine the ray between each top view vertex and the origin, where i is an integer greater than or equal to 1 and less than or equal to N;

[0037] The maximum angle between each ray is determined as the blocking angle corresponding to the i-th target.

[0038] Optionally, in another possible implementation of the first aspect, the aforementioned third determining unit is specifically used for:

[0039] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the coordinates of the center point corresponding to each target object;

[0040] The distance between the center point coordinates of each target object and the origin coordinates is determined as the center distance for each target object.

[0041] Optionally, in another possible implementation of the second aspect, the fourth determining unit is specifically used for:

[0042] Obtain the maximum sensing distance corresponding to the sensing device;

[0043] The importance of each target is determined based on its center distance and maximum sensing distance.

[0044] Optionally, in another possible implementation of the second aspect, the number of the aforementioned target objects is N, where N is an integer greater than 1; correspondingly, the aforementioned filtering module includes:

[0045] The sorting unit is used to sort the target objects in descending order according to the importance of each target object;

[0046] The fifth determining unit is used to determine the angle intersection-union ratio between the i-th target object and the j-th target object based on the occlusion angles corresponding to the i-th target object and the j-th target object, wherein the importance of the i-th target object is greater than the importance of the j-th target object, and the importance of both the i-th and j-th target objects is not 0, i is an integer greater than or equal to 1 and less than N, and j is an integer greater than i and less than or equal to N;

[0047] The sixth determining unit is used to set the importance of the j-th target to 0 when the angle intersection-union ratio between the i-th target and the j-th target is greater than the intersection-union ratio threshold.

[0048] The seventh determination unit is used to determine each target object with a non-zero importance as an unobstructed target.

[0049] Optionally, in another possible implementation of the second aspect, the fifth determining unit described above is specifically used for:

[0050] Transform the occlusion angles corresponding to the i-th target and the j-th target into polar coordinates to represent them, so as to determine the starting angle and ending angle of the first occlusion angle corresponding to the i-th target, and the starting angle and ending angle of the second occlusion angle corresponding to the j-th target.

[0051] Based on the starting angle of the first occlusion angle, the ending angle of the first occlusion angle, the starting angle of the second occlusion angle, and the ending angle of the second occlusion angle, determine the overlap angle and the angle union between the i-th target and the j-th target;

[0052] The angle intersection-union ratio between the i-th and j-th target objects is determined based on the ratio between the overlap angle and the angle union between the i-th and j-th target objects.

[0053] Thirdly, embodiments of this application provide an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned method for screening unobstructed targets.

[0054] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned method for screening unobstructed targets.

[0055] Fifthly, embodiments of this application provide a computer program product that, when run on an electronic device, causes the electronic device to perform the aforementioned method for screening unobstructed targets.

[0056] The beneficial effects of this application embodiment compared with the prior art are as follows: by determining the occlusion angle and importance of each target object according to the position of the 3D detection box corresponding to each target object in the 3D perception data, and then filtering out unoccluded targets according to the occlusion angle and importance of each target object, the accuracy of target detection is improved by distinguishing between occluded and unoccluded targets in the 3D target detection results, and the reliability of subsequent system decisions or data analysis that rely on the target detection results is improved. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a flowchart illustrating a method for screening unobstructed targets according to an embodiment of this application;

[0059] Figure 2 This is a schematic diagram of target detection results for three-dimensional perception data provided in an embodiment of this application;

[0060] Figure 3 This is a schematic diagram of target detection results and coordinate system for three-dimensional perception data provided in an embodiment of this application;

[0061] Figure 4 This is a top view schematic diagram of the three-dimensional detection frame corresponding to the target object provided in an embodiment of this application;

[0062] Figure 5 This is a schematic diagram illustrating an embodiment of this application for solving the occlusion angle corresponding to a target object;

[0063] Figures 6-12 This is a schematic diagram showing the transformation of the occlusion angle corresponding to the target object in polar coordinates according to an embodiment of this application;

[0064] Figure 13This is a schematic diagram of the structure of the unobstructed target screening device provided in the embodiments of this application;

[0065] Figure 14 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0066] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0067] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0068] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0069] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0070] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0071] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0072] The following description, with reference to the accompanying drawings, details the method, apparatus, electronic device, storage medium, and computer program for screening unobstructed targets provided in this application.

[0073] Figure 1 The diagram shows a flowchart of a method for screening unobstructed targets provided in an embodiment of this application.

[0074] like Figure 1 As shown, the method for filtering unobstructed targets includes the following steps:

[0075] Step 101: Acquire the 3D sensing data collected by the sensing device.

[0076] It should be noted that the unobstructed target screening method of this application embodiment can be executed by the unobstructed target screening device of this application embodiment. The unobstructed target screening device of this application embodiment can be configured in any electronic device to execute the unobstructed target screening method of this application embodiment. For example, the unobstructed target screening device of this application embodiment can be configured in the onboard unit (OBU) of a vehicle equipped with 3D perception devices such as LiDAR and depth cameras to process the range perception data collected by the perception devices in the vehicle, so as to distinguish between occluded targets and unobstructed targets among the identified targets, thereby improving the reliability of vehicle target recognition and driving decision-making, and improving vehicle driving safety.

[0077] The sensing device can refer to any type of sensing device that captures three-dimensional spatial data. For example, the sensing device in the embodiments of this application can be a lidar, a depth camera, or a device that renders and models two-dimensional data into three-dimensional spatial data.

[0078] The three-dimensional perception data can refer to any type of perception data containing three-dimensional spatial information. For example, in the embodiments of this application, the three-dimensional perception data can be laser point cloud data collected by lidar, depth image data collected by a depth camera, or three-dimensional image data generated after three-dimensional reconstruction of a two-dimensional image, etc. The embodiments of this application do not limit this.

[0079] In this embodiment of the application, the three-dimensional sensing data collected by the sensing device can be acquired in real time during the process of the sensing device collecting three-dimensional sensing data, and the acquired three-dimensional sensing data can be processed in real time; or, the three-dimensional sensing data collected by the sensing device can be acquired by the user according to the actual usage needs.

[0080] For example, the unobstructed target selection method of this application embodiment is integrated into the vehicle's OBU, and the vehicle is equipped with a LiDAR. The vehicle's OBU can then acquire and process the LiDAR point cloud data collected in real time. Alternatively, the unobstructed target selection method of this application embodiment can be integrated into a roadside monitoring system, and the roadside monitoring system includes a LiDAR installed on the roadside. The roadside monitoring system can then acquire and process the LiDAR point cloud data collected in real time. Furthermore, the unobstructed target selection method of this application embodiment can also be integrated into any type of electronic device as an application, and can acquire and process user-uploaded or selected 3D perception data.

[0081] Step 102: Perform target detection on the 3D perception data to determine the 3D detection box corresponding to each target object contained in the 3D perception data.

[0082] The type of target object may include vehicles, pedestrians, road facilities, etc., which are related to the actual application scenario. This application embodiment does not limit this.

[0083] The 3D detection bounding box corresponding to the target object can be determined based on the edges of the target object in the 3D sensing data, and is used to mark the position of the target object in the 3D sensing data. As an example, the shape of the 3D detection bounding box corresponding to the target object can be a cuboid containing 8 vertices.

[0084] In this embodiment of the application, after obtaining the three-dimensional perception data, any type of three-dimensional target detection algorithm can be used to perform three-dimensional target detection on the three-dimensional perception data in order to determine the three-dimensional detection box corresponding to each target object in the three-dimensional perception data.

[0085] For example, such as Figure 2The diagram shown is a schematic of the result after target detection on a frame of 3D perception data. The 3D perception data contains two targets, target 1 and target 2. The 3D detection box corresponding to target 1 is 3D detection box 1, and the 3D detection box corresponding to target 2 is 3D detection box 2.

[0086] It should be noted that in actual use, a suitable 3D target detection algorithm can be selected to perform target detection on the 3D sensing data based on the type of 3D sensing data, the actual detection requirements, and the specific application scenario. This application embodiment is not intended to limit this. For example, if the 3D sensing data is laser point cloud data, a deep learning point cloud detection algorithm (including but not limited to PointPillar) can be used for detection. Furthermore, the shape of the 3D detection box in this application embodiment is related to the target detection algorithm used, and this application embodiment does not limit this aspect.

[0087] Step 103: Determine the occlusion angle and importance of each target object based on the vertex coordinates of the 3D detection box corresponding to each target object.

[0088] The occlusion angle corresponding to the target object can refer to the maximum angular range occupied by the target object in the three-dimensional sensing data from the perspective of the sensing device, and can be used to characterize the maximum area occupied by the target object in the three-dimensional sensing data.

[0089] The importance of the target object can be used to measure its importance to target identification, subsequent decision-making, and data analysis.

[0090] In the embodiments of this application, since there may be mutual occlusion between the various target objects in a frame of three-dimensional perception data, some of the target objects are occluded. Therefore, it can be determined whether each target object is occluded based on the degree of overlap of each target object in the three-dimensional perception data and the distance relationship between the target object and the sensing device.

[0091] As one possible implementation, the presence or absence of occlusion of a target object can be comprehensively determined based on the positional relationship between the top views of the 3D detection boxes corresponding to each target object, and the distance between each target object and the sensing device. Therefore, the occlusion angle of the target object can be determined based on the top view of the 3D detection box corresponding to the target object; and since the closer the target object is to the sensing device, the higher its importance to target detection and subsequent system decisions, and when there is some overlap in the positions of the target objects, the occluded target object can be identified by combining the front-back relationship of the target objects. Therefore, in one possible implementation of this application embodiment, step 103 above may include:

[0092] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the top view vertex coordinates of the 3D detection box corresponding to each target object.

[0093] Based on the top view vertex coordinates and origin coordinates of the 3D detection box corresponding to each target object, the occlusion angle corresponding to each target object is determined. The origin coordinates refer to the coordinates of the sensing device in the 3D sensing data.

[0094] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the center distance corresponding to each target object;

[0095] The importance of each target object is determined based on its center distance.

[0096] Among them, the center distance corresponding to the target object can be used to represent the distance between the target object and the sensing device.

[0097] As one possible implementation, such as Figure 3 As shown, the coordinate system corresponding to the 3D sensing data is as follows: Figure 3 As shown, assuming the coordinates of four vertices A, B, C, and D in the 3D detection box 1 are A(x1,y1,z1), B(x2,y2,z1), C(x3,y3,z1), and D(x4,y4,z1) respectively, the Z-axis coordinates of these four vertices can be ignored, and only the X-axis and Y-axis coordinates of these four points are obtained as the top-view vertex coordinates of the 3D detection box 1. That is, the top-view vertex coordinates of the 3D detection box 1 are (x1,y1), (x2,y2), (x3,y3), and (x4,y4). Using the same method, the top-view vertex coordinates of the 3D detection box corresponding to each target object can be determined. Figure 4 As shown, Figure 3 The top view of the 3D detection frame corresponds to target object 1 and target object 2 in the image.

[0098] After determining the top-view vertex coordinates of the 3D detection box corresponding to each target object, for a given target object, the occlusion angle corresponding to that target object can be determined based on the top-view vertex coordinates and origin coordinates of the corresponding 3D detection box. In one possible implementation of this application embodiment, the number of the above-mentioned target objects is N, where N is an integer greater than 1; correspondingly, the above-mentioned determination of the occlusion angle corresponding to each target object based on the top-view vertex coordinates and origin coordinates of the 3D detection box corresponding to each target object includes:

[0099] Based on the coordinates of each top view vertex and the origin corresponding to the i-th target object, determine the ray between each top view vertex and the origin, where i is an integer greater than or equal to 1 and less than or equal to N;

[0100] The maximum angle between each ray is determined as the blocking angle corresponding to the i-th target.

[0101] As one possible implementation, such as Figure 5 As shown, the origin O is the location of the sensing device. The top view of the 3D detection frame 1 corresponding to the target object 1 contains four vertices A, B, C, and D. When determining the occlusion angle corresponding to the target object 1, rays can be drawn between the origin O and the four vertices A, B, C, and D respectively. Figure 5 As we know, the maximum included angle formed by rays OA, OB, OC, and OD is ∠BOC. Therefore, ∠BOC can be determined as the occlusion angle corresponding to target object 1. The top view of the 3D detection frame 2 corresponding to target object 2 contains four vertices E, F, G, and H. When determining the occlusion angle corresponding to target object 2, rays can be drawn between the origin O and the four vertices E, F, G, and H respectively. Figure 5 As is known, the maximum included angle formed by rays OE, OF, OG, and OH is ∠EOG. Therefore, ∠EOG can be determined as the occlusion angle corresponding to target object 2. Following the same method, the occlusion angles corresponding to each target object can be determined.

[0102] As one possible implementation, since the distance between the target and the sensing device can reflect the importance of the target in target recognition and subsequent system decisions, and when the occlusion angles of two targets overlap, the front-back relationship of the two targets can help determine which target is occluded, the importance of the target can be determined based on the distance between the target and the sensing device. Therefore, for a target, the center point coordinates of the target can be determined based on the vertex coordinates of the corresponding 3D detection box, thus determining the distance between the target and the sensing device. That is, in one possible implementation of this application embodiment, determining the center distance of each target based on the vertex coordinates of the corresponding 3D detection box can include:

[0103] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the coordinates of the center point corresponding to each target object;

[0104] The distance between the center point coordinates of each target object and the origin coordinates is determined as the center distance for each target object.

[0105] As an example, for a target object, the average coordinates of each vertex of the 3D detection box corresponding to the target object can be determined as the coordinates of the center point of the target object. Based on the distance formula between two points and the coordinates of the center point of the target object and the origin coordinates, the distance between the coordinates of the center point of the target object and the origin coordinates can be determined. Then, the distance between the coordinates of the center point of the target object and the origin coordinates can be determined as the center distance of the target object.

[0106] For example, if the coordinates of the eight vertices of the 3D bounding box corresponding to a target object are (x1, y1, z1), (x2, y2, z1), (x3, y3, z1), (x4, y4, z1), (x1, y1, z2), (x2, y2, z2), (x3, y3, z2), and (x4, y4, z2), then the coordinates of the center point of the target object can be determined as follows: That is, the coordinates of the center point corresponding to the target object are

[0107] As one possible approach, since the closer the target is to the sensing device, the higher its importance, the importance of the target can be determined based on the center distance of the target.

[0108] As an example, since the importance of a target is negatively correlated with the distance between the target and the sensing device, the reciprocal of the center distance of the target can be used to determine the importance of the target.

[0109] As an example, the center distance corresponding to the target object can be normalized to a specific interval (such as [0,1], [0,10], [0,100]), and the difference between the maximum value of the specific interval and the normalized center distance can be determined as the importance of the target object. For example, the center distance can be normalized to the interval [0,10]. If the normalized center distance is 4, then the importance of the target object can be 10-4=6.

[0110] It should be noted that in actual use, the above-mentioned specific range can be set according to actual needs and specific application scenarios, and this application embodiment does not limit this.

[0111] As an example, since sensing devices typically have a maximum sensing distance, they can only effectively sense targets within that maximum sensing distance. Therefore, the importance of a target can also be determined based on the relationship between the center distance of the target and the maximum sensing distance. That is, in one possible implementation of this application embodiment, determining the importance of each target based on its center distance includes:

[0112] Obtain the maximum sensing distance corresponding to the sensing device;

[0113] The importance of each target is determined based on its center distance and maximum sensing distance.

[0114] In one possible implementation of this application, for a target object, the difference between the maximum sensing distance of the sensing device and the importance of the target object can be determined first, and the ratio of this difference to the maximum sensing distance can be determined as the importance of the target object. That is, the importance of the target object can be determined by the following formula:

[0115]

[0116] Where s is the importance of the target object, and d max d represents the maximum sensing distance of the sensing device, and d represents the center distance of the target object.

[0117] It should be noted that the methods for determining the importance of a target object listed above are merely exemplary and should not be considered as limitations on this application. In actual use, an appropriate method can be selected to determine the importance of a target object according to actual needs and specific application scenarios, and this application embodiment does not limit this.

[0118] Step 104: Based on the occlusion angle and importance of each target object, filter out unoccluded targets from all target objects.

[0119] In the embodiments of this application, such as Figure 5 As shown, if the occlusion angles of two targets overlap, it indicates that, from the perspective of the sensing device, one target is occluding the other. Since the importance of a target reflects its distance from the sensing device, when determining if the occlusion angles of two targets overlap, their importance can be used to determine which target is occluded; the target with lower importance is considered occluded. Furthermore, the greater the overlap in the occlusion angles, the more the occluded target is, and the lower its effectiveness for target recognition and subsequent decision-making. Therefore, when the occluded portion of a target exceeds a certain threshold, it can be identified as an occluded target.

[0120] One possible approach is to compare the occlusion angles and importance of each target object pairwise when filtering for unoccluded targets. This determines whether occlusion exists between any two targets, and the unoccluded targets are identified based on the degree of occlusion and their relative positions. Specifically, for the first target object, we can first determine whether there is an overlap between the occlusion angles of the first and second targets, and whether the overlap exceeds an overlap threshold. If there is no overlap or the overlap is less than or equal to the overlap threshold, then it can be determined that the first target object does not occlude the second target object, and the second target object does not occlude the first target object. Therefore, both the first and second targets can be marked as unoccluded targets. If the occlusion angles of the first and second targets overlap and the overlap exceeds the overlap threshold, then it can be further determined whether the importance of the first target is greater than that of the second target. If the importance of the first target is greater than that of the second target, meaning the first target is closer to the sensing device and in front of the second target, then the first target can be marked as an unoccluded target, and the second target as an occluded target. Afterward, all targets can be sequentially compared with the first target in the same manner described above. The process of comparing targets continues until all targets have been traversed, or until the first target is marked as an occluded target. If the importance of the first target is less than or equal to the importance of the second target (meaning the second target is closer to the sensing device and in front of the first), then the first target can be marked as an occluded target, and the second target as an unoccluded target. At this point, it's confirmed that the first target is occluded, and there's no need to compare other targets with it further; the process of judging the first target can then end. After judging the first target, the remaining targets can be judged in the same way. It should be noted that if a target has already been marked as an occluded target, it is unnecessary to compare it with any other target in subsequent judgments. Finally, after traversing all targets in the manner described above, all targets marked as unobstructed can be identified as the selected unobstructed targets.

[0121] As one possible implementation, since only devices closer to the sensing device will obscure targets farther from the sensing device, the targets can be sorted by importance before filtering. Furthermore, to improve the computational efficiency of overlapping occlusion angles, the intersection-union ratio (IUU) of occlusion angles can be used to determine whether and to what extent the occlusion angles of two targets overlap, thereby further improving the reliability and efficiency of filtering unobstructed targets. Specifically, in one possible implementation of this application embodiment, the number of targets is N, where N is an integer greater than 1; correspondingly, step 104 may include:

[0122] Based on the importance of each target, sort the targets in descending order;

[0123] Based on the occlusion angles corresponding to the i-th target and the j-th target, determine the angle intersection-union ratio between the i-th target and the j-th target. Here, the importance of the i-th target is greater than the importance of the j-th target, and the importance of both the i-th and j-th target is not 0. i is an integer greater than or equal to 1 and less than N, and j is an integer greater than i and less than or equal to N.

[0124] If the angle cross-union ratio between the i-th target and the j-th target is greater than the cross-union ratio threshold, the importance of the j-th target is set to 0.

[0125] Each target object with a non-zero importance value is identified as an unobstructed target.

[0126] In one possible implementation of this application, the target objects can be sorted in descending order according to their importance, that is, the target object with the highest ranking is the target object closest to the sensing device. Then, it can be determined whether each target object is occluded based on the intersection-union ratio of the occlusion angles between each pair of target objects.

[0127] For the first target object, we can first compare it with the second target object. Specifically, we calculate the angle intersection-over-union (OCU) ratio between the occlusion angles of the first and second targets. If this OCU is greater than a threshold, we can determine that the first and second targets have a high degree of overlap, and the importance of the first target object is greater than that of the second target object (the first target object is in front of the second target object). Therefore, the importance of the second target object can be set to 0, thus identifying it as an occluded target. If the OCU is less than or equal to the threshold, we can determine that the first target object does not occlude the second target object, or that their overlap is low, and therefore no action needs to be taken against either the first or second target object. Then, we can use the same method to compare the first target object with the third to Nth targets sequentially to identify each occluded target blocked by the first target and set its importance to 0. After processing the first target, the same process can be applied to the second through Nth targets that have a non-zero importance value to distinguish between occluded targets and object-occluded targets among the N targets. After traversing all targets, all targets with a non-zero importance value can be identified as unoccluded targets.

[0128] It should be noted that when comparing target objects pairwise to determine whether there are obstructions between them, target objects with an importance of 0 do not need to be processed. For example, when comparing the first target with the second to Nth targets, if the importance of the second target has already been set to 0 (i.e., the second target is occluded by the first target, and the second target has been identified as an occluded target), and the importance of the third target is not 0 (not occluded by the first target), then after processing the first target, there is no need to process the second target. Instead, the third target is directly compared with the fourth to Nth targets whose importance is not 0 to identify the occluded targets among the fourth to Nth targets that are occluded by the third target. In addition, if there is a target with an importance of 0 among the fourth to Nth targets, then there is no need to compare the third target with the first target. The target can be skipped and the next target with a non-zero importance is compared with the third target.

[0129] Furthermore, to facilitate the calculation of the intersection-over-union ratio (IoU) of the occlusion angles, the occlusion angles corresponding to the target objects can be transformed into polar coordinates for representation, thereby further improving the reliability and efficiency of unobstructed target screening. Specifically, in one possible implementation of this application embodiment, determining the angular IoU between the i-th and j-th target objects based on the occlusion angles corresponding to the i-th and j-th target objects includes:

[0130] Transform the occlusion angles corresponding to the i-th target and the j-th target into polar coordinates to represent them, so as to determine the starting angle and ending angle of the first occlusion angle corresponding to the i-th target, and the starting angle and ending angle of the second occlusion angle corresponding to the j-th target.

[0131] Based on the starting angle of the first occlusion angle, the ending angle of the first occlusion angle, the starting angle of the second occlusion angle, and the ending angle of the second occlusion angle, determine the overlap angle and the angle union between the i-th target and the j-th target;

[0132] The angle intersection-union ratio between the i-th and j-th target objects is determined based on the ratio between the overlap angle and the angle union between the i-th and j-th target objects.

[0133] Wherein, the starting angle of the first occlusion angle is less than the ending angle of the first occlusion angle, and the starting angle of the second occlusion angle is less than the ending angle of the second occlusion angle.

[0134] As one possible implementation, for the i-th target object, the top view vertex coordinates of the two 3D detection boxes that form the occlusion angle corresponding to the i-th target object with respect to the origin can be determined first. These two top view vertex coordinates can then be transformed into polar coordinates to determine the polar radius and polar angle of the two top view vertex coordinates respectively. The smaller polar angle is determined as the starting angle of the first occlusion angle, and the larger polar angle is determined as the ending angle of the first occlusion angle. Similarly, the starting angle of the second occlusion angle and the ending angle of the second occlusion angle corresponding to the j-th target object can be determined in the same way.

[0135] After that, as Figure 6 As shown, if the starting angle of the first occlusion angle is greater than or equal to the ending angle of the second occlusion angle, then the overlap angle between the i-th target and the j-th target can be determined to be 0, and the difference between the ending angle of the first occlusion angle and the starting angle of the second occlusion angle can be determined as the union of the angles between the i-th target and the j-th target; or, as... Figure 7 As shown, if the starting angle of the second occlusion angle is greater than or equal to the ending angle of the first occlusion angle, the overlap angle between the i-th target and the j-th target can be determined to be 0, and the difference between the ending angle of the second occlusion angle and the starting angle of the first occlusion angle can be determined as the union of the angles between the i-th target and the j-th target.

[0136] like Figure 8As shown, if the starting angle of the first occlusion angle is less than the ending angle of the second occlusion angle and greater than or equal to the starting angle of the second occlusion angle, and the ending angle of the first occlusion angle is greater than or equal to the ending angle of the second occlusion angle, then the difference between the ending angle of the second occlusion angle and the starting angle of the first occlusion angle can be determined as the overlap angle between the i-th target and the j-th target, and the difference between the ending angle of the first occlusion angle and the starting angle of the second occlusion angle can be determined as the angle union between the i-th target and the j-th target; or, as... Figure 9 As shown, if the starting angle of the second occlusion angle is less than the ending angle of the first occlusion angle and greater than or equal to the starting angle of the first occlusion angle, and the ending angle of the second occlusion angle is greater than or equal to the ending angle of the first occlusion angle, then the difference between the ending angle of the first occlusion angle and the starting angle of the second occlusion angle can be determined as the overlap angle between the i-th target and the j-th target, and the difference between the ending angle of the second occlusion angle and the starting angle of the first occlusion angle can be determined as the angle union between the i-th target and the j-th target.

[0137] like Figure 10 As shown, if the starting angle of the first occlusion angle is greater than the starting angle of the second occlusion angle, and the ending angle of the first occlusion angle is less than the ending angle of the second occlusion angle, then the difference between the ending angle and the starting angle of the first occlusion angle can be determined as the overlap angle between the i-th target and the j-th target, and the difference between the ending angle and the starting angle of the second occlusion angle can be determined as the angle union between the i-th target and the j-th target; or, as... Figure 11 As shown, if the starting angle of the second occlusion angle is greater than the starting angle of the first occlusion angle and the ending angle of the second occlusion angle is less than the ending angle of the first occlusion angle, then the difference between the ending angle of the second occlusion angle and the starting angle of the second occlusion angle can be determined as the overlap angle between the i-th target and the j-th target, and the difference between the ending angle of the first occlusion angle and the starting angle of the first occlusion angle can be determined as the angle union between the i-th target and the j-th target.

[0138] After determining the overlap angle and the union of angles between the i-th target and the j-th target, the ratio between the overlap angle and the union of angles can be determined as the angle intersection-union ratio between the i-th target and the j-th target.

[0139] It should be noted that, Figures 6-11 The solid lines in the diagram represent the starting angle and ending angle of the first occlusion angle corresponding to the i-th target object, while the dashed lines represent the starting angle and ending angle of the second occlusion angle corresponding to the j-th target object.

[0140] by Figure 5 For example, let's consider the occlusion angle ∠BOC corresponding to object 1. Assume the coordinates of vertex B in the top view are (x...).B ,y B The coordinates of vertex C in the top view are (x...). C ,y C Then, the top view vertex B(x) can be... B ,y B ) and top view vertex C(x) c ,y c Transform the coordinates to polar coordinates to obtain the polar coordinates (ρ) of vertex B in the top view. B ,θ B ), and the polar coordinates (ρ) of vertex C in the top view. C ,θ C ), where ρ B Let θ be the polar radius of vertex B in the top view. B Let ρ be the polar angle of vertex B in the top view. C The polar radius θ of vertex C in the top view C This is the polar angle of vertex C in the top view. For example... Figure 12 As shown, if θ C Less than θ B Then the starting angle of the first occlusion angle corresponding to the occlusion angle ∠BOC can be θ. C The termination angle of the first occlusion angle corresponding to the occlusion angle ∠BOC can be θ. B Following the same method, the starting angle of the second occlusion angle corresponding to the occlusion angle ∠EOG of target object 2 can be determined as θ. G and the termination angle of the second blocking angle is θ E From 12, we know that the starting angle of the first blocking angle is θ. C Less than the second blocking angle termination angle θ E And greater than the starting angle θ of the second occlusion angle G And the first blocking angle terminates at angle θ B Greater than the second blocking angle termination angle θ E Then it can be determined that the occlusion angles of the two intersect, thus allowing θ to be determined. E -θ C The overlap angle between target object 1 and target object 2 is determined, and θ can be... B -θ G The angle union between target object 1 and target object 2 is determined. Therefore, the angle intersection-union ratio between target object 1 and target object 2 can be determined as follows:

[0141] The method for screening unobstructed targets provided in this application determines the occlusion angle and importance of each target object based on the position of the 3D detection box corresponding to each target object in the 3D perception data. Then, it filters out unobstructed targets based on the occlusion angle and importance of each target object. By distinguishing between occluded and unobstructed targets in the 3D target detection results, the accuracy of target detection is improved, and the reliability of subsequent system decisions or data analysis that rely on the target detection results is enhanced.

[0142] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0143] Corresponding to the unobstructed target screening method described in the above embodiments, Figure 13 A schematic diagram of the structure of the target screening device without obstruction provided in the embodiment of this application is shown. For ease of explanation, only the parts related to the embodiment of this application are shown.

[0144] Reference Figure 13 The device 1300 includes:

[0145] The acquisition module 1301 is used to acquire the three-dimensional sensing data collected by the sensing device;

[0146] The first determining module 1302 is used to perform target detection on the three-dimensional perception data to determine the three-dimensional detection box corresponding to each target object contained in the three-dimensional perception data.

[0147] The second determining module 1303 is used to determine the occlusion angle and importance of each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object.

[0148] The filtering module 1304 is used to filter out unobstructed targets from each target object based on the occlusion angle and importance of each target object.

[0149] The unobstructed target screening device provided in this application determines the occlusion angle and importance of each target object based on the position of the 3D detection box corresponding to each target object in the 3D perception data. Then, it filters out unobstructed targets based on the occlusion angle and importance of each target object. By distinguishing between occluded and unobstructed targets in the 3D target detection results, the accuracy of target detection is improved, and the reliability of subsequent system decisions or data analysis that rely on the target detection results is enhanced.

[0150] In one possible implementation of this application, the second determining module 1303 includes:

[0151] The first determining unit is used to determine the top view vertex coordinates of the three-dimensional detection box corresponding to each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object.

[0152] The second determining unit is used to determine the occlusion angle corresponding to each target object based on the top view vertex coordinates and origin coordinates of the three-dimensional detection box corresponding to each target object. The origin coordinates refer to the coordinates of the sensing device in the three-dimensional sensing data.

[0153] The third determining unit is used to determine the center distance of each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object.

[0154] The fourth determining unit is used to determine the importance of each target object based on the center distance of each target object.

[0155] Furthermore, in another possible implementation of this application, the quantity of the aforementioned target objects is N, where N is an integer greater than 1; correspondingly, the aforementioned second determining unit is specifically used for:

[0156] Based on the coordinates of each top view vertex and the origin corresponding to the i-th target object, determine the ray between each top view vertex and the origin, where i is an integer greater than or equal to 1 and less than or equal to N;

[0157] The maximum angle between each ray is determined as the blocking angle corresponding to the i-th target.

[0158] Furthermore, in another possible implementation of this application, the aforementioned third determining unit is specifically used for:

[0159] Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the coordinates of the center point corresponding to each target object;

[0160] The distance between the center point coordinates of each target object and the origin coordinates is determined as the center distance for each target object.

[0161] Furthermore, in yet another possible implementation of this application, the aforementioned fourth determining unit is specifically used for:

[0162] Obtain the maximum sensing distance corresponding to the sensing device;

[0163] The importance of each target is determined based on its center distance and maximum sensing distance.

[0164] Furthermore, in yet another possible implementation of this application, the quantity of the aforementioned target objects is N, where N is an integer greater than 1; correspondingly, the aforementioned filtering module 1304 includes:

[0165] The sorting unit is used to sort the target objects in descending order according to the importance of each target object;

[0166] The fifth determining unit is used to determine the angle intersection-union ratio between the i-th target object and the j-th target object based on the occlusion angles corresponding to the i-th target object and the j-th target object, wherein the importance of the i-th target object is greater than the importance of the j-th target object, and the importance of both the i-th and j-th target objects is not 0, i is an integer greater than or equal to 1 and less than N, and j is an integer greater than i and less than or equal to N;

[0167] The sixth determining unit is used to set the importance of the j-th target to 0 when the angle intersection-union ratio between the i-th target and the j-th target is greater than the intersection-union ratio threshold.

[0168] The seventh determination unit is used to determine each target object with a non-zero importance as an unobstructed target.

[0169] Furthermore, in another possible implementation of this application, the aforementioned fifth determining unit is specifically used for:

[0170] Transform the occlusion angles corresponding to the i-th target and the j-th target into polar coordinates to represent them, so as to determine the starting angle and ending angle of the first occlusion angle corresponding to the i-th target, and the starting angle and ending angle of the second occlusion angle corresponding to the j-th target.

[0171] Based on the starting angle of the first occlusion angle, the ending angle of the first occlusion angle, the starting angle of the second occlusion angle, and the ending angle of the second occlusion angle, determine the overlap angle and the angle union between the i-th target and the j-th target;

[0172] The angle intersection-union ratio between the i-th and j-th target objects is determined based on the ratio between the overlap angle and the angle union between the i-th and j-th target objects.

[0173] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0174] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0175] To implement the above embodiments, this application also proposes an electronic device.

[0176] Figure 14 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application.

[0177] like Figure 14 As shown, the above-mentioned electronic device 200 includes:

[0178] The system includes a memory 210 and at least one processor 220, and a bus 230 connecting the different components (including the memory 210 and the processor 220). The memory 210 stores a computer program, which, when executed by the processor 220, implements the method for screening unobstructed targets as described in the embodiments of this application.

[0179] Bus 230 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0180] Electronic device 200 typically includes a variety of electronic device readable media. These media can be any available media that can be accessed by electronic device 200, including volatile and non-volatile media, removable and non-removable media.

[0181] Memory 210 may also include computer system readable media in the form of volatile memory, such as random access memory (RAM) 240 and / or cache memory 250. Electronic device 200 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 260 may be used to read and write non-removable, non-volatile magnetic media (… Figure 14 Not shown; usually referred to as a "hard drive"). Although Figure 14 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 230 via one or more data media interfaces. Memory 210 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this application.

[0182] A program / utility 280 having a set (at least one) of program modules 270 may be stored in, for example, memory 210. Such program modules 270 include—but are not limited to—an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 270 typically perform the functions and / or methods described in the embodiments of this application.

[0183] Electronic device 200 can also communicate with one or more external devices 290 (e.g., keyboard, pointing device, display 291, etc.), and with one or more devices that enable a user to interact with electronic device 200, and / or with any device that enables electronic device 200 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 292. Furthermore, electronic device 200 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 293. As shown, network adapter 293 communicates with other modules of electronic device 200 via bus 230. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 200, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0184] The processor 220 performs various functional applications and data processing by running programs stored in the memory 210.

[0185] It should be noted that the implementation process and technical principles of the electronic device in this embodiment are explained in the foregoing description of the screening method for unobstructed targets in the embodiments of this application, and will not be repeated here.

[0186] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0187] This application provides a computer program product that, when run on an electronic device, enables the electronic device to perform the steps described in the various method embodiments above.

[0188] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a device / electronic device, a recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0189] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0190] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0191] In the embodiments provided in this application, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings or direct couplings or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0192] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0193] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for screening unobstructed targets, characterized in that, include: Acquire 3D sensing data collected by sensing devices; Target detection is performed on the three-dimensional perception data to determine the three-dimensional detection box corresponding to each target object contained in the three-dimensional perception data; Based on the vertex coordinates of the 3D detection box corresponding to each target object, the occlusion angle and importance of each target object are determined respectively; Based on the occlusion angle and importance of each target object, unoccluded targets are selected from the target objects.

2. The method as described in claim 1, characterized in that, The step of determining the occlusion angle and importance of each target object based on the vertex coordinates of the 3D detection bounding box corresponding to each target object includes: Based on the vertex coordinates of the 3D detection box corresponding to each target object, determine the top view vertex coordinates of the 3D detection box corresponding to each target object. Based on the top view vertex coordinates and origin coordinates of the three-dimensional detection box corresponding to each target object, the occlusion angle corresponding to each target object is determined respectively, wherein the origin coordinates refer to the coordinates of the sensing device in the three-dimensional sensing data; Based on the vertex coordinates of the three-dimensional detection box corresponding to each target object, the center distance corresponding to each target object is determined respectively; The importance of each target object is determined based on the center distance of each target object.

3. The method as described in claim 2, characterized in that, The number of target objects is N, where N is an integer greater than 1. The step of determining the occlusion angle corresponding to each target object based on the top-view vertex coordinates and origin coordinates of the 3D detection box corresponding to each target object includes: Based on the coordinates of each top view vertex and the origin corresponding to the i-th target object, determine the ray between each top view vertex and the origin, where i is an integer greater than or equal to 1 and less than or equal to N; The maximum angle between each of the rays is determined as the blocking angle corresponding to the i-th target.

4. The method as described in claim 2, characterized in that, The step of determining the center distance corresponding to each target object based on the vertex coordinates of the 3D detection box corresponding to each target object includes: Based on the vertex coordinates of the 3D detection box corresponding to each target object, the center point coordinates corresponding to each target object are determined respectively; The distance between the center point coordinates of each target object and the origin coordinates is determined as the center distance for each target object.

5. The method as described in claim 2, characterized in that, The step of determining the importance of each target object based on its center distance includes: Obtain the maximum sensing distance corresponding to the sensing device; The importance of each target object is determined based on the center distance and the maximum sensing distance.

6. The method according to any one of claims 1-5, characterized in that, The number of target objects is N, where N is an integer greater than 1. The step of filtering unobstructed targets from among the target objects based on the occlusion angle and importance of each target object includes: Based on the importance of each target object, the target objects are sorted in descending order; Based on the occlusion angles corresponding to the i-th and j-th target objects, the angle intersection-union ratio between the i-th and j-th target objects is determined, wherein the importance of the i-th target object is greater than the importance of the j-th target object, and the importance of both the i-th and j-th target objects is not 0, i is an integer greater than or equal to 1 and less than N, and j is an integer greater than i and less than or equal to N; If the angle intersection-union ratio between the i-th target and the j-th target is greater than the intersection-union ratio threshold, the importance of the j-th target is set to 0. Each of the target objects whose importance is not 0 is determined as the unobstructed target.

7. The method as described in claim 6, characterized in that, The step of determining the angle intersection-over-union ratio between the i-th target object and the j-th target object based on the occlusion angles corresponding to the i-th target object and the j-th target object includes: The occlusion angles corresponding to the i-th and j-th targets are transformed into polar coordinates for representation, so as to determine the starting angle and ending angle of the first occlusion angle corresponding to the i-th target, and the starting angle and ending angle of the second occlusion angle corresponding to the j-th target. Based on the first occlusion angle start angle, the first occlusion angle end angle, the second occlusion angle start angle, and the second occlusion angle end angle, determine the overlap angle and angle union between the i-th target object and the j-th target object; The angle intersection-union ratio between the i-th target object and the j-th target object is determined based on the ratio between the overlap angle between the i-th target object and the j-th target object and the union of the angles.

8. A screening device for unobstructed targets, characterized in that, include: The acquisition module is used to acquire the 3D sensing data collected by the sensing device; The first determining module is used to perform target detection on the three-dimensional perception data to determine the three-dimensional detection box corresponding to each target object contained in the three-dimensional perception data. The second determining module is used to determine the occlusion angle and importance of each target object based on the vertex coordinates of the three-dimensional detection box corresponding to each target object. The filtering module is used to filter out unobstructed targets from each of the target objects based on the occlusion angle and importance of each target object.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

11. A computer program product, characterized in that, Includes a computer program that, when run on an electronic device, causes the electronic device to perform the method as described in any one of claims 1-7.

Images